阅读说明：本技术 基于掠射角声线修正的水声传感器网络深海目标定位方法 (Deep sea target positioning method of underwater acoustic sensor network based on grazing angle sound ray correction ) 是由 申晓红 康玉柱 王海燕 闫永胜 贾天一 申莹 于 2021-08-16 设计创作，主要内容包括：本发明提供了一种基于掠射角声线修正的水声传感器网络深海目标定位方法,在声速剖面和节点接收端掠射角已知的情况下,可以获得精确的节点和目标间水平距离,进而提高目标定位精度。本发明解决了传感器节点和目标时钟异步情况下基于时间的距离测量方法和传统声线修正方法无法使用的技术难题,修正了声线弯曲给距离测量和目标定位带来的影响,明显提高了目标的定位精度,基于掠射角的声线修正方法不仅可以用于水声传感器网络的深海目标定位,还可以用于其他水声定位系统的深海目标定位。(The invention provides a glancing angle sound ray correction-based deep sea target positioning method for an underwater acoustic sensor network, which can obtain accurate horizontal distance between a node and a target under the condition that a sound velocity profile and a glancing angle of a node receiving end are known, thereby improving the target positioning accuracy. The method solves the technical problem that a distance measurement method based on time and a traditional sound ray correction method cannot be used under the condition that a sensor node and a target clock are asynchronous, the influence of sound ray bending on distance measurement and target positioning is corrected, the positioning accuracy of a target is obviously improved, and the sound ray correction method based on the grazing angle not only can be used for deep sea target positioning of an underwater sound sensor network, but also can be used for deep sea target positioning of other underwater sound positioning systems.)

1. A deep sea target positioning method of an underwater acoustic sensor network based on grazing angle sound ray correction is characterized by comprising the following steps:

(1) the sensor node measures the grazing angle alpha of the sound ray emitted from the target at the sensor node_N；

(2) Determining the depth of the node and the depth of the target, dividing the sound velocity profile into N layers of equal sound velocity gradient distribution according to the sound velocity change condition, wherein the sound velocity value at the boundary of each layer is c_i，i＝0，1，2，…，N；

(3) According to the grazing angle alpha at the node_NAnd Snell's law, calculating the grazing angle alpha at each layer by back-stepping_iAs shown in the following formula:

α_iis the glancing angle of the ith layer, c_iIs the value of the speed of sound at the i-th layer boundary,

(4) using formulasFinding out horizontal distance Deltax of sound ray propagation of each layer_i，i＝1，2，…，N；Z_iRefers to the depth value at each layer;

(5) using formulasSolving the horizontal distance x of total sound ray propagation;

(6) the total number of M nodes in target positioning is carried out, the steps (1) to (5) are repeated, and the horizontal distances x from the target to the M nodes are respectively obtained_j，j＝1.2，…，M；

(7) At sensor node location and horizontal distance x_jUnder the known condition, a linear equation set is formed in a simultaneous mode, and a two-dimensional horizontal coordinate of a target is obtained by using a least square method;

(8) and obtaining the three-dimensional coordinate of the target by combining the known target depth, and obtaining the horizontal coordinate and the target depth.

2. The deep sea target positioning method of the underwater acoustic sensor network based on grazing angle sound ray correction according to claim 1, characterized in that:

the interval of the layers is set according to the speed of sound velocity change, and can be set to be unequal interval layers.

Technical Field

The invention relates to the technical field of underwater sound positioning, in particular to a method for accurately positioning a deep sea target in an underwater sound sensor network.

Background

The ocean accounts for approximately 71% of the earth's surface area, and the space for human activities gradually expands from land to the ocean, the 21 st century also known as the marine century. The ocean contains abundant mineral and fish resources and is a blue bridge and an important portal which are connected with the world in China, so that the ocean development has important political significance, civil value and military value. The underwater acoustic sensor network has the advantages of simple infrastructure, small node size, low equipment cost, convenience in deployment, strong viability and the like, is widely applied to various marine activities including the aspects of marine environment monitoring, seabed resource detection, disaster early warning, auxiliary navigation, homeland defense and the like, and has basic functions of target positioning and tracking.

The sound velocity distribution in the ocean is not uniform and can be influenced by pressure, temperature and salinity, and due to the refraction effect, sound waves are bent during the propagation process of ocean channels. The underwater target positioning is mainly based on the geometric principle, for a sensor network which uses time delay distance measurement intersection to position, the average sound velocity is multiplied by the propagation time to calculate the linear distance of sound ray propagation, so that a certain positioning precision requirement can be met, but if the average sound velocity is inaccurate, a large positioning error can be caused.

In order to ensure the required positioning accuracy, sound ray correction must be performed. The existing target positioning method based on sound ray bending correction aims at a clock accurate synchronization network, and the commonly used methods mainly comprise three methods: average sound velocity method, iterative method for calculation according to ray acoustics principle, and table look-up method. In deep sea, the sound ray correction method is very suitable for positioning deep sea targets due to the fact that the sound ray types are simple, the number is small, and the sound ray types are easy to distinguish. The underwater environment is complex, such as uneven sound velocity, prolonged propagation time, easy node drifting, high communication energy consumption and the like, which make it difficult for underwater nodes and targets to achieve strict clock synchronization and make the cost of maintaining clock synchronization very high. Under the condition that the node and the target clock are asynchronous, the measured propagation time of the sound ray is inaccurate, and the sound ray propagation distance obtained by the sound ray correction method applicable to the clock accurate synchronous network has a large error, so that the error of target positioning is increased. Therefore, it is necessary to research a deep sea target accurate positioning method based on an underwater acoustic sensor network under the conditions of node, target clock asynchronization and sound line bending.

With the development of electronic components and hydroacoustics technology, the node is provided with a direction-finding functional device, and the grazing angle of sound rays at the node can be measured. When the grazing angle at the node is known, the horizontal propagation distance of the sound ray can be obtained according to the ray acoustics principle of the layered medium.

Disclosure of Invention

In order to overcome the defects of the prior art, the invention provides a deep sea target positioning method of an underwater acoustic sensor network based on grazing angle sound ray correction. Aiming at the problems that the existing target positioning method based on sound ray bending correction aims at a clock accurate synchronization network, but the complexity of the actual underwater environment causes that strict clock synchronization between an underwater node and a target is difficult to realize, and the cost for maintaining the clock synchronization is very high, the invention provides a deep sea target accurate positioning method based on grazing angle sound ray correction. According to the method, under the condition that the sound velocity profile and the node receiving end grazing angle are known, the accurate horizontal distance between the node and the target can be obtained, and therefore the target positioning accuracy is improved.

The technical scheme adopted by the invention for solving the technical problem comprises the following steps:

(1) the sensor node measures the grazing angle alpha of the sound ray emitted from the target at the sensor node_N；

(2) Determining the depth of the node and the depth of the target, dividing the sound velocity profile into N layers of equal sound velocity gradient distribution according to the sound velocity change condition, wherein the sound velocity value at the boundary of each layer is c_i，i＝0，1，2，…，N；

(3) According to the grazing angle alpha at the node_NAnd Snell's law, calculating the grazing angle alpha at each layer by back-stepping_iAs shown in the following formula:

α_iis the glancing angle of the ith layer, c_iIs the value of the speed of sound at the i-th layer boundary,

(4) using formulasFinding out horizontal distance Deltax of sound ray propagation of each layer_i，i＝1，2，…，N；Z_iRefers to the depth value at each layer;

(5) using formulasSolving the horizontal distance x of total sound ray propagation;

(6) the total number of M nodes in target positioning is carried out, the steps (1) to (5) are repeated, and the horizontal distances x from the target to the M nodes are respectively obtained_j，j＝1，2，…，M；

(7) At sensor node location and horizontal distance x_jUnder the known condition, a linear equation set is formed in a simultaneous mode, and a two-dimensional horizontal coordinate of a target is obtained by using a least square method;

(8) and obtaining the three-dimensional coordinate of the target by combining the known target depth, and obtaining the horizontal coordinate and the target depth.

The interval of the layering is set according to the speed of sound velocity change, and can be set to be unequal-interval layering, and the specific method comprises the following steps:

a. for the part with faster sound velocity change, the interval width is set to be smaller, and the layering is denser;

b. for the part with slower change of the sound velocity, the interval width can be set to be larger, and the layering can be sparse.

The method has the advantages that under the condition that the underwater acoustic sensor network node and the target clock are asynchronous, the sound ray correction method based on the grazing angle is adopted according to the characteristics of deep sea sound rays and the current-stage underwater acoustic technology development level, so that the accurate horizontal distance between the node and the target is obtained, and the positioning accuracy of the target is further improved. The invention provides an underwater acoustic sensor network deep sea mesh based on grazing angle sound ray correctionThe accurate positioning method well solves the technical problems that a distance measurement method based on time and a traditional sound ray correction method cannot be used under the condition that a sensor node and a target clock are asynchronous, the influence of sound ray bending on distance measurement and target positioning is corrected, and the mean square error of the horizontal distance of the sound ray is about 0.32m²And the positioning precision of the target is obviously improved. In the method provided by the invention, the grazing angle-based sound ray correction method can be used for positioning the deep sea target of the underwater sound sensor network and can also be used for positioning the deep sea target of other underwater sound positioning systems.

Drawings

FIG. 1 is a schematic diagram of the geometric principle of deep sea target positioning of an underwater acoustic sensor network according to the present invention;

FIG. 2 is a flow chart of the deep sea target accurate positioning method based on grazing angle sound ray correction of the invention;

FIG. 3 is a sound velocity and sound diagram of a layered media model of the present invention;

FIG. 4 is a Munk sound velocity profile of the present invention;

FIG. 5 is a graph of the effect of the transit time measurement error of the present invention on the conventional sound ray correction method.

Detailed Description

The invention is further illustrated with reference to the following figures and examples.

In order to more specifically describe the present invention, the following detailed description is provided for the technical solutions of the present invention with reference to the accompanying drawings and specific embodiments.

The geometric principle schematic diagram of deep sea target positioning of an underwater acoustic sensor network is shown in fig. 1, wherein the diagram comprises M sensor nodes and 1 target to be positioned. True position u of the target⁰＝[x⁰,y⁰,z⁰]^TIs unknown, the true location s of the sensor node j_j＝[x_j,y_j,z_j]^T,j＝1,2,…,M，τ_jIs the propagation time of the signal from the target to node j, c is the speed of sound, then:

in the above formula, r_jFor the distance measurement from the target to the node j, M nodes obtain M r_jThe positions of the targets are determined simultaneously. In order to obtain a more accurate target position, the measured value r must be made_jAnd is more accurate. The invention provides a method for accurately positioning deep sea targets of an underwater acoustic sensor network based on grazing angle sound ray correction, which aims to obtain more accurate r_jAnd further, the target position is estimated more accurately.

The deep sea target accurate positioning method of the underwater acoustic sensor network based on grazing angle sound ray correction under the asynchronous condition of the clock comprises the following steps:

the traditional sound ray correction method requires clock synchronization between the sensor node and the target, and is not applicable when the clocks of the node and the target are asynchronous. Under the condition that node and target clocks are asynchronous, the invention provides an underwater acoustic sensor network deep sea target accurate positioning method based on grazing angle sound ray correction, the positioning method is shown as figure 2, and the detailed positioning process is as follows:

(1) m sensor nodes are arranged in seawater at a preset depth below the sea surface, an underwater acoustic sensor network is formed in a self-organizing mode, and the position s of each node is obtained_j(j ═ 1,2, …, M), the target is stationary or moving at a greater depth in the water, the target depth is known as z;

(2) at a certain moment, the target emits sound signals, M nodes respectively measure grazing angles of sound rays emitted from the target at the nodes

(3) For each node, the horizontal distance x passed by the sound ray is respectively solved by using the method for accurately positioning the deep sea target of the underwater sound sensor network based on grazing angle sound ray correction provided by the invention₁,x₂,…,x_M. Taking a certain node as an example, the grazing angle of reception measured by the node is alpha_NThe sound ray correction steps are as follows:

step 1: determining the depth of the node and the target according to the change of the sound velocityThe sound velocity profile is approximately equal to N layers of sound velocity gradient distribution, and the sound velocity value at the boundary of each layer is c_i(i＝0,1,2,…,N)；

As shown in fig. 3, the layering interval may be set according to the speed of sound velocity change, and may not be layered at equal intervals, and the specific method is as follows:

a. for the part with faster sound velocity change, the interval width is set to be smaller, and the layering is denser;

b. for the part with slower sound velocity change, the interval width can be set to be larger, and the layering can be sparse;

step 2: according to the grazing angle alpha at the node_NAnd Snell's law, the glancing angle at each layer is found by back-stepping as shown in the following equation:

and step 3: application typeFinding out horizontal distance Deltax of sound ray propagation of each layer_i(i＝1,2,…,N)；

And 4, step 4: application typeSolving the horizontal distance x of total sound ray propagation;

(4) at sensor node position s_j(j ═ 1,2, …, M) and horizontal distance x_j(j ═ 1,2, …, M) is known, and then a linear equation system is formed simultaneously, and the two-dimensional horizontal coordinates (x, y) of the target are obtained by applying the least square method;

(5) the three-dimensional coordinates (x, y, z) of the target are obtained in combination with the known target depth z.

The effect of the present invention is verified by three examples below.

The invention provides an underwater acoustic sensor network deep sea target accurate positioning method based on grazing angle sound ray correction, which requires a node to obtain the grazing angle alpha of a target emitted sound ray at a receiving end_NFor the purpose of verifying the inventionThe effectiveness of the method is that the glancing angle alpha of the receiving end is obtained by using BELLHOP model simulation_N. The method mainly aims at sound ray correction of deep sea target positioning, so a Munk _ ray. env environment file is selected as an input file for modeling simulation, a Munk sound velocity profile is shown in figure 4, and the water depth is 5000 m.

Example 1: and (3) simulating the influence of the sound ray propagation time measurement value error on the performance of the traditional sound ray correction method.

The target positioning method based on the traditional sound ray correction requires clock synchronization between the node and the target, and when the clocks between the node and the target are asynchronous, errors exist in the measured value of the sound ray propagation time. And when the measured value of the sound ray propagation time has errors, simulating the effect of the traditional sound ray correction method.

The node position coordinates are (50m,100m,35m), and the target position coordinates are (1000m,800m,1000 m). Setting parameters required to be modified in MunkB _ ray.env, and then operating a bellhop command to generate a MunkB _ ray.arr file to obtain the propagation time of the sound ray. When the propagation time measurement value is error-free and error variance is 0.0005s²、0.001s²、0.002s²、0.003s²、0.004s²、0.005s²Meanwhile, the number of monte carlo simulation experiments is 5000, and the effect of the conventional sound ray correction method is shown in fig. 5.

In fig. 5, when the measured travel time is error-free, the RMSE of the estimated horizontal distance traveled by the sound ray is 0.2736m, because the travel time of the sound ray simulated by bell hop is approximately accurate, and then the sound ray is corrected according to the travel time, and the sound velocity profile is also approximately distributed in N-layer equal sound velocity gradients, so that the error exists between the horizontal distance obtained by the sound ray correction and the true horizontal distance, but the error is small and can be ignored, and the feasibility of the traditional sound ray correction method when the travel time is error-free is also verified.

When the variance of the error of the measured value of the propagation time is 0.0005s²The estimated RMSE of the horizontal distance traveled by the sound ray is 43.0609m, and as the error variance gradually increases, the corresponding RMSE also gradually increases. It can be seen that conventional sound is used when the time of flight measurement error is presentThe line correction method may generate a large distance error, thereby increasing the target positioning error. Therefore, when the node and the target clock are asynchronous, the traditional sound ray correction method is not applicable.

Example 2: under the condition of asynchronous clock, the horizontal distance is fixed, and when the target is at different sound velocity profile depths, the performance simulation of the deep sea target positioning method based on grazing angle sound ray correction is carried out.

The underwater environment is complex, so that strict clock synchronization between an underwater node and a target is difficult to achieve, and when small errors caused by asynchronization of the node and the target clock exist in sound ray propagation time measurement values, the performance of a positioning method based on traditional sound ray correction and a deep sea target positioning method based on grazing angle sound ray correction provided by the invention is compared.

The number of the nodes of the sensor network is 6, and the position coordinates of the nodes are s in sequence₁＝(50m,100m,35m)、s₂＝(103m,600m,22m)、s₃＝(500m,670m,30m)、s₄＝(620m,300m,40m)、s₅＝(400m,60m,20m)、s₆(280m,390m,18 m). The simulated environment file still selects MunkB _ ray.env, the receiving grazing angle and the real sound ray propagation time of each node are obtained through bellhop, and the variance of the measured value error of the sound ray propagation time is assumed to be 0.0005s²Then, the horizontal distance passed by the sound ray is obtained by two methods respectively, the position of the target is calculated, and the horizontal distance and the positioning effect corrected by the two methods are compared. The sound ray direction is zero degree with the horizontal axis, and when towards the horizontal axis top, the grazing angle is negative, and when towards the below, the grazing angle is positive.

As can be seen from the Munk sound velocity profile of fig. 4, the sound velocity profile can be roughly divided into two sections, the upper half section is a negative gradient sound velocity, and the lower half section is a positive gradient sound velocity. The horizontal coordinates of the target position are (1000m,800m), the depths are 600m and 2700m respectively, and the simulation results are shown in table 1 and table 2 respectively.

TABLE 1 negative gradient sonic velocity segment sound ray correction and target location effect

TABLE 2 sound ray correction and target location effect in positive gradient sound velocity section

In Table 1, the variance of the error of the measurement of the propagation time caused by the clock asynchronization is 0.0005s²In the process, the traditional sound ray correction method is adopted for correction, the mean square error of horizontal distance estimation is over 30m, and the obtained horizontal distance is used for positioning, so that a large positioning error is brought. However, by using the glancing angle of the receiving end and adopting the deep sea target accurate positioning method based on glancing angle sound ray correction provided by the invention, the errors are all below 1 m. And finally, obtaining an equation set according to a circle intersection principle, further converting the equation set into a linear equation set, solving by adopting a least square method to obtain the horizontal coordinate of the target position (1000.52m,799.52m), and obtaining the three-dimensional coordinate of the target position (1000.52m,799.52m,600m) by combining the known depth of the target. As a result, table 2 shows that the three-dimensional coordinates of the target position are finally obtained as (1000.23m,800.49m,2700 m).

From the above analysis, when the clock is asynchronous, the horizontal distance error obtained by the traditional sound ray correction method is large, and the requirement on the distance accuracy in positioning cannot be met, but the deep sea target accurate positioning method based on grazing angle sound ray correction provided by the invention has a good effect, the horizontal distance errors are within 1m, the influence of sound ray bending on the positioning of the underwater acoustic sensor network target can be corrected well, and the positioning accuracy is improved.

Example 3: under the condition of asynchronous clock, when the depth is fixed and the target is at different horizontal distances, the performance simulation of the deep sea target positioning method based on grazing angle sound ray correction is carried out.

In order to verify that the horizontal distances between the target and the nodes are different, the method for accurately positioning the deep-sea target based on grazing angle sound ray correction provided by the invention has the advantages that the horizontal distance between the target and the nodes is gradually increased on the assumption that the depth of the target is constant. From the simulation and analysis in the embodiment 2, it is known that when the target and the node clock are asynchronous, the conventional sound ray correction method cannot meet the requirement of precision, and the positioning method provided by the invention can obtain high-precision distance estimation, so that the positioning precision of the target is remarkably improved, and therefore, the part only simulates the horizontal distance correction effect based on the grazing angle sound ray correction method provided by the invention.

The coordinates of the node position are (50m,100m and 35m), the target is positioned at the depth of 1000m and gradually far away from the node, and the horizontal coordinates of the target are u in sequence₁＝(100m,150m)、u₂＝(300m,350m)、u₃＝(500m,550m)、u₄＝(800m,850m)、u₅＝(1200m,1250m)、u₆＝(1600m,1650m)、u₇＝(2000m,2050m)、u₈＝(2500m,2550m)、u₉＝(3000m,3050m)、u₁₀Table 3 shows simulation results (3600m,3650 m).

TABLE 3 sound ray correction effect for different horizontal distances with a certain target depth

In table 3, the target is located at a depth of 1000m, the horizontal distance between the target and the node is gradually increased, and by using the grazing angle-based sound ray correction method provided by the present invention, the horizontal distance error is mostly within 1m, and the mean square error is about 0.32m²And the requirement of positioning precision can be met. Meanwhile, the grazing angle of the sound ray at the node is gradually reduced along with the gradual increase of the horizontal distance between the target and the node, and the initial grazing angle at the target and the horizontal distance passed by the sound ray can be obtained by adopting the sound ray correction method provided by the invention, and the initial grazing angle at the target is also along with the horizontal distance between the target and the sound rayThe increase of (b) is gradually reduced and the reception grazing angle is smaller than the initial grazing angle, which can be explained from the geometrical positional relationship of the node and the target and the degree of bending of the sound ray.

In conclusion, due to the fact that the ocean sound velocity is not uniformly distributed, the sound ray is bent, and the underwater sound positioning error is increased. The invention provides a grazing angle sound ray correction-based method for accurately positioning a deep sea target of an underwater acoustic sensor network, aiming at the problems of asynchronous clocks among targets and nodes and sound ray bending in deep sea target positioning of the underwater acoustic sensor network. According to the method, when the node and the target clock are asynchronous, the accurate horizontal distance between the target and the node can be obtained through the receiving grazing angle measured by the node, and the target positioning accuracy is improved. The effectiveness of the method is verified through BELLHOP and MATLAB simulation, and the method for accurately positioning the deep sea target of the underwater acoustic sensor network based on grazing angle sound ray correction provided by the invention can enable the positioning accuracy of the deep sea target of the underwater acoustic sensor network to be higher.