Rapidly convergent underwater acoustic target passive positioning method

Abstract

The invention relates to the technical field of target motion analysis methods, in particular to a fast convergence underwater acoustic target passive positioning method, which thoroughly breaks through the limitation of a traditional least square estimation technology in a target broadside angle range, and meanwhile, an RBOTMA method also has shorter convergence time, farther estimable target distance and higher estimation precision, is wide in application, and can be applied to the field of underwater acoustic target passive positioning. Especially in modern military actions, target positioning is more accurate, time is estimated faster, and more opportunities and selection possibilities can be won. The convergence time is about 10 min under the conditions that the root-mean-square error of the target azimuth is 0.5 degree and the initial distance is 20 km according to the condition that the root-mean-square error of 5% is taken as the convergence standard; the target initial side angle is 0 degree, which indicates that the RBOTMA is not limited by the range of the target side angle when estimating the target initial position parameter; and the same convergence time is selected as a judgment standard, and the RBOTMA has a farther estimable distance.

Description

Technical Field

[0001] The invention relates to the technical field of target motion analysis methods, and in particular to a fast-converging underwater acoustic target passive positioning method. Background Art

[0002] Bearing-Only Targei Motion Anylisis (BOTMA) is a method that uses only the time series of target bearings observed by passive sonar and combines the platform's maneuvering information to estimate the distance, heading and speed of a target in uniform linear motion.

[0003] In view of the reality that there are observation errors in the target azimuth time series, the variance or mean square error of the passive sonar observation azimuth error (hereinafter referred to as noise) is minimized to achieve the optimal estimation of the target motion parameters; or the influence of the noise variance on the least squares method is corrected. The performance index and stability of the target motion parameter estimation have been greatly improved, and the least squares method has become more mature. However, the main problems that still exist in the above methods are: (1) For targets with small side angles, the target motion parameter estimation time is too long. The industry generally recognizes that the range of small side angles is 0°-30°. For example, when the target side angle is 20° and the distance is 20km, the target motion parameter estimation time takes more than half an hour, or even longer. This has lost its practical significance for some military applications. At the same time, targets with small side angles are the most threatening targets in modern warfare and are also the targets that need to be defended or attacked the most during the confrontation process. If the estimation time is too long, it will miss the opportunity to fight.

[0004] (2) For long-distance targets, the target motion parameter estimation time is too long. For example, for targets with a distance greater than 20 km, the stable estimation time is more than half an hour, or even longer. This time performance is far from meeting the requirements of modern weapons, such as submarine-launched anti-ship attack missiles and wire-guided torpedoes.

[0005] In summary, in order to meet the needs of modern military operations, it is urgent to develop better BOTMA technology that can break through the limitations of side angle, distance and convergence time. Summary of the invention

[0006] The purpose of the present invention is to provide a fast-converging method for passive positioning of underwater acoustic targets, which breaks through the limitation of traditional least squares estimation technology in the target side angle range, and at the same time the method can achieve shorter convergence time, farther estimable target distance and higher estimation accuracy.

[0007] In order to solve the above technical problems, the technical solution provided by the present invention is: providing a fast-converging method for passive positioning of underwater acoustic targets, comprising the following steps: Step 1, formation of copy position sequence: assuming that the target motion parameter is any possible set of values, and generating a copy position sequence according to the change of the relative position of the target and the platform over time; Step 2: Construction of objective function: Establishing an objective function that describes the matching degree between the measured position sequence and the copied position sequence; Step 3, calculating the objective function; using adaptive signal processing technology to calculate the objective function and realize adaptive estimation of DC component power; Step 4: Optimization estimation of target motion parameters; taking the target initial position parameters as the variables to be searched, the objective function is optimized using a numerical optimization algorithm, and the variable coordinates corresponding to the maximum value of the objective function correspond to the estimated values ​​of the target motion parameters.

[0008] The present invention provides a fast-converging method for passive underwater acoustic target positioning (hereinafter referred to as RBOTMA), which completely breaks through the limitation of the traditional least squares estimation technology in the target side angle range. At the same time, the RBOTMA method also has a shorter convergence time, a farther estimable target distance and a higher estimation accuracy, and is widely used, especially in modern military operations. More accurate target positioning and faster estimation time can gain more opportunities and selection possibilities.

[0009] Based on the above technical solution, it also includes: Step 5: Constructing the estimation convergence conditions of the target motion parameters; judging the convergence of the parameters to be estimated based on the time series of the estimated parameters outputted in the above steps, as the judgment condition for ending the estimation process.

[0010] Based on the above technical solution, the formation of the copy position sequence in step 1 includes the following process: Assume that the target moves in a uniform straight line with a speed of V. X is the target initial side angle, D 0 is the initial distance of the target, where the three parameters are each set to a possible value; For the motion platform at t i The target position value measured at the moment, the initial position , time starting point , the passive sonar observation platform performs a turning maneuver; like Figure 1 As shown in the figure, according to the motion situation of the passive sonar observation platform and the target, we can get (1); in, , , are the horizontal and vertical components of the underwater observation platform’s range integral; v i and H i t iThe speed and heading of the underwater observation platform at the moment are both measured time series; therefore, the copy azimuth sequence can be calculated by formula (1): .

[0011] Based on the above technical solution, the construction of the objective function in step 2 includes the following process: Assume that the copy position sequence of passive sonar observation is F i , initial orientation , build sequence as follows: (2); In formula (2), , It is the parameter variable that contributes to the derivative. The parameter can be selected according to different situations. and derivative order N The value of , each order derivative can be realized in differential form; Inspection sequence It can be seen that when D is set 0 ,V,X value is equal to the true value, the sequence In 0, 1, ..., until N The derivatives of all orders are 0, so in the sequence The expected signal in is 0; while in D 0 ,V,X value is not equal to the true value, the expected signal is not all 0; this is the difference between D 0 ,V,X whether it is the expected signal feature of the true value; Based on the characteristics of the expected signal in the matching process, the sequence The power of the DC component in Constructed as the objective function; by sequence The analysis of the expected signal in the equation is: (3); Traverse and search for possible D 0 ,V,X value, corresponding to The minimum DC component power in the target motion parameters .

[0012] Based on the above technical solution, step 2 also includes optimization of the objective function, including the following process: Since the initial orientation value There is an error. The initial orientation is also taken as a parameter to be estimated. The initial orientation and the other three target initial position parameters are estimated simultaneously to make the entire target initial position parameter estimation system optimal. The formula for solving the target motion parameters is optimized as follows: (3)'.

[0013] More preferably, in practical applications, the initial azimuth value of the passive sonar observation is There are also errors, giving the position copy sequence The selection of the initial value brings difficulties. If the selection is not good, it may cause systematic errors. To solve this problem, the initial orientation can also be taken as a parameter to be estimated, so that the initial orientation and the other three target initial position parameters can be estimated simultaneously to make the entire target initial position parameter estimation system optimal.

[0014] Based on the above technical solution, the adaptive estimation process of the DC component power in step 3 includes the following process: for each set of D 0 ,V,X value, solve the DC component power after azimuth matching After the orientation is matched, the parameter is set. and order N After the value of , the sequence on the left side of formula (2) is a known sequence; Set sequence The DC signal component in is s , then (4); where n i Represents the noise component; In order to reduce the impact of noise, the adaptive linear weighted minimum variance distortion-free response (MVDR) method is used; Let the filter output sequence be: (5); In formula (5), n i for The noise component in, k is the filter order; w k is the weight vector; To ensure that the filter does not affect the sequence The DC signal (desired signal) component in (6); then (7); The filter output power is (8); The condition that the expected signal is uncorrelated with the noise is used in equation (8).

[0015] The traditional methods for estimating the DC component power are the least square method and the Fourier transform method. Theoretically, it can be proved that the two are equivalent. Related work shows that the estimation results of the target initial position parameters using the least square method are completely consistent with the traditional BOTMA least square method when estimating the DC component power. RBOTMA uses adaptive signal processing technology to achieve Estimates.

[0016] Based on the above technical solution, the optimization process of the filter desired signal power output in step 3 is: It can be seen from the above formula (8) that since the signal power is constant during the filter adjustment process, the minimum noise power is equivalent to the minimum total output power P; therefore, (9); In formula (9), , For sequence The correlation matrix of W has the dimension (K x K), H The superscript H in stands for transposed conjugate; Considering the constraints of equation (9) on the filter, the design problem of optimizing the adaptive filter is: (10); st , Formula (10) is a constrained optimization formula; using the Lagrange multiplier method, the optimized expected signal power output with minimal noise influence can be obtained as (11); where 1 is a row vector whose K elements are all 1.

[0017] The above analysis gives a detailed adaptive estimation process for optimizing the DC component power. In the optimization process, in order to ensure that the noise power in the filter output power has the smallest effect, the filter can be adjusted to minimize the second noise power on the right side of equation (8). However, the noise is unknown and the noise power is difficult to estimate. However, it can be seen from equation (8) that the signal power in the process of adjusting the filter is a constant, so the minimum noise power is equivalent to the minimum total output power P. For real matrices, the superscript H only represents the transpose.

[0018] The Lagrange multiplier method can be used to obtain the optimal filter solution and minimum output power, that is, the optimized expected signal power output with minimal noise influence. For detailed theoretical derivation, please refer to relevant signal processing books on MVDR adaptive filters.

[0019] On the basis of the above technical solution, the optimization estimation method of the target motion parameters in step 4 adopts a particle swarm algorithm, firstly initializes a group of random particles, and then searches for the optimal solution through iteration; in each iteration, the particles update their own speed and position by tracking the individual optimal value and the global optimal value; this process continues until the end condition is met, such as reaching the maximum number of iterations or finding the optimal solution that meets the accuracy requirements.

[0020] In steps 1 to 3, for a possible set of D 0,V,X values, the matching orientation sequence is generated, the objective function and calculation method describing the matching degree are established. On this basis, the optimization search method for target motion parameter estimation is further given in step 4, that is, the method for optimizing the search for the optimal (3)' estimate, and the solution of (3) or (3)' is obtained.

[0021] In this application, RBOTMA uses Particle Swarm Optimization (PSO). PSO optimizes search by simulating group intelligent behaviors such as bird flocks foraging. In PSO, each particle flies in the target search space and knows its current position (point in the search space) and the objective function value. During the position update process of each particle, it will learn from two position points: one is its own historical optimal position point pbest, and the other is the historical optimal position value gbest of the group. The particle will update its own position based on the difference between these two points and the current position, in the hope of finding a better solution. The objective function is a criterion for evaluating the quality of each particle's position.

[0022] During the execution of the algorithm, a group of random particles are first initialized, and then the optimal solution is found through iteration. In each iteration, the particles update their speed and position by tracking the individual optimal value and the global optimal value. This process continues until the end condition is met, such as reaching the maximum number of iterations or finding the optimal solution that meets the accuracy requirements. In our application, the position of the particle is determined by D 0 ,V,X,F 0 The coordinate points in the four-dimensional space composed of the four search variables are determined, and the objective function is calculated by formula (11).

[0023] Based on the above technical solution, the optimization estimation method of the target motion parameters in step 4 includes the following steps: wherein the position of the particle is determined by D 0 ,V,X,F 0 The coordinate points in the four-dimensional space composed of the four search variables are determined, and the objective function is calculated by formula (11); Step 1: Parameter setting: set the number of particles N, the maximum number of iterations and the search range of each variable to be searched, and the search range is the maximum value-minimum value; Step 2: Initialize the position of each particle; randomly determine the position of each particle in the four-dimensional space and calculate their respective objective functions; particle i (i=1,2,…,N) records its initial position as pbest, and the position of the optimal value of the group's objective function is gbest; Step 3: Particle position update; during the search process of the i-th particle, the optimal position value pbest it has experienced, and the optimal position value gbest experienced by all particles in the group; the update speed of particle i is expressed as It represents the update step of the particle in the four-dimensional space. For each update, its dth dimension (1 ≤ d ≤ D, where D is the dimension of the problem space) changes according to the following formula: (12a); (12b); among which, is the inertia weight, c1 and c2 are acceleration constants, r1 and r2 are random values ​​varying in the range [0,1], and p i and v i are the new position and update step of the ith particle, respectively. The superscript - indicates the previous time. At the updated position, recalculate the objective function value of each particle, update pbest and gbest, and record the objective function value Tbest corresponding to gbest. As the number of iterations increases, , c1 and c2 are gradually reduced to reduce the possibility of skipping the extreme value; at the updated position, the objective function value of each particle is recalculated, pbest and gbest are updated, and the objective function value Tbest corresponding to gbest can also be recorded as needed; the first part of formula (12a) is the inertia of the particle's previous behavior, part 2 The third part is the cognitive part, which represents the thinking of the particles themselves. is the social part, which represents the information sharing and mutual cooperation among particles; this formula describes the search behavior of each particle in the multidimensional space and gradually approaches the global optimal solution in each iteration.

[0024] Step 4: Determine the end of the search; when the number of searches reaches the maximum number of iterations, exit the iteration and return the objective function value Tbest of gbest; when the number of iterations M increases and gbest and Tbest remain unchanged, you can choose to exit the iteration and return the objective function value Tbest of gbest.

[0025] The M value can be set according to the problem requirements.

[0026] Based on the above technical solution, the convergence condition of the target motion parameter estimation in step 5 is: after multiple target motion parameters D 0 ,V,X,F 0 After the estimation process, it is determined whether the relative root mean square error of the N points reaches the set accuracy. If so, the estimation process is terminated; if not, continue with steps 1 to 4.

[0027] The above steps 1 to 4 complete an estimation of ; due to the random errors in the time-azimuth sequence observed by passive sonar, it is difficult to make a judgment on the estimation performance of the estimated value of . Therefore, as the length of the observed time-azimuth sequence continues to increase, multiple estimates of can be made, and according to the actual application, the convergence criteria for judging each estimation parameter are constructed, such as the relative root mean square error of N points reaching a certain accuracy, as the judgment criterion for ending the estimation process.

[0028] The beneficial effects of the technical solution provided by the present invention are: The present invention provides a fast-converging method for passive underwater acoustic target positioning (hereinafter referred to as RBOTMA), which completely breaks through the limitation of the traditional least squares estimation technology in the target side angle range. At the same time, the RBOTMA method also has a shorter convergence time, a farther estimable target distance and a higher estimation accuracy, and is widely used, especially in modern military operations. More accurate target positioning and faster estimation time can gain more opportunities and selection possibilities.

[0029] The technical effect produced by the RBOTMA of the present application is as follows: according to the root mean square error of 5% as the convergence standard, under the conditions of the root mean square error of the target azimuth of 0.5° and the initial distance of 20km, the convergence time is about 10min, which is much better than the convergence time of the least squares method of not less than 40min; the initial side angle of the target is 0°, indicating that RBOTMA is not limited by the target side angle range when estimating the initial position parameters of the target. Compared with the limitation of the least squares method on the initial side angle of the target, RBOTMA completely removes this limitation; by selecting the same convergence time as the judgment standard, such as 20min, RBOTMA can estimate a farther distance, even up to 30km. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a motion situation diagram of the passive sonar observation platform and the target in the present invention; Figure 2 is a flow chart of a target motion parameter estimation process according to an embodiment of the present invention, wherein the dotted line portion is an initialization process of an estimation; Figure 3 : is a time convergence curve of the target initial distance estimation when the target initial side angle is 30° in an embodiment of the present invention, wherein (a) is the RBOTMA estimation result, and (b) is the least squares method estimation result; Figure 4 1 is a time convergence curve of the target initial side angle estimation when the target initial side angle is 30° in an embodiment of the present invention, wherein Figure (a) is the RBOTMA estimation result; Figure (b) is the least squares method estimation result; Figure 53 is a time convergence curve of target velocity estimation when the target initial side angle is 30° in an embodiment of the present invention, wherein (a) is the RBOTMA estimation result, and (b) is the least squares method estimation result; Figure 6 is a time convergence curve of target initial distance estimation when the target initial side angle is 0° in an embodiment of the present invention; Figure 7 is a time convergence curve of target initial side angle estimation when the target initial side angle is 0° in an embodiment of the present invention; Figure 8 is a time convergence curve of target speed estimation when the target initial side angle is 0° in an embodiment of the present invention; DETAILED DESCRIPTION The present invention will be further described below in conjunction with the accompanying drawings and embodiments: In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be directly connected or indirectly connected through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0031] In the description of the present invention, it is necessary to understand that the terms "left", "right", "front", "back", "top", "bottom", etc., indicating directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as a limitation on the present invention.

[0032] Example 1 like Figures 2 to 8 As shown, this embodiment provides a specific implementation process of a fast-converging underwater acoustic target passive positioning method, including the following steps: Input the time-azimuth sequence of passive sonar measurement and the corresponding heading and speed sequence of the sonar observation platform, where the sonar observation platform needs to have at least one change of direction or speed; if it is a simulation, it is necessary to simulate and generate the aforementioned measurement data sequence. After turning or changing speed, start the target motion parameter estimation solution process: Set four variables D to be searched 0 ,V,X,F 0The search range is set; the number of particles in the particle swarm and the maximum number of iterations are set; the inertia constant and acceleration constant are set, and both decrease with the increase of the number of iterations. The speed of effectiveness varies depending on the problem, and needs to be set by trial and error; the derivative order and corresponding parameter value in the objective function are set. Within the search range of each variable, the initial position coordinates of each particle are randomly set, that is, D 0 ,V,X,F 0 The initial value of . According to formula (1), the D of each particle is 0 ,V,X,F 0 A set of values ​​is brought into The copy position sequence can be obtained, and then the sequence required to construct the objective function is determined by formula (2) .

[0033] Using equations (4)-(11), we calculate the objective function value, i.e., the sequence The DC component power in the sparse matrix; record the position of each particle as its pbest and the corresponding objective function value, and find the position of the particle with the minimum objective function value as gbest. The above steps complete the optimization search D 0 ,V,X,F 0 Initialization. Record the number of iterations as 1.

[0034] For the kth iteration, use equation (12) to update the position coordinates of each particle, and substitute them into equations (4)--(11) to calculate the objective function value. For each particle, each compares its current objective function value with the objective function value corresponding to the historical record pbest. If the current objective function value is less than or equal to the previous historical value, update the current position to pbest and record the current objective function value; if the current objective function value is greater than the previous historical value, keep pbest and its corresponding objective function value unchanged. Obtain the minimum value of the objective function value among all particles' pbests, update the corresponding particle's pbest to gbest, and record its corresponding objective function value. Record the number of iterations k.

[0035] Repeat the above steps to achieve multiple iterations, and the corresponding objective function value of gbest will gradually decrease. Record the number of iterations.

[0036] You can set an iteration number M in advance. When gbest is not updated for M times during the iteration, the search is exited and the location coordinates of gbest are returned, which is the estimated If this setting is not available, when the maximum number of iterations is reached, the position coordinates of gbest are returned, which is the estimated As an option, the corresponding minimum objective function value can also be returned. When the number of iterations reaches the set maximum number of iterations, the iteration process ends automatically and the position coordinates of gbest are returned. As an option, the corresponding minimum objective function value can also be returned.

[0037] As the length of the observation data increases, multiple The estimation, that is, the output of each parameter to be estimated, each forms a time series. Using these sequences, a function to determine the convergence criteria of these sequences can be established. 0 ,V,X,F 0 After the estimation process, it is determined whether the relative RMS error of the N points reaches the set accuracy. When all the estimated parameters meet the convergence conditions, the optimization estimation process of (3) or (3') can be ended.

[0038] Example 2 In this embodiment, two simulation examples are provided, and for the convenience of comparison, the estimation results of the least square method under the same conditions are also given.

[0039] During the simulation, the target azimuth root mean square error (including the initial azimuth) was set to 0.5° white noise, and the azimuth sampling rate was s. -1 The observation platform has a speed of 6 knots and starts to follow the initial heading. After 2 minutes, it turns 90 degrees to the direction of the target's heading change until the estimation process is completed. The initial distance to the target is 20 km, and the speed is 15 knots (about 7.7 m / s -1 ).

[0040] (1) When the target's initial side angle is 30°: Figures 3 to 5 The changes of the estimated values ​​of the initial distance, initial beam angle and speed of the target over time under this beam angle are given. For the convenience of comparison, the estimation results of the least squares method are also given.

[0041] (2) Target initial side angle 0°: Figures 6 to 8 The changes of the estimated values ​​of the initial distance, initial beam angle and speed of the target over time under this beam angle are given. Under the conditions of combat significance, for the target initial beam angle of 0°, the least squares estimation method can no longer achieve the solution, so no comparison is given.

[0042] Example 3 A large number of numerical simulations were performed according to the steps in the above embodiment. With a root mean square error of 5% as the convergence standard, under the conditions of a target azimuth root mean square error of 0.5° and an initial distance of 20 km, the main technical indicators of the RBOTMA method were obtained as follows: (1) Shorter convergence time The convergence time is about 10 minutes, while the convergence time of the least squares method is no less than 40 minutes.

[0043] (2) Completely remove the target initial side angle restriction In the above simulation example, the initial beam angle of the target is 0°, which shows that RBOTMA is not limited by the target beam angle range when estimating the initial position parameters of the target. Compared with the limitation of the least squares method on the initial beam angle of the target, the new method RBOTMA completely removes this limitation.

[0044] (3) Longer target distance Using a meaningful convergence time as the criterion, such as 20 minutes, RBOTMA can estimate a greater distance, even up to 30 km.

[0045] The basic principles and main features of the present invention are shown and described above. For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments. Therefore, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is limited by the appended claims rather than the above description. Therefore, it is intended to include all changes within the meaning and scope of the equivalent elements of the claims in the present invention.

[0046] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A fast-converging method for passive positioning of underwater acoustic targets, characterized in that: The following steps are involved: Step 1, formation of copy position sequence: assuming that the target motion parameter is any possible set of values, and generating a copy position sequence according to the change of the relative position of the target and the platform over time; Step 2: Construction of objective function: Establishing an objective function that describes the matching degree between the measured position sequence and the copied position sequence; Step 3, calculating the objective function; using adaptive signal processing technology to calculate the objective function and realize adaptive estimation of DC component power; Step 4: Optimization estimation of target motion parameters; taking the target initial position parameters as the variables to be searched, the objective function is optimized using a numerical optimization algorithm, and the variable coordinates corresponding to the maximum value of the objective function correspond to the estimated values ​​of the target motion parameters.

2. The method for rapidly converging passive positioning of underwater acoustic targets according to claim 1, characterized in that: Also includes: Step 5: Constructing the estimation convergence conditions of target motion parameters; The convergence of the estimated parameters is judged based on the time series of the estimated parameters outputted from the above steps, which serves as the judgment condition for terminating the estimation process.

3. The method for rapidly converging passive positioning of underwater acoustic targets according to claim 1, characterized in that: The formation of the copy position sequence in step 1 includes the following processes: Assume that the target moves in a uniform straight line with a speed of V. X is the target initial side angle, D0 is the target initial distance, and the three parameters are each set to a possible value; For the motion platform at t i The target position value measured at the moment, the initial position , time starting point , the passive sonar observation platform performs a turning maneuver; According to the motion situation of the passive sonar observation platform and the target, we can get (1); in, , , are the horizontal and vertical components of the underwater observation platform’s range integral; v i and H i t i The speed and heading of the underwater observation platform at the moment are both measured time series; therefore, the copy azimuth sequence can be calculated by formula (1): .

4. The method for rapidly converging passive underwater acoustic target positioning according to claim 1 is characterized in that: The construction of the objective function in step 2 includes the following process: Assume that the copy position sequence of passive sonar observation is F i , initial orientation , build sequence as follows: (2); In formula (2), , It is the parameter variable that contributes to the derivative. The parameter can be selected according to different situations. and derivative order N The value of , each order derivative can be realized in differential form; Inspection sequence It can be seen that when the set D0, V, X values ​​are equal to the true values, the sequence In 0, 1, ..., until N The derivatives of all orders are 0, so in the sequence The expected signal in is 0; when the value of D0, V, X is not equal to the true value, the expected signal is not all 0; this is the expected signal feature that distinguishes whether D0, V, X is the true value; Based on the characteristics of the expected signal in the matching process, the sequence The power of the DC component in Constructed as the objective function; by sequence The analysis of the expected signal in the equation is: (3); After searching through the possible values ​​of D0, V, and X, the corresponding The minimum DC component power in the target motion parameters .

5. The method for rapidly converging passive underwater acoustic target positioning according to claim 4 is characterized in that: The step 2 also includes the optimization of the objective function. The process includes: Since the initial orientation value There is an error. The initial orientation is also taken as a parameter to be estimated. The initial orientation and the other three target initial position parameters are estimated simultaneously to make the entire target initial position parameter estimation system optimal. The formula for solving the target motion parameters is optimized as follows: (3)’。 6. A fast-converging underwater acoustic target passive positioning method according to claim 4, characterized in that: The adaptive estimation process of the DC component power in step 3 includes the following process: for each set of D0, V, X values, solving the DC component power after azimuth matching After the orientation is matched, the parameter is set. and order N After the value of , the sequence on the left side of formula (2) is a known sequence; Set sequence The DC signal component in is s , then (4); where n i is the noise component; in order to reduce the impact of noise, the adaptive linear weighted minimum variance distortionless response (MVDR) method is adopted; the filter output sequence is assumed to be: (5); In formula (5), n i for The noise component in, k is the filter order; w k is the weight vector; To ensure that the filter does not affect the sequence The DC signal (desired signal) component in (6); then (7)； The filter output power is (8); Formula (8) uses the condition that the expected signal is uncorrelated with the noise.

7. A fast-converging underwater acoustic target passive positioning method according to claim 6, characterized in that: The optimization process of the filter desired signal power output in step 3 is: It can be seen from the above formula (8) that since the signal power is constant during the filter adjustment process, the minimum noise power is equivalent to the minimum total output power P; therefore, (9); In formula (9), , For sequence The correlation matrix of W has the dimension (K x K), H The superscript H in stands for transposed conjugate; Considering the constraints of equation (9) on the filter, the design problem of optimizing the adaptive filter is: (10)； st , Formula (10) is a constrained optimization formula; using the Lagrange multiplier method, the optimized expected signal power output with minimal noise influence can be obtained as (11); where 1 is a row vector whose K elements are all 1.

8. The method for fast-converging passive positioning of underwater acoustic targets according to claim 7, characterized in that: The optimization estimation method of the target motion parameters in step 4 adopts a particle swarm algorithm, firstly initializes a group of random particles, and then searches for the optimal solution in an iterative manner; in each iteration, the particles update their own speed and position by tracking the individual optimal value and the global optimal value; this process continues until the end condition is met, such as reaching the maximum number of iterations or finding the optimal solution that meets the accuracy requirements.

9. A fast-converging underwater acoustic target passive positioning method according to claim 8, characterized in that: The optimization estimation method of the target motion parameters in step 4, The method comprises the following steps: wherein the position of the particle is determined at the coordinate point in the four-dimensional space composed of the four search variables D0, V, X, and F0, and the objective function is calculated by equation (11); Step 1: Parameter setting: set the number of particles N, the maximum number of iterations and the search range of each variable to be searched, and the search range is the maximum value-minimum value; Step 2: Initialize the position of each particle; randomly determine the position of each particle in the four-dimensional space and calculate their respective objective functions; particle i (i=1,2,…,N) records its initial position as pbest, and the position of the optimal value of the group's objective function is gbest; Step 3: Particle position update; during the search process of the i-th particle, the optimal position value pbest it has experienced, and the optimal position value gbest experienced by all particles in the group; the update speed of particle i is expressed as It represents the update step of the particle in the four-dimensional space. For each update, its dth dimension (1 ≤ d ≤ D, where D is the dimension of the problem space) changes according to the following formula: (12a)； (12b); among which, is the inertia weight, c1 and c2 are acceleration constants, r1 and r2 are random values ​​varying in the range [0,1], and p i and v i are the new position and update step of the ith particle, respectively. The superscript - indicates the previous time. At the updated position, recalculate the objective function value of each particle, update pbest and gbest, and record the objective function value Tbest corresponding to gbest. Step 4: Determine the end of the search; when the number of searches reaches the maximum number of iterations, exit the iteration and return the objective function value Tbest of gbest; when the number of iterations M increases and gbest and Tbest remain unchanged, you can choose to exit the iteration and return the objective function value Tbest of gbest.

10. The method for fast-converging passive positioning of underwater acoustic targets according to claim 1, characterized in that: The estimation convergence condition of the target motion parameter in step five is: after multiple estimation processes of the target motion parameters D0, V, X, F0, determine whether the relative root mean square error of the N points reaches the set accuracy. If so, end the estimation process; if not, continue steps one to four.

Images