A multi-source coupled skateboard chassis system and its multi-level fault diagnosis method

Abstract

This invention discloses a multi-source coupled skateboard chassis system and its multi-level fault diagnosis method. The system includes: a drive-by-wire subsystem, an electro-hydraulic composite steering-by-wire subsystem, an electrical composite braking-by-wire subsystem, a multi-level global status observation module, a multi-level fault diagnosis module, an architecture-level fault severity identification module, and a fault alarm module. This invention performs multi-level fault diagnosis on the skateboard chassis system, solving the problem of mutual interference and inaccurate diagnosis when cross-subsystem faults occur in the chassis. The fault diagnosis of each level of the chassis is independent and coordinated, significantly improving the overall fault diagnosis efficiency of the chassis.

Description

Technical Field

[0001] This invention belongs to the technical field of skateboard chassis systems, specifically relating to a multi-source coupled skateboard chassis system and its multi-level fault diagnosis method. Background Technology

[0002] With the continued and stable high-quality development of the automotive industry, the trends of vehicle electrification and intelligentization are becoming increasingly apparent. As a result, skateboard chassis, characterized by high-level modularity, platformization, and drive-by-wire, have emerged as a significant future trend in automotive chassis development. Given the highly integrated nature of skateboard chassis, designing a highly integrated and coordinated fault diagnosis method remains a crucial research topic for designers.

[0003] Existing fault diagnosis technology research mainly focuses on single subsystem and single-level fault diagnosis research due to the deep mechanical coupling of traditional automotive chassis. However, with the emergence of highly modular skateboard chassis, the mechanical decoupling of the chassis has been completed, providing hardware support for multi-level fault diagnosis of the sensing layer, control layer, execution layer, and architecture layer from the perspective of high chassis integration. Furthermore, existing single-subsystem and single-level fault diagnosis is prone to interference between subsystems and inaccurate diagnosis when multiple levels of faults occur simultaneously, resulting in decreased fault diagnosis efficiency and difficulty in meeting the ASIL-D level of multiple safety requirements.

[0004] Currently, research on skateboard chassis is relatively limited, and most chassis-related fault diagnosis studies focus on single systems and single levels. For example, Chinese invention patent application CN201510516435.6 discloses a fault diagnosis and fault-tolerant control method for sensors in an automotive electronic stability control system. It determines whether each sensor has malfunctioned by comparing the residuals and thresholds of target parameters of each sensor in the stability control system. This invention patent only performs fault diagnosis from the sensor layer perspective, and may not be effective in diagnosing faults when multiple levels of faults occur simultaneously. Similarly, Chinese invention patent application CN202110818429.1 discloses a steer-by-wire system and fault diagnosis method. Based on a bidirectional long short-term memory network, it estimates the state of a highly complex, nonlinear, and strongly coupled system. It uses a two-layer state machine to cumulatively diagnose the range and signal difference of the front wheel steering angle signal. This invention patent only performs fault diagnosis from the sensor layer perspective of the steering subsystem, ignoring the influence of other levels of the steering subsystem. Therefore, it is essential to accurately establish a chassis system model and adopt appropriate control architecture and algorithms to perform multi-level integrated fault diagnosis of the chassis system. Summary of the Invention

[0005] In view of the shortcomings of the prior art, the purpose of this invention is to provide a multi-source coupled skateboard chassis system and its multi-level fault diagnosis method, so as to overcome the lack of a highly integrated and coordinated multi-level fault diagnosis method in the prior art to accurately and efficiently complete the fault diagnosis of the skateboard chassis, and avoid the problem of diagnostic interference coupling when multiple levels of faults occur at the same time.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] The present invention provides a multi-source coupled skateboard chassis system, comprising: a drive-by-wire subsystem, an electro-hydraulic composite steering-by-wire subsystem, an electrical composite braking-by-wire subsystem, a multi-level global status observation module, a multi-level fault diagnosis module, an architecture-level fault severity identification module, and a fault alarm module; wherein...

[0008] The drive-by-wire subsystem includes: an accelerator pedal, an accelerator pedal position sensor, a left front wheel hub motor, a right front wheel hub motor, a left rear wheel hub motor, a right rear wheel hub motor, a left front wheel speed sensor, a right front wheel speed sensor, a left rear wheel speed sensor, a right rear wheel speed sensor, a left front wheel, a right front wheel, a left rear wheel, a right rear wheel, and a drive control module; the accelerator pedal position sensor is fixedly mounted on the accelerator pedal and is used to send accelerator pedal position signals to the drive control module; the left front wheel hub motor and the left front wheel speed sensor are installed inside the left front wheel; the right front wheel hub motor and the right front wheel speed sensor... The sensor is installed in the right front wheel; the left rear wheel hub motor and the left rear wheel speed sensor are installed in the left rear wheel; the right rear wheel hub motor and the right rear wheel speed sensor are installed in the right rear wheel; the rotational motion output by the left front wheel hub motor, the right front wheel hub motor, the left rear wheel hub motor, and the right rear wheel hub motor is converted into rotational motion of the left front wheel, the right front wheel, the left rear wheel, and the right rear wheel, respectively, driving the vehicle; the input terminal of the drive control module is connected to the left front wheel speed sensor, the right front wheel speed sensor, the left rear wheel speed sensor, the right rear wheel speed sensor, and the accelerator pedal position sensor through the vehicle communication line;

[0009] The electro-hydraulic composite steer-by-wire subsystem includes: a steering wheel module, a steering motor module, a hydraulic module, a mechanical transmission module, and a steering control module;

[0010] The steering wheel module includes: a steering wheel, a first steering column, a steering wheel angle sensor, and a steering wheel torque sensor; the steering wheel is fixedly connected to one end of the first steering column, and the other end of the first steering column is connected to a clutch; the steering wheel angle sensor and the steering wheel torque sensor are both fixedly connected to the first steering column, respectively collecting the steering wheel angle and torque, and sending the collected steering wheel angle signal and steering wheel torque signal to the steering control module;

[0011] The steering motor module includes: a steering actuator motor, a steering actuator motor speed sensor, a first reduction mechanism, and a second steering column; the output end of the steering actuator motor is connected to the steering actuator motor speed sensor and transmits electric power assistance to the second steering column through the first reduction mechanism to provide power assistance to the steering system;

[0012] The hydraulic module includes: a hydraulic pump, a hydraulic pump drive motor, a second reduction mechanism, an oil tank, a servo proportional solenoid valve, oil pipes, a first pressure sensor, and a second pressure sensor. The hydraulic pump drive motor is connected to the hydraulic pump through the second reduction mechanism, pumping hydraulic oil from the oil tank into the recirculating ball steering gear of the mechanical transmission module through the servo proportional solenoid valve, creating an oil pressure differential in the recirculating ball steering gear, which provides power assistance to the steering system under the action of the oil pressure differential. The first pressure sensor is installed on the oil pipe of the recirculating ball steering gear inlet, and the second pressure sensor is installed on the oil pipe of the recirculating ball steering gear outlet, respectively used to detect the hydraulic power assistance on both sides of the recirculating ball steering gear and send hydraulic power assistance signals.

[0013] The mechanical transmission module includes: a recirculating ball steering gear, a steering rocker arm, a steering tie rod, a left steering knuckle, a left steering knuckle arm, a left steering trapezoidal arm, a steering tie rod, a right steering trapezoidal arm, a right steering knuckle arm, and a right steering knuckle. The recirculating ball steering gear consists of two stages of transmission pairs: the first stage is a screw and nut transmission pair; the second stage is a rack and pinion transmission pair. One end of the second steering column is connected to the clutch, and the other end is connected to the input end of the recirculating ball steering gear through the screw and nut transmission pair. The output end of the recirculating ball steering gear is connected to one end of the steering rocker arm through the rack and pinion transmission pair. The other end of the steering rocker arm is connected to the left steering knuckle arm through the steering tie rod, driving the left steering knuckle and the left front wheel to deflect. The left steering knuckle arm is connected to one end of the steering tie rod through the left steering trapezoidal arm. The other end of the steering tie rod is connected to the right steering trapezoidal arm, and the right steering trapezoidal arm is connected to the right steering knuckle through the right steering knuckle arm, driving the right front wheel to steer.

[0014] The input terminal of the steering control module is electrically connected to the steering wheel angle sensor, the steering wheel torque sensor, and the steering actuator motor speed sensor, respectively, and sends control signals to the steering motor module and the hydraulic module to coordinate the steering of the front wheels.

[0015] The electrical composite line control braking subsystem includes: a brake pedal, a brake pedal position sensor, a motor regenerative braking module, a pneumatic braking module, and a brake control module;

[0016] The regenerative braking module includes: a left front wheel brake motor, a left front wheel brake motor speed sensor, a left front wheel reduction mechanism, a left front wheel ball screw, a left front wheel brake cylinder, a right front wheel brake motor, a right front wheel brake motor speed sensor, a right front wheel reduction mechanism, a right front wheel ball screw, a right front wheel brake cylinder, a left rear wheel brake motor, a left rear wheel brake motor speed sensor, a left rear wheel reduction mechanism, a left rear wheel ball screw, a left rear wheel brake cylinder, a right rear wheel brake motor, a right rear wheel brake motor speed sensor, a right rear wheel reduction mechanism, a right rear wheel ball screw, and a right rear wheel brake cylinder;

[0017] The brake pedal position sensor is fixedly mounted on the brake pedal and is used to send brake pedal position signals to the brake control module. The output of the left front wheel brake motor is connected to one end of the left front wheel ball screw via a left front wheel brake motor speed sensor and a left front wheel reduction mechanism. The other end of the left front wheel ball screw is connected to the left front wheel brake cylinder. The output of the right front wheel brake motor is connected to one end of the right front wheel ball screw via a right front wheel brake motor speed sensor and a right front wheel reduction mechanism. The other end of the right front wheel ball screw is connected to the right front wheel brake cylinder. The output of the left rear wheel brake motor is connected to one end of the left rear wheel ball screw via a left rear wheel brake motor wheel speed sensor and a left rear wheel reduction mechanism. The other end of the left rear wheel ball screw is connected to the left rear wheel brake cylinder. The output of the right rear wheel brake motor is connected to the right rear wheel brake cylinder via a right rear wheel reduction mechanism. The motor speed sensor and the right rear wheel reduction mechanism are connected to one end of the right rear wheel ball screw, and the other end of the right rear wheel ball screw is connected to the right rear wheel brake cylinder. The rotational motion of the left front wheel brake motor, right front wheel brake motor, left rear wheel brake motor, and right rear wheel brake motor is sequentially converted into displacement motion of the screw ends of the left front wheel ball screw, right front wheel ball screw, left rear wheel ball screw, and right rear wheel ball screw through the left front wheel reduction mechanism and left front wheel ball screw, right front wheel reduction mechanism and right rear wheel ball screw, and right rear wheel reduction mechanism and right rear wheel ball screw, respectively. This displacement motion acts on the left front wheel brake cylinder, right front wheel brake cylinder, left rear wheel brake cylinder, and right rear wheel brake cylinder to generate braking torque and complete the vehicle braking operation.

[0018] The pneumatic braking module includes an air compressor, an air tank, and a brake valve, which are connected in sequence. The air compressor compresses air into the air tank, and then inflates the left front wheel brake cylinder, right front wheel brake cylinder, left rear wheel brake cylinder, and right rear wheel brake cylinder through the brake valve to achieve braking.

[0019] The output of the braking control module is connected to the brake pedal position sensor, the left front wheel speed sensor, the right front wheel speed sensor, the left rear wheel speed sensor, and the right rear wheel speed sensor, respectively, and is used to adjust the braking force of the motor regenerative braking force module and the air pressure braking module according to the torque distribution strategy of the electric composite line control braking system.

[0020] The multi-level global state observation module includes: a sensing layer state observation module, a control layer state observation module, and an execution layer state observation module;

[0021] The sensor layer state observation module is connected to the accelerator pedal position sensor, left front wheel speed sensor, right front wheel speed sensor, left rear wheel speed sensor, right rear wheel speed sensor, steering wheel angle sensor, steering motor speed sensor, brake pedal position sensor, left front wheel brake motor speed sensor, right front wheel brake motor speed sensor, left rear wheel brake motor speed sensor, and right rear wheel brake motor speed sensor via vehicle communication lines. This is used to acquire the signals collected by each sensor in real time. At the same time, the sensor layer state observation module acquires the three Hall signals from each sensor and calculates the signals estimated by the algorithm for each sensor.

[0022] The control layer status observation module is used to acquire the current transfer ratio signals of the drive control module, the steering control module, the braking control module, and the communication status signals of each node on the CAN bus.

[0023] The execution layer state observation module is used to acquire the current signals and bearing vibration frequency signals of the two axles of the left front wheel hub motor, the two axles of the right front wheel hub motor, the two axles of the left rear wheel hub motor, the two axles of the right rear wheel hub motor, the two axles of the steering actuator motor, the two axles of the left front wheel brake motor, the two axles of the right front wheel brake motor, the two axles of the left rear wheel brake motor, and the two axles of the right rear wheel brake motor.

[0024] The input of the multi-level fault diagnosis module is connected to the output of the multi-level global state observation module. It is used to acquire real-time state signals of the sensing layer, control layer, and execution layer, perform multi-level fault state diagnosis, and output real-time fault state signals of the sensing layer, control layer, and execution layer.

[0025] The input end of the architecture layer fault degree identification module is connected to the output end of the multi-level fault diagnosis module. It is used to acquire real-time fault status signals of the sensing layer, control layer, and execution layer, identify the fault degree of each subsystem, and establish an architecture layer fault degree database to obtain the fault degree of each subsystem under different multi-level states in the initial preset.

[0026] The input of the fault alarm module is connected to the output of the architecture layer fault severity identification module. It is used to obtain the architecture layer fault severity signal and output fault alarms for the fault subsystem and each level of fault within the subsystem.

[0027] The present invention provides a multi-level fault diagnosis method for a multi-source coupled skateboard chassis system, based on the aforementioned system, comprising the following steps:

[0028] (1) Establish a multi-level global state observation method: Establish a multi-level global state observation module consisting of a sensing layer state observation module, a control layer state observation module, and an execution layer state observation module to obtain relevant real-time state parameters and relevant estimated state parameters of the sensing layer, control layer, and execution layer.

[0029] (2) Establish a multi-level fault diagnosis method based on residual theory: Establish a multi-level fault diagnosis module, calculate the multi-level fault feature matrix according to the state parameters of each level obtained in step (1), diagnose the fault signals of each level, and determine the failure parameters and fault level.

[0030] (3) Establish a fault degree identification method for the architecture layer: Define the fault degree of the architecture layer, and combine the multi-level fault feature matrix obtained in step (2) to establish a fault degree identification method for the architecture layer based on the gradient spectrum entropy of the higher-order differential mathematical morphology, and identify the fault degree of the drive subsystem, steering subsystem and braking subsystem.

[0031] (4) Based on the fault feature matrix and fault degree of the drive subsystem, steering subsystem and braking subsystem obtained in step (3), set the reference fault degree of the architecture layer. When the fault degree of a certain subsystem is greater than the reference fault degree of the architecture layer, the corresponding subsystem and the faults at each level in the subsystem are output through the fault alarm module.

[0032] Furthermore, the multi-level global state observation method in step (1) is as follows:

[0033] (11) Establish a sensor layer state observation module:

[0034] Set the real-time state vector X of the sensing layer s With the estimated state vector

[0035] X s=[X s_ωd X s_ωb x s_θs x s_yd x s_yb x s_ωs ] T

[0036]

[0037] Among them, X s_ωd , Let X represent the measured state vector and the estimated state vector of the wheel speed sensor, respectively. s_ωb , Let x represent the measured state vector and the estimated state vector of the wheel brake motor speed sensor, respectively. s_θs , x represents the measured state parameters and the estimated state parameters of the steering wheel angle sensor, respectively. s_yd , x represents the measured state parameters and the estimated state parameters of the accelerator pedal position sensor, respectively. s_yb , x represents the measured state parameters and the estimated state parameters of the brake pedal position sensor, respectively. s_ωs , These represent the measured state parameters of the steering actuator motor speed sensor and the estimated state parameters of the steering actuator motor speed sensor, respectively.

[0038] X s_ωd =[ω s_ωd11 ω s_ωd12 ω s_ωd21 ω s_ωd22 ]

[0039]

[0040] X s_ωb =[ω s_ωb11 ω s_ωb12 ω s_ωb21 ω s_ωb22 ]

[0041]

[0042] Where, ω s_ωdij This indicates the wheel speed measured by the ij wheel speed sensor. This indicates that the wheel speed is estimated by the ij wheel speed sensor, ω s_ωbij This indicates the actual measured speed of the ij wheel brake motor by the speed sensor. The value represents the estimated rotational speed of the wheel brake motor speed sensor. ij = 11 indicates front left; ij = 12 indicates front right; ij = 21 indicates rear left; and ij = 22 indicates rear right.

[0043] An algorithm for estimating the state parameters of the sensing layer based on the frequency measurement method and the period method is established.

[0044]

[0045] Where N0 represents the unit time vector T F Each internal sensor uses the level of three Hall signals to trigger the DSP's CAP1 / CAP2 / CAP3 interrupt vector, where p is the pole pair number (6p interrupts represent one revolution of the wheel). T represents the estimated state parameter value of the sensor layer frequency measurement method. T This represents the time interval vector between two consecutive interruptions of CAP1 / CAP2 / CAP3 in the electronic control unit (ECU) triggered by the level changes of the three Hall signals from each sensor. A ω The scaling factor is... This represents the estimated state parameters of the sensing layer using the periodic method, where τ is the time constant and S is the time vector of the interruption interval for each sensor. This represents the estimated state vector of the sensing layer;

[0046] (12) Establish a control layer state observation module:

[0047] Set the real-time state vector of the control layer:

[0048] X c =[X c_ctr X c_can ] T

[0049] Among them, X c_ctr X represents the real-time state vector of the control module's current transfer ratio. c_can This represents the real-time communication status vector of each node on the CAN bus (the total number of nodes is n).

[0050] X c_ctr =[ctr d ctr s ctr b ]

[0051] X c_can =[s c_can1 s c_can2 ...s c_cann ]

[0052] Among them, ctr d ,ctr s ,ctr bThese represent the real-time current transfer ratios of the drive, steering, and braking control modules, respectively. c_cani This represents the communication status characteristic value of the i-th node on the CAN bus, where i = 1, 2, ..., n. A value of 0 indicates that the node is working normally, and a value of 1 indicates that the node's communication has failed.

[0053] (13) Establish an execution layer state observation module:

[0054] Define the real-time state vector of the execution layer:

[0055]

[0056] in, Represents the state vector for estimating the internal resistance of the motor in the execution layer, X a_f This represents the state vector of the motor bearing vibration signal in the execution layer.

[0057]

[0058] X a_f =[X a_fd X a_fs x a_fb ]

[0059] in, This represents the state vector for estimating the internal resistance of the hub motor. This represents the estimated state vector of the brake motor's internal resistance. X represents the estimated state parameter of the steering actuator motor's internal resistance. a_fd X represents the state vector of the bearing vibration signal of the hub motor. a_fs x represents the state vector of the vibration signal of the brake motor bearing. a_fb This indicates the vibration signal status parameters of the steering actuator motor bearing;

[0060]

[0061]

[0062] in, These represent the estimated internal resistances of the hub motor and brake motor of wheel ij, respectively. ij = 11 represents the front left; ij = 12 represents the front right; ij = 21 represents the rear left; and ij = 22 represents the rear right.

[0063] An algorithm for estimating motor internal resistance parameters based on the extended Kalman filter algorithm is established.

[0064] Set the target estimate for the algorithm:

[0065]

[0066] Among them, R ms_iThis represents the internal resistance of each motor in the execution layer (i = 1, 2, ..., 9);

[0067] Assume an execution layer motor model:

[0068]

[0069] Among them, X k+1 G represents the system state vector at time k+1. k W is the driving matrix. k For process noise, Z k V represents the system state observation vector at time k. k For observing noise, f is the relationship function between the system state vector at time k and the system state vector at time k+1, and h is the relationship function between the system state observation vector at time k and the system state observation vector at time k+1.

[0070] The internal resistance of the execution layer motor is estimated using the Kalman filter algorithm.

[0071]

[0072] in, This represents the prior prediction of the system state at time k+1. P represents the posterior predicted value of the system state at time k. k+1|k P represents the covariance of the prior estimate error of the system state at time k+1. k|k φ represents the covariance of the posterior estimation error of the system state at time k. k+1|k Let Q be the prior Jacobian matrix of the function f at time k+1. k+1 K represents the covariance of the process noise at time k+1. k+1 H represents the filter gain matrix at time k+1. k+1 R represents the Jacobian matrix of the function h at time k+1. k+1 P represents the covariance of the measurement noise at time k+1. k+1 Let I represent the covariance of the estimation error of the system state at time k+1, where I is the identity matrix;

[0073] Establish the discrete difference equation for the motor:

[0074]

[0075] Where matrix B is the zero matrix, u d u q L represents the voltage across the two axes of the motor in the dq coordinate system. s T represents the inductance of the motor stator winding. s ω represents the stator torque of the motor, ω represents the rotor angular velocity of the motor, and ψ represents the stator torque of the motor. f Indicates the magnetic flux linkage of the motor;

[0076] Set the system state vector and initial matrix:

[0077]

[0078]

[0079]

[0080]

[0081] Where X is the system state vector, i d i q Let P0, Q0, and R0 represent the two-axis currents of the motor in the dq coordinate system, respectively. P0, Q0, and R0 represent the covariance of the system state estimation error at the initial moment, the covariance of the process noise at the initial moment, and the covariance of the measurement noise at the initial moment, respectively.

[0082] Furthermore, the multi-level fault diagnosis method in step (2) is as follows:

[0083] (21) Define the state residual vector R of the sensing layer s :

[0084]

[0085] For single-channel and dual-channel Hall signal jamming faults, threshold vectors for single-channel and dual-channel jamming faults are set, taking into account estimation errors and wheel speed influence factors such as tire sideslip, and a threshold reserve vector is set:

[0086]

[0087] Among them, TH s1 TH represents the threshold vector for single-path stuck faults in the sensing layer. s2 ΔTH represents the wheel speed threshold vector for a dual-path jamming fault in the sensing layer. s1 ΔTH represents the covariant vector of the wheel speed threshold during a single-path jamming fault in the sensing layer. s2 The vector represents the covariant vector of the wheel speed threshold in a dual-path stuck fault at the sensing layer. sg_i The threshold value g represents the fault value of sensor i in the sensing layer, where g=1 indicates a single-path jamming fault; g=2 indicates a dual-path jamming fault; i=ωd represents the wheel speed sensor; i=ωb represents the wheel brake motor speed sensor; i=θs represents the steering wheel angle sensor; i=yd represents the accelerator pedal position sensor; i=yb represents the brake pedal position sensor; and i=ωs represents the steering actuator motor speed sensor.

[0088] Constructing the fault feature matrix of the sensing layer:

[0089]

[0090] Among them, s s_i This represents the fault characteristic value of the corresponding sensor in the sensing layer. A value of 0 indicates that the sensor is working normally, while a value greater than 1 indicates that the sensor has a single-channel / dual-channel jamming fault. This parameter is the failure parameter. g = 1 indicates a single-channel jamming fault; g = 2 indicates a dual-channel jamming fault; i = ωd represents the wheel speed sensor; i = ωb represents the wheel brake motor speed sensor; i = θs represents the steering wheel angle sensor; i = yd represents the accelerator pedal position sensor; i = yb represents the brake pedal position sensor; i = ωs represents the steering actuator motor speed sensor.

[0091] (22) Construct the fault feature matrix of the control layer:

[0092]

[0093] Among them, S c_ctr S represents the fault feature vector of the optocoupler failure in the control module. c_can s represents the fault feature vector of the control layer CAN bus. ctra This represents the fault characteristic value of the optocoupler in control module A. A value of 0 indicates that the optocoupler in the control module is working normally, while a value greater than 1 indicates that the optocoupler in the control module has failed. (ctr) a_t The current transfer ratio of control module a at time t is represented by , a = d represents the drive control module, a = s represents the steering control module, and a = b represents the braking control module; s c_cani This represents the communication status characteristic value of the i-th node on the CAN bus, where i = 1, 2, ..., n. A value of 0 indicates that the node is working normally, and a value of 1 indicates that the node's communication has failed.

[0094] (23) An algorithm for estimating bearing fault characteristic frequencies based on amplitude spectrum analysis:

[0095]

[0096]

[0097] Where x(f) i ) represents the signal amplitude spectrum, {(x i Let X be a time-domain signal, i = 1, 2, ..., N. a_fi f represents the state vector of the vibration signal of the motor bearing in the execution layer obtained from the i-th signal acquisition. Z f O f I These represent the characteristic frequencies of failure in the bearing rolling elements, outer ring, and inner ring, respectively, where d is the diameter of the rolling element, D is the bearing pitch diameter, and f is the bearing pitch diameter. rWhere z is the rotational frequency, z is the number of rolling elements, α is the rolling bearing contact angle, j is the imaginary unit, and π is pi.

[0098] Construct the execution layer fault feature matrix:

[0099]

[0100] Among them, s a_raij This represents the fault characteristic value of the stator winding internal resistance of the corresponding execution motor. When the absolute value of the deviation between the algorithm-estimated internal resistance and the initial internal resistance is greater than 20% of the initial value, its value is 1, indicating that the execution motor has a stator winding fault; this parameter is the failure parameter. When the algorithm-estimated internal resistance is relatively stable, its value is 0, indicating that the stator winding of the execution motor is working normally. a_faij This represents the fault characteristic value of the corresponding actuator motor bearing in the actuator layer. When a fault characteristic frequency is identified in the bearing vibration frequency, a value of 1 indicates that the actuator motor has a fault in the bearing rolling element / outer ring / inner ring; conversely, a value of 0 indicates that the actuator motor bearing is working normally. a = d represents the actuator motor of the ij wheel drive subsystem; a = b represents the actuator motor of the ij wheel brake subsystem; a = s represents the actuator motor of the steering subsystem; ij = 11 represents the left front; ij = 12 represents the right front; ij = 21 represents the left rear; and ij = 22 represents the right rear.

[0101] Furthermore, the method for identifying the degree of architecture-level failure in step (3) is as follows:

[0102] Construct fault characteristic matrices for the drive subsystem, steering subsystem, and braking subsystem:

[0103]

[0104] Among them, S D S S S B These represent the fault characteristic matrices of the drive subsystem, steering subsystem, and braking subsystem, respectively; S sd S cd S ad These represent the fault characteristic matrices of the sensing layer, control layer, and execution layer of the drive subsystem, respectively; S ss S cs S as These represent the fault characteristic matrices of the sensing, control, and execution layers of the steering subsystem, respectively; S sb S cb S ab These represent the fault characteristic matrices of the sensing layer, control layer, and execution layer of the braking subsystem, respectively.

[0105] Define the architecture layer fault severity vector:

[0106]

[0107] Among them, F d , θ represents the actual average driving force of the four wheels and the ideal average driving force of the four wheels, respectively; s , These represent the actual steering angle of the front and rear wheels, and the ideal steering angle of the front and rear wheels, respectively; F b , These represent the actual average braking force of the four wheels and the ideal average braking force of the four wheels, respectively; ξ d ξ s ξ b These represent the compensation coefficients for the drive subsystem, steering subsystem, and braking subsystem, respectively; γ a This indicates the degree of failure of subsystem a. The greater the degree of failure of the subsystem, the greater the performance loss of the subsystem caused by the failure. a = d represents the drive subsystem; a = s represents the steering subsystem; a = b represents the braking subsystem.

[0108] A method for identifying the degree of fault in the architecture layer based on the gradient spectrum entropy of the higher-order differential mathematical morphology is established: the fault feature matrix signals of the drive subsystem, steering subsystem, and braking subsystem under different fault conditions and the corresponding degree of fault in the architecture layer are collected in advance by the architecture layer fault degree identification module, and the fault degree of the architecture layer is preset and stored in the architecture layer fault degree database.

[0109] Define the original input signal to identify the fault level at the architecture layer:

[0110]

[0111] Where f(i) represents the original input signal, S D_i+1 S S_i+1 S B_i+1 This represents the fault feature matrix of the drive subsystem, steering subsystem, and braking subsystem obtained from the i-th signal acquisition.

[0112] Determine the analytical scale range of the original input signal structuring elements, taking λ∈[1,50], and calculate the mathematical morphological gradient spectral entropy:

[0113]

[0114] Where f(n) represents the original input signal defined as a discrete function F = (0,1,...,N-1); g(m) represents the original signal structuring element defined as a discrete function G = (0,1,...,M-1), where N≥M; and λ is the analysis scale. fΘg represents the dilation and erosion operations, respectively; Grad represents the mathematical morphology gradient operator; PGS represents the mathematical morphology gradient spectrum; PGSE represents the mathematical morphology gradient spectrum entropy.

[0115] Determine the optimal analysis scale range [1, λ] for the structuring elements of the original input signal. op ]:

[0116] PGS(f(n),λ op +1,g(m))-PGS(f(n),λ op ,g(m))≤10 -2

[0117] Where, λ op The maximum value of the optimal analysis scale for the original input signal structuring element;

[0118] Calculate the spectral entropy of the mathematical morphology of higher-order differential equations;

[0119]

[0120] Where β is the sampling interval;

[0121] Establish the subsystem fault discrimination index and establish the subsystem fault discrimination index fitting curve;

[0122]

[0123] Among them, G_PGSE i This represents the higher-order mathematical morphological gradient spectral entropy calculated after the i-th signal acquisition;

[0124] The fault feature matrix signals of the drive subsystem, steering subsystem, and braking subsystem under different fault conditions in the architecture layer fault severity database are used to generate multiple subsystem fault discrimination fitting curves along with the corresponding architecture layer fault severity. Simultaneously, the subsystem fault discrimination fitting curve for the current moment is generated. Similarity curves are determined, and a quantitative drive subsystem fault severity γ is obtained. d Steering subsystem failure level γ s γ degree of failure in the braking subsystem b And output the architecture layer failure level vector γ arc .

[0125] Furthermore, step (4) is specifically as follows:

[0126] Define an architecture-level reference failure severity vector:

[0127]

[0128] in, These represent the reference fault levels of the drive subsystem, steering subsystem, and braking subsystem, respectively.

[0129] The subsystem fault alarm output is determined as follows:

[0130]

[0131] The beneficial effects of this invention are:

[0132] 1. This invention integrates all-wheel drive, steering, braking and other systems into one, enabling independent development of the upper body and the lower chassis, greatly enhancing the convenience of vehicle manufacturing. At the same time, it proposes a multi-level integrated and coordinated fault-tolerant control method, which improves the safety of the whole vehicle and enables it to meet the multiple safety requirements of ASIL-D level.

[0133] 2. This invention performs multi-level fault diagnosis on the skateboard chassis system, which solves the problem that when a fault occurs across subsystems of the chassis, the subsystems tend to interfere with each other and the diagnosis is inaccurate. The fault diagnosis of each level of the chassis is independent and coordinated with each other, which significantly improves the overall fault diagnosis efficiency of the chassis.

[0134] 3. The multi-source coupled skateboard chassis system and its multi-level fault diagnosis method proposed in this invention can be applied to various vehicle types, such as commercial vehicles and passenger vehicles, and have high market value and practical significance. Attached Figure Description

[0135] Figure 1 This is a schematic diagram of the multi-source coupled skateboard chassis system of the present invention.

[0136] Figure 2 This is a flowchart of the multi-level fault diagnosis process for the multi-source coupled chassis system of the present invention;

[0137] Figure 3 This is a block diagram illustrating the principle of a multi-level fault diagnosis method.

[0138] Figure 4 A schematic diagram illustrating the principle of the global fault severity identification method;

[0139] In the diagram, 1-steering wheel, 2-steering wheel angle sensor, 3-steering wheel torque sensor, 4-first steering column, 5-clutch, 6-left front wheel brake motor, 7-left front wheel brake motor speed sensor, 8-left front wheel speed sensor, 9-left front wheel hub motor, 10-left front wheel brake cylinder, 11-left front wheel reduction mechanism, 12-left front wheel ball screw, 13-left front wheel, 14-left rear wheel brake motor, 15-left rear wheel brake motor speed sensor, 16-left rear wheel speed sensor, 17-left rear wheel hub motor, 18-left rear wheel brake cylinder, 19-left rear wheel reduction mechanism, 2 0-Left rear wheel ball screw, 21-Left rear wheel, 22-Steering control module, 23-Brake control module, 24-Right rear wheel, 25-Right rear wheel ball screw, 26-Right rear wheel reduction mechanism, 27-Right rear wheel brake cylinder, 28-Right rear wheel hub motor, 29-Right rear wheel speed sensor, 30-Right rear wheel brake motor speed sensor, 31-Right rear wheel brake motor, 32-Right front wheel, 33-Right front wheel ball screw, 34-Right front wheel reduction mechanism, 35-Right front wheel brake cylinder, 36-Right front wheel hub motor, 37-Right front wheel speed sensor, 38-Right front wheel brake motor speed sensor 39-Right front wheel brake motor; 40-Second steering column; 41-Screw and nut transmission pair; 42-Rack and pinion transmission pair; 43-Recirculating ball hydraulic power steering system; 44-Steering rocker arm; 45-Second reduction mechanism; 46-Left steering knuckle; 47-Left steering knuckle arm; 48-Left steering trapezoidal arm; 49-Steering tie rod; 50-Steering actuator motor; 51-First reduction mechanism; 52-Hydraulic module; 53-Servo proportional solenoid valve; 54-Steering tie rod; 55-Right steering trapezoidal arm; 56-Right steering knuckle arm; 57-Right steering knuckle; 58-Steering actuator motor speed sensor. 59-Oil tank, 60-Hydraulic pump drive motor, 61-Hydraulic pump, 62-Drive control module, 63-First pressure sensor, 64-Second pressure sensor, 65-Brake pedal position sensor, 66-Brake pedal, 67-Accelerator pedal, 68-Accelerator pedal position sensor, 69-Brake valve, 70-Air reservoir, 71-Air compressor, 72-Air pressure pipeline, 73-Sensing layer status observation module, 74-Control layer status observation module, 75-Execution layer status observation module, 76-Multi-level fault diagnosis module, 77-Architecture layer fault degree identification module, 78-Fault alarm module;

[0140] A - Steering wheel angle signal; B - Steering wheel torque signal; C - Hydraulic power assist signal; D - Brake pedal position signal; E - Accelerator pedal position signal; F1 - Left front wheel speed signal; F2 - Left rear wheel speed signal; F3 - Right rear wheel speed signal; F4 - Right front wheel speed signal; G1 - Left front wheel brake motor speed signal; G2 - Left rear wheel brake motor speed signal; G3 - Right rear wheel brake motor speed signal; G4 - Right front wheel brake motor speed signal; H - Steering actuator motor speed signal; J - Drive, steering, and braking control module electrical signal. The signal includes the transmission ratio, communication status signals of each node on the K-CAN bus, M-current signals of the two shafts of the left front wheel brake motor, left rear wheel brake motor, right rear wheel brake motor, right front wheel brake motor, left front wheel hub motor, left rear wheel hub motor, right rear wheel hub motor, right front wheel hub motor, and steering actuator motor, and N-vibration frequency signals of the bearings of the left front wheel brake motor, left rear wheel brake motor, right rear wheel brake motor, right front wheel brake motor, left front wheel hub motor, left rear wheel hub motor, right rear wheel hub motor, right front wheel hub motor, and steering actuator motor. Detailed Implementation

[0141] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to embodiments and accompanying drawings. The content mentioned in the embodiments is not intended to limit the present invention.

[0142] Reference Figure 1 As shown, a multi-source coupled skateboard chassis system of the present invention includes: a drive-by-wire subsystem, an electro-hydraulic composite steering-by-wire subsystem, an electrical composite braking-by-wire subsystem, a multi-level global status observation module, a multi-level fault diagnosis module, an architecture-level fault severity identification module, and a fault alarm module; wherein,

[0143] The drive-by-wire subsystem includes: an accelerator pedal 67, an accelerator pedal position sensor 68, a left front wheel hub motor 9, a right front wheel hub motor 36, a left rear wheel hub motor 17, a right rear wheel hub motor 28, a left front wheel speed sensor 8, a right front wheel speed sensor 37, a left rear wheel speed sensor 16, a right rear wheel speed sensor 29, a left front wheel 13, a right front wheel 32, a left rear wheel 21, a right rear wheel 24, and a drive control module 62; the accelerator pedal position sensor 68 is fixedly mounted on the accelerator pedal 67 and is used to send accelerator pedal position signals to the drive control module 62; the left front wheel hub motor 9 and the left front wheel speed sensor 8 are installed inside the left front wheel 13; the right front wheel hub motor 36 and the right front wheel speed sensor 29... The device 37 is installed inside the right front wheel 32; the left rear wheel hub motor 17 and the left rear wheel speed sensor 16 are installed inside the left rear wheel 21; the right rear wheel hub motor 28 and the right rear wheel speed sensor 29 are installed inside the right rear wheel 24; the rotational motion output by the left front wheel hub motor 9, the right front wheel hub motor 36, the left rear wheel hub motor 17, and the right rear wheel hub motor 28 is converted into the rotational motion of the left front wheel 13, the right front wheel 32, the left rear wheel 21, and the right rear wheel 24, respectively, driving the vehicle to move; the input terminal of the drive control module 62 is connected to the left front wheel speed sensor 8, the right front wheel speed sensor 37, the left rear wheel speed sensor 16, the right rear wheel speed sensor 29, and the accelerator pedal position sensor 68 through the vehicle communication line;

[0144] The electro-hydraulic composite steer-by-wire subsystem includes: a steering wheel module, a steering motor module, a hydraulic module, a mechanical transmission module, and a steering control module 22;

[0145] The steering wheel module includes: a steering wheel 1, a first steering column 4, a steering wheel angle sensor 2, and a steering wheel torque sensor 3; the steering wheel 1 is fixedly connected to one end of the first steering column 4, and the other end of the first steering column 4 is connected to a clutch 5; the steering wheel angle sensor 2 and the steering wheel torque sensor 3 are both fixedly connected to the first steering column 4, respectively collecting the steering wheel angle and torque of the steering wheel 1, and sending the collected steering wheel angle signal and steering wheel torque signal to the steering control module 22;

[0146] The steering motor module includes: a steering actuator motor 50, a steering actuator motor speed sensor 58, a first reduction mechanism 51, and a second steering column 40; the output end of the steering actuator motor 50 is connected to the steering actuator motor speed sensor 58 and transmits electric power assistance to the second steering column 40 through the first reduction mechanism 51 to provide power assistance to the steering system;

[0147] The hydraulic module includes: a hydraulic pump 61, a hydraulic pump drive motor 60, a second reduction mechanism 45, an oil tank 59, a servo proportional solenoid valve 53, oil pipes, a first pressure sensor 63, and a second pressure sensor 64. The hydraulic pump drive motor 60 is connected to the hydraulic pump 61 through the second reduction mechanism 45, pumping hydraulic oil from the oil tank 59 into the recirculating ball steering gear of the mechanical transmission module through the servo proportional solenoid valve 53, forming an oil pressure difference in the recirculating ball steering gear, which provides power assistance to the steering system under the action of the oil pressure difference. The first pressure sensor 63 is installed on the oil pipe of the recirculating ball steering gear inlet, and the second pressure sensor 64 is installed on the oil pipe of the recirculating ball steering gear outlet, respectively used to detect the hydraulic power assistance on both sides of the recirculating ball steering gear and send hydraulic power assistance signals.

[0148] The mechanical transmission module includes: a recirculating ball steering gear, a steering rocker arm 44, a steering tie rod 49, a left steering knuckle 46, a left steering knuckle arm 47, a left steering trapezoidal arm 48, a steering tie rod 54, a right steering trapezoidal arm 55, a right steering knuckle arm 56, and a right steering knuckle 57; the recirculating ball steering gear consists of two stages of transmission pairs: the first stage is a screw and nut transmission pair 41; the second stage is a rack and pinion transmission pair 42; one end of the second steering column 40 is connected to the clutch 5, and the other end is connected to the input end of the recirculating ball steering gear through the screw and nut transmission pair 41. The output end of the recirculating ball steering gear is connected to one end of the steering rocker arm 44 via a rack and pinion gear transmission pair 42. The other end of the steering rocker arm 44 is connected to the left steering knuckle arm 48 via a steering tie rod 49, which drives the left steering knuckle 46 and the left front wheel 13 to deflect. The left steering knuckle arm 47 is connected to one end of the steering tie rod 54 via the left steering trapezoidal arm 48. The other end of the steering tie rod 54 is connected to the right steering trapezoidal arm 55. The right steering trapezoidal arm 55 is connected to the right steering knuckle 57 via the right steering knuckle arm 56. The right steering knuckle 57 drives the right front wheel 32 to turn.

[0149] The input terminal of the steering control module 22 is electrically connected to the steering wheel angle sensor 2, the steering wheel torque sensor 3, and the steering actuator motor speed sensor 58, respectively, and sends control signals to the steering motor module and the hydraulic module to coordinate the steering of the front wheels.

[0150] The electrical composite line control braking subsystem includes: brake pedal 66, brake pedal position sensor 65, motor regenerative braking module, air pressure braking module and brake control module 23;

[0151] The regenerative braking module includes: a left front wheel brake motor 6, a left front wheel brake motor speed sensor 7, a left front wheel reduction mechanism 11, a left front wheel ball screw 12, a left front wheel brake cylinder 10, a right front wheel brake motor 39, a right front wheel brake motor speed sensor 38, a right front wheel reduction mechanism 34, a right front wheel ball screw 33, a right front wheel brake cylinder 35, a left rear wheel brake motor 14, a left rear wheel brake motor speed sensor 15, a left rear wheel reduction mechanism 19, a left rear wheel ball screw 20, a left rear wheel brake cylinder 18, a right rear wheel brake motor 31, a right rear wheel brake motor speed sensor 30, a right rear wheel reduction mechanism 26, a right rear wheel ball screw 25, and a right rear wheel brake cylinder 27.

[0152] The brake pedal position sensor 65 is fixedly mounted on the brake pedal 66 and is used to send brake pedal position signals to the brake control module 23. The output end of the left front wheel brake motor 6 is connected to one end of the left front wheel ball screw 12 through the left front wheel brake motor speed sensor 7 and the left front wheel reduction mechanism 11 in sequence. The other end of the left front wheel ball screw 12 is connected to the left front wheel brake cylinder 10. The output end of the right front wheel brake motor 39 is connected to one end of the right front wheel ball screw 33 through the right front wheel brake motor speed sensor 38 and the right front wheel reduction mechanism 34 in sequence. The other end of the right front wheel ball screw is connected to the right front wheel brake cylinder 35. The output end of the left rear wheel brake motor 14 is connected to one end of the left rear wheel ball screw 20 through the left rear wheel brake motor wheel speed sensor 15 and the left rear wheel reduction mechanism 19 in sequence. The other end of the left rear wheel ball screw 20 is connected to the left rear wheel brake cylinder 18. The output end of the right rear wheel brake motor 31 is connected to the right rear wheel brake motor speed sensor 7 and the left rear wheel reduction mechanism 19 in sequence. Speed ​​sensor 30 and right rear wheel reduction mechanism 26 are connected to one end of right rear wheel ball screw 25, and the other end of right rear wheel ball screw 25 is connected to right rear wheel brake cylinder 27; the rotational motion of left front wheel brake motor 6, right front wheel brake motor 39, left rear wheel brake motor 14, and right rear wheel brake motor 31 is respectively passed through left front wheel reduction mechanism 11 and left front wheel ball screw 12, right front wheel reduction mechanism 34 and right front wheel ball screw 33, and left rear wheel reduction mechanism 19 and... The left rear wheel ball screw 20, the right rear wheel reduction mechanism 26, and the right rear wheel ball screw 25 are converted into the displacement movements of the screw ends of the left front wheel ball screw 12, the right front wheel ball screw 33, the left rear wheel ball screw 20, and the right rear wheel ball screw 25, respectively. These movements act on the left front wheel brake cylinder 10, the right front wheel brake cylinder 35, the left rear wheel brake cylinder 18, and the right rear wheel brake cylinder 27 to generate braking torque and complete the vehicle braking operation.

[0153] The pneumatic braking module includes an air compressor 71, an air reservoir 70, and a brake valve 69, which are connected in sequence. The air compressor 71 compresses air into the air reservoir 70, and then inflates the left front wheel brake cylinder 10, the right front wheel brake cylinder 35, the left rear wheel brake cylinder 18, and the right rear wheel brake cylinder 27 through the brake valve 69 to achieve braking.

[0154] The output of the braking control module 23 is connected to the brake pedal position sensor 65, the left front wheel speed sensor 8, the right front wheel speed sensor 37, the left rear wheel speed sensor 16, and the right rear wheel speed sensor 29, respectively, and is used to adjust the braking force of the motor regenerative braking force module and the air pressure braking module according to the torque distribution strategy of the electric composite line control braking system.

[0155] The multi-level global state observation module includes: a sensing layer state observation module 73, a control layer state observation module 74, and an execution layer state observation module 75.

[0156] The sensor layer state observation module 73 is connected to the accelerator pedal position sensor 68, the left front wheel speed sensor 8, the right front wheel speed sensor 37, the left rear wheel speed sensor 16, the right rear wheel speed sensor 29, the steering wheel angle sensor 2, the steering actuator motor speed sensor 58, the brake pedal position sensor 65, the left front wheel brake motor speed sensor 7, the right front wheel brake motor speed sensor 38, the left rear wheel brake motor speed sensor 15, and the right rear wheel brake motor speed sensor 30 via the vehicle communication line. It is used to acquire the signals collected by each sensor in real time. At the same time, the sensor layer state observation module acquires the three Hall signals of each sensor and calculates the signals estimated by the algorithm of each sensor.

[0157] The control layer status observation module 74 is used to acquire the current transmission ratio signal of the drive control module, the current transmission ratio signal of the steering control module, the current transmission ratio signal of the braking control module, and the communication status signals of each node on the CAN bus.

[0158] The execution layer state observation module 75 is used to acquire the current signals and bearing vibration frequency signals of the two shafts of the left front wheel hub motor, the two shafts of the right front wheel hub motor, the two shafts of the left rear wheel hub motor, the two shafts of the right rear wheel hub motor, the two shafts of the steering actuator motor, the two shafts of the left front wheel brake motor, the two shafts of the right front wheel brake motor, the two shafts of the left rear wheel brake motor, and the two shafts of the right rear wheel brake motor.

[0159] The input terminal of the multi-level fault diagnosis module 76 is connected to the output terminal of the multi-level global state observation module. It is used to acquire the real-time state signals of the sensing layer, the control layer, and the execution layer, perform multi-level fault state diagnosis, and output the real-time fault state signals of the sensing layer, the control layer, and the execution layer.

[0160] The input of the architecture layer fault degree identification module 77 is connected to the output of the multi-level fault diagnosis module. It is used to acquire real-time fault status signals of the sensing layer, control layer, and execution layer, identify the fault degree of each subsystem, and establish an architecture layer fault degree database to obtain the fault degree of each subsystem under different multi-level states in the initial preset.

[0161] The input terminal of the fault alarm module 78 is connected to the output terminal of the architecture layer fault degree identification module, which is used to obtain the architecture layer fault degree signal and output fault alarms for the fault subsystem and each level of fault within the subsystem.

[0162] Reference Figures 2-4 As shown, the present invention provides a multi-level fault diagnosis method for a multi-source coupled skateboard chassis system, based on the aforementioned system, comprising the following steps:

[0163] (1) Establish a multi-level global state observation method: Establish a multi-level global state observation module consisting of a sensing layer state observation module, a control layer state observation module, and an execution layer state observation module to obtain relevant real-time state parameters and relevant estimated state parameters of the sensing layer, control layer, and execution layer.

[0164] The multi-level global state observation method in step (1) is as follows:

[0165] (11) Establish a sensor layer state observation module:

[0166] Set the real-time state vector X of the sensing layer s With the estimated state vector

[0167] X s =[X s_ωd X s_ωb x s_θs x s_yd x s_yb x s_ωs ] T

[0168]

[0169] Among them, X s_ωd , Let X represent the measured state vector and the estimated state vector of the wheel speed sensor, respectively. s_ωb , Let x represent the measured state vector and the estimated state vector of the wheel brake motor speed sensor, respectively. s_θs , x represents the measured state parameters and the estimated state parameters of the steering wheel angle sensor, respectively. s_yd , x represents the measured state parameters and the estimated state parameters of the accelerator pedal position sensor, respectively. s_yb , x represents the measured state parameters and the estimated state parameters of the brake pedal position sensor, respectively. s_ωs , These represent the measured state parameters of the steering actuator motor speed sensor and the estimated state parameters of the steering actuator motor speed sensor, respectively.

[0170] X s_ωd =[ω s_ωd11 ω s_ωd12 ω s_ωd21 ω s_ωd22 ]

[0171]

[0172] X s_ωb =[ω s_ωb11 ω s_ωb12 ω s_ωb21 ω s_ωb22 ]

[0173]

[0174] Where, ω s_ωdij This indicates the wheel speed measured by the ij wheel speed sensor. This indicates that the wheel speed is estimated by the ij wheel speed sensor, ω s_ωbij This indicates the actual measured speed of the ij wheel brake motor by the speed sensor. The value represents the estimated rotational speed of the wheel brake motor speed sensor. ij = 11 indicates front left; ij = 12 indicates front right; ij = 21 indicates rear left; and ij = 22 indicates rear right.

[0175] An algorithm for estimating the state parameters of the sensing layer based on the frequency measurement method and the period method is established.

[0176]

[0177] Where N0 represents the unit time vector T FEach internal sensor uses the level of three Hall signals to trigger the DSP's CAP1 / CAP2 / CAP3 interrupt vector, where p is the pole pair number (6p interrupts represent one revolution of the wheel). T represents the estimated state parameter value of the sensor layer frequency measurement method. T This represents the time interval vector between two consecutive interruptions of CAP1 / CAP2 / CAP3 by each sensor using the level changes of the three Hall signals. A ω The scaling factor is... This represents the estimated state parameters of the sensing layer using the periodic method, where τ is the time constant and S is the time vector of the interruption interval for each sensor. This represents the estimated state vector of the sensing layer;

[0178] (12) Establish a control layer state observation module:

[0179] Set the real-time state vector of the control layer:

[0180] X c =[X c_ctr X c_can ] T

[0181] Among them, X c_ctr X represents the real-time state vector of the control module's current transfer ratio. c_can This represents the real-time communication status vector of each node on the CAN bus (the total number of nodes is n).

[0182] X c_ctr =[ctr d ctr s ctr b ]

[0183] X c_can =[s c_can1 s c_can2 ...s c_cann ]

[0184] Among them, ctr d ,ctr s ,ctr b These represent the real-time current transfer ratios of the drive, steering, and braking control modules, respectively. c_cani This represents the communication status characteristic value of the i-th node on the CAN bus, where i = 1, 2, ..., n. A value of 0 indicates that the node is working normally, and a value of 1 indicates that the node's communication has failed.

[0185] (13) Establish an execution layer state observation module:

[0186] Define the real-time state vector of the execution layer:

[0187]

[0188] in, Represents the state vector for estimating the internal resistance of the motor in the execution layer, X a_f This represents the state vector of the motor bearing vibration signal in the execution layer.

[0189]

[0190] X a_f =[X a_fd X a_fs x a_fb ]

[0191] in, This represents the state vector for estimating the internal resistance of the hub motor. This represents the estimated state vector of the brake motor's internal resistance. X represents the estimated state parameter of the steering actuator motor's internal resistance. a_fd X represents the state vector of the bearing vibration signal of the hub motor. a_fs x represents the state vector of the vibration signal of the brake motor bearing. a_fb This indicates the vibration signal status parameters of the steering actuator motor bearing;

[0192]

[0193]

[0194] in, These represent the estimated internal resistances of the hub motor and brake motor of wheel ij, respectively. ij = 11 represents the front left; ij = 12 represents the front right; ij = 21 represents the rear left; and ij = 22 represents the rear right.

[0195] An algorithm for estimating motor internal resistance parameters based on the extended Kalman filter algorithm is established.

[0196] Set the target estimate for the algorithm:

[0197]

[0198] Among them, R ms_i This represents the internal resistance of each motor in the execution layer (i = 1, 2, ..., 9);

[0199] Assume an execution layer motor model:

[0200]

[0201] Among them, X k+1 G represents the system state vector at time k+1. k W is the driving matrix. k For process noise, Z k V represents the system state observation vector at time k.k For observing noise, f is the relationship function between the system state vector at time k and the system state vector at time k+1, and h is the relationship function between the system state observation vector at time k and the system state observation vector at time k+1.

[0202] The internal resistance of the execution layer motor is estimated using the Kalman filter algorithm.

[0203]

[0204] in, This represents the prior prediction of the system state at time k+1. P represents the posterior predicted value of the system state at time k. k+1|k P represents the covariance of the prior estimate error of the system state at time k+1. k|k φ represents the covariance of the posterior estimation error of the system state at time k. k+1|k Let Q be the prior Jacobian matrix of the function f at time k+1. k+1 K represents the covariance of the process noise at time k+1. k+1 H represents the filter gain matrix at time k+1. k+1 R represents the Jacobian matrix of the function h at time k+1. k+1 P represents the covariance of the measurement noise at time k+1. k+1 Let I represent the covariance of the estimation error of the system state at time k+1, where I is the identity matrix;

[0205] Establish the discrete difference equation for the motor:

[0206]

[0207] Where matrix B is the zero matrix, u d u q L represents the voltage across the two axes of the motor in the dq coordinate system. s T represents the inductance of the motor stator winding. s ω represents the stator torque of the motor, ω represents the rotor angular velocity of the motor, and ψ represents the stator torque of the motor. f Indicates the magnetic flux linkage of the motor;

[0208] Set the system state vector and initial matrix:

[0209]

[0210]

[0211]

[0212]

[0213] Where X is the system state vector, i d i q Let P0, Q0, and R0 represent the two-axis currents of the motor in the dq coordinate system, respectively. P0, Q0, and R0 represent the covariance of the system state estimation error at the initial moment, the covariance of the process noise at the initial moment, and the covariance of the measurement noise at the initial moment, respectively.

[0214] (2) Establish a multi-level fault diagnosis method based on residual theory: Establish a multi-level fault diagnosis module, calculate the multi-level fault feature matrix according to the state parameters of each level obtained in step (1), diagnose the fault signals of each level, and determine the failure parameters and fault level.

[0215] The multi-level fault diagnosis method in step (2) is as follows:

[0216] (21) Define the state residual vector R of the sensing layer s :

[0217]

[0218] For single-channel and dual-channel Hall signal jamming faults, threshold vectors for single-channel and dual-channel jamming faults are set, taking into account estimation errors and wheel speed influence factors such as tire sideslip, and a threshold reserve vector is set:

[0219]

[0220] Among them, TH s1 TH represents the threshold vector for single-path stuck faults in the sensing layer. s2 ΔTH represents the wheel speed threshold vector for a dual-path jamming fault in the sensing layer. s1 ΔTH represents the covariant vector of the wheel speed threshold during a single-path jamming fault in the sensing layer. s2 The vector represents the covariant vector of the wheel speed threshold in a dual-path stuck fault at the sensing layer. sg_i The threshold value g represents the fault value of sensor i in the sensing layer, where g=1 indicates a single-path jamming fault; g=2 indicates a dual-path jamming fault; i=ωd represents the wheel speed sensor; i=ωb represents the wheel brake motor speed sensor; i=θs represents the steering wheel angle sensor; i=yd represents the accelerator pedal position sensor; i=yb represents the brake pedal position sensor; and i=ωs represents the steering actuator motor speed sensor.

[0221] Constructing the fault feature matrix of the sensing layer:

[0222]

[0223] Among them, s s_iThis represents the fault characteristic value of the corresponding sensor in the sensing layer. A value of 0 indicates that the sensor is working normally, while a value greater than 1 indicates that the sensor has a single-channel / dual-channel jamming fault. This parameter is the failure parameter. g = 1 indicates a single-channel jamming fault; g = 2 indicates a dual-channel jamming fault; i = ωd represents the wheel speed sensor; i = ωb represents the wheel brake motor speed sensor; i = θs represents the steering wheel angle sensor; i = yd represents the accelerator pedal position sensor; i = yb represents the brake pedal position sensor; i = ωs represents the steering actuator motor speed sensor.

[0224] (22) Construct the fault feature matrix of the control layer:

[0225]

[0226] Among them, S c_ctr S represents the fault feature vector of the optocoupler failure in the control module. c_can s represents the fault feature vector of the control layer CAN bus. ctra This represents the fault characteristic value of the optocoupler in control module A. A value of 0 indicates that the optocoupler in the control module is working normally, while a value greater than 1 indicates that the optocoupler in the control module has failed. (ctr) a_t The current transfer ratio of control module a at time t is represented by , a = d represents the drive control module, a = s represents the steering control module, and a = b represents the braking control module; s c_cani This represents the communication status characteristic value of the i-th node on the CAN bus, where i = 1, 2, ..., n. A value of 0 indicates that the node is working normally, and a value of 1 indicates that the node's communication has failed.

[0227] (23) An algorithm for estimating bearing fault characteristic frequencies based on amplitude spectrum analysis:

[0228]

[0229]

[0230] Where x(f) i ) represents the signal amplitude spectrum, {(x i Let X be a time-domain signal, i = 1, 2, ..., N. a_fi f represents the state vector of the vibration signal of the motor bearing in the execution layer obtained from the i-th signal acquisition. Z f O f I These represent the characteristic frequencies of failure in the bearing rolling elements, outer ring, and inner ring, respectively, where d is the diameter of the rolling element, D is the bearing pitch diameter, and f is the bearing pitch diameter. r Where z is the rotational frequency, z is the number of rolling elements, α is the rolling bearing contact angle, j is the imaginary unit, and π is pi.

[0231] Construct the execution layer fault feature matrix:

[0232]

[0233] Among them, s a_raij This represents the fault characteristic value of the stator winding internal resistance of the corresponding execution motor. When the absolute value of the deviation between the algorithm-estimated internal resistance and the initial internal resistance is greater than 20% of the initial value, its value is 1, indicating that the execution motor has a stator winding fault; this parameter is the failure parameter. When the algorithm-estimated internal resistance is relatively stable, its value is 0, indicating that the stator winding of the execution motor is working normally. a_faij This represents the fault characteristic value of the corresponding actuator motor bearing in the actuator layer. When a fault characteristic frequency is identified in the bearing vibration frequency, a value of 1 indicates that the actuator motor has a fault in the bearing rolling element / outer ring / inner ring; conversely, a value of 0 indicates that the actuator motor bearing is working normally. a = d represents the actuator motor of the ij wheel drive subsystem; a = b represents the actuator motor of the ij wheel brake subsystem; a = s represents the actuator motor of the steering subsystem; ij = 11 represents the left front; ij = 12 represents the right front; ij = 21 represents the left rear; and ij = 22 represents the right rear.

[0234] (3) Establish a fault degree identification method for the architecture layer: Define the fault degree of the architecture layer, and combine the multi-level fault feature matrix obtained in step (2) to establish a fault degree identification method for the architecture layer based on the gradient spectrum entropy of the higher-order differential mathematical morphology, and identify the fault degree of the drive subsystem, steering subsystem and braking subsystem.

[0235] The specific method for identifying the degree of architecture-level failure in step (3) is as follows:

[0236] Construct fault characteristic matrices for the drive subsystem, steering subsystem, and braking subsystem:

[0237]

[0238] Among them, S D S S S B These represent the fault characteristic matrices of the drive subsystem, steering subsystem, and braking subsystem, respectively; S sd S cd S ad These represent the fault characteristic matrices of the sensing layer, control layer, and execution layer of the drive subsystem, respectively; S ss S cs S as These represent the fault characteristic matrices of the sensing, control, and execution layers of the steering subsystem, respectively; S sb S cb S ab These represent the fault characteristic matrices of the sensing layer, control layer, and execution layer of the braking subsystem, respectively.

[0239] Define the architecture layer fault severity vector:

[0240]

[0241] Among them, F d , θ represents the actual average driving force of the four wheels and the ideal average driving force of the four wheels, respectively; s , These represent the actual steering angle of the front and rear wheels, and the ideal steering angle of the front and rear wheels, respectively; F b , These represent the actual average braking force of the four wheels and the ideal average braking force of the four wheels, respectively; ξ d ξ s ξ b These represent the compensation coefficients for the drive subsystem, steering subsystem, and braking subsystem, respectively; γ a This indicates the degree of failure of subsystem a. The greater the degree of failure of the subsystem, the greater the performance loss of the subsystem caused by the failure. a = d represents the drive subsystem; a = s represents the steering subsystem; a = b represents the braking subsystem.

[0242] A method for identifying the degree of fault in the architecture layer based on the gradient spectrum entropy of the higher-order differential mathematical morphology is established: the fault feature matrix signals of the drive subsystem, steering subsystem, and braking subsystem under different fault conditions and the corresponding degree of fault in the architecture layer are collected in advance by the architecture layer fault degree identification module, and the fault degree of the architecture layer is preset and stored in the architecture layer fault degree database.

[0243] Define the original input signal to identify the fault level at the architecture layer:

[0244]

[0245] Where f(i) represents the original input signal, S D_i+1 S S_i+1 S B_i+1 This represents the fault feature matrix of the drive subsystem, steering subsystem, and braking subsystem obtained from the i-th signal acquisition.

[0246] Determine the analytical scale range of the original input signal structuring elements, λ∈[1,50], and calculate the mathematical morphological gradient spectral entropy:

[0247]

[0248] Where f(n) represents the original input signal defined as a discrete function F = (0,1,...,N-1); g(m) represents the original signal structuring element defined as a discrete function G = (0,1,...,M-1), where N≥M; and λ is the analysis scale. fΘg represents the dilation and erosion operations, respectively; Grad represents the mathematical morphology gradient operator; PGS represents the mathematical morphology gradient spectrum; PGSE represents the mathematical morphology gradient spectrum entropy.

[0249] Determine the optimal analysis scale range [1, λ] for the structuring elements of the original input signal. op ]:

[0250] PGS(f(n),λ op +1,g(m))-PGS(f(n),λ op ,g(m))≤10 -2

[0251] Where, λ op The maximum value of the optimal analysis scale for the original input signal structuring element;

[0252] Calculate the spectral entropy of the mathematical morphology of higher-order differential equations;

[0253]

[0254] Where β is the sampling interval;

[0255] Establish the subsystem fault discrimination index and establish the subsystem fault discrimination index fitting curve;

[0256]

[0257] Among them, G_PGSE i This represents the higher-order mathematical morphological gradient spectral entropy calculated after the i-th signal acquisition;

[0258] The fault feature matrix signals of the drive subsystem, steering subsystem, and braking subsystem under different fault conditions in the architecture layer fault severity database are used to generate multiple subsystem fault discrimination fitting curves along with the corresponding architecture layer fault severity. Simultaneously, the subsystem fault discrimination fitting curve for the current moment is generated. Similarity curves are determined, and a quantitative drive subsystem fault severity γ is obtained. d Steering subsystem failure level γ s γ degree of failure in the braking subsystem b And output the architecture layer failure level vector γ arc .

[0259] (4) Based on the fault feature matrix and fault degree of the drive subsystem, steering subsystem and braking subsystem obtained in step (3), set the reference fault degree of the architecture layer. When the fault degree of a certain subsystem is greater than the reference fault degree of the architecture layer, the corresponding subsystem and the faults at each level in the subsystem are output through the fault alarm module.

[0260] Specifically, step (4) is as follows:

[0261] Define an architecture-level reference failure severity vector:

[0262]

[0263] in, These represent the reference fault levels of the drive subsystem, steering subsystem, and braking subsystem, respectively.

[0264] The subsystem fault alarm output is determined as follows:

[0265]

[0266] This invention has many specific applications. The above description is only a preferred embodiment of this invention. It should be noted that for those skilled in the art, several improvements can be made without departing from the principle of this invention, and these improvements should also be considered within the scope of protection of this invention.

Claims

1. A multi-level fault diagnosis method for a multi-source coupled skateboard chassis system, characterized in that, Includes the following steps: (1) Establish a multi-level global state observation method: Establish a multi-level global state observation module consisting of a sensing layer state observation module, a control layer state observation module, and an execution layer state observation module to obtain relevant real-time state parameters and estimated state parameters of the sensing layer, control layer, and execution layer; establish a sensing layer state parameter estimation algorithm based on the frequency measurement method and the period method: Where N0 represents the unit time vector T F Each internal sensor triggers the DSP's CAP1 / CAP2 / CAP3 interrupt vector using the levels of three Hall signals, where p is the pole pair number. T represents the estimated state parameter value of the sensor layer frequency measurement method. T This represents the time interval vector between two consecutive interruptions of the electronic control unit's CAP1 / CAP2 / CAP3, triggered by the level changes of the three Hall signals from each sensor. A ω The scaling factor is... This represents the estimated state parameters of the sensing layer using the periodic method, where τ is the time constant and S is the time vector of the interruption interval for each sensor. This represents the estimated state vector of the sensing layer; (2) Establish a multi-level fault diagnosis method based on residual theory: Establish a multi-level fault diagnosis module, calculate the multi-level fault feature matrix according to the state parameters of each level obtained in step (1), diagnose the fault signals of each level, and determine the failure parameters and fault level. (3) Establish a fault degree identification method for the architecture layer: Define the fault degree of the architecture layer, and combine the multi-level fault feature matrix obtained in step (2) to establish a fault degree identification method for the architecture layer based on the gradient spectrum entropy of the higher-order differential mathematical morphology, and identify the fault degree of the drive subsystem, steering subsystem and braking subsystem. (4) Based on the fault feature matrix and fault degree of the drive subsystem, steering subsystem and braking subsystem obtained in step (3), set the reference fault degree of the architecture layer. When the fault degree of a certain subsystem is greater than the reference fault degree of the architecture layer, the corresponding subsystem and the faults at each level in the subsystem are output through the fault alarm module.

2. The multi-level fault diagnosis method for a multi-source coupled skateboard chassis system according to claim 1, characterized in that, Step (4) is as follows: Define an architecture-level reference failure severity vector: in, These represent the reference fault levels of the drive subsystem, steering subsystem, and braking subsystem, respectively. The subsystem fault alarm output is determined as follows:

Images