A mecanum wheel automatic guided vehicle fault detection method and system

Abstract

The application discloses a kind of Mecanum wheel automatic guided vehicle fault detection method and system, it is related to industrial automation fault diagnosis technical field.The present application is based on differential inclusion theory and Euler-Lagrange equation, establishes the Lur'e differential inclusion type state space equation of the nonlinear of representation system, parameter uncertainty and external disturbance;Design adaptive event trigger mechanism containing dynamic adjustment parameter and exponential decay term, optimize data transmission, reduce resource consumption and strictly exclude Zeno behavior;Robust interval estimation of system state is realized by constructing adaptive event trigger interval observer, and actuator fault detection is realized by residual interval zero-value inclusion.The present application solves the problem that traditional method is difficult to consider fault detection accuracy and resource utilization, and is suitable for state monitoring and fault diagnosis of Mecanum wheel automatic guided vehicle in industrial scene.

Description

Technical Field

[0001] This invention relates to the field of industrial automation fault diagnosis technology, and in particular to a fault detection method and system for Mecanum wheel automated guided vehicles (AGVs), which is suitable for condition monitoring and fault diagnosis of AGVs in high-precision, dynamic industrial environments. Background Technology

[0002] Mecanum wheel automated guided vehicles (AGVs), as core equipment in modern industrial automation and intelligent manufacturing, are widely used in material handling, logistics transportation, and flexible production scenarios. They significantly improve production and operational efficiency and flexibility, becoming an indispensable key piece of equipment in high-precision dynamic industrial environments. However, when operating in complex industrial scenarios, AGVs are susceptible to failures in key components such as actuators and sensors, leading to problems such as inaccurate positioning and accidental collisions. This can result in production interruptions, safety hazards, and economic losses. Therefore, reliable fault diagnosis technology is a core requirement for ensuring the stable, safe, and efficient operation of AGVs.

[0003] Existing fault diagnosis methods mainly include neural network methods and state observer methods. Neural network methods utilize nonlinear fitting capabilities to achieve fault detection, but they do not rely on accurate system models. However, their generalization ability is limited under the complex dynamic characteristics of Mecanum wheel automated guided vehicles (AGVs). Traditional state observer methods construct residuals by comparing estimated outputs with measured outputs to achieve fault identification, but they struggle to handle the inherent nonlinearity, parameter uncertainty, and external disturbances of Mecanum wheel AGVs. Distributed interval observer methods improve the diagnostic reliability in multi-agent collaborative scenarios, but they do not consider resource constraints.

[0004] Since its inception, the interval observer has become an effective tool for state estimation under uncertainty due to its ability to provide deterministic boundary estimates of the system state rather than point estimates, and has been applied in the field of state estimation for Mecanum wheel automated guided vehicles (AGVs). However, traditional interval observer designs suffer from excessive computational and communication resource consumption, limiting their practical application in resource-constrained MAV systems. Event-triggered mechanisms effectively optimize resource utilization by transmitting data only when preset conditions are met. Among them, dynamic event-triggered mechanisms introduce dynamic variables into the triggering conditions, while adaptive event-triggered mechanisms are better at handling system uncertainties and dynamic changes, and their performance is superior to dynamic event-triggered mechanisms.

[0005] However, existing technologies have limited research on the integration of interval observers and Lur'e differential inclusion systems. Furthermore, traditional event-triggered mechanisms in Mecanum wheel automated guided vehicles (AGVs) suffer from Zeno behavior risks, triggering an unlimited number of times within a finite timeframe. Additionally, there is a lack of dedicated adaptive event-triggered interval observer designs for the nonlinear and resource-constrained characteristics of Mecanum wheel AGVs, making it difficult to balance fault diagnosis accuracy with resource utilization. Therefore, a method is urgently needed that can accurately model the complex dynamics of Mecanum wheel AGVs, optimize resource consumption, and reliably detect faults. Summary of the Invention

[0006] This invention aims to overcome the shortcomings of existing technologies and provide a fault detection method and system for Mecanum wheel automated guided vehicles (AGVs), solving the core problems of nonlinear modeling, uncertainty handling, and resource optimization for Mecanum wheel AGVs, thereby improving the accuracy of fault detection and the system's resource utilization.

[0007] To address the aforementioned technical problems, this invention provides a method for detecting faults in a Mecanum wheel automated guided vehicle (AGV), the method comprising the following steps: Step S1: Based on differential inclusion theory and Euler-Lagrange equations, kinematic and dynamic models of the Mecanum wheel automated guided vehicle are established sequentially, and Lur'e differential inclusion state-space equations characterizing the nonlinear characteristics, parameter uncertainties and external disturbances of the system are derived. Step S2: Design an adaptive event triggering mechanism with dynamic adjustment parameters and exponential decay term. Set triggering conditions based on system output error and adjust the data transmission triggering frequency through adaptive updating of dynamic parameters. Step S3: Construct an interval observer for estimating the upper and lower bounds of the system state. Use the discrete trigger signal output by the adaptive event triggering mechanism as the input of the interval observer to form an adaptive event triggering interval observer, thereby realizing robust interval estimation of the state of the Mecanum wheel automated guided vehicle system. Step S4: Construct a residual interval based on the upper and lower bounds of the system state estimate output by the adaptive event-triggered interval observer. By determining whether the zero value is included in the residual interval, the detection and identification of actuator faults of the Mecanum wheel automated guided vehicle can be realized.

[0008] Preferably, in step S1, the specific process of establishing the kinematic model of the Mecanum wheel automated guided vehicle is as follows: Considering the transformation relationship between the inertial coordinate system and the relative coordinate system of the Mecanum wheel automated guided vehicle, and based on the motion characteristics of the Mecanum wheel, the correlation between the wheel angular velocity and the vehicle's motion state is derived, and the kinematic equations are established as follows: ; In the formula, For the wheel radius, , , , The angular velocities of the four Mecanum wheels, , For the vehicle body in the inertial coordinate system , directional linear velocity, The yaw rate of the vehicle body. The coordinate system transformation matrix is ​​indicated by the superscript. This indicates the matrix transpose.

[0009] Preferably, in step S1, the specific process of establishing the kinematic and dynamic models of the Mecanum wheel automated guided vehicle and deriving the Lur'e differential inclusion-type state-space equations characterizing the system's nonlinear characteristics, parameter uncertainties, and external disturbance characteristics is as follows: The dynamic equations of the Mecanum wheel automated guided vehicle are established based on the Euler-Lagrange equations. Introducing parameter uncertainties, external unknown disturbances, and dry friction set-valued functions, the dynamic equations are as follows: ; In the formula, For system control input, This is due to an unknown external disturbance. For the system inertia matrix, For the uncertain terms of the inertia matrix, The matrix of Coriolis force and centrifugal force For its uncertain terms, For dry friction, For its uncertain terms, The angular velocity of the wheel. This refers to the wheel's angular acceleration; By defining the system state vector, combining the kinematic and dynamic equations, and introducing the actuator fault signal, the Lur'e differential inclusion-type state-space equation is derived as follows: ; In the formula, Let be the system state vector. for The first derivative, , , , , , The system coefficient matrix, For set-valued mapping terms, For dry friction set-valued functions that satisfy the monotonicity condition, Nonlinear terms to satisfy the global Lipschitz condition. This represents the lumped uncertainty and disturbance terms. For unknown actuator fault vectors, This is the system's real-time output.

[0010] Preferably, in step S2, the triggering condition and dynamic parameter update law of the adaptive event triggering mechanism are as follows: ; In the formula, For the first Next trigger time For the first Next trigger time For system output error, Triggering time The output discrete trigger signal, For real-time output by the system, , , , , All are preset positive numbers. To adaptively and dynamically adjust parameters, for The first derivative, For exponentially decaying terms, superscript Indicates matrix transpose. Indicates the infimum.

[0011] Preferably, in step S3, the specific process of constructing the interval observer for estimating the upper and lower bounds of the system state is as follows: The system state-space equations are equivalently transformed using linear transformations. Upper and lower bound observers for the system state are designed, and observer gain and design matrices are introduced. The discrete trigger signal output by the adaptive event-triggered mechanism is used as the input to the interval observer. The constructed interval observer dynamic equations include upper and lower bound estimation dynamics, ensuring that the true system state is always constrained between the upper and lower bound estimates. The interval observer dynamic equations are expressed as follows: ; In the formula, , The state estimates output by the interval observers are respectively The upper and lower bounds, , They are respectively , The first derivative, , , , , , This is the transformed system coefficient matrix. For the system's control input, , Nonlinear functions upper and lower boundaries, To design the matrix, , These are the upper and lower bounds of the disturbance, respectively. The observer gain matrix is... , , For system output error, , , Triggering time The output discrete trigger signal, , Set-valued mappings upper and lower boundaries, For dry friction set-valued functions that satisfy the monotonicity condition, This is the system's real-time output.

[0012] Preferably, the interval observer is designed using Metzler matrices and nonnegative matrices to ensure the positivity of the observer system, so that the initial state satisfies... At any given moment, there is ,in This is the initial state of the interval observer. , These are the lower and upper bounds of the initial state, respectively. This is the state estimate output by the interval observer. , These represent the upper and lower bounds of the state estimate output by the interval observer, respectively. At the same time, by solving the linear matrix inequality, the state estimation error of the interval observer is guaranteed to be systematically consistent and eventually bounded, thus achieving robust interval estimation of the system state.

[0013] Preferably, the interval observer needs to meet the following conditions: Condition one: It is a Metzler matrix; Condition two: It is a non-negative matrix; Condition 3: Nonlinear function Globally differentiable and Lipschitz continuous, with a constant. Make ,in , Let be the system state vector. Represents the norm; Condition 4: Disturbance Bounded, satisfied ,in , Disturbance The upper and lower bounds.

[0014] Preferably, the adaptive event triggering mechanism strictly excludes Zeno behavior by deriving a positive minimum event triggering interval; the minimum event triggering interval satisfies: ; In the formula, Minimum event trigger interval, For matrix norm, For positive integers, To preset positive numbers, For the first The next trigger time, and when The inequality always holds true, where These are preset positive numbers.

[0015] Preferably, in step S4, the residual interval is constructed based on the upper and lower bounds of the system state estimate output by the adaptive event-triggered interval observer. The method for detecting and identifying actuator faults in the Mecanum wheel automated guided vehicle by determining whether the zero value is included in the residual interval is as follows: Upper bound of state estimation based on interval observer output and the lower world Construct upper bounds for residuals respectively. and the lower bound of residuals To form the residual interval ,in: ; ; In the formula, Triggering time The output discrete trigger signal; This is the transformed system coefficient matrix; When zero is contained in the residual interval When the value is zero, the system is determined to have no actuator fault; when the zero value is not included in the residual interval... If the system detects an actuator malfunction, a fault alarm is triggered.

[0016] This invention also provides a fault detection system for an automated guided vehicle (AGV) with Mecanum wheels. This system is used to implement the aforementioned fault detection method for AGVs with Mecanum wheels, specifically including: The system modeling module is used to establish the kinematic and dynamic models of the Mecanum wheel automated guided vehicle based on differential inclusion theory and Euler-Lagrange equations, and to derive the Lur'e differential inclusion state-space equations that characterize the system's nonlinear characteristics, parameter uncertainties and external disturbances. The adaptive event triggering mechanism module is used to design an adaptive event triggering mechanism with dynamically adjustable parameters and an exponential decay term. It sets the triggering conditions based on the system output error and adjusts the data transmission triggering frequency through adaptive updates of the dynamic parameters. The adaptive event-triggered interval observer module is used to construct an interval observer for estimating the upper and lower bounds of the system state. The discrete trigger signal output by the adaptive event-triggered mechanism is used as the input of the interval observer to form an adaptive event-triggered interval observer, thereby realizing robust interval estimation of the state of the Mecanum wheel automated guided vehicle system. The fault detection and identification module is used to construct the residual interval based on the upper and lower bounds of the system state estimate output by the adaptive event-triggered interval observer. By determining whether the zero value is included in the residual interval, the module can detect and identify the actuator faults of the Mecanum wheel automated guided vehicle.

[0017] As can be seen from the above technical solutions, this invention application has the following beneficial effects: (1) Based on differential inclusion theory and Euler-Lagrange equation, this invention constructs Lur'e differential inclusion state space equation, which fully characterizes the nonlinear dynamic characteristics, parameter perturbation, external unknown disturbance and dry friction set-valued nonlinear characteristics of Mecanum wheel AGV. It solves the problem that traditional modeling methods are difficult to adapt to the complex dynamic characteristics of AGV, provides an accurate model basis for state estimation and fault detection, and greatly improves the robustness of the algorithm in complex industrial environments.

[0018] (2) This invention integrates an adaptive event triggering mechanism with dynamic adjustment parameters and exponential decay terms with an interval observer. It can adaptively adjust the data transmission frequency according to the system output error, greatly reducing invalid data transmission and redundant calculation. At the same time, it rigorously proves the minimum positive event triggering interval through theoretical derivation, completely eliminating Zeno behavior. Under the premise of ensuring the accuracy of state estimation and the reliability of fault detection, it significantly reduces the system communication and computing resource occupation, perfectly adapting to the resource-constrained characteristics of AGV vehicle-mounted embedded systems.

[0019] (3) The present invention constructs the residual interval based on the upper and lower bounds of the state output of the interval observer, and realizes the actuator fault detection through the inclusion relationship between the zero value and the residual interval. It does not require complex fault feature extraction and data preprocessing, and can quickly and reliably identify various types of actuator faults such as motor output attenuation, jamming, bias, and intermittent failure. At the same time, the interval estimation characteristics naturally shield the system uncertainty and disturbance interference, and there are no false alarms during fault-free periods. When a fault occurs, an alarm can be quickly triggered, which effectively improves the safety and stability of AGV operation and reduces the risk of production interruption and safety accidents in industrial scenarios. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly described below. Referring to the drawings will make the features and advantages of the present invention clearer. The drawings are illustrative and should not be construed as limiting the present invention in any way. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein: Figure 1 This is a flowchart of a fault detection method for an automated guided vehicle (AGV) with Mecanum wheels provided by the present invention; Figure 2 This is a structural block diagram of the Mecanum wheel automated guided vehicle fault detection method in this invention; Figure 3 This is a schematic diagram of the geometric structure of the Mecanum wheel automated guided vehicle in this invention; Figure 4 This is a graphical illustration of set-valued mapping in this invention; Figure 5 is a comparison diagram of the actual state of the Mecanum wheel automated guided vehicle system and the estimated boundary of the section observer in an embodiment of the present invention; Figure 6 is a distribution diagram of the triggering time and triggering interval of the adaptive event triggering mechanism in an embodiment of the present invention; Figure 7 is a graph showing the change in the residual range before and after the actuator failure in an embodiment of the present invention; Figure 8 This is a block diagram of a fault detection system for an automated guided vehicle (AGV) with Mecanum wheels provided by the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] This invention discloses a fault detection method and system for Mecanum wheel automated guided vehicles (AGVs), which is applicable to the condition monitoring and fault diagnosis of AGVs in high-precision, dynamic industrial environments. It aims to solve the core problems of nonlinear modeling, uncertainty handling, and resource optimization of AGVs, and significantly reduce the system's computing and communication resource consumption while ensuring the accuracy of fault detection.

[0023] Example 1: First, combine with the appendix Figure 1 and Figure 2 The overall process of the Mecanum wheel automated guided vehicle (AGV) fault detection method of the present invention is described, and the method specifically includes the following core steps: Step S1: Based on differential inclusion theory and Euler-Lagrange equations, kinematic and dynamic models of the Mecanum wheel automated guided vehicle are established sequentially, and Lur'e differential inclusion state-space equations characterizing the nonlinear characteristics, parameter uncertainties and external disturbances of the system are derived. Step S2: Design an adaptive event triggering mechanism with dynamic adjustment parameters and exponential decay term. Set triggering conditions based on system output error and adjust the data transmission triggering frequency through adaptive updating of dynamic parameters. Step S3: Construct an interval observer for estimating the upper and lower bounds of the system state. Use the discrete trigger signal output by the adaptive event triggering mechanism as the input of the interval observer to form an adaptive event triggering interval observer, thereby realizing robust interval estimation of the state of the Mecanum wheel automated guided vehicle system. Step S4: Construct a residual interval based on the upper and lower bounds of the system state estimate output by the adaptive event-triggered interval observer. By determining whether the zero value is included in the residual interval, the detection and identification of actuator faults of the Mecanum wheel automated guided vehicle can be realized.

[0024] The specific implementation methods for each of the above steps are described in complete detail below: In step S1, the state-space equations of the Mecanum wheel automated guided vehicle (AGV) are constructed. This step, based on differential inclusion and Euler-Lagrange equations, sequentially establishes the kinematic and dynamic models, ultimately deriving the Lur'e differential inclusion state-space equations. This accurately characterizes the complex dynamic characteristics of the Mecanum wheel AGV, and is specifically divided into the following sub-steps: Step S11: Establish the kinematic model of the Mecanum wheel automated guided vehicle. (Refer to Appendix) Figure 2 The diagram shows the geometry of an automated guided vehicle (AGV) with Mecanum wheels. This AGV is equipped with four 45° Mecanum wheels. First, the inertial coordinate system XOY and the vehicle's relative coordinate system are defined, with the angle between the two coordinate systems representing the vehicle's yaw angle. Considering the transformation relationship between the inertial coordinate system and the relative coordinate system, and based on the motion characteristics of the Mecanum wheel, the correlation between the wheel angular velocity and the vehicle's motion state is derived, and the kinematic equations are established as follows: ; In the formula, For the wheel radius, , , , The angular velocities of the four Mecanum wheels, , For the vehicle body in the inertial coordinate system , directional linear velocity, The yaw rate of the vehicle body. The coordinate system transformation matrix is ​​indicated by the superscript. This indicates the matrix transpose.

[0025] The specific form of the coordinate system transformation matrix is ​​defined as follows: ; In the formula, It is the longitudinal half-distance from the geometric center of the vehicle body to the center of the wheel. This is the lateral half-distance from the geometric center of the vehicle body to the center of the wheel. Simultaneously, define... Transformation matrix The pseudo-inverse matrix, and satisfying ,in It is a 3-order identity matrix.

[0026] Step S12: Establish the dynamic model of the Mecanum wheel automated guided vehicle. Based on the Euler-Lagrange equations, establish the dynamic equations of the Mecanum wheel automated guided vehicle. Simultaneously, introduce parameter uncertainties, external unknown disturbances, and dry friction set-valued functions to fully characterize the system's parameter uncertainties, external disturbances, and nonlinear friction characteristics. The dynamic equations are: ; In the formula, For system control input, This is due to an unknown external disturbance. For the system inertia matrix, For the uncertain terms of the inertia matrix, The matrix of Coriolis force and centrifugal force For its uncertain terms, For dry friction, For its uncertain terms, The angular velocity of the wheel. This refers to the wheel's angular acceleration.

[0027] The specific definitions of each core matrix and variable are as follows: Control input vector: ; Wheel angular velocity vector: ; Inertia Matrix Specific form: ; In the formula, , , The total mass of the vehicle body and These represent the moment of inertia of the wheel about its center of rotation and the moment of inertia of the Mecanum wheel automated guided vehicle about its center of rotation, respectively. dry friction The set-valued function acts on each wheel, and its specific form is: ; In the formula, This is the critical amplitude of the Coulomb dry friction torque.

[0028] The dynamic model of this invention simultaneously considers nonlinear factors such as wheel slippage, center of mass displacement, mass change, and friction, and includes unknown external disturbances. and parameter uncertainty terms , and Define the lumped uncertainty term. In practical industrial applications, the DC motor output of the Mecanum wheel automated guided vehicle (AGV) is a finite value; the motor speed and its time derivatives are all bounded. .because , and and and Related, therefore , and All satisfy the boundedness condition: , and ,in , and Since all are positive numbers, the boundedness of the lumped uncertainty term can be derived from this: .

[0029] Simultaneously define: unknown external disturbance It is bounded and satisfies ,in .

[0030] Combining the above kinematic and dynamic equations, the differential equations of the vehicle's motion state can be derived: ; make , Then the above equation can be simplified to the following form: ; ; In the formula, This is the lumped disturbance term.

[0031] Step S13: Derive the Lur'e differential inclusion-type state-space equations. Define the system state vector. Combining the above kinematic and dynamic equations, an unknown actuator fault vector is introduced. Finally, the Lur'e differential inclusion-type state-space equation is derived as follows: ; In the formula, Let be the system state vector. for The first derivative, , , , , , The system coefficient matrix, For set-valued mapping terms, For dry friction set-valued functions that satisfy the monotonicity condition, Nonlinear terms to satisfy the global Lipschitz condition. This represents the lumped uncertainty and disturbance terms. For unknown actuator fault vectors, This is the system's real-time output.

[0032] The specific form of the coefficient matrix for each system is as follows: ; ; ; ; ; ; Nonlinear terms The specific expression is: .

[0033] Actuator fault signal Defined as Let represent the unknown actuator fault vector of the Mecanum wheel automated guided vehicle, where each element corresponds to an actuator fault in a critical degree of freedom of the drive system, covering actuator fault types such as motor output attenuation, sudden jamming, bias faults, and intermittent failures, and the fault vector satisfies . , This is the upper bound of the known fault.

[0034] Set-valued mapping The specific definitions are as follows (see appendix) Figure 4 (A graphical representation of the value mapping of this set) ; set-valued function The following monotonicity condition must be met: for all All are true .

[0035] At the same time, this step makes the following basic definitions for the core variables of the system: Definition 1: The initial state boundary satisfies ,in , These are the initial states of the system. The lower and upper bounds; Definition 2: Nonlinear term For the global Lipschitz function, there exists a constant. Such that for any system state vector , All satisfy: ; Definition 3: Let and Representing nonlinear functions respectively Given lower and upper bounds, perturbation The known lower and upper bounds are denoted as . and Then the nonlinear function and disturbance The following boundedness conditions must be met: .

[0036] In step S2, an adaptive event triggering mechanism is designed. This step designs an adaptive event triggering mechanism with dynamically adjusted parameters and an exponential decay term to optimize data transmission timing, save network and computing resources while ensuring system performance, and strictly eliminate Zeno behavior. The specific implementation is as follows: First, define the trigger time sequence. , ,in The initial trigger time, For the first Next trigger time For the first Trigger time; define system output error ,in For triggering time The output discrete trigger signal, This is the system's real-time output.

[0037] The triggering condition and dynamic parameter update law of the adaptive event triggering mechanism designed in this invention are as follows: ; In the formula, Indicates the infimum, , , , , All are preset positive numbers. To adaptively and dynamically adjust parameters, for The first derivative, For exponentially decaying terms, superscript This indicates the matrix transpose.

[0038] The adaptive event triggering mechanism works as follows: data transmission is triggered only when either event 1 or event 2 is met, updating the observer's input signal. Otherwise, the observer uses the signal from the previous trigger moment, eliminating the need for continuous real-time data transmission. This is achieved through adaptive dynamic parameter adjustment. The system provides real-time updates and dynamically adjusts the trigger frequency based on system output errors: when the system output error is small, the trigger frequency is reduced to minimize invalid data transmission; when the system output error is large, the trigger frequency is increased to ensure the accuracy of state estimation and fault detection. Simultaneously, an exponential decay term is introduced. This can further optimize the triggering frequency during steady-state operation of the system and reduce resource consumption.

[0039] In step S3, an adaptive event-triggered interval observer is constructed.

[0040] This step constructs an interval observer for estimating the upper and lower bounds of the system state. The discrete trigger signal output by the adaptive event-triggered mechanism is used as the input to the interval observer, forming an adaptive event-triggered interval observer. This enables robust interval estimation of the state of the Mecanum wheel automated guided vehicle system. The specific implementation method is as follows: Step S31: System Linear Transformation. To facilitate the design of the interval observer, the system state-space equations are first equivalently transformed through linear transformation. Let... For linear transformation, where Let be the invertible matrix to be designed. Then the transformed system state-space equation is: ; In the formula, the transformed coefficient matrices are: , , , , , .

[0041] Step S32: Design the dynamic equations for the upper and lower bounds of the interval observer. After the above linear transformation, design the upper and lower bound observers for the system state, and introduce the observer gain matrix. and design matrix Discrete trigger signals output by the adaptive event triggering mechanism As input to the interval observer, the constructed interval observer dynamic equation includes upper bound estimation dynamics and lower bound estimation dynamics, ensuring that the true state of the system is always constrained between the upper bound estimate and the lower bound estimate. The interval observer dynamic equation is expressed as: ; In the formula, , The state estimates output by the interval observers are respectively The upper and lower bounds, , They are respectively , The first derivative, , , , , , This is the transformed system coefficient matrix. For the system's control input, , Nonlinear functions upper and lower boundaries, To design the matrix, , These are the upper and lower bounds of the disturbance, respectively. The observer gain matrix is... , , For system output error, , , For triggering time The output discrete trigger signal, , Set-valued mappings upper and lower boundaries, For dry friction set-valued functions that satisfy the monotonicity condition, This is the system's real-time output.

[0042] Step S33: Design Constraints and Performance Guarantees of the Interval Observer. The interval observer designed in this invention guarantees the positivity of the observer system through the design of Metzler matrices and non-negative matrices, and simultaneously guarantees that the state estimation error system is uniformly and eventually bounded by solving linear matrix inequalities. The specific design constraints and performance proof are as follows: 1. Positive guarantee condition for interval observers To ensure that the initial state is satisfied At any given moment, there is That is, the true state of the system is always surrounded by the estimated interval, and the interval observer must satisfy the following core conditions: Condition one: It is a Metzler matrix (i.e., all off-diagonal elements of the matrix are non-negative). Condition two: It is a non-negative matrix (i.e., all elements of the matrix are non-negative); Condition 3: Nonlinear function Globally differentiable and Lipschitz continuous, with a constant. Make ,in , Let be the system state vector. Represents the norm; Condition 4: Disturbance Bounded, satisfied ,in , Disturbance The upper and lower boundaries.

[0043] 2. Proof of stability and bounded error of the interval observer First, we present the two core lemmas required for this proof: Lemma 1: If and ,but ,in , , , .

[0044] Lemma 2: For the system If nonlinear function and If the matrix is ​​Metzler, then the system is a positive system, meaning that at any time... .

[0045] Based on the above lemma, the following core theorem is given: Theorem 1: When no faults occur ( If definitions 1 to 3 above hold, It is a Metzler matrix and If the matrix is ​​nonnegative, then the observer boundary satisfies .

[0046] Proof: By definition 2 and definition 3, combined with Lemma 1, we can obtain , , , ,and , ; Due to the monotonicity of set-valued mappings, if If it is a non-negative matrix, then , ,Right now , Therefore , ; By definition 1 and lemma 2, if Since the error system is a Metzler matrix, it is a positive system. , ,Right now Theorem 1 is proved.

[0047] Definition of estimation error Lower estimation error ,as well as , Combining the transformed system equations with the dynamic equations of the interval observer, the error system equations can be obtained as follows: ; Further define the augmented error vector , Then the error system can be rewritten as: ; In the formula, ; ; ; ; ; ; ; .

[0048] Theorem 2: When no fault occurs ( ), given constant and If a positive definite matrix exists The following linear matrix inequalities must be satisfied: ; in, , , And matching conditions: If this holds true, then the error system is uniformly bounded.

[0049] Proof: Consider candidate Lyapunov functions Calculate its derivative: .

[0050] From the matching conditions, we can obtain: ; According to Cauchy's matrix inequality: ; ; .

[0051] then: ; in: ; .

[0052] like If the error system is uniformly bounded, then Theorem 2 is proved.

[0053] Theorem 3: The minimum event triggering interval of the adaptive event triggering mechanism designed in this invention is: ; In the formula, Minimum event trigger interval, For matrix norm, It is a positive number, and , To preset positive numbers, For the first The next trigger time; and when When the minimum event trigger interval is always positive, this inequality always holds true, which theoretically strictly excludes Zeno's behavior and ensures the feasibility of this mechanism in actual industrial systems.

[0054] Proof: By It can be known that: .

[0055] And because We can obtain: .

[0056] From the above two equations, we can obtain: .

[0057] right Differentiating, we get: .

[0058] because Therefore, we can conclude that: ; From the above formula, we can obtain: ; .

[0059] According to the above formula, we have and Multiply both sides of the above equation by have to: .

[0060] like Its establishment means: .

[0061] Taking the logarithm of both sides of the above equation, we get: ; Finally, the formula for the minimum triggering interval was derived, and Theorem 3 was proved.

[0062] In step S4, actuator fault detection and identification are performed based on the residual interval. This step constructs the residual interval based on the upper and lower bounds of the system state estimate output by the adaptive event-triggered interval observer. By determining whether the zero value is contained within the residual interval, the actuator faults of the Mecanum wheel automated guided vehicle are detected and identified. The specific implementation method is as follows: Step S41: Residual Interval Construction. Upper bound of the state estimate based on the interval observer output. and the lower world Construct upper bounds for residuals respectively. and the lower bound of residuals To form the residual interval ,in: ; ; In the formula, For triggering time The output discrete trigger signal; This is the transformed system coefficient matrix; Step S42: Fault Judgment Rules. Based on the above residual intervals, the following fault judgment rules are formulated: When zero is contained in the residual interval Inner time, that is If the system is determined to have no actuator faults, then... ; When zero is not contained in the residual interval Inner time, that is The system is determined to have an actuator failure. This immediately triggers a fault alarm.

[0063] This fault detection method does not require complex fault feature extraction. It can quickly identify faults simply by the inclusion relationship between the residual interval and the zero value. It can reliably respond to various actuator faults such as motor output attenuation, sudden jamming, bias faults and intermittent failures, and greatly improve the timeliness and reliability of fault diagnosis.

[0064] To further illustrate the specific implementation methods and technical effects of the present invention, this embodiment verifies the method of the present invention through numerical simulation. The specific implementation process is as follows: 1. System core parameter settings See attached document Figure 3 The geometry of the Mecanum wheel automated guided vehicle, setting key parameters: wheel radius Vehicle body structural parameters , Wheel rotational inertia The moment of inertia of the Mecanum wheel automated guided vehicle about its own center of rotation Total mass of vehicle body Critical amplitude of Coulomb dry friction torque .

[0065] Define state vector Choose the linear transformation matrix for: ; Construct the Lur'e differential inclusion-type state-space equations, where the unknown perturbation and nonlinear terms are defined as follows: , ; 2. Adaptive event triggering mechanism parameter settings Configure adaptive event triggering mechanism parameters: , , , , Initial values ​​of dynamic parameters The trigger timing determination rule is as follows: when event 1 or event 2 is met, data transmission is triggered, and the observer input signal is updated. Otherwise, the observer will use the previous trigger signal.

[0066] 3. Parameter settings for the interval observer Set the initial state of the system and the initial boundary of the observer: ; ; .

[0067] The observer gain matrix is ​​obtained by solving the linear matrix inequalities. and design matrix for: , ; After verification, The Metzler matrix, It is a non-negative matrix that satisfies the positive design conditions for interval observers.

[0068] 4. Actuator fault setting The actuator fault vector is set as a piecewise time-varying signal to simulate different types of actuator faults: ; in, Simulate actuator bias fault during the time period. Simulates actuator output attenuation fault during a specific time period. The time period simulates an actuator jamming fault, while the other time periods are fault-free.

[0069] 5. Simulation Results and Verification The simulation results of this embodiment are as follows: (1) Verification of the state estimation effect: Appendix Figure 5 The comparison between the position and velocity state of the Mecanum wheel automated guided vehicle and the estimated boundary of the observer is shown. The solid line represents the original state of the system, and the dashed line represents the upper and lower bounds of the observer's estimation. It can be seen that the true state of the system is always completely surrounded by the estimated interval, which verifies the effectiveness of the state estimation of the interval observer. (2) Verification of resource optimization effect: Appendix Figure 6 The event triggering time and triggering interval are shown. It can be seen that the triggering interval is strictly positive, and there are no infinite triggers within a finite time, which verifies the successful elimination of Zeno's behavior. At the same time, compared with the traditional time-triggered mechanism, the adaptive event triggering mechanism of this invention significantly reduces the number of data transmissions, reduces the number of triggers, and greatly saves communication and computing resources. (3) Verification of fault detection effectiveness: (See attached document) Figure 7This shows the changes in the residual range before and after the fault occurred, during the fault-free period ( , , , ), residual interval The system is considered fault-free if it always contains a zero value; during the fault period ( , , When the residual interval no longer contains zero values, a fault alarm is immediately triggered. The fault detection accuracy is high and the fault response delay is small, which verifies the accuracy and timeliness of the fault detection of this invention.

[0070] Example 2: like Figure 8 As shown, this invention provides a fault detection system for an automated guided vehicle (AGV) with Mecanum wheels. This system is used to implement the fault detection method for an AGV with Mecanum wheels described in Embodiment 1 above, and specifically includes: The system modeling module is used to establish the kinematic and dynamic models of the Mecanum wheel automated guided vehicle based on differential inclusion theory and Euler-Lagrange equations, and to derive the Lur'e differential inclusion state-space equations that characterize the system's nonlinear characteristics, parameter uncertainties and external disturbances. The adaptive event triggering mechanism module is used to design an adaptive event triggering mechanism with dynamically adjustable parameters and an exponential decay term. It sets the triggering conditions based on the system output error and adjusts the data transmission triggering frequency through adaptive updates of the dynamic parameters. The adaptive event-triggered interval observer module is used to construct an interval observer for estimating the upper and lower bounds of the system state. The discrete trigger signal output by the adaptive event-triggered mechanism is used as the input of the interval observer to form an adaptive event-triggered interval observer, thereby realizing robust interval estimation of the state of the Mecanum wheel automated guided vehicle system. The fault detection and identification module is used to construct the residual interval based on the upper and lower bounds of the system state estimate output by the adaptive event-triggered interval observer. By determining whether the zero value is included in the residual interval, the module can detect and identify the actuator faults of the Mecanum wheel automated guided vehicle.

[0071] Each module of the above system corresponds to each step of the above fault detection method, and its specific implementation method is completely consistent with the specific implementation method of the method, so it will not be repeated here.

[0072] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0073] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0074] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The functions specified in one or more boxes. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0075] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for fault detection in a Mecanum wheel automated guided vehicle, characterized in that, Includes the following steps: Step S1: Based on differential inclusion theory and Euler-Lagrange equations, the kinematic and dynamic models of the Mecanum wheel automated guided vehicle are established sequentially, and the Lur'e differential inclusion state-space equations characterizing the system's nonlinear characteristics, parameter uncertainties, and external disturbance characteristics are derived; the specific process of establishing the kinematic model of the Mecanum wheel automated guided vehicle is as follows: Considering the transformation relationship between the inertial coordinate system and the relative coordinate system of the Mecanum wheel automated guided vehicle, and based on the motion characteristics of the Mecanum wheel, the correlation between the wheel angular velocity and the vehicle's motion state is derived, and the kinematic equations are established as follows: ； In the formula, For the wheel radius, , , , The angular velocities of the four Mecanum wheels, , For the vehicle body in the inertial coordinate system , directional linear velocity, The yaw rate of the vehicle body. The coordinate system transformation matrix is ​​indicated by the superscript. Indicates matrix transpose; The specific process of establishing the kinematic and dynamic models of the Mecanum wheel automated guided vehicle and deriving the Lur'e differential inclusion-type state-space equations characterizing the system's nonlinear characteristics, parameter uncertainties, and external disturbances is as follows: The dynamic equations of the Mecanum wheel automated guided vehicle are established based on the Euler-Lagrange equations. Introducing parameter uncertainties, external unknown disturbances, and dry friction set-valued functions, the dynamic equations are as follows: ； In the formula, For system control input, This is due to an unknown external disturbance. For the system inertia matrix, For the uncertain terms of the inertia matrix, The matrix of Coriolis force and centrifugal force For its uncertain terms, For dry friction, For its uncertain terms, The angular velocity of the wheel. This refers to the wheel's angular acceleration; By defining the system state vector, combining the kinematic and dynamic equations, and introducing the actuator fault signal, the Lur'e differential inclusion-type state-space equation is derived as follows: ； In the formula, Let be the system state vector. for The first derivative, , , , , , The system coefficient matrix, For set-valued mapping terms, For dry friction set-valued functions that satisfy the monotonicity condition, Nonlinear terms to satisfy the global Lipschitz condition. This represents the lumped uncertainty and disturbance terms. For unknown actuator fault vectors, For real-time output by the system; Step S2: Design an adaptive event triggering mechanism with dynamic adjustment parameters and exponential decay term. Set triggering conditions based on system output error and adjust the data transmission triggering frequency through adaptive updating of dynamic parameters. Step S3: Construct an interval observer for estimating the upper and lower bounds of the system state. Use the discrete trigger signal output by the adaptive event triggering mechanism as the input of the interval observer to form an adaptive event triggering interval observer, thereby realizing robust interval estimation of the state of the Mecanum wheel automated guided vehicle system. Step S4: Construct a residual interval based on the upper and lower bounds of the system state estimate output by the adaptive event-triggered interval observer. By determining whether the zero value is included in the residual interval, the detection and identification of actuator faults of the Mecanum wheel automated guided vehicle can be realized.

2. The method for fault detection of an automated guided vehicle (AGV) with Mecanum wheels according to claim 1, characterized in that, In step S2, the triggering condition and dynamic parameter update law of the adaptive event triggering mechanism are as follows: ； In the formula, For the first Next trigger time For the first Next trigger time For system output error, Triggering time The output discrete trigger signal, For real-time output by the system, , , , , All are preset positive numbers. To adaptively and dynamically adjust parameters, for The first derivative, For exponentially decaying terms, superscript Indicates matrix transpose. Indicates the infimum.

3. The method for fault detection of an automated guided vehicle with Mecanum wheels according to claim 1, characterized in that, In step S3, the specific process of constructing the interval observer for estimating the upper and lower bounds of the system state is as follows: The system state-space equations are equivalently transformed using linear transformations. Upper and lower bound observers for the system state are designed, and observer gain and design matrices are introduced. The discrete trigger signal output by the adaptive event-triggered mechanism is used as the input to the interval observer. The constructed interval observer dynamic equations include upper and lower bound estimation dynamics, ensuring that the true system state is always constrained between the upper and lower bound estimates. The interval observer dynamic equations are expressed as follows: ； In the formula, , The state estimates output by the interval observers are respectively The upper and lower bounds, , They are respectively , The first derivative, , , , , , This is the transformed system coefficient matrix. For the system's control input, , Nonlinear functions upper and lower boundaries, To design the matrix, , These are the upper and lower bounds of the disturbance, respectively. The observer gain matrix is... , , For system output error, , , Triggering time The output discrete trigger signal, , Set-valued mappings upper and lower boundaries, For dry friction set-valued functions that satisfy the monotonicity condition, This is the system's real-time output.

4. The method for fault detection of an automated guided vehicle with Mecanum wheels according to claim 3, characterized in that, The interval observer is designed using Metzler matrices and nonnegative matrices to ensure the positivity of the observer system, so that the initial state satisfies... At any given moment, there is ,in This is the initial state of the interval observer. , These are the lower and upper bounds of the initial state, respectively. This is the state estimate output by the interval observer. , These represent the upper and lower bounds of the state estimate output by the interval observer, respectively. At the same time, by solving the linear matrix inequality, the state estimation error of the interval observer is guaranteed to be systematically consistent and eventually bounded, thus achieving robust interval estimation of the system state.

5. The method for fault detection of an automated guided vehicle with Mecanum wheels according to claim 4, characterized in that, The interval observer must meet the following conditions: Condition one: It is a Metzler matrix; Condition two: It is a non-negative matrix; Condition 3: Nonlinear function Globally differentiable and Lipschitz continuous, with a constant. Make ,in , Let be the system state vector. Represents the norm; Condition 4: Disturbance Bounded, satisfied ,in , Disturbance The upper and lower bounds.

6. The method for fault detection of an automated guided vehicle with Mecanum wheels according to claim 3, characterized in that, The adaptive event triggering mechanism strictly excludes Zeno behavior by deriving a positive minimum event triggering interval; the minimum event triggering interval satisfies: ； In the formula, Minimum event trigger interval, For matrix norm, For positive integers, To preset positive numbers, For the first The next trigger time, and when The inequality always holds true, where These are preset positive numbers.

7. The method for fault detection of an automated guided vehicle with Mecanum wheels according to claim 1, characterized in that, In step S4, a residual interval is constructed based on the upper and lower bounds of the system state estimate output by the adaptive event-triggered interval observer. The method for detecting and identifying actuator faults in the Mecanum wheel automated guided vehicle is as follows: Upper bound of state estimation based on interval observer output and the lower world Construct upper bounds for residuals respectively. and the lower bound of residuals To form the residual interval ,in: ； ； In the formula, Triggering time The output discrete trigger signal; This is the transformed system coefficient matrix; When zero is contained in the residual interval When the value is zero, the system is determined to have no actuator fault; when the zero value is not included in the residual interval... If the system detects an actuator malfunction, a fault alarm is triggered.

8. A fault detection system for a Mecanum wheel automated guided vehicle, characterized in that, The system is used to implement the Mecanum wheel automated guided vehicle fault detection method according to any one of claims 1 to 7, specifically including: The system modeling module is used to establish the kinematic and dynamic models of the Mecanum wheel automated guided vehicle based on differential inclusion theory and Euler-Lagrange equations, and to derive the Lur'e differential inclusion state-space equations that characterize the system's nonlinear characteristics, parameter uncertainties and external disturbances. The adaptive event triggering mechanism module is used to design an adaptive event triggering mechanism with dynamically adjustable parameters and an exponential decay term. It sets the triggering conditions based on the system output error and adjusts the data transmission triggering frequency through adaptive updates of the dynamic parameters. The adaptive event-triggered interval observer module is used to construct an interval observer for estimating the upper and lower bounds of the system state. The discrete trigger signal output by the adaptive event-triggered mechanism is used as the input of the interval observer to form an adaptive event-triggered interval observer, thereby realizing robust interval estimation of the state of the Mecanum wheel automated guided vehicle system. The fault detection and identification module is used to construct the residual interval based on the upper and lower bounds of the system state estimate output by the adaptive event-triggered interval observer. By determining whether the zero value is included in the residual interval, the module can detect and identify the actuator faults of the Mecanum wheel automated guided vehicle.

Images