Video tracking control method and device for non-linear moving target based on adaptive sliding mode control

Abstract

The invention discloses a video tracking control method and device for a non-linear moving target based on adaptive sliding mode control, which are suitable for accurate tracking of a dynamic target. The system is composed of a visual sensing module, a main control module and an execution mechanism, the visual sensing module collects position information of a target object in real time and transmits the position information to a main control chip through serial port communication, the main control chip calculates the error between the target position and the current position of the system and the error change rate, and a sliding mode surface is designed based on sliding mode control. Meanwhile, sliding mode control parameters are dynamically adjusted in combination with a self-adaptive control algorithm, so that the robustness of the system to uncertainty and external interference is enhanced, and meanwhile the chattering phenomenon is restrained. A stepping motor control signal is optimized by introducing a dynamic acceleration and deceleration algorithm, so that control smoothness and response precision are ensured, and real-time accurate tracking of a target object is realized. The method has strong robustness, high dynamic performance and control precision on the premise of not depending on a target kinematic model, can be widely applied to the fields of unmanned aerial vehicle tracking, intelligent monitoring, robot vision and the like, and effectively solves the problems of easy buffeting, response hysteresis, insufficient control precision and the like during nonlinear target tracking in the prior art.

Description

Technical Field

[0001] The present invention relates to the fields of automatic control and computer vision, and specifically to a video tracking control method and device for non-linear moving targets based on sliding mode control, which can be applied to fields such as UAV tracking, intelligent monitoring, and robot vision. Background Technique

[0002] With the development of society, the application of target tracking has become increasingly widespread. Unmanned driving systems need to identify and track other vehicles, pedestrians, and road signs; UAV autonomous reconnaissance and target strikes also rely on precise video target tracking technology to track moving targets in real time. As the non-linearity, dynamic characteristics of target motion and the complexity of the environment continue to increase, the requirements for the sensitivity and control performance of the target tracking system are becoming more and more stringent. However, although traditional target control methods (such as those based on the PID control algorithm) have the characteristics of simple structure, strong robustness and adaptability, and are widely used in linear target tracking systems, their performance in non-linear target tracking is often not satisfactory: for example, in the face of a more complex environment and dynamic changes, due to the limited computing power of the PID controller, its response sensitivity is low, it is difficult to effectively respond to the rapid changes of the target, resulting in slow system response and difficult rapid convergence of the target deviation. Therefore, improving the sensitivity and control performance of target control has become an urgent problem to be solved.

[0003] In recent years, sliding mode control has been widely used in high-precision control and dynamic systems due to its excellent robustness and dynamic performance. For example, in the field of robotics, sliding mode control can ensure system stability, improve positioning accuracy and adaptability, and achieve precise motion control in complex environments. Since sliding mode control has good control ability in non-linear systems, it also has important application value in the field of video tracking. Its core principle is to design a sliding mode surface and its control law so that the system state can quickly converge and stabilize on the sliding mode surface. However, the performance of traditional sliding mode control highly depends on the design of the sliding mode surface and the control law. Fixed gain parameters may lead to excessive or insufficient control rigidity, thus affecting the convergence speed and control accuracy, and may cause high-frequency chattering problems.

[0004] To overcome these deficiencies, the present invention introduces adaptive sliding mode control. By adjusting the control gain in real time, the system can optimize the control parameters under different working conditions, improve the adaptability to external disturbances, and suppress chattering while ensuring fast convergence. Compared with traditional sliding mode control, adaptive sliding mode control can balance the response speed and steady-state accuracy, enhance the robustness of the system, and is particularly suitable for nonlinear target tracking scenarios. With the characteristics of strong robustness, high sensitivity, and easy implementation, adaptive sliding mode control can effectively make up for the deficiencies of traditional PID control in nonlinear target tracking, especially maintaining precise control in complex environments. Therefore, in-depth research on the application of sliding mode control in video target tracking can not only improve the real-time performance and stability of the system but also enhance its adaptability to complex dynamic targets, which has important practical significance in fields such as UAV tracking, intelligent monitoring, and robot vision.

[0005] Aiming at the deficiencies of video tracking of moving targets, the present invention proposes a nonlinear moving target video tracking control method and device based on adaptive sliding mode control. The device consists of three parts: a visual sensing module, a main control module, and an actuator. The visual sensing module uses an intelligent vision processing unit to collect the position information of the target in real time and perform preprocessing of target recognition and coordinate extraction. The main control module is implemented by an embedded processor and is responsible for error calculation, sliding mode control law design, and the execution of dynamic acceleration and deceleration algorithms. The actuator uses a stepper motor and its driver to achieve target tracking through control signals.

[0006] Specifically, the present invention achieves high-precision target tracking through the following steps: The visual sensing module collects the spatial position information of the target in real time and transmits the data to the main control module through serial communication; the main control module designs a sliding mode surface and generates a control signal based on the error between the target position and the current system position and its rate of change. During the calculation of the control signal, an adaptive control algorithm is introduced to dynamically adjust the sliding mode parameters, enhance the robustness of the system to external disturbances, and effectively suppress the chattering phenomenon caused by sliding mode switching control; at the same time, a dynamic acceleration and deceleration algorithm is used to optimize the motion curve of the stepper motor, ensuring both smooth control and improving the response speed and positioning accuracy; finally, the actuator adjusts according to the signal to accurately complete target tracking.

[0007] Compared with traditional methods, the present invention combines sliding mode control, adaptive control, and dynamic acceleration and deceleration algorithms, significantly improving the control accuracy and stability of the system and optimizing the control algorithm. Using a stepper motor drive system also greatly improves the response accuracy and adaptability, ensuring precise dynamic tracking of the target.

[0008] The present invention is widely applied in fields such as UAV tracking, robot vision, industrial automation, etc. Especially in the high-precision visual positioning and real-time tracking of non-linear dynamic targets, it has successfully solved problems such as insufficient control accuracy, low sensitivity, and poor robustness in the prior art. Summary of the Invention

[0009] The technical problem to be solved by the present invention: The present invention discloses a video tracking control method and device for non-linear moving targets based on adaptive sliding mode control, which solves the above problems existing in the existing tracking control.

[0010] To solve the above technical problems, the present invention provides the following technical solutions:

[0011] A video tracking control method and device for non-linear moving targets based on adaptive sliding mode control, characterized in that it consists of a vision sensing module, a main control module, and an actuator. Among them, the vision sensing module real-time collects the position information of the target object, transmits it to the main control chip through serial communication, calculates the error and the rate of change of the error between the target position and the current position of the system, designs a sliding mode surface based on sliding mode control, combines an adaptive control algorithm, and dynamically adjusts the sliding mode control parameters to enhance the robustness of the system to uncertainties and external disturbances, and at the same time suppress the chattering phenomenon. By introducing a dynamic acceleration and deceleration algorithm to optimize the control signal of the stepping motor, the control smoothness and response accuracy are ensured, and the real-time and precise tracking of the target object is realized.

[0012] Preferably, in this device, the vision sensing module mainly uses an intelligent vision processing unit (such as OpenMV-H7Plus) to realize the real-time acquisition of the graphic data of the target object. By setting the resolution, the processing speed is improved while ensuring the image quality. The camera continuously captures the real-time pictures in the scene, which serves as the basis for subsequent target recognition. Before each tracking starts, a picture is taken and stored in an image variable. Then, through target detection based on color and shape: If the target has specific color characteristics, the threshold range of the color is predefined, and then the regions that meet these characteristics are screened out from the image; If the target has obvious geometric shapes, the regions that meet the conditions are found through the shape detection algorithm, and their center positions are determined. Finally, the center point of the target is calculated as the position coordinate, and this data is transmitted to the main control chip through the serial port for subsequent calculation.

[0013] Through this vision sensing module, efficient and flexible real-time target detection and tracking are realized, the target position is accurately located, and data is transmitted through the serial port, improving the system response speed and accuracy.

[0014] Preferably, the adaptive sliding mode control module is the core point of the present invention. First, define the non-linear state space form of the target object: (where x is the state vector of the system, Denote the derivative of \(x\), \(f(x)\) is the non - linear dynamic part of the system, \(g(x)\) is the input gain, and \(u\) is the control input). Then, clarify the control objective of the system (\(x\) target , \(y\) target ), such that the system state (\(x\) current , \(y\) current ) converges to the target state, and reduce the output error (\(e\) x = \(x\) target - \(x\) current , \(e\) y = \(y\) target - \(y\) current ) to zero. Furthermore, design the sliding mode surface: (where \(S(t)\) is the state of the sliding mode surface at time \(t\), \(e\) is the control error, \(\lambda\) is the slope constant of the sliding mode surface, \(\lambda>0\), representing a positive parameter that determines the convergence rate of the system and is mainly used to adjust the dynamic characteristics of the sliding mode surface), so that the system state converges to \(S(t)=0\). Furthermore, design the control law \(U\) to ensure the stability and rapidity of the system by dynamically adjusting the control gain:

[0015]

[0016] where \(K(e)\) is an adaptive gain function with respect to the error \(e\), which changes with the change of the error \(e\) to realize the dynamic adjustment of the sensitivity of the controller to the error; \(K\) 0 is the basic gain; \(\rho\) is the gain adjustment coefficient used to control the change range of the gain with the error; \(sat(S)\) is a saturation function (used to limit the amplitude of the input signal) to replace the sign function \(sgn(S)\) (outputting the sign of the input signal, indicating the positive or negative of the signal) to suppress the chattering of the system; \(U\) x , \(U\) y are control input variables; \(\varPhi\) is a control parameter used to ensure that the system does not deviate too much from the predetermined trajectory during the sliding mode control process.

[0017] The adaptive sliding mode control algorithm ensures that the system can quickly and stably converge to the target state in the face of external disturbances or uncertainties, and is applicable to non - linear dynamic target tracking scenarios. And this method only needs to construct a sliding mode surface and an approaching law, without building an accurate kinematic model of the stepping motor, avoiding complex model derivation and parameter identification, solving the problem of difficult accurate modeling in traditional methods, and realizing more accurate and convenient closed - loop control. At the same time, to improve the robustness and adaptability of the system, the present invention optimizes the adaptive parameter \(K\) to make it change from a fixed parameter to a dynamic function with respect to the error \(e\), enabling it to more effectively cope with uncertainties and external disturbances.

[0018] Preferably, in the present invention, a dynamic acceleration - deceleration algorithm is adopted to enhance the stability and accuracy of target tracking. First, determine the target position \(x\) target and the initial position \(x\) current, set the maximum speed v max and the maximum acceleration a max , and the allowable error range ∈; then in the dynamic acceleration stage, according to the current error |x target -x current |, dynamically adjust the acceleration and then update the speed v new = min(v current + a·Δt, v max ), gradually increase the speed v until it reaches v max , synchronously update the displacement x new = x current + v new ·Δt, so that the system position continuously approaches the target; in the constant-speed operation stage, when the error is large, maintain the maximum speed v new = v max to move and quickly reduce the distance x new , x new = x current + v new ·Δt; in the dynamic deceleration stage, when the system state approaches the target, gradually reduce the speed and decelerate in a smooth manner to avoid overshoot; when the error |x target -x current | ≤ ∈, the control signal stops and the target position is reached.

[0019] Through the dynamic acceleration algorithm, the vibration of the stepping motor is effectively reduced, and the stability and accuracy of target tracking are improved.

[0020] A method for target tracking uses the above-mentioned video tracking control device for non-linear moving targets based on adaptive sliding mode control to accurately track the target.

[0021] Beneficial effects obtained by the present invention:

[0022] The present invention proposes a video tracking control method and device for non-linear moving targets based on adaptive sliding mode control. The device has a simple structure and is mainly composed of a vision sensing module, a main control module and an actuator. The actuator consists of two motors to realize rotation in the horizontal and vertical directions respectively. The innovative adoption of the design of adaptive sliding mode control greatly enhances the robustness of the system to external disturbances, reduces the chattering phenomenon of the traditional sign function, and effectively improves the smoothness of the driving signal of the stepping motor through the dynamic acceleration and deceleration algorithm, thereby enhancing the stability of the curve fitting of target tracking, realizing a fast response to the target motion trajectory, and ensuring high-precision tracking. It can be widely applied in fields such as UAV tracking, robot vision, and industrial automation. Description of the Drawings

[0023] Figure 1 This is the block diagram of the video tracking control system for nonlinear moving targets, which includes a visual sensor module, a main control module and an actuator.

[0024] Figure 2 The flowchart describes the algorithm from target information acquisition to error calculation, sliding mode control and stepper motor control signal output.

[0025] Figure 3 This is the horizontal error curve diagram of real-time tracking of target tracking achieved through PID control.

[0026] Figure 4 The horizontal error curve diagram of real-time tracking of target tracking achieved through adaptive sliding mode control. DETAILED DESCRIPTION

[0027] The specific implementation modes of the present invention are further explained in detail below through the description of embodiments, so as to help those skilled in the art to have a more complete, accurate and in-depth understanding of the concept and technical solution of the present invention.

[0028] Embodiment 1:

[0029] The above-mentioned video tracking control method and device for nonlinear motion target based on adaptive sliding mode control are used to automatically track human faces:

[0030] This example uses the face as the tracking target and adopts the adaptive sliding mode control method to achieve the function of real-time camera capture and tracking of the face position. The system dynamically adjusts the direction of the camera to keep it aligned with the target face, and improves the tracking accuracy and stability through the smooth drive of the stepper motor.

[0031] The system device is mainly composed of a visual sensor module, a main control module and an actuator.

[0032] Visual sensing module: Use OpenMV-H7 Plus to capture video images in real time, with built-in machine vision algorithms, such as finding color blocks, face detection, edge detection, etc. In this example, the built-in face detection algorithm is used to obtain the coordinate information of the center point of the face. By performing visual processing tasks on OpenMV-H7 Plus, the computing resources of the main control board can be saved, and the time delay of communication between modules can be reduced, thereby improving the overall operation speed of the system. Among them, the image resolution is selected as 320x240 (QVGA) to meet the real-time requirements.

[0033] Main control module: In this example, the STMicroelectronics STM32F103C8T6 chip can be used to receive the face coordinates transmitted by the vision module, calculate the error and error change rate between the target position and the current camera direction, and execute precise sliding mode control and dynamic acceleration and deceleration algorithms.

[0034] Actuator: A pan-tilt unit driven by a stepper motor. By controlling the rotation of the pan-tilt unit, dynamic adjustment of the camera in the horizontal and vertical directions is achieved, ensuring that the target face is always at the center of the camera's field of view. The stepper motor has a fixed step angle, usually within the accuracy range of 0.9°, 1.8°, etc., which can meet the accuracy and speed requirements of face tracking.

[0035] First, the vision sensing module captures video images in real time through the camera for the positioning and tracking of face targets. The Haar cascade algorithm built into the OpenMV-H7 Plus is used to obtain the detected face area, and then the center point coordinates of this area are calculated. This coordinate represents the position of the target face in the image, and this data is stored in the variables (x target , y target ) as the target position. This data will be transmitted to the main control chip through the serial port for subsequent processing.

[0036] After receiving the position data of the target face, the main control chip uses the image center point (160, 120) as the system position and calculates the error value e(t) between the current system position and the center of the target face area in real time. The error value is calculated by the following formula:

[0037] e x (t) = x target - 160 e y (t) = y target - 120

[0038] where e x and e y are the errors in the horizontal and vertical directions respectively.

[0039] At the same time, the main control chip also calculates the error change rate , that is, the change speed of the error with time, which is expressed as:

[0040]

[0041] Based on these errors and change rates, the main control chip designs a sliding mode surface S(t). This sliding mode surface helps the control system quickly drive the error to approach zero and maintain it in a low-error state. The sliding mode surface is designed as:

[0042]

[0043] (It is described that the value of Lambda depends on) To further improve the system's adaptability to uncertainties and external disturbances, the system uses an adaptive sliding mode control method. The adaptive controller dynamically adjusts the control gain according to the magnitude of the target error |e|, enabling the system to ensure real-time response performance while achieving high-precision tracking. The control signal is designed as:

[0044] U(t) = -K(e)*sat(S(t))

[0045] where, is the adaptive function.

[0046] After calculating the control signal, the system adjusts the speed and acceleration of the stepper motor according to the control signal through a dynamic acceleration and deceleration algorithm. This algorithm dynamically adjusts the movement of the stepper motor according to the change of the target position to ensure a smooth response to the dynamic change of the target and avoid mechanical vibration or accuracy degradation caused by sharp acceleration or deceleration. The stepper motor drives the pan-tilt to adjust the horizontal and vertical directions of the camera according to the calculated control signal. The precise step control provided by the stepper motor can help the system quickly adjust the angle of the camera within a small error range, thereby maintaining continuous tracking of the target.

[0047] Finally, based on the above steps, the system determines in real time whether the error is less than the preset threshold ε. If the error is already less than the threshold and the adjustment amplitude of the control signal is also small enough, the system will stop adjusting and complete the stable tracking of the target. The judgment condition is:

[0048] |e x | < ε and |e y | < ε

[0049] Otherwise, the system will continue to collect the target position information, recalculate the error and the control signal, and repeat the above control process to ensure that the target always remains at the center of the camera's field of view.

[0050] This system can capture and track the target face in real time in complex scenarios. Even if the target moves quickly or is briefly occluded, it can achieve smooth tracking through the sliding mode control and the dynamic acceleration and deceleration algorithm. And the simulation test results show that, such as Figure 3 (the error of using the traditional target tracking system), Figure 4 (the error of using the present invention to achieve target tracking), the accuracy of achieving target tracking by using the present invention is much greater than that of the traditional target tracking system. Thus, it can be seen that this system has high robustness and low latency performance when dealing with target switching and environmental interference, and has important practical application value.

[0051] Example 2:

[0052] Apply the above video tracking control method and device for non-linear moving targets based on adaptive sliding mode control to a high-power laser damage system:

[0053] This example introduces the application of the video tracking control method and device for non-linear moving targets based on adaptive sliding mode control in a high-power laser damage system (such as anti-drone C-UAS systems, etc.). This technology can achieve real-time and precise laser aiming and damage operations on dynamic moving targets such as drones and vehicles. The core goal of the system is to achieve automatic tracking and energy projection of the target through visual recognition and precise control technologies.

[0054] The system device mainly consists of a visual sensing module, a main control module, and an actuator.

[0055] Visual sensing module: Use OpenMV-H7 Plus to capture video images in real time. It has built-in a variety of machine vision algorithms, such as the template matching algorithm based on NCC. In this example, the position of the drone is identified and located in the image through the NCC algorithm. By performing visual processing tasks on the OpenMV-H7 Plus, the computational burden on the main control board can be reduced, the time delay of communication between modules can be decreased, and the overall operating speed of the system can be improved. The image resolution is selected as 640x480 (VGA) to meet the requirements of real-time processing.

[0056] Main control module: In this example, the STM32F407 chip of STMicroelectronics can be used to receive the coordinates of the drone transmitted by the visual module, calculate the error and the rate of change of the error between the position of the drone and the target direction of the current actuator, and at the same time execute the precise sliding mode control and the dynamic acceleration and deceleration algorithm.

[0057] Actuator: The stepping motor drives the pan-tilt with a laser aiming device and a camera. The laser is fixedly pointed at the center of the camera image. By controlling the camera to track the drone, the dynamic adjustment of the laser in the horizontal and vertical directions is linked to ensure that the laser always points at the drone. The stepping motor has a fixed stepping angle, usually within the accuracy range of 0.9°, 1.8°, etc., which can meet the accuracy and speed of tracking.

[0058] First, the system obtains the position information of the target in real time through the visual sensing module. To achieve the identification of the drone, the system uses the NCC template matching algorithm. In this process, the system compares the pre-set drone template image with the real-time captured image. By calculating the matching degree of each position in the image with the template image, the best matching area is found. The position with the highest matching degree corresponds to the position of the drone in the image. Through this position, the system further calculates the center coordinates of the drone image. These coordinates will be stored in the variable (x target ,y target)Taking the center point of the image (320, 240) as the target position and the system position, the subsequent target tracking and aiming operations are the same as those in Embodiment 1.

[0059] This example mainly demonstrates the advantages and practical effects of the video tracking method and device based on sliding mode control in complex military scenarios, such as precise damage to dynamic targets and good anti-interference ability in complex environments, providing strong support for high-precision and high-robustness applications.

[0060] The above embodiments are only used to illustrate the technical idea of the present invention, and the protection scope of the present invention cannot be limited thereby. Any changes made on the basis of the technical solution according to the technical idea proposed by the present invention shall fall within the protection scope of the present invention; the technologies not involved in the present invention can all be realized through the prior art.

Claims

1. A video tracking control method and device for a nonlinear moving target based on adaptive sliding mode control, characterized in that: It consists of a visual sensor module, a main control module and an actuator. The visual sensor module collects the position information of the target object in real time, transmits it to the main control chip through serial communication, calculates the error and error change rate between the target position and the current position of the system, and designs the sliding surface based on the sliding mode control. Combined with the adaptive control algorithm, the sliding mode control parameters are dynamically adjusted to enhance the robustness of the system to uncertainty and external interference, while suppressing the chattering phenomenon. By introducing a dynamic acceleration and deceleration algorithm to optimize the stepper motor control signal, the control smoothness and response accuracy are guaranteed, and the real-time and accurate tracking of the target object is achieved.

2. The method and device for video tracking control of a nonlinear moving target based on adaptive sliding mode control according to claim 1, characterized in that: include: Visual sensor module: that is, image acquisition and processing module, mainly used to collect the position information of the target in real time; Main control chip: It is an embedded microcontroller with CPU processing function, which mainly implements error calculation module, sliding mode control module, adaptive control module and dynamic acceleration and deceleration algorithm; Actuator: It is a high-precision stepper motor pan / tilt that executes the control instructions of the main control module to complete target tracking.

3. The method and device for video tracking control of a nonlinear moving target based on adaptive sliding mode control according to claim 1, characterized in that: By adopting the visual sensing module in the present invention, the position and motion information of the target object can be accurately collected in real time, meeting the high-precision and flexible tracking requirements and ensuring the system's rapid response to dynamic targets.

4. The method and device for video tracking control of a nonlinear moving target based on adaptive sliding mode control according to claim 1, characterized in that: This method adopts sliding mode control, sets a suitable sliding surface based on the system state, designs a reaching law, and replaces the sign function with a saturation function to achieve the sliding of the system state on the sliding surface and eventually approach the target value, that is, to make the system error approach zero and make the system state quickly approach the sliding surface, thereby achieving stable control of the system.

5. The method and device for video tracking control of a nonlinear moving target based on adaptive sliding mode control according to claim 1, characterized in that: Adaptive control is added to assist sliding mode control. Dynamic adjustment of sliding mode control gain according to error and sliding surface state solves the contradiction between response speed and accuracy caused by fixed gain in traditional sliding mode control, avoids high-frequency jitter caused by frequent signal switching of sliding mode controller, and optimizes the balance between tracking speed and steady-state accuracy.

6. The method and device for video tracking control of a nonlinear moving target based on adaptive sliding mode control according to claim 1, characterized in that: The dynamic acceleration and deceleration algorithm makes the system control smoother and more responsive by dynamically adjusting the acceleration and deceleration curves, solves the control delay and oscillation problems of the actuator, and improves the stability of the system during target acceleration or deceleration.

7. The method and device for video tracking control of a nonlinear moving target based on adaptive sliding mode control according to claim 1, characterized in that: In this device, a stepper motor is used as an actuator, which is responsible for accurately adjusting the motion path or direction of the system to ensure real-time tracking of the target.

8. A target tracking device, characterized in that: The video tracking control method for nonlinear motion targets based on adaptive sliding mode control described in claims 1 to 7 is used to accurately track the target.

Images