Design method, device and equipment for feedback controller of inspection robot and medium

Abstract

The invention discloses an inspection robot feedback controller design method, device and equipment and a medium, and the method comprises the steps: building a kinetic model of an inspection robot based on the kinetic parameters of the inspection robot; based on the dynamical model, combining with a preset condition and utilizing a neural network to construct a control block diagram based on event triggering; constructing an ideal controller according to the dynamic model; the ideal controller does not consider event triggering; based on a neural network and an event triggering mechanism, a control law and an adaptive law of the inspection robot are determined, and then a target controller of the inspection robot is determined according to the control law and the adaptive law. According to the invention, through combination of dynamic modeling, a neural network and event trigger control, an efficient, accurate and adaptive control strategy is provided for the inspection robot. The automation level and the intelligence level of inspection operation are remarkably improved, and the method can be widely applied to the technical field of data processing.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a design method, device, equipment and medium for a feedback controller of an inspection robot. Background Art

[0002] At present, with the continuous expansion of the scale of inspection racks and servers, maintenance costs continue to rise, posing a challenge to manufacturers to reduce costs while ensuring quality. Traditional manual operation and maintenance methods have gradually become incapable of meeting the needs of large-scale and high-frequency equipment operation and maintenance, and cannot meet the requirements of accurate and real-time operation and maintenance. In addition, the operation and maintenance requirements of large-scale inspection equipment have increasingly higher requirements for accuracy and efficiency, and traditional methods can no longer meet these requirements, which has brought adverse effects on inspection and maintenance.

[0003] With the rapid development of robot technology, digital intelligent inspection and intelligent computer room have become a reality. Robots are used to replace manual operation and maintenance. Through automated and intelligent technical means, they can work continuously and stably in special environments, realize the replacement of human labor in operation and maintenance work, have extensive economic and social benefits, and can effectively improve work efficiency.

[0004] However, as a new generation of automation equipment, robots face many challenges in their control methods. Robot systems are multivariable, time-varying, strongly coupled, nonlinear, and have many uncertain factors, which make robots difficult to control. In order to obtain fast response, good stability, and high-precision trajectory tracking control, it is necessary to comprehensively consider the dynamic modeling error of the robot system, uncertain external interference, and the influence of unknown parameters, which requires the designed controller to have a certain degree of robustness. Although the traditional time-triggered control method is predictable and easy to implement, from the perspective of resource utilization, periodic execution of control tasks will waste computing and equipment energy resources. Summary of the invention

[0005] The present invention proposes a patrol robot feedback controller design method, device, equipment and medium, aiming to solve one of the technical problems in the related art at least to a certain extent. The embodiments of the present invention can efficiently design a patrol robot feedback controller.

[0006] On the one hand, an embodiment of the present invention provides a method for designing a feedback controller of an inspection robot, comprising:

[0007] Based on the dynamic parameters of the inspection robot, a dynamic model of the inspection robot is established;

[0008] Based on the dynamic model, the control block diagram based on event triggering is constructed by using neural network in combination with predetermined conditions;

[0009] According to the dynamic model, an ideal controller is constructed; the ideal controller does not consider event triggering;

[0010] Based on the neural network and event-triggered mechanism, the control law and adaptive law of the inspection robot are determined, and then the target controller of the inspection robot is determined according to the control law and adaptive law.

[0011] Optionally, based on the dynamic parameters of the inspection robot, a dynamic model of the inspection robot is established, comprising the following steps:

[0012] The generalized joint position and velocity vector, inertia matrix, centrifugal force and Coriolis moment, gravity vector, external disturbance torque and joint driving torque of the inspection robot are obtained as dynamic parameters;

[0013] According to the dynamic parameters, the state space model of the inspection robot is established;

[0014] According to the dynamic parameters and the state space model, a dynamic model of the inspection robot system is generated.

[0015] Optionally, based on the dynamic model and in combination with predetermined conditions, a control block diagram based on event triggering is constructed using a neural network, including the following steps:

[0016] Using the dynamic model, the tracking error and the filtered tracking error are defined based on the expected parameters of the inspection robot; the expected parameters include the expected output and the expected state vector;

[0017] Based on the inspection robot, a neural network is used to build an event-triggered adaptive neural network controller, and then the weight estimation error and tracking error of the neural network are set to be consistent and ultimately bounded, and the control block diagram based on event triggering is obtained.

[0018] Optionally, according to the dynamic model, an ideal controller is constructed, including the following steps:

[0019] According to the dynamic model, the preset ideal control input is obtained without considering the event trigger;

[0020] Based on the ideal control input, the ideal controller is constructed according to the Lyapunov stability principle.

[0021] Optionally, based on the neural network and event-triggered mechanism, determining the control law and the adaptive law of the inspection robot includes the following steps:

[0022] By using the approximation properties of the neural network, the actual control input of the inspection robot is constructed, and then the control law of the inspection robot is determined;

[0023] Based on the event triggering mechanism, determine the updating rule of the control variables in the actual control input at the moment of event triggering;

[0024] The adaptive law of the inspection robot is determined according to the updating law.

[0025] Optionally, the method further comprises the following steps:

[0026] The dynamic equation analysis of the filter tracking error is carried out according to the control law and adaptive law.

[0027] Optionally, the method further comprises a step of performing stability verification on the target controller, the step comprising the following steps:

[0028] Based on the control law and adaptive law, determine the event triggering conditions of the inspection robot;

[0029] Verify the boundedness of the control parameters of the target controller based on the stability function and the preset stability algorithm;

[0030] Based on the verification results of the boundedness of the control parameters, it is determined whether the target controller is stable.

[0031] On the other hand, an embodiment of the present invention provides a design device for a feedback controller of an inspection robot, comprising:

[0032] The first module is used to establish a dynamic model of the inspection robot based on the dynamic parameters of the inspection robot;

[0033] The second module is used to construct a control block diagram based on event triggering by using a neural network based on a dynamic model and predetermined conditions;

[0034] The third module is used to construct an ideal controller based on the dynamic model; the ideal controller does not consider event triggering;

[0035] The fourth module is used to determine the control law and adaptive law of the inspection robot based on the neural network and event-triggered mechanism, and then determine the target controller of the inspection robot according to the control law and adaptive law.

[0036] Optionally, the device further comprises:

[0037] The fifth module is used to analyze the dynamic equation of the filter tracking error according to the control law and the adaptive law.

[0038] Optionally, the device further comprises:

[0039] The sixth module is used to verify the stability of the target controller; wherein the sixth module is specifically used for:

[0040] Based on the control law and adaptive law, determine the event triggering conditions of the inspection robot;

[0041] Verify the boundedness of the control parameters of the target controller based on the stability function and the preset stability algorithm;

[0042] Based on the verification results of the boundedness of the control parameters, it is determined whether the target controller is stable.

[0043] On the other hand, an embodiment of the present invention provides an electronic device, including: a processor and a memory; the memory is used to store programs; the processor executes the program to implement the above-mentioned inspection robot feedback controller design method.

[0044] On the other hand, an embodiment of the present invention provides a computer storage medium, which stores a program executable by a processor. When the program executable by the processor is executed by the processor, it is used to implement the above-mentioned inspection robot feedback controller design method.

[0045] The embodiment of the present invention establishes a dynamic model of the inspection robot based on the dynamic parameters of the inspection robot; based on the dynamic model, a neural network is used to construct a control block diagram based on event triggering in combination with predetermined conditions; an ideal controller is constructed according to the dynamic model; the ideal controller does not consider event triggering; based on the neural network and event triggering mechanism, the control law and adaptive law of the inspection robot are determined, and then the target controller of the inspection robot is determined according to the control law and adaptive law. The embodiment of the present invention includes at least the following beneficial effects:

[0046] Improve control accuracy and response speed: By establishing an accurate dynamic model based on the dynamic parameters of the inspection robot, the robot's motion characteristics can be described more accurately, thus providing a solid foundation for the design of the control strategy. This helps to improve control accuracy, enabling the robot to perform tasks more accurately, while speeding up the response speed and responding to various emergencies in a timely manner.

[0047] Optimize resource utilization and reduce energy consumption: The event-triggered control block diagram constructed by neural network combined with predetermined conditions can dynamically adjust the control timing according to actual needs and avoid unnecessary calculations and equipment energy consumption compared to traditional time-triggered control. This intelligent control method not only improves resource utilization efficiency, but also helps reduce operation and maintenance costs.

[0048] Enhanced controller robustness: Although the ideal controller constructed does not consider event triggering, it provides a benchmark for designing uncertainties in actual operation. On this basis, by introducing neural networks and event triggering mechanisms, the determined control laws and adaptive laws enable the target controller to better adapt to various changes in the robot's motion process, enhance the robustness of the controller, and improve the stability and reliability of the system.

[0049] In summary, the technical solution of the present invention provides an efficient, accurate and adaptive control strategy for the inspection robot through the combination of dynamic modeling, neural network and event-triggered control. This not only significantly improves the automation and intelligence level of inspection operations, but also provides technical support for more complex and efficient inspection tasks in the future, which helps to promote scientific and technological progress and industrial upgrading in the inspection industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The accompanying drawings are used to provide a further understanding of the technical solution of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the technical solution of the present invention and do not constitute a limitation on the technical solution of the present invention.

[0051] Figure 1 A schematic diagram of an implementation environment for designing a feedback controller for an inspection robot provided by an embodiment of the present invention;

[0052] Figure 2 A schematic diagram of a flow chart of a method for designing a feedback controller for an inspection robot provided by an embodiment of the present invention;

[0053] Figure 3 A schematic diagram of an overall architecture example of a control block diagram provided by an embodiment of the present invention;

[0054] Figure 4 A schematic diagram of the overall process of a design method for a feedback controller for an inspection robot provided by an embodiment of the present invention;

[0055] Figure 5 A schematic diagram of an example structure of an initial RBF neural network provided in an embodiment of the present invention;

[0056] Figure 6 A schematic diagram of an example of a system output tracking curve provided by an embodiment of the present invention;

[0057] Figure 7 A schematic diagram of an example of a tracking error curve provided by an embodiment of the present invention;

[0058] Figure 8 A schematic diagram of an example of a control input curve provided by an embodiment of the present invention;

[0059] Fig. 9 A schematic diagram of an example of an adaptive weight norm provided by an embodiment of the present invention;

[0060] Fig.10 A schematic diagram of an example of a triggering moment and an interval between events of a system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0062] It should be noted that, although the functional modules are divided in the system schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the system or the order in the flowchart. The terms "first / S100", "second / S200", etc. in the specification, claims and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0063] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0064] To facilitate the understanding of the technical solution of the present invention, the technical feature terms that may appear in the embodiments of the present invention are first explained:

[0065] The Artificial Neural Networks (ANN) system is composed of a large number of neurons connected by adjustable connection weights. It has the characteristics of large-scale parallel processing, distributed information storage, and good self-organization and self-learning capabilities.

[0066] A radial basis function network (RBF network) is an artificial neural network that uses a radial basis function as an activation function. The output of a radial basis function network is a linear combination of the input radial basis function and the neuron parameters.

[0067] It is understandable that the inspection robot feedback controller design method provided in the embodiment of the present invention can be applied to any computer device with data processing and computing capabilities, and this computer device can be various types of terminals or servers. When the computer device in the embodiment is a server, the server is an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. Optionally, the terminal is a smart phone, a tablet computer, a laptop computer, a desktop computer, etc., but is not limited to this.

[0068] like Figure 1 FIG. 1 is a schematic diagram of an implementation environment provided by an embodiment of the present invention. Figure 1 The implementation environment includes at least one terminal 102 and a server 101. The terminal 102 and the server 101 can be connected to a network wirelessly or wired to complete data transmission and exchange.

[0069] Server 101 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), as well as big data and artificial intelligence platforms.

[0070] In addition, the server 101 can also be a node server in the blockchain network. Blockchain is a new application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, and encryption algorithm.

[0071] The terminal 102 may be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, etc., but is not limited thereto. The terminal 102 and the server 101 may be directly or indirectly connected via wired or wireless communication, which is not limited in the embodiment of the present invention.

[0072] Based on the example Figure 1 In the implementation environment shown, an embodiment of the present invention provides a patrol robot feedback controller design method. The following is explained using the patrol robot feedback controller design method applied in the terminal 102 as an example. It can be understood that the patrol robot feedback controller design method can also be applied to the server 101.

[0073] Reference Figure 2 , Figure 2 The flowchart of the inspection robot feedback controller design method applied to a computer system provided by the embodiment of the present invention, the execution subject of the inspection robot feedback controller design method can be any of the aforementioned computer devices (including servers or terminals). Figure 2 , the method comprises the following steps:

[0074] S100, establishing a dynamic model of the inspection robot based on the dynamic parameters of the inspection robot;

[0075] It should be noted that, in some embodiments, establishing a dynamic model of the inspection robot based on the dynamic parameters of the inspection robot may include the following steps:

[0076] The generalized joint position and velocity vector, inertia matrix, centrifugal force and Coriolis moment, gravity vector, external disturbance torque and joint driving torque of the inspection robot are obtained as dynamic parameters;

[0077] According to the dynamic parameters, the state space model of the inspection robot is established;

[0078] According to the dynamic parameters and the state space model, a dynamic model of the inspection robot system is generated.

[0079] For example, in some specific implementations, a dynamic equation model of an n-DOF inspection robot may be established first:

[0080]

[0081] in are the generalized joint position and velocity vectors, is the inertia matrix, Including centrifugal force and Coriolis moment, is the gravity vector, is the external disturbance torque, and the control input vector is the joint driving torque. At the same time, the joint position and velocity of the fan detection robot operating arm are defined as the state vector

[0082] The dynamic equation (1) of the n-DOF inspection robot system introduced above is written in state space as follows:

[0083]

[0084] in is the state vector, and is an unknown bounded smooth nonlinear function. d(t) is an external disturbance with an unknown upper bound. and Represents the input and output of the system, and only the output y is measurable.

[0085] S200, based on the dynamics model, a control block diagram based on event triggering is constructed by using a neural network in combination with predetermined conditions;

[0086] It should be noted that in some embodiments, based on the dynamic model and in combination with predetermined conditions, a control block diagram based on event triggering is constructed using a neural network, which may include the following steps: using the dynamic model, the tracking error and the filtered tracking error are defined based on the expected parameters of the inspection robot; the expected parameters include the expected output and the expected state vector; based on the inspection robot, a neural network is used to construct an adaptive neural network controller based on event triggering, and then the weight estimation error and tracking error of the neural network are set to be consistent and ultimately bounded, to obtain a control block diagram based on event triggering.

[0087] For example, in some specific implementations, the construction of a control block diagram can be realized by describing the system filtering error and the system control process, which can be realized as follows:

[0088] The following assumptions (i.e., predetermined conditions) are made for system (2):

[0089] Assumption 1: There exists a known positive constant b m , b M Make satisfy where I is the identity matrix of appropriate dimension.

[0090] Assumption 2: There exists a constant τ dB >0 makes ||τ d (t)||≤τ dB .

[0091] Assumption 3: The neural network activation function ξ(·) satisfies Where l is the number of neurons in the hidden layer, and there are w M >0 and ε M >0 makes the ideal weight of the neural network satisfy ‖w * ‖≤w M , the approximation error satisfies ‖ε‖≤ε M .

[0092] Assume y m is the expected output and is the desired state vector. At the same time, define the tracking error e = y m -y and the tracking error vector is The filtered tracking error r is defined as:

[0093]

[0094] in And α>0 is a design parameter. The derivative of the filtering tracking error r with respect to time can be obtained:

[0095]

[0096] in

[0097] Under the assumptions 1-3 and the state is unmeasurable, and only the output y is measurable, the main purpose of this paper is to design an event-triggered adaptive neural network controller for system (2) without using a state observer but only using a neural network, so that the neural network weight estimation error and tracking error are consistent and ultimately bounded. The control block diagram is designed as follows Figure 3 As shown, ZOH (Zero Order Holder) represents the zero-order holder; Actuator represents the brake; Plant represents the controlled object or system, which can be any physical process or equipment, such as machines, chemical processes, bioreactors, etc.; Sensor represents the sensor; Trigger Mechanism represents the trigger mechanism / mechanism; Controller represents the controller; Neural Network represents the neural network; Adaptive Law represents the adaptive law.

[0098] S300, construct an ideal controller based on the dynamics model;

[0099] Among them, the ideal controller does not consider event triggering;

[0100] It should be noted that, in some embodiments, constructing an ideal controller according to a dynamic model may include the following steps: according to the dynamic model, obtaining a preset ideal control input without considering event triggering; based on the ideal control input, constructing an ideal controller according to the Lyapunov stability principle.

[0101] For example, in some specific implementations, designing an ideal controller may be implemented as follows:

[0102] Assuming a nonlinear function and is known exactly, and without considering external disturbance d(t) and event triggering, there exists an ideal feedback control u * , so that under this control the output y tracks the desired trajectory y m .

[0103] For system (2), if the ideal control input is:

[0104]

[0105] Then the filter tracking error r converges to 0. Where β>0 is the design parameter, and

[0106] Proof: Consider the Lyapunov candidate function And V r The time derivative is:

[0107]

[0108] According to Lyapunov's stability principle, the above formula means that lim t→∞ r = 0. However, for system (2) we consider and Unknown and d(t)≠0, the state vector Not completely measurable, ideal controller u * Therefore, the ideal controller (5) is rewritten as:

[0109]

[0110] in is an unknown function.

[0111] S400, based on the neural network and event-triggered mechanism, determine the control law and adaptive law of the inspection robot, and then determine the target controller of the inspection robot according to the control law and adaptive law.

[0112] It should be noted that in some embodiments, determining the control law and adaptive law of the inspection robot based on the neural network and event-triggered mechanism may include the following steps: utilizing the approximation properties of the neural network to construct the actual control input of the inspection robot, and then determining the control law of the inspection robot; based on the event-triggered mechanism, determining the update law of the control variables in the actual control input at the moment of event triggering; and determining the adaptive law of the inspection robot based on the update law.

[0113] Exemplarily, in some specific implementations, a neural network control law and an adaptive law based on event triggering are designed;

[0114] Considering the universal approximation of neural networks, RBF neural networks are used to approximate unknown functions. We can get:

[0115]

[0116] Where w is the target neural network weight vector of the output layer, is the basis function vector and c i is the center of the ith basis function, σ i is the smoothing parameter, which determines the width of the center point of the i-th basis function. is the input vector of the neural network, d 1 >0 indicates the delay time, ε is the approximation error. The optimal weight w * Defined as:

[0117]

[0118] Design the actual neural network feedback control input:

[0119]

[0120] in is the output of the neural network, is the optimal weight w of the neural network * The estimated value of is a positive definite design matrix. ET (t) is the event trigger sampling error, which is defined as:

[0121]

[0122] e(t k ) is the error at the event triggering time. Since the last hold state and neural network weights are at the event triggering time t = t k Therefore, the control input is also updated at the trigger time and is held by the zero-order holder until the next update. Neural Network Weight Estimation In the interval between two events t k <t≤t k+1 remains unchanged and at the triggering time t=t k Update. Therefore, the update law of the neural network weights during the event interval is defined as:

[0123]

[0124] At the same time, define the triggering time t = t k The neural network weight adaptation law:

[0125]

[0126] in is the neural network weight estimation after the triggering moment, and the design parameter Γ>0 is the neural network learning law.

[0127] Defining Neural Network Weight Estimation Error Using equation (11), the dynamics of the neural network weight estimation error during the event interval can be written as:

[0128]

[0129] At the triggering time t = t k , the neural network weight estimation error obtained by equation (12) dynamically becomes:

[0130]

[0131] in

[0132] Furthermore, in some embodiments, the method may also include the following steps: performing a dynamic equation analysis of the filter tracking error according to the control law and the adaptive law.

[0133] For example, in some specific implementations, the dynamic equation for analyzing the filtering error may be implemented as follows:

[0134] We have already introduced the approximation theorem of neural networks and designed the actual control input based on event triggering. Next, we will analyze the dynamic equation of the filter tracking error by combining the designed control law and adaptive law.

[0135] From equation (4), the time derivative of the filtered tracking error r can be written as:

[0136]

[0137] Substituting formula (7) into (15) yields:

[0138]

[0139] Since the tracking error vector available:

[0140]

[0141] Substituting formula (7) and (9) into (17) yields:

[0142]

[0143] There is also a neural network weight error definition and Formula (18) can be transformed into:

[0144]

[0145] Furthermore, in some embodiments, the method may also include a step of verifying the stability of the target controller, which may include the following steps: determining the event triggering conditions of the inspection robot based on the control law and the adaptive law; verifying the boundedness of the control parameters of the target controller based on the stability function and the preset stability algorithm; and determining whether the target controller is stable based on the verification result of the boundedness of the control parameters.

[0146] For example, in some specific implementations, the Lyapunov stability principle can be used to prove the stability of the system, which can be implemented as follows:

[0147] In this section, we will prepare to present the stability results of the closed-loop system with output feedback based on the previous design of the neural network adaptive law and the analysis of the filter tracking error dynamics.

[0148] Theorem 1: Consider the control input (9) and the filtered tracking error dynamics (19), and by definition at the trigger time t = t k The neural network weight adaptation law (14) assumes that 1-3 holds and triggers an event so that the following conditions hold during the sampling interval:

[0149]

[0150] Then the neural network weight estimation error and tracking error are consistent and ultimately bounded. k ,ψ∈(0,1).

[0151] Proof: This proof can be considered in two cases

[0152] Case 1: First, during the sampling interval, consider the Lyapunov candidate function Since the neural network weights are not updated during the sampling interval, that is, in [t k , t k+1 ) Therefore, according to the previously analyzed filter tracking error dynamics (19), the derivative of the Lyapunov candidate function can be obtained

[0153]

[0154] Using norm operations and Young's inequality equations Formula (21) can be written as:

[0155]

[0156] Because of the definition Then equation (22) becomes:

[0157]

[0158] choose and available:

[0159]

[0160] in

[0161] Case 2: At the event triggering time t = t k Consider the Lyapunov function Since at t = t k When all Just consider the Lyapunov function difference:

[0162]

[0163] Substituting the neural network weight estimation error dynamics (14) into (25), we can obtain:

[0164]

[0165] Use definition Substituting into (26) we can obtain:

[0166]

[0167] Using norm operations, the upper bound of the difference (27) can be written as:

[0168]

[0169] use and Young's inequality equation You can get:

[0170]

[0171] Then use the bounds from Assumption 3 ||w * ||≤w M , we can get:

[0172]

[0173] in

[0174] From the above formula (30), we can get Then ΔL<0. Therefore, it can be concluded from Case 1 and Case 2 that the neural network weight estimation error and tracking error are consistent and ultimately bounded, and the stability proof is completed.

[0175] In order to explain the principle of the technical solution of the present invention in detail, the overall process of the present invention is described below in combination with some specific embodiments. It is easy to understand that the following is an explanation of the technical principle of the present invention and cannot be regarded as a limitation of the present invention.

[0176] First of all, it should be noted that traditional control is executed in a periodic manner, namely the so-called time-triggered control, which is based on the theory of sampled data systems, is predictable and easy to implement. However, from the perspective of resource utilization, time-triggered control that periodically executes control tasks will waste computing and equipment energy resources. Therefore, the present invention adopts an output feedback design method and avoids the design of a state observer. Considering the modeling error and uncertain interference of the inspection robot system, the trajectory tracking control algorithm adopts the output feedback method under the event trigger mechanism, so that the robot can run quickly, stably and accurately, which is of great significance to the theoretical research and practical application of the inspection robot. Under this technical background, the present invention provides a design method for a neural network output feedback controller of an inspection robot based on event triggering.

[0177] In some preferred implementations, in view of the requirements for the design of the inspection robot controller, the entire design process of the present invention focuses on the design of the inspection robot neural network controller based on the event trigger mechanism. Figure 4 , the overall process of the present invention can be implemented as follows:

[0178] 1. First, the dynamic equation model of the n-degree-of-freedom inspection robot is established. 2. The system filtering error and system control process are described. 3. The ideal controller is designed. 4. The neural network control law and adaptive law based on event triggering are designed. 5. The dynamic equation of the filtering error is analyzed. 6. The event triggering conditions are designed, and the system stability is proved by using the Lyapunov stability principle.

[0179] The present invention is a design method for a neural network output feedback controller of an n-degree-of-freedom inspection robot. The specific steps of the method can be implemented as follows:

[0180] Step 1: Establish the dynamic equation model of the n-DOF inspection robot:

[0181]

[0182] in are the generalized joint position and velocity vectors, is the inertia matrix, Including centrifugal force and Coriolis moment, is the gravity vector, is the external disturbance torque, and the control input vector is the joint driving torque. At the same time, the joint position and velocity of the fan detection robot operating arm are defined as the state vector

[0183] The dynamic equation (1) of the n-DOF inspection robot system introduced above is written in state space as follows:

[0184]

[0185] in is the state vector, and is an unknown bounded smooth nonlinear function. d(t) is an external disturbance with an unknown upper bound. and Represents the input and output of the system, and only the output y is measurable.

[0186] Step 2: Describe the system filtering error and system control process:

[0187] The following assumptions (i.e., predetermined conditions) are made for system (2):

[0188] Assumption 1: There exists a known positive constant b m , b M Make satisfy where I is the identity matrix of appropriate dimension.

[0189] Assumption 2: There exists a constant τ dB >0 makes ||τ d (t)||≤τ dB .

[0190] Assumption 3: The neural network activation function ξ(·) satisfies Where l is the number of neurons in the hidden layer, and there are w M >0 and ε M >0 makes the ideal weight of the neural network satisfy ‖w * ‖≤w M , the approximation error satisfies ‖ε‖≤ε M .

[0191] Assume y m is the expected output and is the desired state vector. At the same time, define the tracking error e = y m -y and the tracking error vector is The filtered tracking error r is defined as:

[0192]

[0193] in And α>0 is a design parameter. The derivative of the filtering tracking error r with respect to time can be obtained:

[0194]

[0195] in

[0196] Under the assumptions 1-3 and the state is unmeasurable, and only the output y is measurable, the main purpose of this paper is to design an event-triggered adaptive neural network controller for system (2) without using a state observer but only using a neural network, so that the neural network weight estimation error and tracking error are consistent and ultimately bounded. The control block diagram is designed as follows Figure 3 shown.

[0197] Step 3: Design an ideal controller:

[0198] Assuming a nonlinear function and is known exactly, and without considering external disturbance d(t) and event triggering, there exists an ideal feedback control u * , so that under this control the output y tracks the desired trajectory y m .

[0199] For system (2), if the ideal control input is:

[0200]

[0201] Then the filter tracking error r converges to 0. Where β>0 is the design parameter, and

[0202] Proof: Consider the Lyapunov candidate function And V r The time derivative is:

[0203]

[0204] According to Lyapunov's stability principle, the above formula means that lim t→∞ r = 0. However, for system (2) we consider and Unknown and d(t)≠0, the state vector Not completely measurable, ideal controller u * Therefore, the ideal controller (5) is rewritten as

[0205]

[0206] in is an unknown function.

[0207] Step 4: Design the neural network control law and adaptive law based on event triggering:

[0208] Considering the universal approximation of neural networks, RBF neural networks (such as Figure 5 Approximation of unknown function We can get:

[0209]

[0210] Where w is the target neural network weight vector of the output layer, is the basis function vector and c i is the center of the ith basis function, σ i is the smoothing parameter, which determines the width of the center point of the i-th basis function. is the input vector of the neural network, d 1 >0 indicates the delay time, ε is the approximation error. The optimal weight w * Defined as:

[0211]

[0212] Design the actual neural network feedback control input:

[0213]

[0214] in is the output of the neural network, is the optimal weight w of the neural network * The estimated value of is a positive definite design matrix. ET (t) is the event trigger sampling error, which is defined as:

[0215]

[0216] e(t k ) is the error at the event triggering time. Since the last hold state and neural network weights are at the event triggering time t = t k Therefore, the control input is also updated at the trigger time and is held by the zero-order holder until the next update. Neural Network Weight Estimation In the interval between two events t k <t≤t k+1 remains unchanged and at the triggering time t=t k Update. Therefore, the update law of the neural network weights during the event interval is defined as:

[0217]

[0218] At the same time, define the triggering time t = tk The neural network weight adaptation law:

[0219]

[0220] in is the neural network weight estimation after the triggering moment, and the design parameter Γ>0 is the neural network learning law.

[0221] Defining Neural Network Weight Estimation Error Using equation (11), the dynamics of the neural network weight estimation error during the event interval can be written as:

[0222]

[0223] At the triggering time t = t k , the neural network weight estimation error obtained by equation (12) dynamically becomes:

[0224]

[0225] in

[0226] Step 5: Analyze the dynamic equation of filtering error:

[0227] We have already introduced the approximation theorem of neural networks and designed the actual control input based on event triggering. Next, we will analyze the dynamic equation of the filter tracking error by combining the designed control law and adaptive law.

[0228] From equation (4), the time derivative of the filtered tracking error r can be written as:

[0229]

[0230] Substituting formula (7) into (15) yields:

[0231]

[0232] Since the tracking error vector available:

[0233]

[0234] Substituting formula (7) and (9) into (17) yields:

[0235]

[0236] There is also a neural network weight error definition and Formula (18) can be transformed into:

[0237]

[0238] Step 6: Use Lyapunov stability principle to prove the stability of the system:

[0239] In this section, we will prepare to present the stability results of the closed-loop system with output feedback based on the previous design of the neural network adaptive law and the analysis of the filter tracking error dynamics.

[0240] Theorem 1: Consider the control input (9) and the filtered tracking error dynamics (19), and by definition at the trigger time t = t k The neural network weight adaptation law (14) assumes that 1-3 holds and triggers an event so that the following conditions hold during the sampling interval:

[0241]

[0242] Then the neural network weight estimation error and tracking error are consistent and ultimately bounded. k ,ψ∈(0,1).

[0243] Proof: This proof can be considered in two cases

[0244] Case 1: First, during the sampling interval, consider the Lyapunov candidate function Since the neural network weights are not updated during the sampling interval, that is, in [t k , t k+1 ), Therefore, according to the previously analyzed filter tracking error dynamics (19), the derivative of the Lyapunov candidate function We can get:

[0245]

[0246] Using norm operations and Young's inequality equations Formula (21) can be written as:

[0247]

[0248] Because of the definition Then equation (22) becomes:

[0249]

[0250] choose and available:

[0251]

[0252] in

[0253] Case 2: At the event triggering time t = t k Consider the Lyapunov function Since at t = t k When all Just consider the Lyapunov function difference:

[0254]

[0255] Substituting the neural network weight estimation error dynamics (14) into (25), we can obtain:

[0256]

[0257] Use definition Substituting into (26) we can obtain:

[0258]

[0259] Using norm operations, the upper bound of the difference (27) can be written as:

[0260]

[0261] use and Young's inequality equation You can get:

[0262]

[0263] Then use the bounds from Assumption 3 ||w * ||≤w M , we can get:

[0264]

[0265] in

[0266] From the above formula (30), we can get Then ΔL<0. Therefore, it can be concluded from Case 1 and Case 2 that the neural network weight estimation error and tracking error are consistent and ultimately bounded, and the stability proof is completed.

[0267] The whole design process focuses on the design of the neural network controller of the inspection robot based on the event trigger mechanism. First, in step 1, the dynamic equation model of the inspection robot with n degrees of freedom is established; step 2 describes the system filtering error and system control process; step 3 designs the ideal controller; step 4 designs the neural network control law and adaptive law based on event triggering; step 5 analyzes the dynamic equation of the filtering error; step 6 designs the event triggering conditions and uses the Lyapunov stability principle to prove the system stability; after the above steps, the design is completed.

[0268] The present invention is further described in detail below in conjunction with embodiments in specific application scenarios. The following embodiments are only used to more clearly illustrate the technical method of the present invention, and are not intended to limit the scope of protection of the present invention. Figures 6 to 10 As shown, the embodiment of the present invention discloses a design method for a neural network output feedback controller of an inspection robot based on event triggering. The main purpose is to target the control system of an n-degree-of-freedom inspection robot. In actual engineering, the system state is difficult to measure, the system is frequently updated, the energy consumption is high, and the controller structure is complex. The dynamic equation model of the n-degree-of-freedom inspection robot is studied and established; the system filtering error and the system control process are described; the neural network control law and adaptive law based on event triggering are designed, and a linear parameterized neural network is used to design the adaptive law and control law. The adaptive law model is used to approximate nonlinear unknown dynamic equations. The system only updates the control law and parameter adaptive law when triggered, which greatly reduces the communication frequency and energy consumption, and significantly saves communication resources. The dynamic equation of the filtering error is analyzed; and the system stability is proved by using the Lyapunov stability principle.

[0269] First, consider the Lagrangian equation of the single-link robot manipulator as follows:

[0270]

[0271] where q and are the joint positions and joint velocities of the robot manipulator respectively. M=1kg.m 2 Represents mechanical inertia, F v =1Nm·s / rad is the viscous friction coefficient matrix, the arm mass m=1kg, the gravity acceleration g=10N / kg and the length of the single link L=1m. The control input τ(t) is defined as the joint driving torque and τ d (t) is the external disturbance torque.

[0272] Let u = τ and d(t) = τ d (t), then the dynamics of the single-link robot manipulator can be written as follows:

[0273]

[0274] The following parameters are selected during the simulation. The expected output y of the system m =q m = sin(0.8t), the initial output y(0) = q(0) = 0.5 and the interference signal d(t) = 0.2sin(t). We choose the design parameters α = 4, β = 16, κ = 0.01 and the neural network learning law Γ = 20. Neural network weights is randomly initialized from a uniform distribution in the interval [0,0.1] and the RBF neural network adjusts the initial network structure and parameters online. In addition, the event trigger condition is determined by the parameter ψ=0.005、b M =3 and k s =20I for evaluation.

[0275] For the single-link robot control system, the event-triggered neural network control input and adaptive law are as follows:

[0276]

[0277] Select sampling period h = 0.1s, and get the simulation results as follows Figures 6 to 10 shown. Figure 6 Plots the actual output q and the expected output q m The curve, Figure 7 shows the trajectory of the tracking error. Figure 6 and Figure 8 It can be clearly seen that q can track the desired signal q within a small error m The control signal u is plotted on Figure 8 It can be seen that event-triggered control can greatly reduce the number of actuator updates. Fig. 9 It shows the change process of the adaptive parameters of the neural network weights. Figures 8 to 9 It can be seen that these signals are bounded. Fig.10 The time interval of each event is shown, which verifies the correctness and superiority of the event-triggered control scheme, that is, the proposed event-triggered control method (only 79 events with a sampling interval of 0.1s) shows that it can greatly reduce the amount of transmitted data and save communication resources to some extent.

[0278] In summary, compared with the prior art, the present invention has at least the following beneficial effects:

[0279] 1. Design a linear parameterized neural network algorithm. This method takes into account the balance between performance and efficiency and solves the problem that the embedded hardware platform is limited by the processor performance and cannot run large neural networks. In actual robot inspection and maintenance tasks, the controller model usually needs to be deployed in embedded hardware devices, which cannot meet the hardware requirements of high-precision control feedback models. Therefore, a linear parameterized neural network algorithm is used to construct an event-triggered adaptive law and control law for the n-DOF inspection robot control system, simplifying the controller structure.

[0280] 2. The event trigger mechanism designed based on the n-DOF inspection robot system allows the system to update the control law and adaptive law only at the triggering moment. From the perspective of resource utilization, the time-triggered control that periodically executes control tasks will waste computing and equipment energy resources, greatly reduce the communication frequency, and save communication resources.

[0281] 3. Considering the external interference of the inspection robot system, the designed system has good robustness. The robot controller adopts the feedback method of output feedback, which can effectively solve the problem that the system state is difficult to measure in practice, and does not require a state observer, avoiding the design of a state observer, and further realizing the lightweight and stable network model.

[0282] On the other hand, an embodiment of the present invention provides a design device for a feedback controller of an inspection robot, which may include:

[0283] The first module is used to establish a dynamic model of the inspection robot based on the dynamic parameters of the inspection robot;

[0284] The second module is used to construct a control block diagram based on event triggering by using a neural network based on a dynamic model and predetermined conditions;

[0285] The third module is used to construct an ideal controller based on the dynamic model; the ideal controller does not consider event triggering;

[0286] The fourth module is used to determine the control law and adaptive law of the inspection robot based on the neural network and event-triggered mechanism, and then determine the target controller of the inspection robot according to the control law and adaptive law.

[0287] In some embodiments, the apparatus may further include:

[0288] The fifth module is used to analyze the dynamic equation of the filter tracking error according to the control law and the adaptive law.

[0289] In some embodiments, the apparatus may further include:

[0290] The sixth module is used to verify the stability of the target controller; wherein the sixth module is specifically used for:

[0291] Based on the control law and adaptive law, determine the event triggering conditions of the inspection robot;

[0292] Verify the boundedness of the control parameters of the target controller based on the stability function and the preset stability algorithm;

[0293] Based on the verification results of the boundedness of the control parameters, it is determined whether the target controller is stable.

[0294] The contents of the method embodiments of the present invention are all applicable to the device embodiments. The functions specifically implemented by the device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0295] On the other hand, an embodiment of the present invention further provides an electronic device, the electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the above sensitive information method when executing the computer program. The electronic device can be any intelligent terminal including a tablet computer, a car computer, etc.

[0296] It can be understood that the contents of the above method embodiments are all applicable to the present device embodiments, the functions specifically implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0297] The contents of the method embodiments of the present invention are all applicable to the electronic device embodiments. The functions specifically implemented by the electronic device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0298] Another aspect of an embodiment of the present invention further provides a computer-readable storage medium, wherein the storage medium stores a program, and the program is executed by a processor to implement the above method.

[0299] It should be noted that the computer-readable medium shown in the embodiment of the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (Compact Disc Read to Only Memory, CD to ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than computer readable storage media, which may send, propagate, or transmit programs for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0300] The contents of the method embodiments of the present invention are all applicable to the computer-readable storage medium embodiments. The functions specifically implemented by the computer-readable storage medium embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0301] The embodiment of the present invention also discloses a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. A processor of a computer device can read the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes the above method.

[0302] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0303] It should be noted that, although several modules of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to an embodiment of the present invention, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into being embodied by multiple modules or units.

[0304] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD to ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the implementation of the present invention.

[0305] In some selectable embodiments, the function / operation mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the function / operation involved, the two boxes shown in succession can actually be executed substantially simultaneously or the boxes can sometimes be executed in reverse order. In addition, the embodiment presented and described in the flow chart of the present invention is provided by way of example, for the purpose of providing a more comprehensive understanding of technology. The disclosed method is not limited to the operation and logic flow presented herein. Selectable embodiments are expected, wherein the order of various operations is changed and the sub-operation of a part for which is described as a larger operation is performed independently.

[0306] In addition, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise specified, one or more of the functions and / or features can be integrated into a single physical device and / or software module, or one or more functions and / or features can be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the present invention. More specifically, in view of the properties, functions and internal relationships of the various functional modules in the device disclosed herein, the actual implementation of the module will be understood within the conventional skills of the engineer. Therefore, those skilled in the art can implement the present invention set forth in the claims without excessive experimentation using ordinary techniques. It is also understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present invention, which is determined by the full scope of the appended claims and their equivalents.

[0307] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0308] The logic and / or steps represented in the flowchart or otherwise described herein, for example, may be considered as an ordered list of executable instructions for implementing logical functions, and may be embodied in any computer-readable medium for use by an instruction execution device, apparatus or device (such as a computer-based device, a device including a processor, or other device that can fetch instructions from an instruction execution device, apparatus or device and execute the instructions), or in conjunction with such instruction execution device, apparatus or device. For purposes of this specification, "computer-readable medium" may be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution device, apparatus or device, or in conjunction with such instruction execution device, apparatus or device.

[0309] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0310] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution device. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0311] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0312] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

[0313] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of the present invention.

Claims

1. A design method for a feedback controller of an inspection robot, characterized in that: The following steps are involved: Based on the dynamic parameters of the inspection robot, a dynamic model of the inspection robot is established; Based on the dynamic model, a control block diagram based on event triggering is constructed by using a neural network in combination with predetermined conditions; According to the dynamic model, an ideal controller is constructed; The ideal controller does not consider the event triggering; Based on the neural network and the event-triggered mechanism, the control law and the adaptive law of the inspection robot are determined, and then the target controller of the inspection robot is determined according to the control law and the adaptive law.

2. The inspection robot feedback controller design method according to claim 1, characterized in that: The step of establishing the dynamic model of the inspection robot based on the dynamic parameters of the inspection robot comprises the following steps: Obtaining the generalized joint position and velocity vector, inertia matrix, centrifugal force and Coriolis moment, gravity vector, external interference torque and joint driving torque of the inspection robot as the dynamic parameters; According to the dynamic parameters, a state space model of the inspection robot is established; The dynamic model of the inspection robot system is generated according to the dynamic parameters and the state space model.

3. The inspection robot feedback controller design method according to claim 1, characterized in that: The method of constructing a control block diagram based on event triggering by using a neural network based on the dynamic model and in combination with predetermined conditions includes the following steps: Using the dynamic model, a tracking error and a filtered tracking error are defined based on expected parameters of the inspection robot; the expected parameters include an expected output and an expected state vector; Based on the inspection robot, a neural network is used to construct an adaptive neural network controller based on event triggering, and then the weight estimation error and the tracking error of the neural network are set to be consistent and ultimately bounded, so as to obtain the control block diagram based on event triggering.

4. The inspection robot feedback controller design method according to claim 1, characterized in that: The construction of an ideal controller according to the dynamic model comprises the following steps: According to the dynamic model, obtaining a preset ideal control input without considering event triggering; Based on the ideal control input, the ideal controller is constructed according to the Lyapunov stability principle.

5. The inspection robot feedback controller design method according to claim 1, characterized in that: The control law and the adaptive law of the inspection robot are determined based on the neural network and the event-triggered mechanism, and include the following steps: Using the approximation property of the neural network, the actual control input of the inspection robot is constructed, and then the control law of the inspection robot is determined; Based on the event triggering mechanism, determining the updating rule of the control variable in the actual control input at the moment of the event triggering; The adaptive law of the inspection robot is determined according to the updating law.

6. The inspection robot feedback controller design method according to claim 1, characterized in that: The method further comprises the following steps: The filter tracking error dynamic equation analysis is performed according to the control law and the adaptive law.

7. The inspection robot feedback controller design method according to any one of claims 1 to 6, characterized in that: The method further comprises the step of performing stability verification on the target controller, the step comprising the following steps: Based on the control law and the adaptive law, determining an event triggering condition of the inspection robot; Verifying the boundedness of control parameters of the target controller based on the stability function and the preset stability algorithm; According to the verification result of the boundedness of the control parameter, it is determined whether the target controller has stability.

8. A design device for a feedback controller of an inspection robot, characterized in that: include: The first module is used to establish a dynamic model of the inspection robot based on the dynamic parameters of the inspection robot; The second module is used to construct a control block diagram based on event triggering by using a neural network based on the dynamic model and in combination with predetermined conditions; The third module is used to construct an ideal controller according to the dynamic model; The ideal controller does not consider the event triggering; The fourth module is used to determine the control law and adaptive law of the inspection robot based on the neural network and the event-triggered mechanism, and then determine the target controller of the inspection robot according to the control law and the adaptive law.

9. An electronic device, characterized in that: including a processor and a memory; The memory is used to store programs; The processor executes the program to implement the method according to any one of claims 1 to 7.

10. A computer storage medium storing a program executable by a processor, characterized in that: The program executable by the processor is used to implement the method according to any one of claims 1 to 7 when executed by the processor.

Images