阅读说明：本技术 用于运载工具的雷达测程法 (Radar range finding method for vehicles ) 是由 吴镇炯 莫森·莱克哈尔-阿亚特 于 2019-05-28 设计创作，主要内容包括：本公开提供了“用于运载工具的雷达测程法”。描述了与使用安置在运载工具上的一个或多个雷达的运载工具测程法相关的技术和示例。一种用于雷达测程的方法可涉及由处理器从雷达接收位于运载工具所穿过的环境中的静止对象和移动对象的测量数据。所述方法还可涉及由所述处理器执行随机抽样一致性(RANSAC)计算,以选择所述静止对象的所述测量数据并忽略所述移动对象的所述测量数据。所述方法还可涉及由所述处理器基于所述静止对象的所述测量数据计算所述运载工具的一个或多个动态变量。所提出的方法利用单次RANSAC计算和一个最小二乘问题求解对所述静止对象的所述测量数据进行处理,从而大大降低计算成本和时间并缩短操作延迟以提供所述运载工具测程法。(The present disclosure provides a "radar ranging method for a vehicle". Techniques and examples related to vehicle ranging using one or more radars mounted on a vehicle are described. A method for radar ranging may involve receiving, by a processor, measurement data from a radar of stationary and moving objects located in an environment through which a vehicle passes. The method may also involve performing a random sample consensus (RANSAC) calculation by the processor to select the measurement data of the stationary object and to ignore the measurement data of the moving object. The method may also involve calculating, by the processor, one or more dynamic variables of the vehicle based on the measurement data of the stationary object. The proposed method processes the measurement data of the stationary object using a single RANSAC calculation and a least squares solution, thereby greatly reducing computational cost and time and reducing operational delays to provide the vehicle range.)

1. A method, the method comprising:

Receiving, by a processor, measurement data of stationary and moving objects located in an environment through which a vehicle passes from one or more radars disposed on or in the vehicle;

Performing, by the processor, a random sample consensus (RANSAC) calculation to identify the measurement data of the stationary object; and

Calculating, by the processor, one or more dynamic variables of the vehicle based on the measurement data of the stationary object.

2. The method of claim 1, wherein the measurement data comprises a radial distance, a radial velocity, an azimuth, an elevation, or a combination of two or more thereof, of one of the stationary object and the moving object as detected by one of the one or more radars.

3. The method of claim 1, wherein the environment comprises a two-dimensional (2D) plane, and wherein the one or more dynamic variables of the vehicle comprise a longitudinal speed, a lateral speed, and a yaw rate of the vehicle.

4. The method of claim 1, wherein the environment comprises a three-dimensional (3D) space, and wherein the one or more dynamic variables of the vehicle comprise a longitudinal speed, a lateral speed, a vertical speed, a roll rate, a pitch rate, and a yaw rate of the vehicle.

5. The method of claim 1, wherein the performing the RANSAC calculation to identify the measurement data of the stationary object comprises:

Calculating a plurality of representative points in a mathematical space, each of the representative points being associated with respective measurement data of one of the stationary or moving objects as detected by one of the one or more radars;

finding a best-fit plane in the mathematical space, the best-fit plane covering a majority of representative points within a predetermined threshold;

Identifying one or more representative points located within the predetermined threshold from the best-fit plane as interior points; and

Designating a portion of the measurement data associated with the interior point as the measurement data of the stationary object.

6. The method of claim 1, wherein the calculating the one or more dynamic variables comprises:

Forming a least squares problem for the one or more dynamic variables, the least squares problem comprising a set of first order linear equations, each first order linear equation having coefficients related to the measurement data of the stationary object; and

Solving the least squares problem to obtain values for the one or more dynamic variables.

7. the method of claim 1, wherein the calculating the one or more dynamic variables is further based on one or more sets of installation parameters, wherein each of the one or more sets of installation parameters is associated with a corresponding one of the one or more radars, wherein each of the one or more sets of installation parameters includes an installation angle of the corresponding one of the one or more radars and a position vector extending from a reference point of the vehicle to the corresponding one of the one or more radars, and wherein the calculating the one or more dynamic variables comprises:

forming a least squares problem for the one or more dynamic variables, the least squares problem comprising a set of first order linear equations, each first order linear equation having coefficients related to the measurement data of the stationary object and the one or more sets of mounting parameters; and

solving the least squares problem to obtain values for the one or more dynamic variables.

8. The method of claim 1, further comprising:

Manipulating, by the processor, the vehicle based on the one or more dynamic variables.

9. A radar ranging apparatus implementable in a vehicle, the radar ranging apparatus comprising:

One or more radars disposed in or on the vehicle, each of the one or more radars capable of receiving measurement data of one or more objects in an environment through which the vehicle passes;

A memory capable of storing one or more sets of installation parameters, each of the one or more sets of installation parameters associated with a corresponding one of the one or more radars; and

A processor communicatively coupled with the memory and the one or more radars and capable of calculating one or more linear or angular velocities of the vehicle based on the measurement data and the one or more sets of installation parameters.

10. the radar ranging apparatus of claim 9, wherein the one or more objects comprise one or more stationary objects that are stationary in the environment, and wherein the measurement data of the one or more objects comprises a radial distance, a radial velocity, an azimuth angle, an elevation angle, or a combination of two or more thereof, of the one or more stationary objects.

11. The radar ranging apparatus of claim 9, wherein each of the one or more sets of installation parameters comprises an installation angle of the corresponding one of the one or more radars and a position vector extending from a reference point of the vehicle to the corresponding one of the one or more radars, wherein the one or more linear or angular velocities of the vehicle comprise a longitudinal velocity and a lateral velocity of the vehicle, and wherein the processor is further configured to calculate a sideslip angle based on a ratio between the longitudinal velocity and the lateral velocity.

12. The radar ranging apparatus of claim 9, wherein the environment comprises a two-dimensional (2D) plane, and wherein the one or more linear or angular velocities of the vehicle comprise a longitudinal velocity, a lateral velocity, and a yaw rate of the vehicle.

13. The radar ranging apparatus of claim 9, wherein the environment comprises a three-dimensional (3D) space, and wherein the one or more linear or angular velocities of the vehicle comprise a longitudinal velocity, a lateral velocity, a vertical velocity, a roll rate, a pitch rate, and a yaw rate of the vehicle.

14. The radar ranging apparatus of claim 9, wherein:

The one or more objects include one or more stationary objects that are stationary in the environment and one or more moving objects that move in the environment, and

The processor includes:

A random sample consensus (RANSAC) circuit capable of distinguishing respective measurement data of the one or more stationary objects from respective measurement data of the one or more moving objects;

a least squares circuit capable of calculating the one or more linear or angular velocities of the vehicle based on the respective measurement data of the one or more stationary objects and the one or more sets of mounting parameters; and

An object tracking circuit capable of tracking one or more of the moving objects by setting a bounding box around a cluster containing the one or more of the moving objects based on respective measurement data of the one or more of the moving objects.

15. the radar ranging apparatus as recited in claim 9, further comprising:

An autonomous driving controller, wherein the vehicle is an autonomous vehicle, and wherein the processor is further configured to maneuver the autonomous vehicle through the autonomous driving controller based on the one or more linear or angular velocities.

Technical Field

The present disclosure relates generally to automobiles and, more particularly, to methods and apparatus for ranging a moving vehicle using real-time radio detection and ranging ("RADAR") measurements of stationary objects around the moving vehicle.

Background

Odometry uses data acquired or otherwise sensed by various sensors typically provided in the vehicle to estimate the change in position of the vehicle over time, particularly as the vehicle moves. In particular, estimating the change in position of the vehicle requires some dynamic variables of the vehicle, such as the linear and angular velocity of the vehicle. The closer the estimate is to the actual position of the vehicle, the more accurate the ranging method is. Accurate odometry plays an important role in characterizing various dynamic characteristics of a vehicle, including critical data for controlling or otherwise navigating the vehicle, particularly when the vehicle is unmanned or autonomous. For example, for a vehicle such as an automobile, mobile robot, or aerial drone, accurate ranging methods may provide accurate information about the position, orientation, linear and angular velocity of the vehicle relative to its surrounding two-dimensional (2D) or three-dimensional (3D) environment. Dynamic information is of vital importance in applications such as stability control and accurate navigation of vehicles.

Traditionally, for automobiles, odometry may be implemented using wheel odometry to estimate linear velocity and by measuring angular velocity via an Inertial Measurement Unit (IMU) embedded in the vehicle. However, wheel odometry has limited accuracy due to factors such as tire size uncertainty (e.g., user-modified tires or insufficient tire pressure) and low wheel encoder resolution. IMU also has measurement errors, especially in low speed maneuvers. Mobile robots and drones typically use visual odometry (via a camera equipped on the vehicle) with the IMU. However, visual ranging methods often have integral drift problems; that is, measurement errors of linear and/or angular velocity may integrate or otherwise accumulate over time, resulting in random, unbounded drift of the calculated navigational position and/or altitude.

recently, as more and more vehicles are equipped with Advanced Driving Assistance System (ADAS) sensors, which may include GPS or LIDAR, odometry based on Global Positioning System (GPS) and light detection and ranging ("LIDAR" or "LIDAR") technologies have been developed. However, GPS may not be able to function in places where satellite reception is limited (e.g., in tunnels), and lidar is not able to operate in all weather conditions.

furthermore, many existing odometry methods are limited to providing 2D dynamic variables (i.e., linear and angular velocity of the vehicle on a 2D plane) and cannot provide dynamic variables of the vehicle in 3D space. The 2D variables are insufficient to control and navigate a vehicle (e.g., a drone) that traverses in 3D space, which requires 3D range capability.

disclosure of Invention

It is an object of the present disclosure to provide accurate estimation or calculation of dynamic variables of a vehicle using one or more radar devices or radar collected measurement data equipped in or on the vehicle. In particular, doppler radars are used because of their higher measurement accuracy compared to other types of automotive radars. One method according to the present disclosure involves a processor receiving measurement data of stationary and moving objects located in an environment through which a vehicle passes from one or more radars disposed on or in the vehicle. The method also involves the processor performing a random sample consensus (RANSAC) calculation to identify measurement data for the stationary object. The method also involves the processor calculating one or more dynamic variables of the vehicle based on the measurement data of the stationary object.

drawings

Non-limiting and non-exhaustive embodiments of the present disclosure are described with reference to the following figures, wherein like reference numerals refer to like parts throughout the various views unless otherwise specified.

FIG. 1 is an illustration depicting example dynamic variables of a vehicle traversing 2D and 3D environments.

Fig. 2 is a diagram illustrating example target detection for radar ranging in accordance with the present disclosure.

FIG. 3 is a graph illustrating the results of a random sample consensus algorithm according to the present disclosure.

fig. 4 is a flow chart depicting an example process according to an embodiment of the present disclosure.

Fig. 5 is a flow chart depicting another example process in accordance with an embodiment of the present disclosure.

FIG. 6 illustrates a radar target map with and without a stationarity tag in accordance with an embodiment of the present disclosure.

Fig. 7 is a block diagram depicting an example device according to an embodiment of the present disclosure.

Detailed Description

In the following description, reference is made to the accompanying drawings, which form a part hereof, and in which is shown by way of illustration specific exemplary embodiments in which the disclosure may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the concepts disclosed herein, and it is to be understood that various disclosed embodiments may be modified and that other embodiments may be utilized without departing from the scope of the present disclosure. The following detailed description is, therefore, not to be taken in a limiting sense.

As mentioned above, accurate ranging is critical for vehicles that operate primarily on 2D planes (e.g., automobiles) or in 3D space (e.g., aerial drones). That is, in order to control, adjust, calibrate, navigate, or otherwise maneuver the vehicle, it may be desirable to accurately determine dynamic variables of the vehicle, such as linear and angular velocities as shown in fig. 1, by ranging methods. Fig. 1 illustrates a first vehicle 10 (e.g., an automobile) and three dynamic variables associated with the vehicle 10. Fig. 1 also shows a second vehicle 50 (e.g., an air drone) and six dynamic variables associated with the vehicle 50. In particular, a vehicle 10 that is operated to move primarily on the 2D plane 120 may be characterized by three dynamic variables: longitudinal speed 11, lateral speed 12, and yaw rate 13. Longitudinal velocity 11 represents the linear velocity component of vehicle 10 in a longitudinal direction on 2D plane 120 relative to vehicle 10, the longitudinal direction generally toward the front of vehicle 10 and interchangeably referred to as the "x-axis" or "x-direction". The lateral velocity 12 represents a linear velocity component of the vehicle 10 on the 2D plane 120 in a lateral direction of the vehicle 10, the lateral direction being orthogonal to the longitudinal direction and interchangeably referred to as the "y-axis" or "y-direction". The yaw rate 13 represents the angular velocity of the vehicle 10 about an axis (commonly referred to as the "z-axis" or "z-direction") perpendicular to the 2D plane 120.

Similarly, a vehicle 50 that is operated to move primarily in 3D space 130 may be characterized by six dynamic variables: longitudinal speed 51, lateral speed 52, vertical speed 53, roll rate 54, pitch rate 55, and yaw rate 56. Longitudinal speed 51, lateral speed 52, and yaw rate 56 are defined in 3D space 130 relative to vehicle 50 in a similar manner as longitudinal speed 11, lateral speed 12, and yaw rate 13 are defined in 2D plane 120 relative to vehicle 10. That is, longitudinal velocity 51 represents a linear velocity component of vehicle 50 in 3D space 130 relative to vehicle 50 along a longitudinal direction, which may be interchangeably referred to as the "x-axis" or "x-direction". The lateral velocity 52 represents a linear velocity component of the vehicle 50 in the 3D space 130 in a lateral direction of the vehicle 50 that is orthogonal to the longitudinal direction and interchangeably referred to as the "y-axis" or "y-direction". The vertical velocity 53 represents a linear velocity component of the vehicle 50 in the 3D space 130 along a vertical direction of the vehicle 50 that is orthogonal to the longitudinal direction and the lateral direction in the 3D space 130 and is interchangeably referred to as a "z-axis" or a "z-direction". The roll rate 54, pitch rate 55, and yaw rate 56 each represent the angular velocity of the vehicle 50 in the 3D space 130 about the x-axis, y-axis, and z-axis, respectively.

The present disclosure provides methods and apparatus for implementing radar ranging. That is, the present disclosure provides methods and apparatus for providing accurate estimates or calculations of dynamic variables of a vehicle (such as those shown in fig. 1) using measurement data collected by one or more radar devices or radars equipped in or on the vehicle. Preferably, doppler radar is used because of the higher measurement accuracy compared to other types of automotive radar. Although the radar ranging method according to the present disclosure is described below primarily with respect to the vehicle 10 traversing the 2D plane 120 as an example, the same radar ranging method is also applicable to the vehicle 50 traversing the 3D space 130 as another example.

fig. 2 shows a top view of the vehicle 10 and the radar 40 positioned near the front end of the vehicle 10 as the vehicle 10 traverses the 2D plane 120. The radar 40 is used to detect one or more stationary objects (i.e., stationary relative to the 2D plane 120), such as the target 60 in fig. 2, as the vehicle 10 passes over the plane 120. The radar 40 may have a target direction 44, which generally defines a maximum antenna gain axis of the radar 40. That is, the radar 40 may emit a majority of its radiated electromagnetic power along the sighting direction 44 when detecting the surroundings of the vehicle 10. As described above and shown in fig. 2, the x-direction may be defined to align with the forward direction of the vehicle 10, i.e., the direction the vehicle 10 is facing by default. Some installation parameters of the radar 40 may be used to determine the position and orientation of the radar 40 relative to the vehicle 10. The installation parameters may include a 2D position vector from a reference point of the vehicle 10 (e.g., the center of mass or any point of the vehicle 10) to a reference point of the radar 40 (e.g., the center of mass of the radar 40, the origin of the sensor coordinate system, or any point), represented in fig. 2 asthe reference point for the vehicle 10 is denoted as C in fig. 2. Further, the mounting parameters may include a radar mounting angle, denoted as φ in FIG. 2, defined as the angle between the x-direction of the vehicle 10 and the aim direction 44 of the radar 40. The mounting parameters of the radar 40 have known and fixed values once the radar 40 is mounted or otherwise positioned on the vehicle 10.

as shown in fig. 1, radar ranging may provide estimates of dynamic variables of the vehicle 10, including longitudinal speed 11, lateral speed 12, and yaw rate 13. The yaw rate 13 of the vehicle 10 is shown in FIG. 2 asThe longitudinal speed 11 is expressed asAnd the lateral velocity 12 is expressed asThe velocity of the vehicle 10 at the reference point C as the vehicle 10 traverses the 2D plane 120Can be expressed as:

WhereinRepresents the x-axis unit vector of the vehicle 10, andRepresenting the y-axis unit vector of the vehicle 10.

available 2D vector for velocity of target 60Indicating that the vector satisfies the following relationship:

Whereinrepresents a 2D position vector from the radar 40 to the target 60, as shown in FIG. 2, and thusrepresenting the doppler velocity of the target 60 as measured by the radar 40.

The second term on the right side of equation (2) can be written as:

WhereinandAre respectivelyX and y coordinates of (a) and noteAnd is

Refer to FIG. 2 and noteand isThe third term on the right side of equation (2) can be written as:

Where r represents the radial distance of the target 60 as measured by the radar 40,Represents the radial velocity of the target 60 as measured by the radar 40, and θ (also shown in fig. 2) represents the azimuth angle of the target 60 as measured by the radar 40.

using equations (1), (3), and (4), equation (2) can be written as:

Assuming that the target 60 is a stationary object, it specifiesAnd thus the following equation can be derived:

Notably, equation (8) is a first order linear equation with the linear and angular velocities of the vehicle 10 as variables:(i.e., longitudinal velocity 11),(i.e., lateral velocity 12) and(i.e., yaw rate 13). The coefficients of equation (8) include radar measurements of the target 60 by the radar 40 (i.e.,And θ), and installation parameters of the radar 40 (i.e., whereinAnd phi) are used. For each stationary object detected by the radar 40, the associated equation (8) may be written. Thus, when the radar 40 detects multiple stationary objects, a set of equations (8) may be generated, each associated with one of the stationary objects. The set of equations (8), each being a first order linear equation of linear and angular velocity of the vehicle 10, may thus form a least squares problem in a mathematical sense. The least squares problem may then be solved to obtain the linear and angular velocities of the vehicle 10 to implement a odometry of the vehicle 10.

The radar ranging methods described above may also benefit from the cancellation of inherent noise when applied to multiple stationary objects. That is, when the radar 40 measures multiple stationary objects, such as the target 60, the resulting dynamic variables of the vehicle 10 may inherently be more accurate, resulting in estimation of the linear and angular velocities of the vehicle 10 with greater accuracy, since the measurement errors introduced by noise in the radar 40 and in the surrounding environment will cancel out or otherwise be compensated for in solving the least squares problem.

in addition to detecting multiple stationary objects using one radar disposed on the vehicle 10, the radar ranging methods disclosed above may also be applicable to situations where multiple stationary objects are detected by the vehicle 10 using multiple radar devices. For example, the vehicle 10 may be equipped with N radars, and K stationary objects may be present in the surroundings of the vehicle 10 to be detected by each of the N radars. Respectively orderAndIs the radial velocity and azimuth of the kth object detected by the nth radar. Thus, for the kth object detected by the nth radar, equation (8) can be written as:

Wherein N1, N, and K1. Note that equation (9) includes the installation parameters of N radars; that is to say that the position of the first electrode,andAre the x and y coordinates, respectively, of the nth radar relative to a reference point C of the vehicle 10, and phiⁿIs the radar mounting angle of the nth radar with respect to the x-direction. As mentioned above, all radar mounting parameters are known values in the least squares problem. Equation (9) represents the least squares problem for N radars detecting K stationary objects. Written in matrix form, equation (9) can be varied as:

Equation (10) can be simplified or rewritten in the alternative by defining the following matrix:

Note that each of the matrices X, Y, Z and W includes an N · K × 1 matrix, i.e., a matrix having N · K rows and 1 column. It should also be noted that each element of the matrices X, Y, Z and W includes installation parameters of the radar and/or measurement parameters of the stationary object measured by the radar of the vehicle 10. From matrices X, Y, Z and W defined in equations (11) - (14), equation (10) can be rewritten as:

thus, the linear and angular velocities of the vehicle 10 are(i.e., longitudinal velocity 11),(i.e., lateral velocity 12) and(i.e., yaw rate 13) may be obtained substantially simultaneously by solving a least squares problem represented by equation (15).

In some embodiments, the angular velocity in equation (9)(i.e., yaw rate 13) may be considered a known variable. For example, the vehicle 10 may rely on a gyroscope or other IMU provided in the vehicle 10 to determine the yaw rate 13. In this case, the least squares problem represented by equation (9) can be written in matrix form as follows:

where the matrix Z' (also N · K × 1 matrix) is defined as:

Similar to matrices X, Y, Z and W, each element of matrix Z' also includes installation parameters of the radar and/or measurement parameters of the stationary object measured by the radar of vehicle 10. Thus, the linear velocity of the vehicle 10, i.e. the yaw rate 13, is considered to be a known parameter(i.e., longitudinal velocity 11) and(i.e., lateral velocity 12) can be obtained by solving a least squares problem represented by equation (17).

notably, the above equation is derived based on the assumption that each of the N radar-detected K objects is a stationary object. That is, each of the K objects has zero velocity on the 2D plane 120 or in the 3D space 130, and thus each of the matrices X, Y, Z, Z' and W has a size of N · K elements, where each element corresponds to a respective one of the K objects detected by a respective one of the N radars. However, this assumption is not valid in a real environment. The actual environment may include both stationary objects (i.e., objects that are stationary in the environment) and moving objects (i.e., objects that are moving in the environment). For example, in addition to K objects that are stationary, the actual environment may include P objects that are moving. A radar disposed in or on the vehicle 10 or the vehicle 50 will detect (K + P) objects, whether stationary or moving, in the surroundings of the vehicle 10 or the vehicle 50. That is, the measurement data obtained by the N radars (e.g., radial distance, radial velocity, and azimuth of the detected object) may include data from both K stationary objects (desired) and P moving objects (undesired). While the measurement data from a stationary object may be used to construct the correct least squares problem (e.g., as shown in equations 9, 10 or 15) that can be solved to obtain the dynamic variables of the vehicle 10 or vehicle 50, the measurement data from a moving object is not needed because it interferes with or otherwise distorts the least squares problem, which when solved for the least squares problem, can result in inaccurate estimates of the dynamic variables of the vehicle 10 or vehicle 50. Therefore, a screening method is required to distinguish measurement data of a stationary object from measurement data of a moving object.

To eliminate most, if not all, of the unwanted measurement data from moving objects, a data filtering algorithm or selection process may be used. To this end, one or more different data processing techniques may be used, and these techniques may be implemented using software methods, hardware methods, or a combination of both. In one embodiment below, the random sample consensus (RANSAC) algorithm is used as an example, but is not intended to limit the scope of the present disclosure. RANSAC is an iterative method that can estimate the parameters of a mathematical model from a set of observations that contain outliers. That is, among the measurement data of (K + P) objects detected by the N radars, if the number of detections of a stationary object is dominant, the detection of a moving object will appear as an outlier.

If the measurement data of (K + P) objects detected by N radars are plotted in the mathematical coordinate space, the outliers can be intuitively identified. Take the least squares problem represented by equation (17) as an example. For each detection of one of the (K + P) objects by one of the N radars, a corresponding representative point may be plotted in 3D mathematical space, where the representative point isthe three coordinates of the point have the values of the corresponding elements of the matrices X, Y and Z', respectively. In particular, for the kth object of the (K + P) objects detected or otherwise measured by the nth radar of the N radars, a representative point may be plotted in mathematical space to represent the measurement, where each coordinate of the representative point is a corresponding element of the matrices X, Y and Z', i.e., the matrix X, Y and ZAnd

Fig. 3 shows an example mathematical space 300 and some representative points within the mathematical space 300 that represent, or are otherwise derived from, measurement data of some stationary and moving objects detected by the radar of the vehicle 10. As shown in fig. 3, the representative points of the measurement data for detecting the stationary object are mainly located on the 2D plane 320 (referred to as "optimal plane") or within a predetermined proximity in the vicinity of the plane 320 because the coordinates of these representative points satisfy equation (9) or equivalently equation (17). These representative points located on or near the best-fit plane 320 are called "interior points" and are considered by the RANSAC algorithm to represent measurement data from stationary objects in the environment surrounding the vehicle 10. The interior points are used to construct or otherwise form a least squares problem represented by equation (17), and in which the dynamic variables (i.e., linear velocity) of the vehicle 10And) Can be obtained accordingly by solving equation (17).

On the other hand, other representative points away from the optimal plane 320 in fig. 3 are referred to as "outliers". Outliers are considered by the RANSAC algorithm to represent measurement data from moving objects in the vehicle 10 surroundings. The outliers are not located on or near the plane 320 because they represent measurement data from the moving object and therefore the coordinates of these outliers do not satisfy equation (9) or equation (17). In contrast to the interior points, the exterior points are excluded from being used to construct or otherwise form the least squares problem represented by equation (17).

in short, the RANSAC algorithm aims to find the best-fit plane 320 that fits most of the representative points in fig. 3. Once the optimal plane 320 is found, the least squares problem of the linear and/or angular velocity of the vehicle that the radar ranging method aims at determining is constructed using only the interior points associated with the plane 320. The least squares problem is then solved to determine linear and/or angular velocities.

the RANSAC iterative process described above for the vehicle 10 of fig. 1 may be implemented using the process 400 shown in fig. 4. Process 400 may include one or more operations, actions, or functions as illustrated by blocks 410, 420, 430, 440, 450, 460, 470, 480, and 490, etc. in fig. 4. Although shown as separate blocks, the various blocks of process 400 may be divided into additional blocks, combined into fewer blocks, or eliminated, depending on the desired implementation. Process 400 may begin at block 410.

at 410, process 400 may involve a processor (e.g., a computer or an application specific integrated circuit) setting a RANSAC threshold ζ having a positive value and a maximum number of RANSAC iterations M as a positive integer. Process 400 may continue from 410 to 420.

At 420, process 400 may involve the processor calculating coordinates of representative points in the 3D mathematical space based on the radar measurement data, where each representative point corresponds to detection of one of the K objects by one of the N radars. As described above, each coordinate representing a point may include corresponding elements of matrices X, Y and Z', as defined in equations (11), (12), and (18), i.e.Andprocess 400 may continue from 420 to 430.

At 430, process 400 may involve the processor randomly selecting three representative points and finding a 2D plane that encompasses the three randomly selected representative points. Process 400 may continue from 430 to 440.

At 440, process 400 may involve the processor calculating a distance between each representative point and the 2D plane found at block 430. The calculation of the distance may be performed by a series of vector operations. For example, let the three randomly selected representative points at block 430 beAndComprisesAndOf the 2D planeCan be given by:

the distance between the 2D plane and the representative point in the 3D mathematical space can be found by: first find the vector between the representative point and the 2D plane, then calculate the vector sumDot product between. Specifically, letIs the nth representative point in the 3D mathematical space.And comprisesAndIs located at a distance D from the 2D plane_nGiven by:

process 400 may continue from 440 to 450.

At 450, process 400 may involve the processor calculating or counting the number of representative points that are less than ζ away from the 2D plane. The more representative points within ζ that are close to the 2D plane, the more suitable the 2D plane separates the outer points from the inner points. That is, the optimal 2D plane can most accurately distinguish between outliers and inliers. Thus, the inliers of the measurement data determined by the best-fit 2D plane found by the RANSAC iterative process 400 will form a least squares problem with minimal distortion or interference caused by the measurement data from the outliers (i.e., the measurement data from the moving object). Process 400 may continue from 450 to 460.

At 460, process 400 may involve the processor determining whether the count obtained at block 450 is greater than the previous highest count obtained in the RANSAC iteration. If the count obtained at block 450 is not greater than the previous highest count, the process 400 may continue from 460 to 480 without updating the best 2D plane at block 470. If the count obtained at block 450 is greater than the previous highest count, the process 400 may continue from 460 to 470.

at 470, process 400 may involve the processor updating the best 2D plane to the 2D plane found in the current iteration at block 430. Process 400 may continue from 470 to 480.

At 480, the process 400 may involve the processor checking whether the maximum number of iterations M set at block 410 is reached. If the maximum number of iterations has not been reached, process 400 may continue from 480 to 430 for another iteration round. If the maximum number of iterations is reached, process 400 may continue from 480 to 490.

At 490, process 400 may involve the processor separating the interior and exterior points among the representative points based on the best-fit 2D plane found in the iteration. Then, the interior points can be used to construct a least squares problem, such as the least squares problem represented by equation (17).

although the radar ranging methods described in detail above (including equations (1) - (21) and process 400) are illustrated using a vehicle that traverses a 2D plane (e.g., vehicle 10 traversing 2D plane 120, as shown in fig. 1), the same methods may also be applied to radar ranging methods that implement a vehicle that traverses a 3D space (e.g., vehicle 50 traversing a 3D space, as shown in fig. 1). That is, for a vehicle traversing a 3D space, the radar ranging method may likewise involve detecting multiple stationary and moving objects within the 3D space using multiple radars of the vehicle traversing the 3D space to obtain measurement data, performing RANSAC calculations based on the measurement data to identify the measurement data from the stationary objects, and forming a least squares problem of the dynamic variables of the vehicle using the measurement data from the stationary objects. For a target in 3D space, the measurement data may include elevation (i.e., the angle between the ground and a position vector extending from the radar of the vehicle 50 to the target) in addition to radial distance, radial velocity, and azimuth for the target on the 2D plane. That is, for a 3D embodiment of radar ranging, equations similar to equations (1) - (21) above for a 2D embodiment of radar ranging may be derived, and a RANSAC process similar to process 400 of fig. 4 may be employed to accurately estimate linear and angular velocities of a vehicle traversing a 3D space, such as longitudinal velocity 51, lateral velocity 52, vertical velocity 53, roll rate 54, pitch rate 55, and yaw rate 56 of vehicle 50 as shown in fig. 1. Notably, for 3D embodiments, each representative point may be a point in seven-dimensional hyperspace in view of up to six dynamic variables to be solved for, and thus have up to six coordinate components. The best-fit plane to be found in the RANSAC iteration will be a six-dimensional hyperplane close to most of the representative points.

fig. 5 shows a flowchart depicting an example process 500 of a radar ranging method, in accordance with an embodiment of the present disclosure. Process 500 may include one or more operations, actions, or functions as indicated by the blocks 510, 520, 530, 540, 550, and 560, etc. in FIG. 5. Although shown as separate blocks, the various blocks of process 500 may be divided into additional blocks, combined into fewer blocks, or eliminated, depending on the desired implementation. The process 500 may be applicable to both the vehicle 10 of fig. 1 through its 2D plane 120 and the vehicle 50 of fig. 1 through its 3D space 130. Process 500 may begin at block 510.

At 510, process 500 may involve a processor (e.g., a computer or an application specific integrated circuit) receiving measurement data of a fixed object and/or a moving object detected or measured by one or more radars disposed in or on a vehicle as the vehicle traverses a surrounding environment. As described above, the measurement data may include radial distance, radial velocity, azimuth, elevation, or a combination of two or more thereof, of a stationary object or a moving object (e.g., target 60 of fig. 2). Process 500 may continue from 510 to 520.

at 520, process 500 may involve the processor performing a RANSAC calculation (e.g., process 400 in fig. 4) using the measurement data received at 510. The results of the RANSAC calculation can be used to identify which of the measurement data received at 510 are from stationary objects and which of the measurement data received at 510 are from moving objects. For example, as shown in FIG. 3, measurement data associated with the "interior points" (i.e., representative points located on or near the plane 320) of FIG. 3 may be identified as measurement data from a stationary object. On the other hand, measurement data associated with the "outer points" (i.e., representative points away from plane 320) of fig. 3 may be identified as measurement data from a moving object. Process 500 may continue from 520 to 530.

at 530, process 500 may involve the processor forming a least squares problem for one or more dynamic variables (e.g., linear velocities 11, 12, 51, 52, and 53 and angular velocities 13, 54, 55, and 56 shown in fig. 1) using the measurement data from the stationary object identified at 520. As described above, the least squares problem may include a set of first order linear equations (15) or (17) as shown above) for one or more dynamic variables. Each first order linear equation having a measured data phase with a stationary object detected by the radarthe correlation or related coefficients, and the radar mounting parameters (mounting angle phi and position vector shown in FIG. 2)). For example, each element of the matrices X, Y, Z, Z' and W derived above is a coefficient of a first order linear equation of the least squares problem. Process 500 may continue from 530 to 540.

at 540, the process 500 may involve the processor solving a least squares problem formed at 530 and obtaining vehicle dynamic variables characterizing how the vehicle traverses the vehicle surroundings. For example, the longitudinal speed 11, the lateral speed 12, and the yaw rate 13 of the vehicle 10 in fig. 1 may be obtained substantially simultaneously by solving a least squares problem represented in equation (17) using equation (19).

In some embodiments, at 540, process 500 may also involve the processor calculating one or more secondary variables based on the resulting dynamic variables. For example, the processor may solve a least squares problem and obtain the longitudinal and lateral velocities of the vehicle, and further calculate the sideslip angle of the vehicle (as a secondary variable) directly from the ratio between the longitudinal and lateral velocities. The sideslip angle is defined as the angle between the actual direction of travel of the vehicle and the expected heading of the vehicle. The sideslip angle plays an important role in many areas of vehicle dynamics, including vehicle stability control. However, existing odometry methods are difficult to estimate the sideslip angle, especially due to the inaccuracy of the estimation of the lateral velocity of the vehicle. In contrast, the radar ranging method according to the present disclosure can calculate the slip angle directly from the longitudinal velocity and the lateral velocity in a real-time manner, thereby estimating the slip angle with high accuracy.

In some embodiments, process 500 may end at 540. In some embodiments, process 500 may continue from 540 to perform further estimation, computation, application, or data post-processing. For example, the process 500 may continue from 540 to 550 to maneuver the vehicle, or from 540 to 560 to identify and track moving objects, as described below. It should be noted that 550 and 560 are examples of additional steps that the processor may perform after step 540 and are not intended to be limiting.

At 550, the process 500 may involve the processor manipulating the vehicle based on the dynamic variables and/or other secondary variables obtained at 540, particularly when the vehicle is an autonomous vehicle. That is, the processor may control, adjust, calibrate, or navigate the vehicle based on the linear velocity, angular velocity, and/or sideslip angle obtained at 540, sometimes through autonomous driving controls provided in the vehicle.

at 560, process 500 may involve the processor identifying one or more of the moving objects based on the dynamic variables and/or other secondary variables obtained at 540. Process 500 may also involve the processor tracking one or more moving objects as the vehicle traverses the surrounding environment. In some embodiments, the processor may identify and/or track one or more moving objects by placing a virtual box (referred to as a bounding box) around one or more of the moving objects. The bounding box may be arranged around a cluster of moving objects (i.e. a plurality of moving objects that are close in space), in particular when they are moving at substantially the same speed. A bounding box around the moving object cluster may be considered by the processor as a moving vehicle in the surrounding environment.

Fig. 6 may be used to further illustrate the above-described bounding box concept. As shown in fig. 6, a vehicle equipped with one or more radars (referred to as a "host vehicle" in fig. 6) is represented by a solid color box at the center of each of graphs 61, 62, and 63, the box being around a location having coordinates (0, 0). The graph 61 of fig. 6 shows a radar target graph (where the host vehicle is located at (0,0) of the graph) and a plurality of stationary or moving objects detected by the radar of the host vehicle. Note that a physical entity in the environment, such as a bus, truck, billboard, or building, may generate more than one object marker in chart 61. For example, when the truck is driven by the host vehicle, various portions of the truck at different locations of the truck may each be detected by the radar of the host vehicle, thereby producing corresponding object markers on the radar target map of the graph 61.

Due to already being provided withforming a least squares problem in block 530 of the process 500 and then solving for the host vehicle's dynamic variables in block 540 of the process 500, the processor of the process 500 can determine the stationarity of the object, i.e., identify whether the object detected by the radar is moving or stationary. In particular, equation (8) may be used to determine the stationarity of objects in the environment through which the vehicle 10 in fig. 1 is passing. Order toAndRespectively, a radial velocity measurement and an azimuth angle measurement of the nth radar-detected object of the vehicle 10. Based on the longitudinal velocity obtained in block 540 of process 500Transverse velocityand yaw rate/yaw rateAn object is considered stationary if and only if the following conditions are met:

Whereinis the additive noise of the radial velocity measurements in the nth radar. Note that the left side of (22) depends only onAndThey are available every cycle of radar operation. Note also, condition (22)is at an upper limit of(typically smaller numbers) because doppler radar is the most accurate measurement of the various types of automotive radar.

Once the processor determines the degree of stationarity for each object in graph 61 using conditions (22), the processor may update graph 61 with the stationarity information to produce graph 62 of FIG. 6, where stationary objects are marked differently than moving objects. The processor may also continue to place or otherwise set a bounding box around the moving object cluster, particularly when the moving object cluster is moving in the same direction at a similar speed. As shown in diagram 63 of FIG. 6, bounding boxes 631, 632, 633, 634, 635 and 636 are disposed around or otherwise assigned to several clusters of moving objects therein. Each bounding box may represent a moving physical entity, such as a truck driven by a host vehicle.

Notably, the processor may be capable of performing or otherwise implementing step 510 of process 500 within a relatively short period of time (e.g., a few milliseconds to a few hundred milliseconds) such that process 500 may provide real-time or pseudo real-time radar ranging in an application sense.

Fig. 7 illustrates an example block diagram of a radar ranging apparatus 700 (hereinafter "apparatus 700") that may be implemented with a vehicle, such as the vehicle 10 and the vehicle 50 of fig. 1. As shown in fig. 7, apparatus 700 may include one or more radar devices, such as radars 740(1), 740(2), …, and 740 (N). The device 700 may also include a processor 710 and a memory 720 accessible to the processor 710. Processor 710 may be communicatively coupled to radars 740(1) -740 (N) and to memory 720. The memory 720 may store one or more sets of installation parameters, and each set of installation parameters is associated with a respective one of the radars. For each radar, the associated set of mounting parameters may include a mounting angle of the radar, and a position of the radar relative to a vehicle reference point, which is typically represented by a position vector extending from the vehicle reference point to the radar.

In some embodiments, processor 710 may include RANSAC circuitry 712 capable of performing a RANSAC calculation as implemented in block 520 of process 500 of fig. 5. That is, RANSAC circuit 712 may be able to perform a RANSAC calculation as described in process 400 of fig. 4 and find a best-fit plane (e.g., plane 320 in fig. 3) that may be used to separate or otherwise distinguish between interior points representing measurement data from stationary objects and exterior points representing measurement data from moving objects, as shown in fig. 3.

in some embodiments, the processor 710 may also include a least squares circuit 714 that may be used to obtain the instantaneous linear (e.g., 11, 12, 51, 52, and 53 in fig. 1) and/or angular (e.g., 13, 54, 55, and 56 in fig. 1) velocity of the vehicle during each radar cycle. In particular, the least squares circuit 714 may be capable of forming a least squares problem of linear and/or angular velocity, such as the least squares problem represented by equation (9), (15), or (17), based on measurement data from a stationary object. Further, the least squares circuit 714 may also be capable of solving a least squares problem to obtain linear and/or angular velocity values of the vehicle. In some embodiments, the processor 710 may also be capable of calculating one or more secondary variables, such as a sideslip angle, based on the linear and/or angular velocity values obtained by the least squares circuit 714.

In some embodiments, the processor 710 may further include an object tracking circuit 716 that is capable of tracking one or more moving objects on the radar target map by setting a bounding box (such as the bounding box shown in the graph 63 of fig. 6) that surrounds a cluster of moving objects that are movable in substantially the same direction at substantially the same speed.

In some embodiments, the apparatus 700 may also include an autonomous driving controller 730, particularly when the apparatus 700 is implemented in an autonomous vehicle. With the autonomous driving controller 730, the processor 710 may be capable of steering the autonomous vehicle based on the one or more linear or angular velocities obtained by the least squares circuit 714 and a secondary variable (such as sideslip angle) derived by the processor 710 from the one or more linear or angular velocities.

it is obvious that, although described with respect to one or more radar devices, all the aspects described above are equally applicable to any type of sensor capable of measuring doppler frequencies, such as a lidar device or an ultrasonic sensor.

In the foregoing disclosure, reference is made to the accompanying drawings, which form a part hereof, and in which is shown by way of illustration specific implementations in which the disclosure may be practiced. It is to be understood that other implementations may be utilized and structural changes may be made without departing from the scope of the present disclosure. References in the specification to "one embodiment," "an example embodiment," etc., indicate that the embodiment described may include a particular feature, structure, or characteristic, but every embodiment may not necessarily include the particular feature, structure, or characteristic. Moreover, such phrases are not necessarily referring to the same embodiment. Further, when a particular feature, structure, or characteristic is described in connection with an embodiment, it is submitted that it is within the knowledge of one skilled in the art to affect such feature, structure, or characteristic in connection with other embodiments whether or not explicitly described.

Implementations of the systems, apparatus, and methods disclosed herein may include or utilize a special purpose or general-purpose computer including computer hardware (e.g., one or more processors and system memory as discussed herein). Implementations within the scope of the present disclosure may also include physical and other computer-readable media for carrying or storing computer-executable instructions and/or data structures. Such computer-readable media can be any available media that can be accessed by a general purpose or special purpose computer system. Computer-readable media storing computer-executable instructions are computer storage media (devices). Computer-readable media carrying computer-executable instructions are transmission media. Thus, by way of example, and not limitation, implementations of the present disclosure can include at least two distinct computer-readable media: computer storage media (devices) and transmission media.

Computer storage media (devices) include RAM, ROM, EEPROM, CD-ROM, solid state drives ("SSDs") (e.g., based on RAM), flash memory, phase change memory ("PCM"), other types of memory, other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store desired program code means in the form of computer-executable instructions or data structures and which can be accessed by a general purpose or special purpose computer.

Implementations of the apparatus, systems, and methods disclosed herein may communicate over a computer network. A "network" is defined as one or more data links that enable the transfer of electronic data between computer systems and/or modules and/or other electronic devices. When information is transferred or provided over a network or another communications connection (either hardwired, wireless, or any combination of hardwired or wireless) to a computer, the computer properly views the connection as a transmission medium. Transmission media can include a network and/or data links which can be used to carry desired program code means in the form of computer-executable instructions or data structures and which can be accessed by a general purpose or special purpose computer. Combinations of the above should also be included within the scope of computer-readable media.

Computer-executable instructions comprise, for example, instructions and data which, when executed in a processor, cause a general purpose computer, special purpose computer, or special purpose processing device to perform a certain function or group of functions. The computer-executable instructions may be, for example, binary code, intermediate format instructions (such as assembly language), or even source code. Although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the features and acts are disclosed as example forms of implementing the claims.

those skilled in the art will appreciate that the disclosure may be practiced in network computing environments with many types of computer system configurations, including internal vehicle computers, personal computers, desktop computers, laptop computers, message processors, hand-held devices, multi-processor systems, microprocessor-based or programmable consumer electronics, network PCs, minicomputers, mainframe computers, mobile telephones, PDAs, tablet computers, pagers, routers, switches, various storage devices, and the like. The present disclosure may also be practiced in distributed system environments where local and remote computer systems, which are linked (either by hardwired data links, wireless data links, or by any combination of hardwired and wireless data links) through a network, both perform tasks. In a distributed system environment, program modules may be located in both local and remote memory storage devices.

Further, where appropriate, the functions described herein may be performed in one or more of the following: hardware, software, firmware, digital components, or analog components. For example, one or more Application Specific Integrated Circuits (ASICs) can be programmed to perform one or more of the systems and processes described herein. Certain terms are used throughout the description and claims to refer to particular system components. Those skilled in the art will appreciate that components may be referred to by different names. This document does not intend to distinguish between components that differ in name but not function.

It should be noted that the sensor embodiments discussed above may include computer hardware, software, firmware, or any combination thereof to perform at least a portion of their functions. For example, the sensor may include computer code configured to be executed in one or more processors, and may include hardware logic/circuitry controlled by the computer code. These example devices are provided herein for illustrative purposes and are not intended to be limiting. Embodiments of the present disclosure may be implemented in other types of devices, as known to those skilled in the relevant art.

At least some embodiments of the present disclosure relate to computer program products that include such logic (e.g., in the form of software) stored on any computer-usable medium. Such software, when executed in one or more data processing devices, causes the devices to operate as described herein.

While various embodiments of the present disclosure have been described above, it should be understood that they have been presented by way of example only, and not limitation. It will be apparent to persons skilled in the relevant art that various changes in form and detail can be made therein without departing from the spirit and scope of the disclosure. Thus, the breadth and scope of the present disclosure should not be limited by any of the above-described exemplary embodiments, but should be defined only in accordance with the following claims and their equivalents. The foregoing description has been presented for purposes of illustration and description. It is not intended to be exhaustive or to limit the disclosure to the precise form disclosed. Many modifications and variations are possible in light of the above teaching. Further, it should be noted that any or all of the aforementioned alternative implementations may be used in any combination desired to form additional hybrid implementations of the present disclosure.

According to the invention, a method comprises: receiving, by a processor, measurement data of stationary and moving objects located in an environment through which a vehicle passes from one or more radars disposed on or in the vehicle; performing, by the processor, a random sample consensus (RANSAC) calculation to identify the measurement data of the stationary object; and calculating, by the processor, one or more dynamic variables of the vehicle based on the measurement data of the stationary object.

According to an embodiment, the measurement data comprises a radial distance, a radial velocity, an azimuth angle, an elevation angle, or a combination of two or more thereof, of one of the stationary object and the moving object as detected by one of the one or more radars.

According to one embodiment, the environment comprises a two-dimensional (2D) plane, and wherein the one or more dynamic variables of the vehicle comprise a longitudinal speed, a lateral speed, and a yaw rate of the vehicle.

according to one embodiment, the environment comprises a three-dimensional (3D) space, and wherein the one or more dynamic variables of the vehicle comprise a longitudinal velocity, a lateral velocity, a vertical velocity, a roll rate, a pitch rate, and a yaw rate of the vehicle.

According to one embodiment, the performing RANSAC calculations to identify the measurement data of the stationary object comprises: calculating a plurality of representative points in a mathematical space, each of the representative points being associated with respective measurement data of one of the stationary or moving objects as detected by one of the one or more radars; finding a best-fit plane in the mathematical space, the best-fit plane covering a majority of representative points within a predetermined threshold; identifying one or more representative points located within the predetermined threshold from the best-fit plane as interior points; and designating a portion of the measurement data associated with the interior point as the measurement data of the stationary object.

According to one embodiment, said calculating one or more dynamic variables comprises: forming a least squares problem for the one or more dynamic variables, the least squares problem comprising a set of first order linear equations, each first order linear equation having coefficients related to the measurement data of the stationary object; and solving the least squares problem to obtain values for the one or more dynamic variables.

According to one embodiment, the calculating one or more dynamic variables is further based on one or more sets of installation parameters, and wherein each of the one or more sets of installation parameters is associated with a corresponding one of the one or more radars.

According to one embodiment, each of the one or more sets of mounting parameters includes a mounting angle of the corresponding one of the one or more radars and a position vector extending from a reference point of the vehicle to the corresponding one of the one or more radars.

According to one embodiment, said calculating one or more dynamic variables comprises: forming a least squares problem for the one or more dynamic variables, the least squares problem comprising a set of first order linear equations, each first order linear equation having coefficients related to the measurement data of the stationary object and the one or more sets of mounting parameters; and solving the least squares problem to obtain values for the one or more dynamic variables.

According to one embodiment, the vehicle is manipulated by a processor based on the one or more dynamic variables.

According to the present invention there is provided a radar range finding apparatus implementable in a vehicle, the radar range finding apparatus having: one or more radars disposed in or on the vehicle, each of the one or more radars capable of receiving measurement data of one or more objects in an environment through which the vehicle passes; a memory capable of storing one or more sets of installation parameters, each of the one or more sets of installation parameters associated with a corresponding one of the one or more radars; and a processor communicatively coupled with the memory and the one or more radars and capable of calculating one or more linear or angular velocities of the vehicle based on the measurement data and the one or more sets of installation parameters.

According to one embodiment, the one or more objects comprise one or more stationary objects that are stationary in the environment, and wherein the measurement data of the one or more objects comprises a radial distance, a radial velocity, an azimuth, an elevation, or a combination of two or more thereof, of the one or more stationary objects.

According to one embodiment, each of the one or more sets of mounting parameters includes a mounting angle of the corresponding one of the one or more radars and a position vector extending from a reference point of the vehicle to the corresponding one of the one or more radars.

According to one embodiment, the environment comprises a two-dimensional (2D) plane, and wherein the one or more linear or angular velocities of the vehicle comprise a longitudinal velocity, a lateral velocity, and a yaw rate of the vehicle.

according to one embodiment, the environment comprises a three-dimensional (3D) space, and wherein the one or more linear or angular velocities of the vehicle comprise a longitudinal velocity, a lateral velocity, a vertical velocity, a roll rate, a pitch rate, and a yaw rate of the vehicle.

According to one embodiment, the one or more objects comprise one or more stationary objects that are stationary in the environment and one or more moving objects that move in the environment, and the processor comprises random sample consensus (RANSAC) circuitry capable of distinguishing respective measurement data of the one or more stationary objects from respective measurement data of the one or more moving objects.

According to one embodiment, the processor further comprises a least squares circuit capable of calculating the one or more linear or angular velocities of the vehicle based on the respective measurement data of the one or more stationary objects and the one or more sets of mounting parameters.

According to one embodiment, the processor further comprises an object tracking circuit capable of tracking one or more of the moving objects by placing a bounding box around a cluster containing the one or more of the moving objects based on respective measurement data of the one or more of the moving objects.

According to one embodiment, the one or more linear or angular velocities of the vehicle comprise a longitudinal velocity and a lateral velocity of the vehicle, and wherein the processor is further capable of calculating a sideslip angle based on a ratio between the longitudinal velocity and the lateral velocity.

According to one embodiment, the invention also features an autonomous driving controller, wherein the vehicle is an autonomous vehicle, and wherein the processor is further capable of maneuvering the autonomous vehicle through the autonomous driving controller based on the one or more linear or angular velocities.