A design method for shipborne optoelectronic tracking system with adaptive backstepping sliding mode control
Through the adaptive inverse-step sliding mode control method, combined with the inverse-step method and RBF neural network, the problem of reducing accuracy of the ship-based photoelectric tracking system under the hull shaking conditions is solved, and effective suppression of nonlinear interference and improvement of tracking accuracy is achieved. .
Patent Information
- Application Number
- CN202110594758.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-05-28
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2041-05-28
AI Technical Summary
The disturbing torque and model uncertainty of the ship-based photoelectric tracking system under the hull shaking conditions lead to a reduction in tracking accuracy. The existing control methods require high system model accuracy, making it difficult to effectively suppress nonlinear interference and jitter phenomena.
Adaptive inverse-step sliding mode control method is adopted, the basic controller and sliding mode control law are designed through the inverse-step method, and the RBF neural network is used to estimate perturbation, and the adaptive inverse-step sliding mode control law is designed, combining integral sliding mode and complementary sliding mode to improve the robustness and immunity of the system.
Effectively suppress nonlinear interference in the system, improve photoelectric tracking accuracy, enhance dynamic disturbance ability, reduce jitter phenomenon, improve control performance, and adapt to system model uncertainty and internal disturbance.
Smart Images

Figure CN113156825B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of high-precision control of an optoelectronic tracking system, and in particular to a design method for a shipborne optoelectronic tracking system with adaptive backstepping sliding mode control. Background Art
[0002] The optoelectronic tracking system is a device that integrates optics, electrical, mechanical and control. It can transmit video images of foreign targets in the surveillance area in real time, providing a reliable basis for further judgment or action. Therefore, it is widely used in military and civilian fields.
[0003] Compared with the fixed-base optoelectronic tracking platform, the shipborne optoelectronic tracking system has the maneuverability of transfer, can change location at any time, and can avoid the influence of bad weather that affects target imaging and light transmission efficiency. However, the shipborne optoelectronic tracking platform is installed on the deck and is fixed to the deck. The hull is affected by wind, waves, currents, etc., causing the ship to shake, which will also act on the tracking platform. Under such working conditions, the system will cause mass unbalance torque, friction torque between its own shaft system, cable elastic torque and other interference torques and uncertainty of the controlled object model due to the tilt of the tracking platform; this interference torque and uncertainty greatly reduce the tracking accuracy. In order to improve the tracking accuracy, this requires the system to have strong dynamic anti-interference ability and robustness.
[0004] In recent years, servo control has been mainly based on classical PID, on which control methods such as disturbance observer have been introduced. However, these methods have very high requirements on the accuracy of the model built for the system. In practice, it is difficult to obtain an accurate model, which greatly reduces the tracking accuracy. Summary of the invention
[0005] In view of the problems existing in the prior art, the present invention discloses a design method for a shipborne optoelectronic tracking system with adaptive backstepping sliding mode control, comprising the following steps:
[0006] Build a two-axis shipborne optoelectronic tracking platform;
[0007] According to the transfer function of the DC torque motor in the photoelectric tracking servo system, the state equation of the DC torque motor model is obtained;
[0008] According to the state equation of the DC torque motor model, the basic controller of the DC torque motor is designed by backstepping method.
[0009] Based on the backstepping method and sliding mode control, the backstepping sliding film control law is designed;
[0010] The stability of the basic controller and the backstepping sliding film controller is verified by Lyapunov function I;
[0011] According to the RBF neural network, the disturbance d is estimated and the adaptive backstepping sliding mode control law is designed;
[0012] Lyapunov function II is used to verify the stability of the adaptive backstepping sliding mode controller.
[0013] Furthermore, the process of designing a backstepping synovial film control law based on a synovial film controller using the backstepping method is as follows:
[0014] Design the integral sliding surface S(t) and the complementary sliding surface S c (t) are as follows:
[0015]
[0016]
[0017] According to the integral sliding surface S(t) and the complementary sliding surface S c (t) and according to the sliding mode equivalent control principle, the sliding mode control law u is obtained by the equivalent control law u eq and the switching control law u hit The sliding mode control law is designed as follows:
[0018]
[0019]
[0020]
[0021] Where D is the upper bound of the perturbation d, |d|≤D.
[0022] Furthermore, the expression of the adaptive backstepping sliding mode control law for estimating the disturbance d according to the RBF neural network is as follows:
[0023] Defining perturbations
[0024] d=W *T S(z)+ε (19)
[0025] Among them, z is the neural network input, W * is the ideal weight of RBF, ε is the neural network approximation error, S(z)=[s1,s2,…,s j ] is the RBF neural network basis function vector;
[0026] The weight error is defined as:
[0027] in, is the error between the ideal weight and the estimated weight, Estimate weights for neural networks;
[0028] Assume that the disturbance is estimated as:
[0029]
[0030] in, is the estimated value of ε;
[0031] The adaptive backstepping sliding mode control law is obtained as:
[0032]
[0033] z1(t)=x1(t)-x d (t) (5)
[0034] z2(t)=x2(t)-α1(t) (6)
[0035]
[0036]
[0037]
[0038] Among them, x d (t), is the target given quantity, the first-order derivative and the second-order derivative are given, k1 is the controller parameter, k2 represents the controller parameter, a0 and a1 are both motor parameters, α1 is the virtual control quantity in the backstepping design, is the estimated value of the RBF neural network weight, S(z) is the neural network basis function, ε neural network estimation error, and d is the total disturbance.
[0039] Furthermore, the RBF neural network basis function vector S(z) is expressed as follows:
[0040] S(z)=[s1,s2,...,s l ] T (twenty two)
[0041]
[0042] Where z = [z1 z2] T is the neural network input, W * is the ideal weight of RBF, S(z) is the RBF neural network basis function, c j and σ j are the width and center of the jth Gaussian basis function, and l is the number of nodes in the hidden layer of the neural network.
[0043] Due to the adoption of the above technical scheme, the present invention provides a design method for a shipborne optoelectronic tracking system with adaptive backstepping sliding mode control. Aiming at the problem that the shipborne optoelectronic tracking system needs to have strong dynamic anti-disturbance capability when tracking targets, adaptive sliding mode control is used for the speed loop of the optoelectronic tracking servo system. The core of the method is that the sliding mode control has high robustness to system parameter changes and external disturbances, but it will produce jittering phenomenon. Further using a neural network to approximate disturbances can effectively reduce the jittering phenomenon and improve control performance. The control method not only has good robustness and adaptability to system model uncertainty and internal disturbances, but can also effectively suppress the influence of nonlinear factors in the system, thereby improving the tracking accuracy of the optoelectronic tracking system. The present invention designs a sliding mode controller in combination with the backstepping method to overcome the influence of disturbances and model uncertainty on the shipborne optoelectronic tracking system. BRIEF DESCRIPTION OF THE DRAWINGS
[0044] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0045] Figure 1 This is the working flow chart of the shipborne optoelectronic tracking system described in the present invention;
[0046] Figure 2 It is the model diagram of DC torque motor;
[0047] Figure 3 It is a structural diagram of the adaptive backstepping sliding mode control system of the speed loop of the shipborne optoelectronic tracking system of the present invention;
[0048] Figure 4 This is the simulation diagram of PI speed loop;
[0049] Figure 5 Input simulation diagram for PI control;
[0050] Figure 6 This is the simulation diagram of the back-stepping sliding mode speed loop;
[0051] Figure 7 Input simulation diagram for backstepping sliding mode control;
[0052] Figure 8 This is the simulation diagram of the adaptive backstepping sliding mode speed loop;
[0053] Fig. 9 Input simulation diagram for adaptive backstepping sliding mode control. DETAILED DESCRIPTION
[0054] In order to make the technical solutions and advantages of the present invention more clear, the technical solutions in the embodiments of the present invention are clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention:
[0055] A design method for a shipborne optoelectronic tracking system with adaptive backstepping sliding mode control comprises the following steps:
[0056] Build a two-axis shipborne optoelectronic tracking platform, which includes a host computer, a controller, a first encoder, a second encoder, an azimuth frame, a pitch frame, a camera, a gyroscope, a DC torque motor and a load;
[0057] The camera is mounted on the inner frame of the tilt Figure 1 ,This system uses a desktop computer as the host computer. The real-time data collected by the camera is connected to the host computer through a data line. The controller receives the image information obtained by the host computer and extracts the miss distance of the tracking target. The host computer sends the miss distance to the control chip DSP through the data line. The control chip calculates the control input information of the system according to the encoder information and the gyroscope information, and the torque motor generates the corresponding torque; the gyroscope is installed on the visual axis to measure the angular velocity in the inertial space; the DC torque motor is directly connected to the pitch axis or the azimuth axis;
[0058] In the control process of the shipborne stable tracking platform, the DC torque motor is directly connected to the load. If the influence of mechanical resonance is ignored, the DC torque motor and the load can be regarded as a whole. Therefore, the controlled object of the stable platform system can be regarded as a DC torque motor. The DC torque motor model is as follows: Figure 2 shown.
[0059] According to the transfer function of the DC torque motor in the photoelectric tracking servo system, the state equation of the DC torque motor model is obtained;
[0060] According to the state equation of the DC torque motor model, the basic controller of the DC torque motor is designed by backstepping method.
[0061] If the disturbance is not considered, the transfer function of the controlled object of the photoelectric tracking servo system is:
[0062]
[0063] Convert it to differential form:
[0064] Where: a0 = C m K e / JL, a1=R / L, b0=C m / JL
[0065] When the disturbance is further considered, the differential form of the controlled object model is transformed as follows:
[0066] x1=ω (2)
[0067]
[0068]
[0069] Using the backstepping method, a model-based controller is designed. Based on the controlled object model, the controller is designed as follows:
[0070] z1(t)=x1(t)-x d (t) (5)
[0071] z2(t)=x2(t)-α1(t) (6)
[0072]
[0073]
[0074]
[0075] Among them: z1, z2 represent error variables, K2 represents gain coefficient, x1, x2 represent velocity and acceleration respectively, u represents control input, x d (t), is the target given quantity, given the first-order derivative and the second-order derivative. k1 is the controller parameter, α1(t), are the virtual feedback quantity and the first-order derivative of the virtual feedback quantity, respectively.
[0076] When the disturbance is unknown, the strong robustness of sliding mode control to disturbance and parameter uncertainty is utilized. In order to overcome the problem that the introduction of integral sliding mode may cause system instability, complementary sliding mode is introduced on the basis of integral sliding mode, and the integral sliding mode surface S(t) and complementary sliding mode surface S are designed. c (t) is:
[0077]
[0078]
[0079] Among them, k2 is a normal number greater than 0.
[0080] According to S(t) and S c The relationship between (t) is:
[0081]
[0082]
[0083] The sliding mode control law u is composed of the equivalent control law ueq and the switching control law u hit Composition, according to the RBF neural network to estimate the disturbance d, the sliding mode control law is designed as:
[0084]
[0085]
[0086]
[0087] Where D is the upper bound of the perturbation d, |d|≤D.
[0088] To ensure the stability of the basic controller and the backstepping sliding film controller, the Lyapunov function I is selected as:
[0089]
[0090] Derivative of the above formula (17) yields:
[0091]
[0092] When the system reaches the sliding mode, the state parameters of the system are no longer affected by uncertain factors such as parameter changes and external disturbances, thereby ensuring the strong robustness of the system. However, the chattering problem affects the application of sliding mode control. In order to reduce the chattering problem caused by sliding mode control, according to the approximation characteristics of the RBF neural network, the disturbance d is approximated with a neural network, which can effectively reduce chattering and overshoot. The relationship is:
[0093] d=W *T S(z)+ε (19)
[0094] Among them, z is the neural network input, W * is the ideal weight of RBF, ε is the neural network approximation error, S(z)=[s1,s2,…,s j ] is the RBF neural network basis function vector,
[0095] The weight error is defined as:
[0096] in, is the error between the ideal weight and the estimated weight, Estimate weights for neural networks;
[0097] Assume that the disturbance is estimated as:
[0098]
[0099] in, is the estimated value of ε;
[0100] Adaptive neural network basis functions:
[0101] S(Z)=[s1,s2,...,s l ] T (twenty two)
[0102]
[0103] Where z = [z1 z2] T is the neural network input, W * is the ideal weight of RBF, S(z) is the RBF neural network basis function, c j and σ j are the width and center of the jth Gaussian basis function respectively. S(Z) is the output of the hidden layer of the neural network, l is the number of nodes in the hidden layer of the neural network;
[0104] The adaptive backstepping sliding mode control law is designed as:
[0105]
[0106] in, is an estimate of ε, is the error between the ideal weight and the estimated weight;
[0107] In order to verify the stability of the adaptive backstepping sliding mode controller, the Lyapunov function II is selected as:
[0108]
[0109] Derivative of the above formula:
[0110]
[0111] In order to make the system stable and converge faster, we can get:
[0112]
[0113]
[0114] Among them, λ1 and λ2 are convergence coefficients.
[0115] To verify the effectiveness of the controller designed in this paper, the control effects of the adaptive backstepping sliding mode control method, backstepping sliding mode control and PI control are compared and analyzed when all the parameters of the controller are the same. The azimuth axis of a certain type of shipborne optoelectronic tracking platform is taken as an example, and the pitch axis is the same. The parameters of the controlled object are as follows: L = 18mH, R = 6.5ΩJ = 0.36kg / m 2 C m =6.75N·m / AK e=5.48V / (rad / s)
[0116] first, Figure 3 The structure diagram of the adaptive backstepping sliding mode control system of the speed loop of the shipborne optoelectronic tracking system of the present invention is as follows; the simulation diagram when a step input and a load disturbance of 5N·m are added at 0.5s is as follows:
[0117] Figure 4 This is the simulation diagram of PI speed loop; Figure 5 Input simulation diagram for PI control;
[0118] Figure 6 This is the simulation diagram of the back-stepping sliding mode speed loop; Figure 7 Input simulation diagram for backstepping sliding mode control;
[0119] Figure 8 This is the simulation diagram of the adaptive backstepping sliding mode speed loop; Fig. 9 Input simulation diagram for adaptive backstepping sliding mode control;
[0120] From the above analysis, it can be seen that the adaptive backstepping sliding mode controller designed in this paper has a significant suppression effect on torque disturbance compared with PI control. The above results prove that the control strategy proposed in this paper not only has strong robustness but also has good dynamic performance.
[0121] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.
Claims
1. A design method for a shipborne optoelectronic tracking system with adaptive backstepping sliding mode control. Features: The following steps are involved: Build a two-axis shipborne optoelectronic tracking platform; According to the transfer function of the DC torque motor in the photoelectric tracking servo system, the state equation of the DC torque motor model is obtained; According to the state equation of the DC torque motor model, the basic controller of the DC torque motor is designed by backstepping method. If the disturbance is not considered, the transfer function of the controlled object of the photoelectric tracking servo system is: Convert it to differential form: Where: a 0 = C m K e / JL, a 1 = R / L, b 0 = C m / JL Further considering the disturbance, the differential form of the controlled object model is transformed as follows: x 1 =ω (2) The backstepping method is used to design the backstepping synovial film control law based on the synovial film controller. The controller is designed as follows: z 1 (t)=x 1 (t)-x d (t) (5) z 2 (t)=x 2 (t)-α 1 (t) (6) Where: z 1 、z 2 represents the error variable, K 2 represents the gain factor, x 1 , x 2 They represent velocity and acceleration respectively, u represents the control input, and x d (t), Given the target quantity, given the first-order derivative and the second-order derivative, k 1 is the controller parameter, α 1 (t), are the virtual feedback quantity and the first-order derivative of the virtual feedback quantity, respectively; The stability of the basic controller and the backstepping sliding film controller is verified by Lyapunov function I; According to the RBF neural network, the disturbance d is estimated and the adaptive backstepping sliding mode control law is designed; The Lyapunov function II is used to verify the stability of the adaptive backstepping sliding mode controller; The process of designing the backstepping synovial film control law based on the synovial film controller by using the backstepping method is as follows: Design the integral sliding surface S(t) and the complementary sliding surface S c (t) are as follows: According to the integral sliding surface S(t) and the complementary sliding surface S c (t) and according to the sliding mode equivalent control principle, the sliding mode control law u is obtained by the equivalent control law u eq and the switching control law u hit The sliding mode control law is designed as follows: Where D is the upper bound of the disturbance d, |d|≤D; The S(t) and S c The relationship between (t) is as follows: The expression of the adaptive backstepping sliding mode control law is as follows: Defining perturbations d=W *T S(z)+ε (19) Among them, z is the neural network input, W * is the ideal weight of RBF, ε is the neural network approximation error, S(z)=[s 1 ,s 2 ,···,s j ] is the RBF neural network basis function vector; The weight error is defined as: in, is the error between the ideal weight and the estimated weight, Estimate weights for neural networks; Assume that the disturbance is estimated as: in, is the estimated value of ε; The adaptive backstepping sliding mode control law is obtained as: Among them, x d (t), Given the target quantity, given the first-order derivative and the second-order derivative, k 1 is the controller parameter, k 2 Represents controller parameters, a 0 、a 1 are motor parameters, α 1 is the virtual control quantity in the backstepping design, is the estimated value of RBF neural network weight, S(z) is the neural network basis function, and ε is the neural network estimation error; The RBF neural network basis function S(z) is expressed as follows: S(z)=[s 1 ,s 2 ,...,s l ] T (22) Where z = [z 1 z 2 ] T is the neural network input, W * is the ideal weight of RBF, S(z) is the RBF neural network basis function, c j and σ j are the width and center of the jth Gaussian basis function, S(Z) is the output of the hidden layer of the neural network, and l is the number of nodes in the hidden layer of the neural network; The adaptive backstepping sliding mode control law is designed as: in, is an estimate of ε, is the error between the ideal weight and the estimated weight; In order to verify the stability of the adaptive backstepping sliding mode controller, the Lyapunov function II is selected as: Derivative of the above formula: In order to make the system stable and converge faster, we can get: Among them, λ 1 and λ 2 is the convergence coefficient.
