Real-time moving target tracking and decision-making method based on multi-target optimization algorithm
Through real-time motion target tracking and decision-making methods based on multi-objective optimization algorithm, the problem of insufficient robustness and accuracy of traditional methods in complex dynamic environments is solved, efficient and reliable target tracking is achieved, and the adaptability and practicality of the system are improved.
Patent Information
- Application Number
- CN202511020714.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-24
- Publication Date
- 2025-08-29
AI Technical Summary
Traditional single-objective optimization methods are difficult to meet the actual needs of multi-dimensional and multi-constraints in complex dynamic environments. The robustness and accuracy of motion target tracking are insufficient, and the dynamic adaptability of prediction and observation matching is insufficient.
Using a multi-objective optimization algorithm, by measuring the absolute position, velocity and acceleration of the moving target, combining the dynamic threshold matching algorithm and the multi-objective optimization algorithm, an executable optimal control instruction vector is generated, and the target position is predicted using a uniform acceleration motion model, and the objective function is solved through the gradient descent method to balance tracking errors and control costs.
It improves the robustness and accuracy of the motion target tracking system in complex dynamic scenarios, enhances the adaptability to the target motion trend, reduces energy consumption and mechanical stress, and improves the practicality and sustainability of the system.
Smart Images
Figure CN120560045A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of moving target tracking, and in particular to a real-time moving target tracking and decision-making method based on a multi-objective optimization algorithm. Background Art
[0002] Real-time moving target tracking and decision-making technology is an important research direction in the field of modern science and technology. It is widely used in multiple scenarios such as autonomous driving, robot navigation, drone control, security monitoring, military reconnaissance, and intelligent transportation systems. With the rapid development of sensor technology, computing power, and artificial intelligence algorithms, the performance requirements of real-time moving target tracking and decision-making methods in complex dynamic environments are constantly increasing. Traditional single-objective optimization methods often cannot meet the actual needs of multi-dimensional and multi-constrained conditions. There are also problems such as insufficient robustness and accuracy of moving target tracking; insufficient dynamic adaptability of prediction and observation matching; and difficulty in balancing multi-objective optimization and hardware constraints. Real-time moving target tracking and decision-making methods based on multi-objective optimization algorithms offer new insights and possibilities for solving complex problems by comprehensively considering multiple objectives and constraints. These methods are poised to play a significant role in areas such as autonomous driving, drone navigation, security monitoring, and military reconnaissance. In the future, these methods will find efficient and reliable application in a wider range of scenarios, providing strong support for the further development of intelligent systems. Summary of the Invention
[0003] The present invention provides a real-time moving target tracking and decision-making method based on a multi-objective optimization algorithm to solve the technical problem that traditional moving target tracking methods are difficult to adapt to the rapid movement, direction change or noise interference of the target in complex dynamic scenes, resulting in reduced tracking accuracy or target loss.
[0004] The present invention provides a real-time moving target tracking and decision-making method based on a multi-objective optimization algorithm, which specifically includes the following technical solutions: A real-time moving target tracking and decision-making method based on a multi-objective optimization algorithm includes the following steps: S1. Measure the observed relative position of the moving target and calculate the absolute position of the moving target; based on the absolute position of the moving target, calculate the speed and acceleration of the moving target, and predict the absolute position of the moving target at the next moment to obtain the predicted absolute position of the moving target; S2. Introducing a dynamic threshold matching algorithm based on motion trends, matching the absolute position of the moving target at the next moment with the predicted absolute position of the moving target, generating a matching cost, and determining the target position based on the matching cost; S3. Based on the target position, a multi-objective optimization algorithm is used to output an executable optimal control instruction vector.
[0005] Preferably, the S2 specifically includes: The dynamic threshold matching algorithm based on motion trend introduces Euclidean distance to quantify the deviation between the absolute position of the moving target at the next moment and the predicted absolute position of the moving target, and generates a matching cost in combination with a dynamic matching threshold.
[0006] Preferably, the S2 specifically includes: According to the radar measurement accuracy, a basic threshold is defined; based on the basic threshold, an adjustment factor is introduced to calculate the dynamic matching threshold.
[0007] Preferably, the S2 specifically includes: Based on the speed of the moving target, an exponential decay parameter is introduced, and the adjustment factor is calculated by combining the cosine of the angle between the velocity vector and the acceleration vector of the moving target.
[0008] Preferably, the S2 specifically includes: Set the matching cost threshold and compare the matching cost with the matching cost threshold. When the matching cost is less than or equal to the matching cost threshold, it means that the absolute position of the moving target at the next moment is the target position. Otherwise, the predicted absolute position of the moving target is used as the target position.
[0009] Preferably, the S3 specifically includes: In the implementation process of the multi-objective optimization algorithm, based on the target position, the tracking error term and the control cost term are introduced, combined with the weight coefficient, the objective function is constructed and solved, and the executable optimal control instruction vector is output.
[0010] Preferably, the S3 specifically includes: The tracking error term introduces the Euclidean distance to quantify the deviation between the predicted position of the tracking system and the target position; the predicted position of the tracking system is calculated based on the current position of the tracking system and the displacement generated by the control command within the radar sampling time interval; the control cost term is obtained by calculating the squared Euclidean norm of the control command vector.
[0011] Preferably, the S3 specifically includes: In the implementation process of the multi-objective optimization algorithm, the objective function is solved by the gradient descent method, and velocity constraints and acceleration constraints are imposed on the control instruction vector; when the number of iterations of the objective function reaches the preset maximum number of iterations, the iteration is stopped and the current control instruction vector is output as the optimal control instruction vector for tracking the execution of the system.
[0012] The beneficial effects of the technical solution of the present invention are: 1. Based on the absolute position data of the moving target at continuous moments, the finite difference method is used to calculate the speed and acceleration of the moving target, accurately reflecting the motion state of the moving target. Through simple and efficient numerical calculations, the real-time capture of the dynamic characteristics of the moving target is achieved, providing key input for subsequent predictions based on the uniform acceleration motion model, and improving the tracking system's adaptability to the target's motion trend.
[0013] 2. Based on the calculated absolute position, velocity, and acceleration of the moving target, a uniform acceleration motion model is used to predict the absolute position of the moving target at the next moment. This fully utilizes the current motion state of the moving target and generates highly reliable prediction results, providing a reference for target matching and enhancing the robustness of the tracking system in complex dynamic scenarios.
[0014] 3. By calculating the Euclidean distance between the observed and predicted positions and combining it with a dynamic matching threshold, a matching cost is generated to determine whether the observed position is a moving target. By introducing an adjustment factor, the system dynamically adapts to the target's motion trend, especially in scenarios with high-speed motion or frequent direction changes. This allows the system to effectively distinguish between targets and noise, thereby improving the accuracy and reliability of target tracking.
[0015] 4. A multi-objective optimization algorithm is used to integrate tracking error and control cost to generate executable control instructions. The tracking error term drives the tracking system to approach the target position and maintain tracking accuracy. The control cost term limits the size of control instructions, reducing energy consumption and sudden changes. The objective function is solved by the gradient descent method, and velocity and acceleration constraints are imposed at the same time to ensure that the control instructions are executable within the hardware capabilities. This effectively balances tracking accuracy and tracking system energy consumption, reduces mechanical stress or operating costs, and improves the practicality and sustainability of the tracking system. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] Figure 1 This is a flow chart of a real-time moving target tracking and decision-making method based on a multi-objective optimization algorithm described in the present invention. DETAILED DESCRIPTION
[0017] In order to further illustrate the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.
[0018] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
[0019] The following describes in detail a specific solution of a real-time moving target tracking and decision-making method based on a multi-objective optimization algorithm provided by the present invention with reference to the accompanying drawings.
[0020] Refer to the attached Figure 1 , which shows a flow chart of a real-time moving target tracking and decision-making method based on a multi-objective optimization algorithm provided by an embodiment of the present invention, the method comprising the following steps: S1. Measure the observed relative position of the moving target and calculate the absolute position of the moving target; based on the absolute position of the moving target, calculate the speed and acceleration of the moving target, and predict the absolute position of the moving target at the next moment to obtain the predicted absolute position of the moving target; The radar with sampling frequency ≥10Hz is used to measure the moving target at time The relative position of observation , that is, the two-dimensional Cartesian coordinates of the moving target relative to the tracking system (with the radar position as the coordinate origin), Indicates that the moving target is at time The relative position of the observation is on the horizontal axis ( axis), Indicates that the moving target is at time The relative position of the observation is on the vertical axis ( axis); represents transpose; The moving target is at time The relative position of the observed target is based on the radar measurement at time distance and angles , which is calculated by converting polar coordinates to Cartesian coordinates, is a well-known technical method for those skilled in the art and will not be described in detail here; Recording radar at all times through GPS The absolute position ,in, Indicates that the radar is at time The absolute position of the horizontal axis ( axis), Indicates that the radar is at time The absolute position of the vertical axis ( axis); combined with the moving target at time The relative position of the observation is calculated to calculate the moving target at time The absolute position of is as follows: , in, Indicates that the moving target is at time The absolute position of Indicates that the moving target is at time The absolute position of the horizontal axis ( axis), Indicates that the moving target is at time The absolute position of the vertical axis ( axis); Based on the movement target at the current moment and the previous moment The absolute position difference is divided by the time interval using the finite difference method to calculate the speed of the moving target. ; Also using the finite difference method, based on the moving target at the current moment and the previous moment Divide the speed difference by the time interval to calculate the acceleration of the moving target. , used to reflect the rate of change of the moving target's velocity; the finite difference method is a technical means well known to those skilled in the art and will not be described in detail here; Use motion targets at the moment The absolute position, velocity and acceleration of the moving target are predicted based on the uniform acceleration motion model at the next moment. The absolute position of the moving target is obtained to obtain the predicted absolute position of the moving target, which provides a reference for subsequent target matching. The specific formula is as follows: , in, Indicates that the moving target is at the next moment The predicted absolute position of Indicates the radar sampling time interval, which is determined by the radar sampling frequency. For example, when the radar sampling frequency is 10Hz, the corresponding radar sampling time interval is .
[0021] S2. Introducing a dynamic threshold matching algorithm based on motion trends, matching the absolute position of the moving target at the next moment with the predicted absolute position of the moving target, generating a matching cost, and determining the target position based on the matching cost; The dynamic threshold matching algorithm based on motion trend calculates the Euclidean distance between the absolute position of the moving target observed by the radar at the next moment and the predicted absolute position of the moving target, and combines it with the dynamic matching threshold to generate a matching cost, thereby determining whether the observed absolute position of the moving target belongs to the moving target, thereby improving the robustness and accuracy of moving target tracking; The dynamic matching threshold is obtained by multiplying the basic threshold by the adjustment factor; the basic threshold is the minimum tolerance for matching and is defined according to the radar measurement accuracy to ensure that the matching process can exclude obvious noise even when the moving target is stationary or moving at a low speed; The adjustment factor is calculated based on the speed and acceleration of the moving target and is used to adaptively adjust the dynamic matching threshold to adapt to complex dynamic scenes. Specifically, the speed modulus of the moving target is divided by the speed standard deviation, and then multiplied by an exponential decay parameter as the negative exponent of the exponential function. The output range of the exponential function is The faster the speed of the moving target, the smaller the output of the exponential function, and the slower the speed of the moving target, the larger the output of the exponential function, which is used to reflect the impact of the speed of the moving target on the dynamic matching threshold; further calculate the cosine of the angle between the velocity vector and the acceleration vector of the moving target, and take the absolute value of the cosine of the angle. The range of the absolute value of the cosine of the angle is , to reflect the directional consistency of the acceleration vector and the velocity vector of the moving target. If the angle between the acceleration vector and the velocity vector of the moving target is close to 0 degrees, that is, the absolute value of the cosine of the angle is close to 1, it means that the moving target is accelerating or decelerating, and the dynamic matching threshold is increased to tolerate the deviation. If the angle between the acceleration vector and the velocity vector of the moving target is close to 90 degrees, that is, the absolute value of the cosine of the angle is close to 0, it means that the moving target is turning, and the dynamic matching threshold is reduced to improve the accuracy. The calculation method of the cosine of the angle is a technical means well known to those skilled in the art and will not be described in detail here. The calculation formula for the matching cost is: , in, Indicates that at the current moment The matching cost; Indicates that the radar will The Euclidean distance between the observed absolute position of the moving target and the predicted absolute position of the moving target at the next moment is used to quantify the position deviation to reflect the closeness between the observation and the prediction; Indicates that at the current moment The dynamic matching threshold is calculated as follows: , in, Indicates the basic threshold, which is defined according to the radar measurement accuracy; represents the adjustment factor; Represents the sensitivity parameter, which is used to control the dynamic adjustment range and is set by expert experience. The value range is ; Represents the exponential function term, which is used to ensure that the faster the moving target is, the smaller the exponential function output is; Represents an exponential decay parameter, which is used to control the sensitivity of the moving target's speed to the dynamic matching threshold adjustment. It is set by expert experience and has a value range of ; Represents the velocity modulus, that is, the relative velocity of the moving target; represents the velocity standard deviation, which is used to normalize the velocity modulus and is calculated from the historical velocity data of the moving target, which comes from an existing database; The absolute value of the cosine of the angle between the velocity vector and the acceleration vector of the moving target can reflect the movement trend and is used to adjust the dynamic matching threshold to adapt to the direction change of the moving target; The Euclidean distance is used to quantify the deviation between the absolute position of the observed moving target (observed position) and the predicted absolute position of the moving target at the next moment (predicted position). The matching cost can reflect the size of the position deviation relative to the dynamic matching threshold. The smaller the matching cost, the closer the observed position is to the predicted position, and the higher the matching reliability. Set the matching cost threshold, compare the matching cost with the matching cost threshold, and determine the target location :If at the current moment If the matching cost is less than or equal to the matching cost threshold, the absolute position of the moving target observed at the next moment is confirmed to be the target position Otherwise, the observed absolute position of the moving target is judged to be noise or target loss, and the predicted absolute position of the moving target is used as the target position , to maintain tracking continuity.
[0022] S3, based on the target position, uses a multi-objective optimization algorithm to output an executable optimal control instruction vector; Based on the target position, a multi-objective optimization algorithm is used to introduce tracking error terms and control cost terms. Combined with weight parameters, the objective function is constructed and solved, and an executable optimal control instruction vector is output, that is, the optimal speed vector of the tracking system, to balance the tracking error and control cost. The tracking error term, calculated using the square of the Euclidean distance, measures the deviation between the tracking system's predicted position and the target position, reflecting the spatial proximity of the tracking system and the target position. The predicted position of the tracking system is based on the current tracking system position plus the displacement generated by the control command (calculated by the system dynamics model) within the radar sampling time interval. The control cost term measures the energy consumption of the control command. The weight parameters are used to control the relative importance of the tracking error term and the control cost term. The system dynamics model is a well-known technical means for those skilled in the art and will not be described in detail here. The objective function optimization formula is as follows: , in, Indicates The optimal control command vector at time t; Represents the optimization operation, and finds the control instruction vector that minimizes the objective function as the optimal control instruction vector; Represents the weight parameter of the tracking deviation term, which is used to adjust the importance of the tracking deviation term in multi-objective optimization. It is set according to the expert experience method and the value range is ; represents the tracking error term, which reflects the deviation between the predicted position and the target position of the tracking system after executing the control command vector; represents the squared Euclidean norm; Indicates that the radar is at time The absolute position of represents the control instruction vector in the optimization process, that is, the velocity vector of the tracking system; Represents the weight parameter of the control cost item, which is set according to the expert experience method and has a value range of ; represents the control cost item, which is used to reflect the energy consumption of the control instruction; To ensure that the optimal control instruction vector is within the hardware capability, two physical constraints are imposed: velocity constraint and acceleration constraint; The speed constraint is used to ensure that the modulus of the optimal control instruction vector does not exceed the maximum speed limit; the acceleration constraint is used to ensure that the speed change rate of the optimal control instruction vector does not exceed the maximum acceleration limit, thereby limiting the severity of the speed change; The objective function is solved by the gradient descent method. When the number of iterations of the objective function reaches the maximum number of iterations preset according to the expert experience method, the iteration is stopped and the current control instruction vector is output as the optimal control instruction vector, which is directly used to track the execution of the system.
[0023] In summary, a real-time moving target tracking and decision-making method based on multi-objective optimization algorithm was completed.
[0024] The order in which the embodiments of the invention are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0025] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.
[0026] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the scope of protection of the present invention.
Claims
1. A real-time moving target tracking and decision-making method based on a multi-objective optimization algorithm, characterized in that: The following steps are involved: S1. Measure the observed relative position of the moving target and calculate the absolute position of the moving target; based on the absolute position of the moving target, calculate the speed and acceleration of the moving target, and predict the absolute position of the moving target at the next moment to obtain the predicted absolute position of the moving target; S2. Introducing a dynamic threshold matching algorithm based on motion trends, matching the absolute position of the moving target at the next moment with the predicted absolute position of the moving target, generating a matching cost, and determining the target position based on the matching cost; S3. Based on the target position, a multi-objective optimization algorithm is used to output an executable optimal control instruction vector.
2. A real-time moving target tracking and decision-making method based on a multi-objective optimization algorithm according to claim 1, characterized in that: Said S2 specifically includes: The dynamic threshold matching algorithm based on motion trend introduces Euclidean distance to quantify the deviation between the absolute position of the moving target at the next moment and the predicted absolute position of the moving target, and generates a matching cost in combination with a dynamic matching threshold.
3. The real-time moving target tracking and decision-making method based on a multi-objective optimization algorithm according to claim 2, characterized in that: Said S2 specifically includes: According to the radar measurement accuracy, a basic threshold is defined; based on the basic threshold, an adjustment factor is introduced to calculate the dynamic matching threshold.
4. The real-time moving target tracking and decision-making method based on a multi-objective optimization algorithm according to claim 3 is characterized in that: Said S2 specifically includes: Based on the speed of the moving target, an exponential decay parameter is introduced, and the adjustment factor is calculated by combining the cosine of the angle between the velocity vector and the acceleration vector of the moving target.
5. The real-time moving target tracking and decision-making method based on a multi-objective optimization algorithm according to claim 4 is characterized in that: Said S2 specifically includes: Set the matching cost threshold and compare the matching cost with the matching cost threshold. When the matching cost is less than or equal to the matching cost threshold, it means that the absolute position of the moving target at the next moment is the target position. Otherwise, the predicted absolute position of the moving target is used as the target position.
6. The real-time moving target tracking and decision-making method based on a multi-objective optimization algorithm according to claim 1, characterized in that: Said S3 specifically includes: In the implementation process of the multi-objective optimization algorithm, based on the target position, the tracking error term and the control cost term are introduced, combined with the weight coefficient, the objective function is constructed and solved, and the executable optimal control instruction vector is output.
7. The real-time moving target tracking and decision-making method based on a multi-objective optimization algorithm according to claim 6, characterized in that: Said S3 specifically includes: The tracking error term introduces the Euclidean distance to quantify the deviation between the predicted position of the tracking system and the target position; the predicted position of the tracking system is calculated based on the current position of the tracking system and the displacement generated by the control command within the radar sampling time interval; the control cost term is obtained by calculating the squared Euclidean norm of the control command vector.
8. The real-time moving target tracking and decision-making method based on a multi-objective optimization algorithm according to claim 7, characterized in that: Said S3 specifically includes: In the implementation process of the multi-objective optimization algorithm, the objective function is solved by the gradient descent method, and velocity constraints and acceleration constraints are imposed on the control instruction vector; when the number of iterations of the objective function reaches the preset maximum number of iterations, the iteration is stopped and the current control instruction vector is output as the optimal control instruction vector for tracking the execution of the system.