Train operation control method, medium and device based on double-loop adaptive sliding mode control
By using a dual-loop adaptive sliding mode control method combined with an RBF neural network to simulate the train's kinematics model, the problem of insufficient control accuracy and robustness of rail transit trains under nonlinear systems is solved, enabling efficient and safe operation of trains in complex environments.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA RAILWAY SIYUAN SURVEY & DESIGN GRP CO LTD
- Filing Date
- 2026-04-09
- Publication Date
- 2026-06-09
AI Technical Summary
The operation model of rail transit trains is a nonlinear system. The system parameters are subject to disturbances and are difficult to determine. Traditional PID controllers are not adaptable and stable enough in the face of sudden disturbances and environmental changes, resulting in low control accuracy and efficiency.
A dual-loop adaptive sliding mode control method is adopted, with the outer loop being a displacement control subsystem and the inner loop being a speed control subsystem. The additional resistance in the train kinematic model is simulated using an RBF neural network. The adaptability and robustness of the system are enhanced by integral sliding surfaces and adaptive control laws, and the resultant force of traction and braking is output to control the train operation.
It improves the control precision and robustness of trains under complex operating conditions, ensures the safe and efficient operation of trains under extreme conditions, and enhances the quality of operation services.
Smart Images
Figure CN122166176A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of rail transit train operation control technology, and in particular to a train operation control method, medium, and equipment based on dual-loop adaptive sliding mode control. Background Technology
[0002] With the development of automation and intelligence in rail transit (including high-speed rail, intercity rail, urban rail, and suburban rail), the development of Automatic Train Operation (ATO) systems has become increasingly mature. PID control is commonly used in rail transit train operation control. This method has simple control logic and low resource consumption, but its adaptability, stability, and reliability are insufficient in the face of sudden large disturbances and changes in the operating environment, often requiring manual intervention, thus increasing labor intensity and the uncertainty of human operation. Many domestic scholars have made numerous improvements to the PID control method. The literature "Wang Kang. Research on Intelligent Algorithm for PID Control of Urban Rail Trains [D]. Beijing Jiaotong University, 2019" introduces an improved fruit fly optimization algorithm for online adaptive optimization of PID controller parameters, but the algorithm's optimization direction is uncertain and its accuracy is insufficient. The literature "Shi Li, Wang Jiangtao. Application of Fuzzy Prediction-PID Composite Control in Braking of High-Speed Trains [J]. Computer Engineering and Applications," 2010, 46(31):228-231. "Combining fuzzy prediction with PID control and applying it to the braking and operation control of high-speed trains makes train operation more precise and comfortable. However, fuzzy control has problems such as large overshoot and fast convergence speed when converging to the target value, which easily leads to large steady-state error. The literature "Xiao Long, Liang Xinrong, Wang Xueqi, et al. High-speed train speed tracking based on RBF neural network PID control [J]. Journal of Wuyi University (Natural Science Edition), 2020, 34(02):24-31." uses RBF neural network to integrate PID control and designs a nonlinear feedback controller to control the multi-mass train model with unit displacement. This makes the train control accuracy much better than the traditional method. However, the RBF neural network training convergence time is long and the train tracking control accuracy still needs to be improved. The literature "Wang Jiacheng. Research on high-speed train speed tracking control method based on integral reinforcement learning [D]. Beijing Jiaotong University, The 2020 paper applied LQR control to train speed tracking, but LQR control is only effective for linear systems and has poor control performance on nonlinear systems. The above literature combines traditional PID control algorithms with other intelligent algorithms to improve the performance of traditional PID control. However, the rail transit train operation model is a nonlinear system with fluctuating and difficult-to-determine system parameters, thus requiring control methods with strong robustness and adaptability.
[0003] The operation model of a rail transit train is a nonlinear system, and its parameters are subject to disturbances and are difficult to determine. In actual operation, the basic resistance parameters and the train's mass cannot be measured, and the additional resistance varies with the track, resulting in deviations between the train model and reality, which in turn affects the performance of traditional PID controllers.
[0004] For the operation control of rail transit trains, there is an urgent need to adopt control algorithms with stronger adaptability, robustness and responsiveness to solve the control problems of divergence and low convergence efficiency in train speed tracking control under complex working conditions. Summary of the Invention
[0005] The purpose of this invention is to provide a train operation control method, medium, and equipment based on dual-loop adaptive sliding mode control, aiming to enhance the adaptability and robustness of the operation control system, ensure control accuracy under extreme conditions, and improve the quality of operational services. The specific technical solution is as follows:
[0006] A train operation control method based on dual-loop adaptive sliding mode control includes the following steps: A displacement control subsystem, acting as an outer loop control, controls the train to track a target displacement. The input of the displacement control subsystem is the difference between the desired displacement and the actual displacement, i.e., the displacement error, and the output is the desired speed. A speed control subsystem, acting as an inner loop control, controls the train to follow a speed curve generated by the displacement control subsystem, which is a sliding mode controller. The input of the speed control subsystem is the difference between the desired speed and the actual train speed, i.e., the speed error, and the output is the resultant force of traction and braking. The control law for the resultant force of traction and braking is determined based on the integral sliding surface of the speed control subsystem and the train kinematic model. The additional resistance in the train kinematic model is simulated using an RBF neural network algorithm. The basic resistance parameters, total train mass, ideal weight matrix of the RBF neural network, and approximation error in the train kinematic model are controlled using an adaptive control law based on the integral sliding surface of the speed control subsystem.
[0007] Furthermore, the displacement control subsystem is a sliding mode controller.
[0008] Furthermore, if the output of the displacement control subsystem touches the emergency braking trigger curve, it will be forcibly switched to emergency braking mode; otherwise, the desired speed will be output.
[0009] Furthermore, the kinematic model expression for the train is as follows:
[0010]
[0011] in, This refers to the displacement of the train during its movement; M represents the speed of the train; M represents the total mass of the train during operation. The resultant force of traction and braking along the direction of train travel; This is the first fundamental resistance parameter. This is the second fundamental resistance parameter. These are the third fundamental resistance parameters, all of which are time-varying constants; This adds resistance.
[0012] Furthermore, the kinematic model of the displacement control subsystem is as follows:
[0013]
[0014] in, This refers to the actual speed of the train.
[0015] The displacement error is:
[0016]
[0017] in, For displacement error; The desired displacement;
[0018] The integral sliding surface in the sliding mode controller is designed as follows:
[0019]
[0020] in, The integral sliding surface of the displacement control subsystem; A coefficient that is greater than zero; For displacement error in time The integral over the range [0, t];
[0021] Integral sliding surface The first derivative is:
[0022]
[0023] at this time:
[0024]
[0025] in, , A coefficient that is greater than zero;
[0026] The train displacement control law is as follows:
[0027]
[0028] in, This is the train displacement control law.
[0029] Furthermore, the Lyapunov function of the displacement control subsystem .
[0030] Furthermore, the desired speed is set as The actual speed of the train is Then the speed error is:
[0031]
[0032] in, For speed error; The desired train speed;
[0033] The integral sliding surface of the speed control subsystem is taken as:
[0034]
[0035] in, The integral sliding surface of the speed control subsystem; A coefficient that is greater than zero; For displacement error in time The integral over the range [0, t];
[0036] Then we have:
[0037]
[0038] Also:
[0039]
[0040] in, for The error; for The error; for The error; for The error; for The estimated value; for The estimated value; for The estimated value; for The estimated value;
[0041] Using the RBF neural network algorithm to simulate additional resistance The formula is:
[0042]
[0043] in, For the first The output of each basis function; This serves as the input to the neural network; For the first The center of each basis function; For the first The width parameter of each basis function; This is the output of the Gaussian function; The ideal weight matrix for the network; For network approximation error;
[0044] Take the input of the RBF neural network as The network output will then be: , For real unmodeled dynamics The real-time approximation results; To approximate the error; To estimate the weight matrix; This is the weight error matrix;
[0045] For the speed control subsystem, a Lyapunov function is designed as follows:
[0046]
[0047] in, For the Lyapunov functions of the speed control subsystem; ; ; This is the estimation error of the network approximation error. , This is an estimate of the network approximation error; , , , , , All parameters are configurable parameters of the controller and are greater than 0;
[0048] Therefore, the train speed control law can be obtained as follows:
[0049]
[0050] in, For train speed control laws; This is a configurable coefficient.
[0051] Furthermore, the adaptive control law for each parameter is as follows:
[0052] .
[0053] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the train operation control method of dual-loop adaptive sliding mode control as described above.
[0054] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the train operation control method of dual-loop adaptive sliding mode control as described above.
[0055] The present invention provides a train operation control method, medium, and equipment for dual-loop adaptive sliding mode control, which has the following beneficial effects:
[0056] This invention controls the train to track the target displacement through a displacement control subsystem as the outer loop control. The input of the displacement control subsystem is the difference between the desired displacement and the actual displacement, i.e., the displacement error, and the output is the desired speed. An inner loop speed control subsystem controls the train to follow the speed curve generated by the displacement control subsystem. The speed control subsystem is a sliding mode controller. Its input is the difference between the desired speed and the actual train speed, i.e., the speed error, and its output is the resultant force of traction and braking. The control law for the resultant force of traction and braking is determined based on the integral sliding surface of the speed control subsystem and the train kinematic model. The additional resistance in the train kinematic model is simulated using an RBF neural network algorithm. The basic resistance parameters, total train mass, ideal weight matrix of the RBF neural network, and approximation error in the train kinematic model employ an adaptive control law based on the integral sliding surface of the speed control subsystem. This invention solves the control problems of divergent speed tracking control deviations and low convergence efficiency under complex operating conditions, effectively enhancing the adaptability and robustness of the operation control system, ensuring control accuracy under extreme conditions, and improving the quality of operational services. Attached Figure Description
[0057] Figure 1 A diagram of a dual-loop adaptive sliding mode control architecture provided in an embodiment of the present invention;
[0058] Figure 2 This is a schematic diagram of train resistance force provided in an embodiment of the present invention;
[0059] Figure 3 The expected speed curve provided for the verification embodiments of the present invention;
[0060] Figure 4 The expected displacement curve is provided for the verification embodiment of the present invention;
[0061] Figure 5 The displacement tracking error curve is provided for the verification embodiment of the present invention;
[0062] Figure 6 The speed tracking error curve is provided for the verification embodiment of the present invention;
[0063] Figure 7 This is a structural block diagram of a computer device according to an embodiment of the present invention. Detailed Implementation
[0064] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. The advantages and features of the present invention will become clearer from the following description. It should be noted that the drawings are all in a very simplified form and use non-precise proportions, and are only used to facilitate and clearly illustrate the purpose of the embodiments of the present invention.
[0065] Example 1
[0066] This embodiment provides a train operation control method based on dual-loop adaptive sliding mode control. (See attached document.) Figure 1 As shown, the method includes the following steps: The train is controlled to track the target displacement by a displacement control subsystem acting as an outer loop control. The input of the displacement control subsystem is the difference between the desired displacement and the actual displacement, i.e., the displacement error, and the output of the displacement control subsystem is the desired speed. The train is controlled to follow the speed curve generated by the displacement control subsystem by a speed control subsystem acting as an inner loop control. The speed control subsystem is a sliding mode controller. The input of the speed control subsystem is the difference between the desired speed and the actual train speed, i.e., the speed error, and the output of the speed control subsystem is the resultant force of traction and braking. The control law for the resultant force of traction and braking is determined based on the integral sliding surface of the speed control subsystem and the train kinematic model. The additional resistance in the train kinematic model is simulated using an RBF neural network algorithm. The basic resistance parameters, total train mass, ideal weight matrix of the RBF neural network, and approximation error in the train kinematic model are controlled using an adaptive control law based on the integral sliding surface of the speed control subsystem.
[0067] For details, please refer to Figure 2 As shown, the kinematic model expression for the train is:
[0068]
[0069] in, This refers to the displacement of the train during its movement; M represents the speed of the train; M represents the total mass of the train during operation. The resultant force of traction and braking along the direction of train travel can be regarded as the control force output by the controller; This is the first fundamental resistance parameter. This is the second fundamental resistance parameter. These are the third fundamental resistance parameters, all of which are time-varying constants; This adds resistance.
[0070] Because trains operate in different environments on different routes, and different trains have different drag coefficients, and because the aerodynamic drag they experience is also related to the train's shape, wind speed, and wind direction, therefore... , , These three parameters are subject to some uncertainty.
[0071] Additional resistance This resistance is mainly generated when the train passes through curves, tunnels, and slopes, and these resistances vary with changes in track conditions; additional resistance. That is, the additional resistance is due to the curve. Additional resistance of tunnels Additional resistance of ramps Composition, its expression is:
[0072]
[0073] The formula for additional resistance on a curve is:
[0074]
[0075] in, The length of the curve. As the central angle, For train quality, This is the acceleration due to gravity.
[0076] The formula for additional resistance in a tunnel is:
[0077]
[0078] in, This represents the tunnel length.
[0079] The formula for additional resistance of a ramp is:
[0080]
[0081] in, The slope angle is denoted by .
[0082] The displacement control subsystem, belonging to the outer position loop, is responsible for macroscopic position planning and safety protection. By generating a reasonable desired speed curve, it indirectly constrains the train's operating range and speed limit, making it a crucial link in ensuring train operation safety. Its control objective is: to achieve the desired displacement... With actual displacement The difference The input is the target displacement, and the core objective is to track the target displacement to ensure the train accurately reaches the designated position. Its output characteristic is: the output is the desired speed. An emergency braking trigger curve is introduced as a safety constraint: if the output (desired speed) of the displacement control subsystem touches the emergency braking trigger curve, the system is forced to switch to emergency braking mode (output 2); otherwise, the system outputs the normal desired speed command (output 1).
[0083] Specifically, the kinematic model of the displacement control subsystem is as follows:
[0084]
[0085] in, This represents the train's actual speed.
[0086] The displacement error is:
[0087]
[0088] in, For displacement error; The desired displacement.
[0089] The integral sliding surface in the sliding mode controller is designed as follows:
[0090]
[0091] in, The integral sliding surface of the displacement control subsystem; A coefficient that is greater than zero; For displacement error in time The integral over the range [0, t].
[0092] Integral sliding surface The first derivative is:
[0093]
[0094] at this time:
[0095]
[0096] in, , The coefficient is greater than zero.
[0097] The train displacement control law is as follows:
[0098]
[0099] in, This is the train displacement control law.
[0100] When the displacement control subsystem Lyapunov function ,but Train displacement control law Convergence ensures the stability of the displacement control subsystem.
[0101] Specifically, the desired speed is set as The actual speed of the train is Then the speed error is:
[0102]
[0103] in, For speed error; The desired train speed.
[0104] The integral sliding surface of the speed control subsystem is taken as:
[0105]
[0106] in, The integral sliding surface of the speed control subsystem; A coefficient that is greater than zero; For displacement error in time The integral over the range [0, t].
[0107] Then we have:
[0108]
[0109] in, , , and All are time-varying constants, and , , , All are 0.
[0110] Also:
[0111]
[0112] in, for The error; for The error; for The error; for The error; for The estimated value; for The estimated value; for The estimated value; for The estimated value.
[0113] Using the RBF neural network algorithm to simulate additional resistance The formula is:
[0114]
[0115] in, For the first The output of each basis function; This serves as the input to the neural network; For the first The center of each basis function; For the first The width parameter of each basis function; This is the output of the Gaussian function; The ideal weight matrix for the network; This represents the network approximation error.
[0116] The ideal weight matrix is used by the RBF neural network to approximate the unmodeled real dynamics. At that time, the theoretically optimal weights, used to accurately fit the unmodeled portion, are... ,in It is the inherent approximation error of the RBF neural network, which is a bounded quantity.
[0117] The estimated weight matrix is calculated in real time and used for approximation in actual control. .
[0118] The weight error matrix represents the deviation between the ideal weights and the estimated weights. It is the object that the adaptive law needs to adjust and is used for stability analysis and parameter updates.
[0119] In this control system, the input of the RBF neural network is taken as... The network output will then be: , For real unmodeled dynamics The real-time approximation results are used for compensation in the control law.
[0120] The approximation error represents the deviation between the true unmodeled dynamics and the neural network estimate.
[0121] For the speed control subsystem, a Lyapunov function is designed as follows:
[0122]
[0123] in, For the Lyapunov functions of the speed control subsystem; ; ; This is the estimation error of the network approximation error. , The estimated value for the network approximation error is used to compensate for... The designed online estimation term, updated in real time via an adaptive algorithm, is used to offset the control law. The impact; , , , , , All of these are configurable parameters of the controller and are greater than 0.
[0124] Therefore, the train speed control law can be obtained as follows:
[0125]
[0126] in, For train speed control laws; This is a configurable coefficient.
[0127] The adaptive control law for each parameter is designed as follows:
[0128] .
[0129] At this point, the inner loop speed control subsystem is designed to be asymptotically stable, and the actual speed converges to the desired speed.
[0130] This invention controls the train to track the target displacement through a displacement control subsystem as the outer loop control. The input of the displacement control subsystem is the difference between the desired displacement and the actual displacement, i.e., the displacement error, and the output is the desired speed. An inner loop speed control subsystem controls the train to follow the speed curve generated by the displacement control subsystem. The speed control subsystem is a sliding mode controller. Its input is the difference between the desired speed and the actual train speed, i.e., the speed error, and its output is the resultant force of traction and braking. The control law for the resultant force of traction and braking is determined based on the integral sliding surface of the speed control subsystem and the train kinematic model. The additional resistance in the train kinematic model is simulated using an RBF neural network algorithm. The basic resistance parameters, total train mass, ideal weight matrix of the RBF neural network, and approximation error in the train kinematic model employ an adaptive control law based on the integral sliding surface of the speed control subsystem. This invention solves the control problems of divergent speed tracking control deviations and low convergence efficiency under complex operating conditions, effectively enhancing the adaptability and robustness of the operation control system, ensuring control accuracy under extreme conditions, and improving the quality of operational services.
[0131] Verification of Examples
[0132] Using the control algorithm designed in this invention, the entire process is divided into two acceleration phases, two deceleration phases, and four cruise phases. The maximum speed can reach 90 m / s (324 km / h), the simulation lasts for 1000 seconds, and the total operating distance is 62.67 km.
[0133] Depend on Figures 3-6 As can be seen, the controller designed in this invention operates normally. Whether in the acceleration, cruising, or deceleration phases, the speed and displacement errors of the overall operating system are very small, enabling the train to accurately track the desired speed and displacement curves, thus ensuring the safety and efficiency of train operation.
[0134] Example 2
[0135] This embodiment provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the train operation control method of dual-loop adaptive sliding mode control described above.
[0136] The storage medium may be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc.; the storage medium may also include combinations of the above types of memory.
[0137] Example 3
[0138] This embodiment provides a computer device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps of the train operation control method of dual-loop adaptive sliding mode control described above.
[0139] like Figure 7 As shown, the computer device 70 may include: at least one processor 71, such as a CPU (Central Processing Unit), at least one communication interface 73, a memory 74, and at least one communication bus 72. The communication bus 72 is used to enable communication between these components. The communication interface 73 may include a display screen and a keyboard; optionally, the communication interface 73 may also include a standard wired interface or a wireless interface. The memory 74 may be high-speed RAM (Random Access Memory) or non-volatile memory, such as at least one disk storage device. Optionally, the memory 74 may also be at least one storage device located remotely from the aforementioned processor 71. The memory 74 stores application programs, and the processor 71 calls the program code stored in the memory 74 to execute any of the above-described method steps.
[0140] The communication bus 72 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The communication bus 72 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 7 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.
[0141] The memory 74 may include volatile memory, such as random-access memory (RAM); the memory may also include non-volatile memory, such as flash memory, hard disk drive (HDD) or solid-state drive (SSD); the memory 74 may also include a combination of the above types of memory.
[0142] The processor 71 can be a central processing unit (CPU), a network processor (NP), or a combination of CPU and NP.
[0143] The processor 71 may further include a hardware chip. This hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.
[0144] Optionally, the memory 74 is also used to store program instructions. The processor 71 can call the program instructions to implement the train operation control method of the dual-loop adaptive sliding mode control of the present invention.
[0145] Those skilled in the art should understand that the present invention can be implemented in many other specific forms without departing from the spirit and scope of the invention. Any changes or modifications made by those skilled in the art based on the embodiments of the present invention and the above disclosure shall fall within the protection scope of the claims.
Claims
1. A train operation control method with dual-loop adaptive sliding mode control, characterized in that, The method includes the following steps: A displacement control subsystem, acting as an outer loop control, controls the train to track the target displacement. The input to the displacement control subsystem is the difference between the desired displacement and the actual displacement, i.e., the displacement error. The output of the displacement control subsystem is the desired speed. A speed control subsystem, acting as an inner loop control, controls the train to follow the speed curve generated by the desired speed of the displacement control subsystem. The speed control subsystem is a sliding mode controller. The input to the speed control subsystem is the difference between the desired speed and the actual train speed, i.e., the speed error. The output of the speed control subsystem is the resultant force of traction and braking. The control law for the resultant force of traction and braking is determined based on the integral sliding surface of the speed control subsystem and the train kinematic model. The additional resistance in the train kinematic model is simulated using an RBF neural network algorithm. The basic resistance parameters, total train mass, ideal weight matrix of the RBF neural network, and approximation error in the train kinematic model are controlled using an adaptive control law based on the integral sliding surface of the speed control subsystem.
2. The train operation control method of dual-loop adaptive sliding mode control according to claim 1, characterized in that, The displacement control subsystem is a sliding mode controller.
3. The train operation control method of dual-loop adaptive sliding mode control according to claim 1, characterized in that, If the output of the displacement control subsystem touches the emergency braking trigger curve, it will be forcibly switched to emergency braking mode; otherwise, the desired speed will be output.
4. The train operation control method of dual-loop adaptive sliding mode control according to claim 1, characterized in that, The kinematic model expression for the train is: ； in, This refers to the displacement of the train during its movement; M represents the speed of the train; M represents the total mass of the train during operation. The resultant force of traction and braking along the direction of train travel; This is the first fundamental resistance parameter. This is the second fundamental resistance parameter. These are the third fundamental resistance parameters, all of which are time-varying constants; This adds resistance.
5. The train operation control method of dual-loop adaptive sliding mode control according to claim 4, characterized in that, The kinematic model of the displacement control subsystem is as follows: ； in, This refers to the actual speed of the train. The displacement error is: ； in, For displacement error; The desired displacement; The integral sliding surface in the sliding mode controller is designed as follows: ； in, The integral sliding surface of the displacement control subsystem; A coefficient that is greater than zero; For displacement error in time The integral over the range [0, t]; Integral sliding surface The first derivative is: ； at this time: ； in, , A coefficient that is greater than zero; The train displacement control law is as follows: ； in, This is the train displacement control law.
6. The train operation control method of dual-loop adaptive sliding mode control according to claim 5, characterized in that, Lyapunov function of displacement control subsystem .
7. The train operation control method of dual-loop adaptive sliding mode control according to claim 5, characterized in that, Set the desired speed as The actual speed of the train is Then the speed error is: ； in, For speed error; The desired train speed; The integral sliding surface of the speed control subsystem is taken as: ； in, The integral sliding surface of the speed control subsystem; A coefficient that is greater than zero; For displacement error in time The integral over the range [0, t]; Then we have: ； Also: ； in, for The error; for The error; for The error; for The error; for The estimated value; for The estimated value; for The estimated value; for The estimated value; Using the RBF neural network algorithm to simulate additional resistance The formula is: ； in, For the first The output of each basis function; This serves as the input to the neural network; For the first The center of each basis function; For the first The width parameter of each basis function; This is the output of the Gaussian function; The ideal weight matrix for the network; For network approximation error; Take the input of the RBF neural network as The network output will then be: , For real unmodeled dynamics The real-time approximation results; To approximate the error; To estimate the weight matrix; This is the weight error matrix; For the speed control subsystem, a Lyapunov function is designed as follows: ； in, For the Lyapunov functions of the speed control subsystem; ; ; This is the estimation error of the network approximation error. , This is an estimate of the network approximation error; , , , , , All parameters are configurable parameters of the controller and are greater than 0; Therefore, the train speed control law can be obtained as follows: ； in, For train speed control laws; This is a configurable coefficient.
8. The train operation control method of dual-loop adaptive sliding mode control according to claim 7, characterized in that, The adaptive control law for each parameter is as follows: 。 9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the train operation control method of dual-loop adaptive sliding mode control as described in any one of claims 1-8.
10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the train operation control method of dual-loop adaptive sliding mode control as described in any one of claims 1-8.