A finite-time H-infinity control method under the hopping FlexRay protocol
By designing a finite time H infinite control method under the jump FlexRay protocol, combining GA genetic algorithm and LMI linear matrix inequality, the problem of slow response speed and insufficient robustness of traditional control methods in the face of network attacks is solved, and the system is realized The rapid response and attack resistance reduce the conservativeness of the controller design.
Patent Information
- Application Number
- CN202411102958.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-13
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-08-13
AI Technical Summary
Traditional control methods are slow to respond, insufficient robustness, heavy communication burden when facing network attacks, and nonlinear constraints exist in the controller design process, making it difficult to deal with external network attacks and insufficient bandwidth.
The finite time H infinite control method under the jump FlexRay protocol is adopted. By obtaining the parameters of the single-link robotic arm system, it is divided into static segments and dynamic segments to transmit the sensor output, and a zero-order holder is added to the controller input end. Combining the GA genetic algorithm and the LMI linear matrix inequality, a dynamic output feedback finite time controller is designed to optimize the controller gain matrix to improve system performance.
It improves the system's performance against external attacks, avoids data conflicts, saves communication resources, reduces the conservativeness of controller design, and improves the system's robustness and response speed.
Smart Images

Figure CN118625680B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of networked system control, and in particular to a finite time H-infinity control method under a hopping FlexRay protocol. Background Art
[0002] Unlike traditional point-to-point control systems based on physical connections, networked control systems use shared wireless networks as a bridge for data transmission between sensors and controllers, and between controllers and actuators. Introducing wireless networks into control systems helps the development of remote control and resource sharing, making the system more modular, more convenient, more flexible and less expensive in terms of installation and maintenance. However, the introduction of shared communication networks exposes the system's information transmission to the view of external attackers, and traditional control methods cannot cope well with network attacks against the system, such as denial of service attacks, which may cause system instability or failure. At the same time, the communication resources in the networked system are limited. If a large amount of data is transmitted at the same time, problems such as insufficient bandwidth and data conflicts may occur. Therefore, it is necessary to introduce communication protocols in the communication network for information scheduling.
[0003] In the tradition The controller design conditions of control problems often contain nonlinear terms. If the linear matrix inequality (LMI) method is used to solve the controller parameters, the coupled unknown variables in the nonlinear terms need to be decoupled. However, decoupling usually increases the conservatism of the controller design conditions, and the controller designed is not optimal. Therefore, combining evolutionary algorithms with LMI methods for controller design can reduce conservatism and improve system performance. Summary of the invention
[0004] In view of the shortcomings of the prior art, the present invention provides a finite time H infinite control method under the jump FlexRay protocol, which solves the problem of the traditional The control method has slow response speed, insufficient robustness, heavy communication burden, and difficulty in dealing with external network attacks, and There are technical issues of nonlinear constraints in the controller design process.
[0005] To solve the above technical problems, the present invention provides the following technical solutions: a finite time H infinity control method under a hopping FlexRay protocol, the method comprising the following steps:
[0006] S1. Obtain the parameters of the controlled single-link robotic arm system, and reset the discrete time system parameter matrix of the single-link robotic arm system and the upper bound of the external disturbance based on the parameters;
[0007] S2, dividing the sensors that collect system output information into two groups, and setting the measurement outputs of the two groups of sensors to jump to the static segment and the dynamic segment of the FlexRay protocol for transmission respectively;
[0008] S3, according to the jump FlexRay communication protocol, select the measurement output that needs to be transmitted at each transmission time, and add a zero-order holder ZOH at the input end of the controller to obtain the measurement signal available to the controller ;
[0009] S4. Constructing dynamic output feedback finite time The controller models the system and designs a dynamic output feedback finite time Matrix inequality set for the controller;
[0010] S5, using GA genetic algorithm and LMI linear matrix inequality solution algorithm to obtain dynamic output feedback finite time The gain matrix of the controller ;
[0011] S6. The gain matrix Substituting into the dynamic output feedback finite time Used in the controller to stabilize the single-link robotic arm system.
[0012] Furthermore, in step S1, the specific process includes the following steps:
[0013] S11. Obtain the parameters of the controlled single-link robotic arm system , determine the mathematical model of the single-link robotic arm system, namely:
[0014] ;
[0015] In the above formula, is the angular position of the robot arm; is the control input; is the external disturbance, is the load mass; is the moment of inertia; is the acceleration due to gravity; is the arm length; is the uncertainty coefficient of viscous friction; is the angular velocity of the robot arm;
[0016] S12. Order , , and linearize the mathematical model of the single-link robotic arm system to obtain the continuous-time system state space model, namely:
[0017] ;
[0018] in,
[0019] ;
[0020] ;
[0021] ;
[0022] S13, set the sampling time to , and use the c2d function in MATLAB, select the zero-order hold method as the discretization method, and discretize the continuous-time system state space model to obtain the discrete-time system matrix ;
[0023] S14. Define the external disturbance satisfying condition of the single-link manipulator system, and obtain the upper bound of the external disturbance according to the external disturbance satisfying condition. The external disturbance satisfying condition is:
[0024] ;
[0025] in, is a known positive scalar, i.e., the upper bound of the external disturbance; For external disturbance The transpose of is a finite time parameter.
[0026] Furthermore, in step S2, it specifically includes:
[0027] According to the two groups of sensors, the measurement output of the single-link robot system Expressed as ,Right now:
[0028] The measured output Before Dimension, coming soon The measurement output of the sensor Set the static segment transmission in the FlexRay protocol; output the measurement After The first dimension, The measurement output of the sensor Set up dynamic segment transmission in FlexRay protocol;
[0029] Measurement output as well as Respectively expressed as:
[0030] ;
[0031] ;
[0032] in, , The number of sensors that collect system output; Represents the transmission of the single-link robotic arm system in the static segment The measurement output of each sensor; Represents the transmission of the single-link robotic arm system in the dynamic segment The measurement output of a sensor.
[0033] Furthermore, in step S3, the specific process includes the following steps:
[0034] S31. The sensors are divided into a set of sensors that are attacked by denial of service through the scheduler. and the set of sensors that have not been attacked ;
[0035] S32, setting a scheduling rule for jumping the FlexRay communication protocol;
[0036] S33, selecting sensor information to be transmitted at each transmission time according to the scheduling rule of the jump FlexRay communication protocol;
[0037] like , then calculate the jump step length at the current moment If , then calculate the jump weight matrix at the current moment The value of The value and jump weight matrix The value of determines the scheduling signal at the current moment;
[0038] in, express When the time is the static segment transmission, express The time is when the dynamic segment is transmitted;
[0039] S34. Add a zero-order holder ZOH at the input of the controller to obtain a measurement signal that can be used by the controller. .
[0040] Further, in step S32, the scheduling rule for jumping the FlexRay communication protocol includes a static segment scheduling rule for jumping the FlexRay protocol and a dynamic segment scheduling rule for jumping the FlexRay protocol;
[0041] Among them, the static segment scheduling rules of the FlexRay protocol are:
[0042] ;
[0043] The dynamic segment scheduling rule of the designed jump FlexRay protocol is:
[0044] ;
[0045] ;
[0046] ;
[0047] in, It is the scheduling signal of the static segment; It is the scheduling signal of the dynamic segment; Represents the discrete-time system sampling moments; is the set of static segment transmission times; is the set of dynamic segment transmission times; ; is the Kroneckerdelta function, To express or ,when hour , in other cases , is the dimensionally appropriate identity matrix; For the current The jump weight matrix at time; For the The weight value of each sensor node.
[0048] Further, in step S34, the controller can use the measurement signal It is expressed as:
[0049] ;
[0050] in,
[0051] ;
[0052] ;
[0053] ;
[0054] in, , The number of sensors that collect system output; It is the scheduling signal of the dynamic segment; The dimension is The identity matrix of is the number of sensors transmitting in the static segment; As a switching signal, when When the time is the static segment transmission, that is hour, ,when The moment is when the dynamic segment is transmitted, that is, hour, , , is the Kroneckerdelta function, To express or ,when hour 1, otherwise is 0; , Indicates when hour 1, otherwise is 0; is the dimensionally appropriate identity matrix.
[0055] Further, in step S4, the dynamic output feedback finite time constructed The system model of the controller is:
[0056] ;
[0057] in, is the controller status; , , , is the gain matrix of the controller;
[0058] Dynamic output feedback limited time The matrix inequality system of the controller is expressed as:
[0059] ;
[0060] in,
[0061] ;
[0062] in, is the unknown positive definite matrix in the Lyapunov function; is an unknown positive definite matrix; is a given finite time parameter; ; ; ; ; ; ; ; ; is the dimensionally appropriate identity matrix; ; ; is the dynamic segment scheduling signal, is the Kronecker delta function, To express or ,when hour 1, otherwise is 0; is an unknown positive definite matrix; is an unknown positive definite matrix; for The square of the performance index; is the square of the upper bound of the external disturbance.
[0063] Furthermore, in step S5, the specific process includes the following steps:
[0064] S51, dynamic output feedback limited time The matrix inequality of the controller with nonlinear terms is scaled by applying Schur's complement lemma to obtain the corresponding linear matrix inequality, as follows:
[0065] ;
[0066] Among them, the symbol " ” represents the omission of symmetric elements in a symmetric matrix;
[0067] S52, solving the linear matrix inequality obtained in step S51 and other linear matrix inequalities to obtain a set of controller gain matrices feasible solution;
[0068] S53, determine the initialization range of each individual when the GA genetic algorithm population is initialized ,in, , is the proportionality coefficient given by experience, , is the row vector of controller gain matrix elements;
[0069] S54, randomly generating an initial population from the initialization range obtained in step S53;
[0070] S55, using the LMI tool to solve the linear matrix inequality group for each individual, if there is a solution, calculate the corresponding fitness value; if there is no solution, assign a sufficiently large value to the fitness value of the individual;
[0071] Among them, the fitness value of GA genetic algorithm is expressed as , for Performance indicators;
[0072] S56, executing the selection, crossover, mutation, and elite retention GA operations in the GA genetic algorithm;
[0073] If the maximum number of iterations is reached, a set of optimal individual gain matrices obtained during the iteration process is saved , otherwise return to step S55.
[0074] By means of the above technical solution, the present invention provides a finite time H-infinity control method under the FlexRay protocol jump, which has at least the following beneficial effects:
[0075] 1. The present invention proposes a new communication protocol scheduling rule, which can avoid selecting sensor nodes that are subject to denial of service attacks when selecting sensors, thereby improving the performance of the system when facing external attacks.
[0076] 2. The dynamic output feedback in the present invention has a limited time The network communication problem between the sensor that collects the system status and the controller is considered in the controller design process, and it is proved that the system is bounded in finite time and satisfies the FlexRay protocol scheduling when the communication network is in a hopping state. Performance. Information communication in a communication network scheduled by a communication protocol can avoid data conflicts and save communication resources.
[0077] 3. Dynamic output feedback limited time in the present invention The controller design method combines the traditional LMI method with the GA genetic algorithm, which can reduce the conservatism in the controller design and effectively improve the system performance. The introduction of the GA genetic algorithm can well handle the nonlinear terms in the sufficient conditions for solving the gain matrix. Control can reduce performance indicators, thereby improving system performance. BRIEF DESCRIPTION OF THE DRAWINGS
[0078] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0079] Figure 1 It is a framework diagram of the single-link mechanical arm control system of the present invention;
[0080] Figure 2 It is a fitness value iteration curve diagram of the genetic algorithm of the present invention;
[0081] Figure 3 It is a state trajectory curve diagram of the single-link mechanical arm control system of the present invention;
[0082] Figure 4 It is a schematic diagram of the communication protocol scheduling of the single-link robotic arm control system of the present invention. DETAILED DESCRIPTION
[0083] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific implementation methods, so that the implementation process of how the present application uses technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly.
[0084] Please refer to Figure 1-Figure 4 This embodiment proposes a finite-time H-infinity control method under the jump FlexRay protocol, and uses the genetic algorithm Genetic Algorithm-GA combined with the linear matrix inequality LMI solution algorithm to design the controller to achieve fast response, strong robustness, and anti-attack control effects. At the same time, the GA genetic algorithm is used to optimize Control performance indicators. The method comprises the following steps:
[0085] S1, obtaining the parameters of the controlled single-link robotic arm system, and resetting the discrete time system parameter matrix of the single-link robotic arm system and the upper bound of the external disturbance based on the parameters; in step S1, the specific process includes the following steps:
[0086] S11. Obtain the parameters of the controlled single-link robotic arm system , determine the mathematical model of the single-link robotic arm system, namely:
[0087] ;
[0088] In the above formula, is the angular position of the robot arm; is the control input; is the external disturbance, is the load mass; is the moment of inertia; is the acceleration due to gravity; is the arm length; is the uncertainty coefficient of viscous friction; is the angular velocity of the robot arm;
[0089] S12. Order , , and linearize the mathematical model of the single-link robotic arm system to obtain the continuous-time system state space model, namely:
[0090] ;
[0091] in,
[0092] ;
[0093] ;
[0094] ;
[0095] S13, set sampling time seconds, and use the c2d function in MATLAB, select the zero-order hold method as the discretization method, and discretize the continuous-time system state space model to obtain the discrete-time system matrix ;
[0096] S14. Define the external disturbance satisfying condition of the single-link manipulator system, and obtain the upper bound of the external disturbance according to the external disturbance satisfying condition. The external disturbance satisfying condition is:
[0097] ;
[0098] in, is a known positive scalar, i.e., the upper bound of the external disturbance; For external disturbance The transpose of is a finite time parameter.
[0099] Specifically, this embodiment estimates the external disturbance of the single-link manipulator system to obtain the upper bound of the external disturbance of the system: , and by resetting the upper bound of the external disturbance and the parameters, more accurate information of the single-link robotic arm system can be obtained, which effectively increases the robustness of the single-link robotic arm system and improves the performance of the single-link robotic arm system.
[0100] S2, dividing the sensors for collecting system output information into two groups, and setting the measurement outputs of the two groups of sensors to jump to the static segment and the dynamic segment of the FlexRay protocol for transmission respectively; in step S2, specifically including:
[0101] According to the two groups of sensors, the measurement output of the single-link robot system Expressed as ,Right now:
[0102] The measured output Before Dimension, coming soon The measurement output of the sensor Set the static segment transmission in the FlexRay protocol; output the measurement After The first dimension, The measurement output of the sensor Sets the dynamic segment transmission in the FlexRay protocol.
[0103] The static segment of the FlexRay protocol is scheduled using Round Robin (RR), and the dynamic segment of the FlexRay protocol is scheduled using Try-Once-Discard (TOD).
[0104] Specifically, the measured output as well as Respectively expressed as:
[0105] ;
[0106] ;
[0107] in , The number of sensors that collect system output; Represents the transmission of the single-link robotic arm system in the static segment The measurement output of each sensor; Represents the transmission of the single-link robotic arm system in the dynamic segment The measurement output of a sensor.
[0108] The jump FlexRay protocol designed in this embodiment is a hybrid protocol that has both time triggering and event triggering features. It has the advantages of both and is more flexible. It also collects the measurement output of the single-link robotic arm system through sensors. , and then input into the communication network, and quickly obtain the measurement signal available to the controller under the scheduling of the jump FlexRay protocol .
[0109] S3, according to the jump FlexRay communication protocol, select the measurement output that needs to be transmitted at each transmission time, and add a zero-order holder ZOH at the input end of the controller to obtain the measurement signal available to the controller ; In step S3, the specific process includes the following steps:
[0110] S31. The sensors are divided into a set of sensors that are attacked by denial of service through the scheduler. and the set of sensors that have not been attacked ;
[0111] S32, setting the scheduling rules for jumping the FlexRay communication protocol; the scheduling rules for jumping the FlexRay communication protocol include the static segment scheduling rules for jumping the FlexRay protocol and the dynamic segment scheduling rules for jumping the FlexRay protocol; wherein the static segment scheduling rules for jumping the FlexRay protocol are:
[0112] ;
[0113] The dynamic segment scheduling rule of the designed jump FlexRay protocol is:
[0114] ;
[0115] in, , is the scheduling signal of the static segment; It is the scheduling signal of the dynamic segment; Represents the discrete-time system sampling moments; is the set of static segment transmission times; is the set of dynamic segment transmission times; ; is the Kroneckerdelta function, To express or ,when hour , in other cases , is the dimensionally appropriate identity matrix; For the current The jump weight matrix at time; For the The weight value of each sensor node.
[0116] S33, select the sensor information to be transmitted at each transmission time according to the scheduling rules of the FlexRay communication protocol. , then calculate the jump step length at the current moment If , then calculate the jump weight matrix at the current moment The value of The value and jump weight matrix The value of determines the scheduling signal at the current moment. express When the time is the static segment transmission, express The time is when the dynamic segment is transmitted.
[0117] Among them, the jump step length The initial value of is set to 0. The system makes a judgment each time it performs communication scheduling in the static segment. , then the current scheduling signal is ,like ,but And according to the new jump step length Recalculate Repeat the above steps until the calculated The meaning of the calculation here is to skip the attacked sensor node until the scheduling signal selects a safe sensor node.
[0118] The system calculates the jump weight matrix each time it performs communication scheduling in the dynamic segment The value of , where the jump weight matrix To satisfy The smallest matrix, , footer , , Respectively The weight value of the sensor node, is the dimensionally appropriate identity matrix.
[0119] Dynamic segment scheduling rules middle, , where the jump weight matrix The calculation of the scheduling signal Therefore, each time communication scheduling is performed in the dynamic segment, the jump weight matrix needs to be calculated first. The value of the current scheduling signal .
[0120] The jump step length calculated here is and the jump weight matrix The value of is used to select the scheduling signal at the current moment in the subsequent steps. It should be noted that the traditional FlexRay protocol does not need to calculate the jump step length. and the jump weight matrix . This embodiment adds a jump step and a jump weight matrix to the selection of the scheduling signal, so that when the communication scheduling is performed according to the calculated scheduling signal, the system will not choose to transmit the information collected by the attacked sensor. Therefore, when selecting the sensor, it is possible to avoid selecting the sensor node that is attacked by the denial of service, thereby improving the performance of the system when facing external attacks.
[0121] S34. Add a zero-order holder ZOH at the input of the controller to obtain a measurement signal that can be used by the controller. ; Among them, the function of the zero-order holder ZOH is to save the measurement signal output by the single-link robotic arm system received by the controller at the last transmission moment.
[0122] Measuring signals available to the controller It is expressed as:
[0123] ;
[0124] in,
[0125] ;
[0126] ;
[0127] ;
[0128] in, , The number of sensors that collect system output; It is the scheduling signal of the dynamic segment; The dimension is The identity matrix of is the number of sensors transmitting in the static segment; As a switching signal, when When the time is the static segment transmission, that is hour, ,when The moment is when the dynamic segment is transmitted, that is, hour, , , is the Kroneckerdelta function, To express or ,when hour 1, otherwise is 0; , Indicates when hour 1, otherwise is 0; is the dimensionally appropriate identity matrix.
[0129] S4. Constructing dynamic output feedback finite time The controller models the system and designs a dynamic output feedback finite time The matrix inequality group of the controller; the single-link robotic arm system framework diagram selected in this embodiment is as follows Figure 1 As shown, in the single-link robotic arm system there is The sensor collects the system output, and the collected information is transmitted to the dynamic output feedback controller via the communication network controlled by the FlexRay protocol scheduling. At the same time, denial of service attacks against the sensors may occur in the communication network.
[0130] In this embodiment, the dynamic output feedback finite time The system model of the controller is:
[0131] ;
[0132] in, is the controller status; , , , is the gain matrix of the controller;
[0133] More specifically, this embodiment proves the finite-time boundedness of the dynamic output feedback controller through the Lyapunov equation and satisfies Performance indicators, namely:
[0134] ;
[0135] in, is the Lyapunov function; is an unknown positive definite matrix; , , is the system state vector; , Row dimension , Levi is The zero matrix of The dimension is The identity matrix of Matrix and The matrices are zero matrices and identity matrices of appropriate dimension.
[0136] Further, in step S4, the designed dynamic output feedback finite time The matrix inequality system of the controller is expressed as:
[0137] ;
[0138] in,
[0139] ;
[0140] in, is the unknown positive definite matrix in the Lyapunov function; is an unknown positive definite matrix; is a given finite time parameter; ; ; ; ; ; ; ; ; is the dimensionally appropriate identity matrix; ; ; is the dynamic segment scheduling signal, is the Kronecker delta function, To express or ,when hour 1, otherwise is 0; is an unknown positive definite matrix; is an unknown positive definite matrix; for The square of the performance index; is the square of the upper bound of the external disturbance.
[0141] Note that the matrix With positive definite matrix are unknown, so the matrix in the controller design condition is , , and are all nonlinear and cannot be solved using the LMI tool; when the controller gain matrix When given a priori by GA, is a known matrix. The above matrix inequalities are all linear and can be solved using the LMI tool.
[0142] In this embodiment, the present invention will dynamically output feedback for a limited time The controller is applied to a single-link robotic arm system, which can effectively increase the robustness of the system compared to the traditional PID control method; in addition, Based on the control theory and finite time control, a dynamic output feedback controller is designed to ensure the finite time performance and enhance the rapid response capability of the controller.
[0143] S5, using GA genetic algorithm and LMI linear matrix inequality solution algorithm to obtain dynamic output feedback finite time The gain matrix of the controller As a preferred implementation of step S5, the specific process includes the following steps:
[0144] S51, dynamic output feedback limited time The matrix inequality of the controller with nonlinear terms is scaled by applying Schur's complement lemma to obtain the corresponding linear matrix inequality, as follows:
[0145] ;
[0146] Among them, the symbol " ” represents the omission of symmetric elements in a symmetric matrix;
[0147] S52, solving the linear matrix inequality obtained in step S51 and other linear matrix inequalities to obtain a set of controller gain matrices feasible solution;
[0148] S53, determine the initialization range of each individual when the GA genetic algorithm population is initialized ,in, , is the proportionality coefficient given by experience, , is the row vector of controller gain matrix elements;
[0149] S54, randomly generating an initial population from the initialization range obtained in step S53;
[0150] S55, using the LMI tool to solve the linear matrix inequality group for each individual, if there is a solution, calculate the corresponding fitness value; if there is no solution, assign a sufficiently large value to the fitness value of the individual;
[0151] Among them, the fitness value of GA genetic algorithm is expressed as , for Performance indicators;
[0152] S56, executing the selection, crossover, mutation, and elite retention GA operations in the GA genetic algorithm;
[0153] If the maximum number of iterations is reached, a set of optimal individual gain matrices obtained during the iteration process is saved , otherwise return to step S55.
[0154] The dynamic output feedback in this embodiment has a limited time The controller design method combines the traditional LMI method with the GA genetic algorithm, which can reduce the conservatism in the controller design and effectively improve the system performance. The introduction of the GA genetic algorithm can well handle the nonlinear terms in the sufficient conditions for solving the gain matrix, and can reduce the performance indicators, thereby improving the performance of the single-link robotic arm system.
[0155] S6. The gain matrix Substituting into the dynamic output feedback finite time Used in the controller to stabilize the single-link robotic arm system.
[0156] Specifically, GA-assisted dynamic output feedback is performed for a limited time The fitness value obtained in the controller design algorithm, the iteration curve is as follows Figure 2 As shown in the figure, the blue dotted line represents the optimal individual in the GA genetic algorithm iteration process. Performance Indicators , the existing linear matrix inequality LMI method is solved Performance Indicators The value is 1.1565, which shows that the optimal population in the first iteration of the genetic algorithm in the method of the present invention is The value is 0.1612, which is significantly lower than the solution value of the existing technology. After 52 iterations, our solution value reaches the optimal value, which is 0.1502. The dynamic output after genetic algorithm optimization is fed back to the finite time The controller is applied to a single-link robotic arm system. The state curve of the system is as follows: Figure 3 As shown, it can be seen from the figure that the two state components of the closed-loop system are both controlled near 0, that is, the performance requirement of finite-time stability is achieved within a specified time (14 seconds in this embodiment). Figure 4 The dotted lines in the figure represent sensor nodes that are attacked by external attackers for denial of service, and the solid lines represent the selection of system sensor nodes under the hopping FlexRay protocol. It can be seen that under the hopping FlexRay protocol, the attacked nodes will not be selected for data transmission, ensuring the security and stability of the system.
[0157] Those skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a program, so the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.
[0158] The above implementation methods have been described in detail. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.
Claims
1. A finite time H-infinity control method under a hopping FlexRay protocol, characterized in that: The method comprises the following steps: S1. Obtain the parameters of the controlled single-link robotic arm system, and reset the discrete time system parameter matrix of the single-link robotic arm system and the upper bound of the external disturbance based on the parameters; S2, dividing the sensors that collect system output information into two groups, and setting the measurement outputs of the two groups of sensors to jump to the static segment and the dynamic segment of the FlexRay protocol for transmission respectively; S3, according to the jump FlexRay communication protocol, select the measurement output that needs to be transmitted at each transmission time, and add a zero-order holder ZOH at the input end of the controller to obtain the measurement signal available to the controller S4. Construct dynamic output feedback finite time H ∞ The controller model is used to design a dynamic output feedback finite time H ∞ Matrix inequality set for the controller; S5, using GA genetic algorithm and LMI linear matrix inequality solution algorithm to obtain dynamic output feedback finite time H ∞ The controller gain matrix A c ,B c ,D c ,E c The specific process includes the following steps: S51, feed back the dynamic output for a limited time H ∞ The matrix inequality of the controller with nonlinear terms is scaled by applying Schur's complement lemma to obtain the corresponding linear matrix inequality; S52, solving the linear matrix inequality obtained in step S51 and other linear matrix inequalities to obtain a set of controller gain matrices A feasible solution for S53, determine the initialization range of each individual when the GA genetic algorithm population is initialized as [lb K ×K LMI ,ub K ×K LMI ], where lb K ,ub K is the proportionality coefficient given by experience, K LMI =[A c , B c , D c , E c ], A c , B c , D c , E c is the row vector of controller gain matrix elements; S54, randomly generating an initial population from the initialization range obtained in step S53; S55, using the LMI tool to solve the linear matrix inequality group for each individual, if there is a solution, calculate the corresponding fitness value; if there is no solution, assign a sufficiently large value to the fitness value of the individual; Among them, the fitness value of the GA genetic algorithm is expressed as γ, γ is H ∞ Performance indicators; S56, executing the selection, crossover, mutation, and elite retention GA operations in the GA genetic algorithm; If the maximum number of iterations is reached, the gain matrix A of a group of optimal individuals obtained during the iteration is saved c , B c , D c , E c , otherwise return to step S55; S6. The gain matrix A c ,B c ,D c ,E c Substituting into the dynamic output feedback finite time H ∞ Used in the controller to stabilize the single-link robotic arm system.
2. The finite time H-infinity control method according to claim 1, characterized in that: In step S1, the specific process includes the following steps: S11. Obtain the parameters m, M, J, g, D0, T of the controlled single-link robotic arm system p , and determine the mathematical model of the single-link robotic arm system, namely: In the above formula, θ(t) is the angular position of the robot arm; u(t) is the control input; ω(t) is the external disturbance, is the load mass; J is the moment of inertia; g is the acceleration of gravity; L is the arm length; D(t) is the uncertainty coefficient of viscous friction; is the angular velocity of the robot arm; S12, let x1(t) = θ(t), The mathematical model of the single-link robotic arm system is linearized to obtain the continuous-time system state space model, namely: in, S13, set the sampling time to T p , and use the c2d function in MATLAB, select the zero-order hold method as the discretization method, discretize the continuous-time system state space model to obtain the discrete-time system matrices A, B, C, G, and F; S14. Define the external disturbance satisfying condition of the single-link manipulator system, and obtain the upper bound of the external disturbance according to the external disturbance satisfying condition. The external disturbance satisfying condition is: Where d>0 is a known positive scalar, i.e., the upper bound of the external disturbance; ω(k) T is the transpose of the external disturbance ω(k); K is a finite time parameter.
3. The finite-time H-infinity control method according to claim 1, characterized in that: In step S2, it specifically includes: According to the two groups of sensors, the measurement output y(k) of the single-link robotic arm system is expressed as Right now: The first l dimensions of the measured output y(k), that is, the measured output y of the 1st, ..., lth sensors [1] (k) is set to be transmitted in the static segment of the FlexRay protocol; the last p1 dimensions of the measured output y(k), i.e., the measured output y of the l+1,…,pth sensors [2] (k) Setting the dynamic segment transmission in the FlexRay protocol; Measurement output y [1] (k) and y [2] (k) are respectively expressed as: in, p is the number of sensors output by the acquisition system; y [1] (k) represents the measured output of l sensors transmitted by the single-link manipulator system in the static section; y [2] (k) represents the measured output of pl sensors transmitted by the single-link robotic arm system in the dynamic segment.
4. The finite time H-infinity control method according to claim 1, characterized in that: In step S3, the specific process includes the following steps: S31. The sensors are divided into a set of sensors that are attacked by denial of service through the scheduler. and the set of sensors that have not been attacked S32, setting a scheduling rule for jumping the FlexRay communication protocol; S33, selecting sensor information to be transmitted at each transmission time according to the scheduling rule of the jump FlexRay communication protocol; like Then calculate the jump step length J at the current moment step If Then calculate the value of the jump weight matrix M(k) at the current moment, and according to the jump step length J step The value of and the jump weight matrix M(k) determine the scheduling signal at the current moment; in, Indicates that when time k is a static segment transmission, Indicates that time k is the dynamic segment transmission; S34. Add a zero-order holder ZOH at the input of the controller to obtain a measurement signal that can be used by the controller.
5. The finite time H-infinity control method according to claim 4, characterized in that: In step S32, the scheduling rule for jumping the FlexRay communication protocol includes a static segment scheduling rule for jumping the FlexRay protocol and a dynamic segment scheduling rule for jumping the FlexRay protocol; Among them, the static segment scheduling rules of the FlexRay protocol are: The dynamic segment scheduling rule of the designed jump FlexRay protocol is: in, It is the scheduling signal of the static segment; is the scheduling signal of the dynamic segment; k represents the kth sampling moment of the discrete time system; is the set of static segment transmission times; is the set of dynamic segment transmission times; i =diag{δ(il-1)I,…,δ(iq)I}, δ(a) is the Kronecker delta function, a is used to represent il-1 or iq, when a=1, δ(a)=1, otherwise δ(a)=0, I is the dimensionally appropriate identity matrix; M(k) is the jump weight matrix at the current time k; Q i is the weight value of the i-th sensor node.
6. The finite-time H-infinity control method according to claim 5, characterized in that: In step S34, the controller makes available the measurement signal It is expressed as: in, in, p is the number of sensors output by the acquisition system; τ(k) is the scheduling signal of the dynamic segment; I l is a unit matrix with dimension l×l; l is the number of sensors transmitting in the static segment; ρ(k) is a switching signal. When the kth moment is the static segment transmission, that is, When ρ(k)=0, when time k is the dynamic segment transmission, that is, When ρ(k)=1; Λ σ(k) =diag{δ(σ(k)-1)I, δ(σ(k)-2)I, ..., δ(σ9k)-1)I}, δ(a) is the Kronecker delta function, a is used to represent σ(k)-1 or σ(k)-2, when a=1, δ(a) is 1, otherwise δ(a) is 0; Ψ τ(k) =diag{δ(τ(k)-l-1)I,δ(τ(k)-2)I,...,δ(τ(k)-q)I},δ(a) means that when a=1,δ(a) is 1, otherwise,δ(a) is 0; I is the dimensionally appropriate identity matrix; p is the number of sensors output by the acquisition system; y [1] (k) represents the measured output of l sensors transmitted by the single-link manipulator system in the static section; y [2] (k) represents the measured output of p1 sensors transmitted by the single-link manipulator system in the dynamic segment; y [1] (k-1) represents the measurement output of l sensors transmitted by the single-link manipulator system at time k-1; y [2] (k-1) represents the measurement output of p1 sensors transmitted by the single-link robotic arm system at time k-1.
7. The finite time H-infinity control method according to claim 6, characterized in that: In step S4, the dynamic output feedback finite time H is constructed ∞ The system model of the controller is: in, is the controller status; is the gain matrix of the controller; Dynamic output feedback finite time H ∞ The matrix inequality system of the controller is expressed as: in, Among them, P α(k) is the unknown positive definite matrix in the Lyapunov function; R is the unknown positive definite matrix; μ, K, κ are given finite time parameters; I is the dimensionally appropriate identity matrix; Ψ τ(k) =diag{δ(τ(k)-l-1)I,…,δ(τ(k)-q)I}; τ(k) is the dynamic segment scheduling signal, δ(a) is the Kronecker delta function, a is used to represent τ(k)-l-1 or τ(k)-q, when a=1, δ(a) is 1, otherwise δ(a) is 0; P τ(k) is an unknown positive definite matrix; P σ(k) is an unknown positive definite matrix; γ 2 H ∞ The square of the performance index; d 2 is the square of the upper bound of the external disturbance.
8. The finite-time H-infinity control method according to claim 7, characterized in that: In step S51, the linear matrix inequality is as follows: The symbol "*" represents the omission of symmetric elements in a symmetric matrix.
