阅读说明：本技术 基于多源信息融合的移动物体的测速定位方法和系统 (Speed measurement positioning method and system of moving object based on multi-source information fusion ) 是由 于天剑 刘畅 成庶 向超群 伍珣 陈特放 于 2020-07-15 设计创作，主要内容包括：本发明提供了一种基于多源信息融合的移动物体的测速定位方法和系统,通过采用多个测速定位装置同时采集被测移动物体的运行状态相关信息,以获得各个采集状态量,然后通过构建联邦卡尔滤波器模型来融合各个所述采集状态量,以获得最终的所述移动物体的运行状态的融合状态预测量,各个测速定位装置之间实现了各装置之间的优势互补,使得整个测速定位系统的精准度。(The invention provides a speed measurement positioning method and system of a moving object based on multi-source information fusion, which are characterized in that a plurality of speed measurement positioning devices are adopted to simultaneously acquire running state related information of a moving object to be measured so as to obtain each acquisition state quantity, then a Federal Carl filter model is constructed to fuse each acquisition state quantity so as to obtain the final fusion state prediction quantity of the running state of the moving object, and the advantages of the speed measurement positioning devices are complemented, so that the accuracy of the whole speed measurement positioning system is improved.)

1. A speed measurement positioning method of a moving object based on multi-source information fusion is characterized by comprising the following steps:

a plurality of speed measuring and positioning devices are adopted to respectively acquire the state related information of the moving object in real time so as to obtain each acquisition state quantity, the acquisition state quantity comprises the position information and the speed information of the moving object,

constructing a Federal Kalman filter model, wherein the Federal Kalman filter model comprises sub-filters corresponding to the plurality of speed measurement positioning devices and a main filter,

determining a discrete state equation and a measurement equation corresponding to each speed measurement positioning device through each sub-filter, so as to calculate and obtain each measurement state quantity corresponding to each speed measurement positioning device and used for representing state related information of the moving object at the current time according to each acquisition state quantity at the previous time of the current time and the state equation and the measurement equation corresponding to each acquisition state quantity,

and enabling each sub-filter to respectively calculate each sub-state prediction quantity of the state related information of the moving object at the current moment and each sub-mean square error prediction quantity corresponding to the sub-state prediction quantity according to each measured state information quantity at the current moment, respectively fusing each sub-state prediction quantity and each sub-mean square error prediction quantity at the current moment through the main filter so as to obtain a fusion state prediction quantity and a fusion mean square error prediction quantity of the moving object at the current moment, and obtaining a positioning speed measurement result of the moving object at the current moment according to the fusion state prediction quantity and the fusion mean square error prediction quantity at the current moment.

2. The speed measurement positioning method according to claim 1, further comprising: feeding back the obtained fusion state prediction quantity to at least part of each sub-filter through the main filter, so that the difference between the sub-state prediction quantity generated in each sub-filter and the fusion state prediction quantity is reduced continuously, and feeding back the fusion mean square error prediction quantity P_g(k) Is multiplied by the set weight distribution coefficients and then is respectively fed back to the sub-filter modules so as to be in each sub-filter according to a formula

the higher the speed measurement positioning accuracy of each speed measurement positioning device in the corresponding speed range is, the higher the weight distribution coefficient corresponding to the speed measurement positioning device is.

3. The speed measurement positioning method according to claim 2, further comprising: setting the weight distribution coefficient corresponding to the failed speed measuring and positioning device to be 0 during the fault period of one of the plurality of speed measuring and positioning devices, and distributing the weight distribution coefficient corresponding to the failed speed measuring and positioning device to each sub-filter corresponding to each non-failed speed measuring and positioning device according to the speed of the moving object during the fault period and the speed measuring and positioning accuracy of other non-failed positioning devices within the corresponding speed,

And during the fault period, calculating the fusion state prediction quantity and the fusion mean square error prediction quantity at each moment in the fault period according to the measurement state quantity at the previous moment of the fault, the measurement state quantity at the current moment and the fusion state prediction value through the Federal Kalman filter model.

4. The method according to claim 3, wherein the moving object is a magnetic-levitation train, the plurality of speed-measuring positioning devices comprises 3 positioning-side devices, the first is a Beidou navigation system, the second is an inertial navigation system, and the third is a Doppler radar system.

5. The speed measurement positioning method according to claim 4, wherein the discrete state equation corresponding to the i-th speed measurement positioning device in the plurality of speed measurement positioning devices is as follows: x_i(k)＝Φ_i(k,k-1)X_i(k-1)+_iW_i(k-1), the measurement equation corresponding to the ith speed measurement positioning device is as follows: z_i(k)＝H_iX(k)+M_i(k)，

Wherein, X_i(k) The state variable of the collected state quantity comprises the position information S of the maglev train collected by the ith speed measuring and positioning device_i(k) And velocity v_i(k)，_iA system noise moving matrix corresponding to the ith speed measuring and positioning device is obtained; phi _i(k, k-1) is the ith speed measuring and positioning device pairResponsive to a state one-step transition matrix, W_i(k-1) is the system noise at the previous moment k-1 of the moment k corresponding to the ith speed measurement positioning device, and Z_i(k) The measurement state quantity H corresponding to the ith speed measurement positioning device_iMeasuring matrix for the ith speed measuring and positioning device, M_iAnd measuring noise for the ith speed measuring and positioning device.

6. The speed measurement positioning method according to claim 5, wherein during the data processing of the state related information of the maglev train acquired by the inertial navigation system and the doppler radar system to obtain each acquired state quantity, error compensation is performed on the position information of the maglev train obtained during the processing according to the fed-back fusion state prediction quantity.

7. The method according to claim 6, further comprising initializing a velocity measurement positioning system for executing the velocity measurement positioning method, and calculating the fusion state prediction quantity at the current time by the Federal Kalman filter model according to the initial value of each sub-state prediction quantity, the initial value of each mean square error prediction quantity, and the measurement state quantity at the current time, by giving the initial value of each sub-state prediction quantity at the initial time and the initial value of each mean square error prediction quantity.

8. The utility model provides a speed measurement positioning system of moving object based on multisource information fusion which characterized in that includes: a speed measuring and positioning module, a communication module, an information processing module, a data storage module and a positioning output module,

the speed measuring and positioning module comprises a plurality of speed measuring and positioning devices, the plurality of speed measuring and positioning devices respectively collect the state related information of the moving object in real time to obtain each collected state quantity, the collected state quantity comprises the position information and the speed information of the moving object,

the information processing module determines a discrete state equation and a measurement equation corresponding to each speed measurement positioning device by constructing a Federal Kalman filtering model comprising a sub-filter and a main filter corresponding to the plurality of speed measurement positioning devices, so as to calculate and obtain each measurement state quantity corresponding to each speed measurement positioning device and used for representing state related information of the moving object at the current moment according to each acquisition state quantity at the previous moment at the current moment and the state equation and the measurement equation corresponding to each acquisition state quantity,

the information processing module respectively calculates each sub-state prediction quantity of the state related information of the moving object at the current moment and each sub-mean square error prediction quantity corresponding to the sub-state prediction quantity through each sub-filter according to each measured state information quantity at the current moment, and respectively fuses each sub-state prediction quantity and each sub-mean square error prediction quantity at the current moment through the main filter so as to obtain a fusion state prediction quantity and a fusion mean square error prediction quantity,

The positioning output module outputs the positioning speed measurement result of the moving object according to the information generated by the information processing module,

the data storage module is used for storing various data transmitted by the communication module and various data generated in the processing process of the information processing module so as to provide data for the information processing module to process information.

9. The system according to claim 8, wherein the main filter feeds back the obtained fused state prediction to at least some of the sub-filters, so that the difference between the sub-state prediction and the fused state prediction generated in each sub-filter is reduced, and the fused mean square error prediction P is obtained_g(k) Is multiplied by the set weight distribution coefficients and then is respectively fed back to the sub-filter modules so as to be in each sub-filter according to a formulaTo update and optimize each of the sub-mean square error predictors P_i(k) I is the serial number of N speed measuring and positioning devices, i is more than or equal to 1 and less than or equal to N, beta_iDistributing coefficients for the weights corresponding to the ith speed measuring and positioning device, The k is the number of the time instant,

the higher the speed measurement positioning accuracy of each speed measurement positioning device in the corresponding speed range is, the higher the weight distribution coefficient corresponding to the speed measurement positioning device is,

the speed measuring and positioning system also comprises a working state detection module which is used for detecting whether each speed measuring and positioning device is in a fault state or not so as to output a fault indication signal to the main filter,

the main filter sets the weight distribution coefficient corresponding to the failed speed measurement positioning device to be 0, and distributes the weight distribution coefficient corresponding to the failed speed measurement positioning device to each sub-filter corresponding to each non-failed speed measurement positioning device according to the speed of the moving object during the failure period and the speed measurement positioning accuracy of other non-failed positioning devices within the corresponding speed,

and during the fault period, the information processing module calculates the fusion state prediction quantity and the fusion mean square error prediction quantity at each moment in the fault period according to the measurement state quantity at the moment before the fault occurs, the measurement state quantity at the current moment and the fusion state prediction value.

10. The system of claim 9, wherein the moving object is a magnetic levitation train, the motion model is a uniform acceleration motion model, the first one of the plurality of speed measurement positioning devices is a Beidou navigation system, the second one is an inertial navigation system, and the third one is a Doppler radar system, wherein the inertial navigation system is a strapdown inertial navigation system,

and the information processing module feeds the fusion state prediction quantity back to the inertial navigation system and the Doppler radar system so as to perform error compensation on the position information of the magnetic-levitation train obtained in the processing process according to the fed-back fusion state prediction quantity in the process of obtaining each acquisition state quantity by the inertial navigation system and the Doppler radar system.

Technical Field

The invention belongs to the technical field of train speed measurement and positioning, and particularly relates to a speed measurement and positioning method and system of a moving object based on multi-source information fusion.

Background

With the rapid development of the transportation industry in China, the railway transportation industry represented by high-speed and heavy-load trains has become a supporting industry in China and is the comprehensive embodiment of China advancing to the world of the strong modernization science and technology. The magnetic suspension train is an advanced rail transportation mode and is one of the future development directions of high-speed railways in China and even the world. In the daily railway operation process, the precision and the reliability of the real-time position and speed information provided by the train speed measuring and positioning system for the train not only ensure the normal operation of the train, but also directly influence the operation core criterion of 'train safety', and the importance of the system in the field of high-speed special passenger trains is particularly remarkable. Therefore, how to develop an accurate and reliable novel speed measurement positioning system to adapt to the rapid innovation of the modern high-speed railway industry and how to lay a good foundation for the wide-range popularization of future magnetic suspension trains becomes the leading research direction of scientific researchers in various industries at home and abroad.

The magnetic levitation train is different from a common wheel-rail contact train, and the non-contact guidance between the train and a rail is realized by generating electromagnetic force through a linear motor based on the electromagnetic induction principle. Therefore, the maglev train cannot perform speed measurement and positioning of the train by a wheel-rail contact method similar to a method of installing a tachometer on an axle. The current speed measurement positioning method of the maglev train is divided according to the reference of position information and can be divided into two categories of a relative positioning method and an absolute positioning method, and the method for calculating the real-time relative position information of the maglev train based on the initial position and the real-time displacement of the maglev train is called a relative positioning method. The method mainly comprises the following steps: the speed measurement positioning method is based on induction loop, sleeper counting, long stator tooth space detection, Doppler radar and Inertial Navigation System-INS (Inertial Navigation System-INS) and other principles. The method for acquiring the absolute position of the train based on the trackside known position device is called an absolute positioning method, and mainly comprises the following steps: the speed measurement positioning method is based on a Global Navigation Satellite System (Global Navigation Satellite System-GNSS), a query-transponder and a pulse width coding principle.

The precision and the overall reliability of train speed measurement and positioning are key performance indexes for evaluating the system. The speed measurement positioning accuracy of the train is closely related to the accuracy of each type of sensor, but the solution for improving the accuracy of the sensor is not realistic due to the comprehensive consideration of a plurality of factors such as production cost, technology, manufacturing process and the like. Therefore, the existing relative positioning method or absolute positioning method using a single sensor cannot meet the requirement of the high-speed maglev train on speed measurement and positioning due to the limited detection range and low measurement precision. Aiming at the defect that a single sensor has respective application limitation, the prior art provides an integrated speed measuring and positioning system with multi-sensor information fusion, position information is determined through the running distance of a train, the speed information of the train is measured by means of the multi-sensor, but the actual efficiency of the method is limited by various unknown external environmental factors, and meanwhile, the complex system composition causes high cost.

Disclosure of Invention

In view of this, the present invention provides a speed measurement positioning method and system for a mobile object with multi-source information fusion, so as to solve the problems of low speed measurement accuracy and high cost of the existing speed measurement positioning method and system.

A speed measurement positioning method of a moving object based on multi-source information fusion comprises the following steps:

a plurality of speed measuring and positioning devices are adopted to respectively acquire the state related information of the moving object in real time so as to obtain each acquisition state quantity, the acquisition state quantity comprises the position information and the speed information of the moving object,

constructing a Federal Kalman filter model, wherein the Federal Kalman filter model comprises sub-filters corresponding to the plurality of speed measurement positioning devices and a main filter,

determining a discrete state equation and a measurement equation corresponding to each speed measurement positioning device through each sub-filter, so as to calculate and obtain each measurement state quantity corresponding to each speed measurement positioning device and used for representing state related information of the moving object at the current time according to each acquisition state quantity at the previous time of the current time and the state equation and the measurement equation corresponding to each acquisition state quantity,

And enabling each sub-filter to respectively calculate each sub-state prediction quantity of the state related information of the moving object at the current moment and each sub-mean square error prediction quantity corresponding to the sub-state prediction quantity according to each measured state information quantity at the current moment, respectively fusing each sub-state prediction quantity and each sub-mean square error prediction quantity at the current moment through the main filter so as to obtain a fusion state prediction quantity and a fusion mean square error prediction quantity of the moving object at the current moment, and obtaining a positioning speed measurement result of the moving object at the current moment according to the fusion state prediction quantity and the fusion mean square error prediction quantity at the current moment.

Preferably, the speed measurement positioning method further includes: feeding back the obtained fusion state prediction quantity to at least part of each sub-filter through the main filter, so that the difference between the sub-state prediction quantity generated in each sub-filter and the fusion state prediction quantity is reduced continuously, and feeding back the fusion mean square error prediction quantity P_g(k) Is multiplied by the set weight distribution coefficients and then is respectively fed back to the sub-filter modules so as to be in each sub-filter according to a formula To update and optimize each of the sub-mean square error predictors P_i(k) I is the serial number of N speed measuring and positioning devices, i is more than or equal to 1 and less than or equal to N, beta_iDistributing coefficients for the weights corresponding to the ith speed measuring and positioning device,

the k is the time number, and the k is the time number,

the higher the speed measurement positioning accuracy of each speed measurement positioning device in the corresponding speed range is, the higher the weight distribution coefficient corresponding to the speed measurement positioning device is.

Preferably, the speed measurement positioning method further comprises: setting the weight distribution coefficient corresponding to the failed speed measuring and positioning device to be 0 during the fault period of one of the plurality of speed measuring and positioning devices, and distributing the weight distribution coefficient corresponding to the failed speed measuring and positioning device to each sub-filter corresponding to each non-failed speed measuring and positioning device according to the speed of the moving object during the fault period and the speed measuring and positioning accuracy of other non-failed positioning devices within the corresponding speed,

and during the fault period, calculating the fusion state prediction quantity and the fusion mean square error prediction quantity at each moment in the fault period according to the measurement state quantity at the previous moment of the fault, the measurement state quantity at the current moment and the fusion state prediction value through the Federal Kalman filter model.

Preferably, the moving object is a magnetic-levitation train, the plurality of speed-measuring positioning devices include 3 positioning-side devices, the first is a beidou navigation system, the second is an inertial navigation system, and the third is a doppler radar system.

Preferably, the discrete state equation corresponding to the ith speed measurement positioning device in the plurality of speed measurement positioning devices is as follows: x_i(k)＝Φ_i(k,k-1)X_i(k-1)+_iW_i(k-1), the measurement equation corresponding to the ith speed measurement positioning device is as follows: z_i(k)＝H_iX(k)+M_i(k)，

Wherein, X_i(k) The state variable of the collected state quantity comprises the position information S of the maglev train collected by the ith speed measuring and positioning device_i(k) And velocity v_i(k)，_iA system noise moving matrix corresponding to the ith speed measuring and positioning device is obtained; phi_i(k, k-1) is a one-step transition matrix of the corresponding state of the ith speed measuring and positioning device, W_i(k-1) is the ith tachymeterSystem noise, Z, at time k-1 prior to time k corresponding to the bit device_i(k) The measurement state quantity H corresponding to the ith speed measurement positioning device_iMeasuring matrix for the ith speed measuring and positioning device, M_iAnd measuring noise for the ith speed measuring and positioning device.

Preferably, the state related information of the magnetic-levitation train acquired by the inertial navigation system and the doppler radar system is subjected to data processing to obtain the data processing process of each acquired state quantity, and the position information of the magnetic-levitation train obtained in the processing process is subjected to error compensation according to the fed-back fusion state prediction quantity.

Preferably, the speed measurement positioning method further includes initializing a speed measurement positioning system for executing the speed measurement positioning method, and calculating, by using an initial value of each of the sub-state predictors at a given initial time and an initial value of each of the mean square error predictors, the federal kalman filter model according to the initial value of each of the sub-state predictors, the initial value of each of the mean square error predictors, and the measurement state metric at the current time to obtain the fusion state predictor at the current time.

A speed measurement positioning system of a moving object based on multi-source information fusion comprises: a speed measuring and positioning module, a communication module, an information processing module, a data storage module and a positioning output module,

the speed measuring and positioning module comprises a plurality of speed measuring and positioning devices, the plurality of speed measuring and positioning devices respectively collect the state related information of the moving object in real time to obtain each collected state quantity, the collected state quantity comprises the position information and the speed information of the moving object,

the information processing module determines a discrete state equation and a measurement equation corresponding to each speed measurement positioning device by constructing a Federal Kalman filtering model comprising a sub-filter and a main filter corresponding to the plurality of speed measurement positioning devices, so as to calculate and obtain each measurement state quantity corresponding to each speed measurement positioning device and used for representing state related information of the moving object at the current moment according to each acquisition state quantity at the previous moment at the current moment and the state equation and the measurement equation corresponding to each acquisition state quantity,

The information processing module respectively calculates each sub-state prediction quantity of the state related information of the moving object at the current moment and each sub-mean square error prediction quantity corresponding to the sub-state prediction quantity through each sub-filter according to each measured state information quantity at the current moment, and respectively fuses each sub-state prediction quantity and each sub-mean square error prediction quantity at the current moment through the main filter so as to obtain a fusion state prediction quantity and a fusion mean square error prediction quantity,

the positioning output module outputs the positioning speed measurement result of the moving object according to the information generated by the information processing module,

the data storage module is used for storing various data transmitted by the communication module and various data generated in the processing process of the information processing module so as to provide data for the information processing module to process information.

Preferably, the main filter feeds back the obtained predicted quantity of the fusion state to at least part of each of the sub-filters, so that the difference between the predicted quantity of the sub-state and the predicted quantity of the fusion state generated in each of the sub-filters is reduced, and the predicted quantity of the fusion mean square error P is obtained _g(k) Is multiplied by the set weight distribution coefficients and then is respectively fed back to the sub-filter modules so as to be in each sub-filter according to a formula

To update and optimize each of the sub-mean square error predictors P_i(k) I is the serial number of N speed measuring and positioning devices, i is more than or equal to 1 and less than or equal to N, beta_iDistributing coefficients for the weights corresponding to the ith speed measuring and positioning device,

k is the number of the time，

The higher the speed measurement positioning accuracy of each speed measurement positioning device in the corresponding speed range is, the higher the weight distribution coefficient corresponding to the speed measurement positioning device is,

the speed measuring and positioning system also comprises a working state detection module which is used for detecting whether each speed measuring and positioning device is in a fault state or not so as to output a fault indication signal to the main filter,

the main filter sets the weight distribution coefficient corresponding to the failed speed measurement positioning device to be 0, and distributes the weight distribution coefficient corresponding to the failed speed measurement positioning device to each sub-filter corresponding to each non-failed speed measurement positioning device according to the speed of the moving object during the failure period and the speed measurement positioning accuracy of other non-failed positioning devices within the corresponding speed,

And during the fault period, the information processing module calculates the fusion state prediction quantity and the fusion mean square error prediction quantity at each moment in the fault period according to the measurement state quantity at the moment before the fault occurs, the measurement state quantity at the current moment and the fusion state prediction value.

Preferably, the motion model is a uniform acceleration motion model, the first of the plurality of speed measurement positioning devices is a Beidou navigation system, the second is an inertial navigation system, and the third is a Doppler radar system, wherein the inertial navigation system is a strapdown inertial navigation system,

and the information processing module feeds the fusion state prediction quantity back to the inertial navigation system and the Doppler radar system so as to perform error compensation on the position information of the magnetic-levitation train obtained in the processing process according to the fed-back fusion state prediction quantity in the process of obtaining each acquisition state quantity by the inertial navigation system and the Doppler radar system.

The invention has the following beneficial effects: according to the invention, the plurality of speed measuring and positioning devices are adopted to simultaneously acquire the relevant information of the running state of the moving object to be measured so as to obtain each acquisition state quantity, then the federal Carl filter model is constructed to fuse each acquisition state quantity so as to obtain the final fused state prediction quantity of the running state of the moving object, and the advantages of each speed measuring and positioning device are complemented, so that the precision of the whole speed measuring and positioning system is improved.

The invention has the following beneficial effects: and distributing the finally obtained fusion state prediction quantity and the fusion mean square error quantity to each sub-filter according to the weight distribution coefficient to optimize each sub-state prediction quantity, thereby finally realizing the optimization of the positioning result.

The invention has the following beneficial effects: the speed measuring and positioning system has the inherent advantages of safety redundancy by adopting a plurality of mobile speed measuring and positioning devices, so that the stable fault tolerance of the system under the fault condition is ensured, the speed measuring and positioning system can stably run under the condition that the train does not stop midway through the state prediction obtained by the prediction estimation method of Kalman filtering, and the comprehensive operation efficiency of the line is improved on the premise of ensuring the running safety of the train.

Drawings

Fig. 1 is a flow chart of a speed measurement and positioning method for a moving object based on multi-source information fusion according to the present invention;

fig. 2 is a schematic diagram of a speed measurement and positioning structure of a moving object based on multi-source information fusion according to the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without any creative effort, shall fall within the protection scope of the present invention. It should be noted that "…" in this description of the preferred embodiment is only for technical attributes or features of the present invention.

The invention provides a speed measurement positioning method and system of a moving object based on multi-source information fusion, fig. 1 is a flow chart of the speed measurement positioning method of the moving object based on multi-source information fusion, and fig. 2 is a schematic diagram of the speed measurement positioning structure of the moving object based on multi-source information fusion. The method and system for speed measurement and positioning of a moving object based on multi-source information fusion provided by the present invention will be specifically described with reference to fig. 1 and fig. 2. In addition, in the embodiment of the present invention, the moving object is a magnetic levitation train, but the speed measurement positioning method and system provided by the present invention are not limited to the application of the magnetic levitation train.

In order to avoid the contradiction between the positioning accuracy and the cost caused by the existing single speed measuring and positioning device, the speed measuring and positioning module in the speed measuring and positioning system of the magnetic-levitation train is formed by combining a plurality of speed measuring and positioning devices, so that the core advantages of each speed measuring and positioning device can be utilized to make up the weakness of other speed measuring and positioning devices, the requirements on the manufacturing and design of each speed measuring and positioning device and the performances of each aspect are not required to be too high, and the manufacturing cost of each speed measuring and positioning device is relatively low.

The speed measurement positioning method of the moving object based on multi-source information fusion mainly comprises the following steps: and respectively acquiring the state related information of the moving object in real time by adopting a plurality of speed measuring and positioning devices to obtain each acquisition state quantity, wherein the acquisition state quantity comprises the position information and the speed information of the moving object. And constructing a motion model of the moving object to determine a discrete state equation and a measurement equation corresponding to each speed measurement positioning device, so as to calculate and obtain each measurement state quantity corresponding to each speed measurement positioning device and used for representing state related information of the moving object at the current moment according to each acquisition state quantity at the previous moment of the current moment and the state equation and the measurement equation corresponding to each acquisition state quantity. Constructing a Federal Kalman filter model, wherein the Federal Kalman filter model comprises sub-filters corresponding to the plurality of speed measurement positioning devices and a main filter, calculating each sub-state prediction quantity of the state related information of the moving object at the current time and each sub-mean square error prediction quantity corresponding to the sub-state prediction quantity respectively according to each measured state information quantity at the current time through each sub-filter, and respectively fusing the sub-state prediction quantity and the sub-mean square error prediction quantity at the current moment through the main filter, to obtain a fusion state prediction quantity and a fusion mean square error prediction quantity of the moving object at the current moment, and obtaining a positioning speed measurement result of the moving object at the current moment according to the fusion state prediction quantity and the fusion mean square error prediction quantity at the current moment.

Specifically, in the embodiment of the present invention, a first device of each speed measurement positioning device is an inertial navigation system, a second device is a beidou navigation system, and a third device is a doppler radar system. In an inertial navigation system, an inertial navigation sensor is used for acquiring relevant information of each state of a magnetic-levitation train, such as acceleration, angular velocity and the like of the magnetic-levitation train, and then a position/data calculation module in the inertial navigation system calculates and obtains position information and speed information of the magnetic-levitation train according to data acquired by the inertial sensor and initial data of the magnetic-levitation train, so as to serve as a first state information quantity corresponding to the inertial navigation system. In the Beidou navigation system, a Beidou navigation sensor is used for acquiring relevant information of each state of the magnetic-levitation train, and then a position/data calculation module in the Beidou navigation system calculates and obtains position information and speed information of the magnetic-levitation train according to data acquired by the inertial sensor and initial data of the magnetic-levitation train to serve as second state information quantity corresponding to the Beidou navigation system. In a doppler radar system, a doppler radar sensor is enabled to collect information related to each state of the magnetic levitation train, and then a position/data calculation module in the doppler radar system calculates and obtains position information and speed information of the magnetic levitation train according to data collected by the doppler radar sensor and initial data of the magnetic levitation train, so as to serve as a third state information quantity corresponding to the doppler radar system.

Because the speed measuring and positioning module is provided with the Beidou satellite navigation system for accurately measuring and controlling the energy in real time, the inertial navigation system can select a quick-connected inertial navigation system which does not need a stable platform and is suitable for being equipped with light weight, so that the system has complete autonomous navigation capability which does not depend on external information and is not influenced by external complex environmental factors. The Beidou navigation system is used for continuously correcting the speed measurement positioning result of the inertial navigation system. The Beidou navigation system can also be replaced by a global navigation satellite system (GPS), but in order to avoid potential dependence risks of the GPS, the Beidou navigation system which is independently researched and developed in China is selected as a second device in the speed measuring and positioning module to make up for the defects of the inertial navigation system. The Beidou navigation system is not limited by space-time limitation, can provide continuous and accurate real-time speed and position information for the maglev train all weather, and eliminates accumulated errors. However, the efficacy of the GNSS of the Beidou navigation system is still limited by the influence of special closed terrains such as tunnels along the railway and uncontrollable factors of satellite signal blind areas exist, so that a Doppler radar system without a working blind area is also introduced into the speed measuring and positioning module, and a vehicle-mounted XS-IQ2 type radar speed measuring sensor (speed measuring range: 0.1-2000 km/h) is adopted to continuously acquire information of the magnetic suspension train, so that the defect of the working blind area of the Beidou navigation system is successfully overcome. On the contrary, the Beidou navigation system and the inertial navigation system also make good reference for correcting the vehicle body vibration error of the Doppler radar. Therefore, the advantages of the speed measuring and positioning devices in the speed measuring and positioning module are complementary, and the accuracy of the whole speed measuring and positioning system is improved.

The structure composition and the selection and matching of the speed measuring and positioning module are a combined method for compensating potential defects of other speed measuring and positioning devices by using the core advantages of one speed measuring and positioning device, the comprehensive efficiency of the speed measuring and positioning system of the magnetic suspension train can be integrally improved, in addition, due to the fact that a plurality of speed measuring and positioning speed measuring systems are arranged, the premise is provided for safety redundancy under the condition of system faults, and a good hardware environment is provided for the realization of a subsequent software fusion algorithm.

After each speed measuring and positioning device obtains each collected state quantity, each state information quantity is transmitted to an information processing module and a data storage module through a communication module to be processed and stored respectively, and each state information quantity is synchronously processed and converted into data meeting the requirements of the information processing module by the communication module and then transmitted to the information processing module. The information processing module processes information in a software mode. Because each acquisition state quantity may contain different noise contents, in the information processing module, information processing is performed on each acquisition state quantity by constructing a software fusion algorithm, and output results of each speed measurement positioning device are fused, so that a final positioning result is determined by finally combining the advantages and disadvantages of each speed measurement positioning device in quick break in a corresponding magnetic suspension train. And constructing a software fusion algorithm mainly comprises constructing the Federal Kalman filter model.

The discrete state equation corresponding to the ith speed measurement positioning device in the plurality of speed measurement positioning devices is as formula (1):

X_i(k)＝Φ_i(k,k-1)X_i(k-1)+_iW_i(k-1) (1)

the measurement equation corresponding to the ith speed measurement positioning device is formula (1):

Z_i(k)＝H_iX(k)+M_i(k) (2)

wherein, X_i(k) The state variable of the collected state quantity comprises the position information S of the maglev train collected by the ith speed measuring and positioning device_i(k) And velocity v_i(k)，_iA system noise moving matrix corresponding to the ith speed measuring and positioning device is obtained; phi_i(k, k-1) is a one-step transition matrix of the corresponding state of the ith speed measuring and positioning device, W_i(k-1) is the system noise at the previous moment k-1 of the moment k corresponding to the ith speed measurement positioning device, and Z_i(k) Is the ithThe measuring state quantity H corresponding to the speed measuring and positioning device_iMeasuring matrix for the ith speed measuring and positioning device, M_iAnd measuring noise for the ith speed measuring and positioning device. The device comprises a compass navigation system, an inertial navigation system, a Beidou navigation system and a Doppler radar navigation system, wherein i is 1, i is 2, and i is 3.

Selecting position information S of magnetic-levitation train in text_i(k) And the train speed v_i(k) Is variable of the measurement state quantity, R, of the ith speed measurement positioning device _i(k) Representing the measured noise covariance, Q_i(k) Representing the system noise covariance, we set in the present invention: r_i(k) Is a positive definite matrix; q_i(k) The method is a non-negative definite matrix, and under an ideal condition, system noise and measurement noise are regarded as Gaussian white noise with a mean value of zero, on the basis, by taking the ith speed measurement positioning device as an example, the information processing and fusion are carried out on the constructed Federal Kalman filter recursion algorithm according to a formula (3) to a formula (7):

P_i(k,k-1)＝Φ_i(k,k-1)P_i(k-1)·Φ^T _i(k,k-1)+_iQ_i(k-1)^T(4)

P_i(k)＝(I_i-K_i(k)H_i)P_i(k,k-1) (7)

wherein, in the formula (3) to the formula (7),for one-step pre-measurement of the state of the magnetic suspension train corresponding to the ith speed measuring and positioning device,the state prediction quantity of the magnetic suspension train at the moment K corresponding to the ith speed measuring and positioning device is K_i(k) A Kalman filtering gain matrix, P, corresponding to the ith speed measurement positioning device_i(k, k-1) isAmount of mean square error of (P)_i(k) The mean square error quantity P of the maglev train at the moment k corresponding to the ith speed measuring and positioning device_i(k-1) mean square error quantity H of magnetic-levitation train corresponding to ith speed-measuring and positioning device at previous moment k_iThe ith sensor measurement matrix corresponding to the speed measurement positioning device, I_iThe unit matrix, Z, corresponding to the ith speed measuring and positioning device _i(k) And the measured state quantity is the measured state quantity corresponding to the ith speed measuring and positioning device.

From the expressions (3) to (7), it can be seen that the prediction amount is predicted in a given stateInitial value of (2)Sum mean square error amount P_i(k) Initial value P_i(0) Can be based on the measured state quantity Z of each sensor at the time k_i(k) Calculating the optimal fusion state prediction quantity of the speed measurement positioning system of the magnetic-levitation train at the moment k

And fuse the mean square error predictors P_g(k) In that respect In fig. 1, for simplicity, we omit the temporal parameter k for each parameter.

In order to efficiently integrate the speed measurement positioning information of each speed measurement positioning device, such as an inertial navigation system, a Beidou navigation system and a Doppler radar system, as shown in fig. 1, the invention provides that the Federal Kalman filter comprises a main filter and speed measurement positioning devices corresponding to the main filterAnd (4) sub-filter formation. Adopting the steps of formulas (3) to (7) to carry out Kalman filtering on a Beidou navigation system, an inertial navigation system and a Doppler radar system in the speed measurement positioning system according to the steps of formulas (3) to (7), and carrying out independent time and measurement updating on the acquired state quantity of each speed measurement positioning device through each sub-filter to obtain a plurality of groups of different train motion state local state prediction quantities keeping small difference And the local mean square error amount P_i(k) And then the main filter as shown in FIG. 1 effectively fuses the sub-state prediction quantity and the sub-mean square error quantity output by the three sub-filters according to the formulas (8) and (9) to obtain the fused state prediction quantity of the motion state of the magnetic-levitation trainAnd fuse the mean square error quantity P_g(k) And the optimal output of the speed measurement positioning system is obtained. Finally, the fusion state of the train motion state is measured in advance

And fuse the mean square error quantity P_g(k) And outputting the positioning result as the optimal positioning result of the speed measuring and positioning system through a positioning output module. The specific expressions of the formulas (8) and (9) are as follows, wherein N in the formula (9) is the number of each speed measuring and positioning device in the speed measuring and positioning module, and i is more than or equal to 1 and less than or equal to N.

In the information processing module, the kalman filtering model constructed by us also feeds back the obtained fusion state prediction quantity to at least part of each sub-filter through the main filter, so that the sub-filter generates the sub-filter in each sub-filterThe difference between the state prediction quantity and the fusion state prediction quantity is continuously reduced, and the fusion mean square error prediction quantity P is obtained_g(k) Is multiplied by the set weight distribution coefficients and then is fed back to the sub-filter modules respectively so as to update and optimize the sub-mean square error prediction quantity P in the sub-filters _i(k) In that respect The higher the speed measurement positioning accuracy of each speed measurement positioning device in the corresponding speed range is, the higher the weight distribution coefficient corresponding to the speed measurement positioning device is. In addition, the speed measurement positioning system provided by the present invention is further provided with a function of detecting the operating state of each speed measurement positioning device to determine whether each current speed measurement positioning device is faulty, during a fault of one of the plurality of speed measurement positioning devices, the weight distribution coefficient corresponding to the faulty speed measurement positioning device is set to 0, the weight distribution coefficient corresponding to the faulty speed measurement positioning device is distributed to each sub-filter corresponding to each non-faulty speed measurement positioning device according to the speed of the moving object during the fault and the speed measurement positioning accuracy of other non-faulty positioning devices within the corresponding speed, and during the fault, the sub-filters are distributed according to the measured state quantity at the previous moment of the fault and the measured state quantity at the current moment through the federal kalman filter model, And the fusion state prediction value is used for calculating the fusion state prediction quantity and the fusion mean square error prediction quantity at each moment in the fault period.

In the following we will use specific equations to further elaborate how to feed back the optimal output of the main filter output to each of the sub-filters according to the weight distribution for optimizing the output of each of the sub-filters. The allocation rule of the feedback according to the weight allocation is as shown in formula (10) to formula (12):

X_i(k)＝X_g(k) (10

wherein, set beta_iIn order to realize optimal information fusion, the weight distribution coefficient beta is distributed according to the speed segmentation of the magnetic-levitation train and the accuracy of speed measurement and positioning of each speed measurement positioning device in each speed range_i. As shown in fig. 1, the distribution situation of a specific embodiment of the present invention for distributing the weight coefficients according to the distribution rule is as follows:

when the speed measurement of the magnetic-levitation train is less than 50km/h, the beta value is enabled to be₁＝0.6，β₂＝0.3，β₃＝0.1；

When the speed measurement of the magnetic-levitation train is more than 50km/h and less than 150km/h, the beta value is enabled to be₁＝0.5，β₂＝0.4，β₃＝0.1；

When the speed measurement of the magnetic-levitation train is more than 150km/h, the beta value is enabled to be₁＝0.4，β₂＝0.4，β₃0.2; wherein the content of the first and second substances,

wherein, beta₁Corresponding to said inertial navigation system, beta₂Corresponding to the Beidou navigation System, beta₃Corresponding to the doppler radar system.

The adjustment of the de-weighting distribution coefficient is analyzed according to the speed period, in the low-speed period, the Beidou system and the Doppler radar system have low speed measurement and positioning accuracy, so that the weight distribution coefficient is small, and the inertial navigation system has good performance in the low-speed period, so that the defect of the speed period can be made up, and the distribution coefficient is also large. In the middle-high speed section, the influence of the error of the collected data of the Beidou navigation system on the accuracy of speed measurement and positioning can be reduced, and the weight distribution coefficient can be increased at the moment. In addition, when the train runs at a high speed, the Doppler radar system can well exert the speed measurement positioning advantage, so the distribution coefficient of the radar should be increased at the moment. In the parameter optimization method, the speed measurement accuracy of each speed measurement positioning device in the speed measurement positioning system in different running time periods is fully considered, and the reliability and the algorithm accuracy of the speed measurement positioning system are further improved by the detailed discussion close to the actual complex situation.

In addition, the combined speed measuring and positioning system provided by the invention also has the inherent advantages of safety redundancy. When a certain speed measuring and positioning device magnetic-levitation train breaks down in the running process, the main filter in the information processing module can distribute the weight distribution coefficient corresponding to the broken speed measuring and positioning device to other non-broken speed measuring and positioning devices according to an average distribution method or according to the current speed of the magnetic-levitation train and the accuracy of the speed measuring and positioning devices in the current speed range of other non-broken speed measuring and positioning devices, so as to ensure the stable fault tolerance of the system under the condition of the breakdown, and at the same time, the temporary and reliable prediction can be made on the system state by means of the state prediction method obtained by the Kalman filtering model, namely, the running state of the magnetic-levitation train in a certain period of time in the future (during the breakdown of the speed measuring and positioning device in the speed measuring and positioning module) can be reasonably predicted according to the measured state quantity and the fused state prediction quantity of the speed measuring and positioning system at the previous moment and the current moment, as shown in equations (13) and (14), thereby achieving a safe transition.

Wherein j is the time when the fault occurs, t is the time from j to k after the fault occurs, phi _i(k, j) is a state transition matrix from time j to time k. The state prediction quantity obtained by the prediction estimation method of Kalman filtering can enable the speed measurement positioning system to stably run under the condition that the train does not stop midway, and the comprehensive operation efficiency of the line is improved on the premise of ensuring the running safety of the train.

Performing position/velocity calculation according to the inertial navigation sensor and the Doppler radar sensor to obtain the data of the first acquisition state quantity and the third acquisition state quantityIn the process of processing, the fusion state prediction quantity is fed back in the process

The inertial navigation system predicts according to the feedback fusion stateAnd carrying out error compensation on the position information of the magnetic-levitation train obtained in the processing process. In addition, before the recursive algorithm of the kalman filter algorithm is performed, the method further includes initializing a velocity measurement positioning system for executing the velocity measurement positioning method, and calculating to obtain the fusion state prediction quantity at the current time according to the initial value of each sub-state prediction quantity, the initial value of each mean square error prediction quantity, and the measurement state quantity at the current time by using the federal kalman filter model by giving the initial value of each sub-state prediction quantity and the initial value of each mean square error prediction quantity at the initial time.

After the software fusion algorithm is completed, the invention also detects the performance of the speed measuring and positioning system, so that a motion model which is close to the actual motion state of the magnetic-levitation train and is suitable for theoretical analysis is established. Reasonably, the maglev train can be regarded as an inertial body in a constant or uniform acceleration motion state, and the acceleration can not change suddenly under the condition of variable acceleration, so that the acceleration of the train is kept constant at a certain moment. In addition, the uniform motion model can be equivalent to a uniform acceleration motion model with zero acceleration mean. Therefore, the invention finally selects the uniform acceleration motion model to describe the motion state of the magnetic-levitation train. The mathematical model is derived as shown in equation (15):

wherein S (k), a (k), v (k) respectively represent displacement, acceleration and speed information of the maglev train at the time k, and omega_S(k)、ω_a(k)、ω_v(k) Are respectively provided withRepresenting the system noise contained in the displacement, acceleration and velocity information of the maglev train at time k.

Let us order:

therefore, the discrete state equation and the measurement equation of the train corresponding to the formulas (1) and (2) can be obtained as shown in the formulas (16), (17) and (18):

in the formulas (16) and (18), n, b and r respectively correspond to the inertial navigation system, the Beidou navigation system and the Kepler radar system, and after preset parameter values are brought into the above formulas, a train discrete state equation and a measurement equation under the uniform acceleration running condition of the magnetic-levitation train can be obtained.

In order to verify whether the speed measurement positioning system provided by the invention can obtain the performances that the speed measurement positioning data refresh rate is not lower than 50Hz, the positioning error precision is higher than 10m, and the speed measurement error is less than 1km/h, the speed measurement positioning system provided by the invention is also subjected to experimental simulation by means of an MATLAB software platform. In order to fit actual conditions better, various different noise standard deviations of a simulation system are added, the simulation time is set to be 500s according to an established accelerated motion model of the magnetic levitation train, the fusion period of a main filter is set to be 1s, real-time train operation data can be generated through simulation, and according to simulation results, the combined speed measurement positioning method of the multi-speed measurement positioning device based on the Federal Kalman filtering and state prediction principle achieves a higher precision level. Under the condition of adding multiple external interference noises, the system can still keep higher correction precision and stable waveform output, which shows that the speed measurement positioning scheme provided by the invention not only has better measurement precision, but also has stronger robustness and autonomous operation capability.

While embodiments in accordance with the invention have been described above, these embodiments are not intended to be exhaustive or to limit the invention to the precise embodiments described. Many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best utilize the invention and various embodiments with various modifications as are suited to the particular use contemplated. The invention is limited only by the claims and their full scope and equivalents.