Chassis intelligent electric control method, system and vehicle

Abstract

This invention discloses a chassis intelligent electronic control method, system, and vehicle. The method includes: S1, acquiring environmental information, vehicle status information, driver intention information, and the actual output status of each actuator; S2, calculating the feedforward compensation yaw moment, determining the boundary constraints for correcting vehicle stability, and calculating the generalized target force and torque required to maintain vehicle stability; S3, determining whether the generalized target force and torque exceed the current available execution capability range of each actuator; if not, proceeding to S4; if exceeding, proceeding to S5; S4, converting the generalized target force and torque into fixed allocation control commands for each wheel according to a preset fixed allocation rule; S5, constructing a multi-objective quadratic programming (QP) model, generating cooperative control commands based on rolling time-domain optimization of model predictive control; S6, controlling each actuator based on the control commands. This invention can prevent command conflicts and has an environmental perception feedforward function.

Description

Technical Field

[0001] This invention relates to the field of drive-by-wire chassis technology, specifically to a chassis intelligent electronic control method, system, and vehicle. Background Technology

[0002] The drive-by-wire chassis technology is a technology that uses electronic signals to replace mechanical connections to control the vehicle chassis. In drive-by-wire chassis technology, existing vehicles are usually equipped with independent electronic steering (EPS), electronic braking (ESC / brake-by-wire), electronic suspension (CDC), and electronic drive (TCU / VCU) subsystems. The mainstream technical solution adopts a distributed control architecture, that is, each subsystem has an independent electronic control unit (ECU). This architecture has the following defects: (1) Under extreme conditions (such as emergency double lane change conditions), the braking system requires yaw stability, while the steering system may require path tracking. Independent decision-making by the two can easily lead to conflicting control commands, and control conflicts can exacerbate the risk of instability. (2) The system only performs feedback control based on the vehicle's own state (such as yaw rate), which is "post-correction". Before the vehicle enters a complex road (such as a curve), it cannot use the information of the road ahead perceived by the camera / radar to perform "pre-control", lacking pre-control capability, resulting in lag when entering a curve. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide a chassis intelligent electronic control method, system and vehicle, which coordinates the execution commands of drive, braking, steering and suspension through a chassis domain controller, can prevent command conflicts and has environmental perception feedforward function.

[0004] A chassis intelligent electronic control method according to the present invention includes:

[0005] S1. Acquire environmental information, vehicle status information, driver intention information, and the actual output status of each actuator;

[0006] S2. Based on the environmental information and vehicle state information, calculate the feedforward compensation yaw moment, determine the boundary constraints for correcting vehicle stability, and calculate the generalized target force and moment required to maintain vehicle stability. The generalized target force and moment include the total desired longitudinal force, the additional desired yaw moment, and the roll moment.

[0007] S3. Obtain the current available execution capability range of each actuator, and determine whether the generalized target force and torque exceed the current available execution capability range of each actuator; if the generalized target force and torque do not exceed the current available execution capability range of each actuator, proceed to S4; if the generalized target force and torque exceed the current available execution capability range of a certain actuator, proceed to S5.

[0008] S4. Convert the generalized target force and torque into fixed distribution control commands for each wheel according to the preset fixed distribution rules;

[0009] S5. Construct a multi-objective quadratic programming (QP) model, and based on the rolling time-domain optimization of model predictive control, redistribute and optimize the target driving torque, target braking torque, target steering angle and target suspension damping force of each wheel to generate cooperative control commands.

[0010] S6. Control each actuator based on the fixed allocation control command in S4 or the cooperative control command in S5, and confirm the command execution status based on the actual output status of each actuator.

[0011] Furthermore, S1 includes:

[0012] S101. Environmental information is collected through an environmental perception module, including road curvature and road surface adhesion coefficient; vehicle status information is collected through a vehicle status perception module, including actual longitudinal vehicle speed, lateral vehicle speed, yaw rate, vehicle three-axis acceleration, vehicle three-axis angular velocity, wheel speed of each wheel, and steering wheel angle; driver intention information is collected through a driver intention perception module, including throttle opening, brake pedal travel, and steering wheel torque; and the actual output status of each actuator is collected through an execution status feedback module, including actual torque of the drive motor, actual pressure of the brake wheel cylinder, actual steering angle, and actual damping force of the suspension.

[0013] S102. Preprocess the signal acquired in S101. The preprocessing includes filtering, noise reduction and calibration.

[0014] S103. Perform a rationality check on the preprocessed signals in S102 and compare each signal with the preset normal working range. If all signals are normal and do not exceed the normal range, proceed directly to S2. If an abnormal signal exceeds the normal range, mark the sensor corresponding to the abnormal signal as faulty and replace each abnormal signal with backup sensor data or estimated data from the vehicle state estimation unit before proceeding to S2.

[0015] Furthermore, S2 includes:

[0016] S201. Establish the state equation based on the two-degree-of-freedom vehicle model. Take the steering wheel angle and actual longitudinal vehicle speed as inputs, and the yaw rate and lateral acceleration as observations. Use an unscented Kalman filter to make the optimal estimate of the centroid sideslip angle and obtain the estimated centroid sideslip angle.

[0017] S202. Calculate the feedforward yaw moment based on the road curvature and actual longitudinal vehicle speed using the following formula: ;

[0018] In the formula, To compensate for yaw moment; For road curvature; The feedforward gain related to the actual longitudinal vehicle speed is pre-calibrated into a MAP table for lookup. The MAP table is a table showing the relationship between the actual longitudinal vehicle speed and... The mapping table is calculated during calibration using the following formula: ,in, For vehicle mass parameters, For vehicle wheelbase parameters, This represents the equivalent lateral stiffness of the vehicle's front axle. This is the equivalent lateral stiffness of the vehicle's rear axle. This refers to the actual longitudinal speed;

[0019] S203. Boundary constraints for vehicle stability correction based on road surface adhesion coefficient, including: dynamically correcting the upper limit of yaw rate and the upper limit of center-of-gravity sideslip angle based on tire friction circle theory and phase plane stability domain analysis method, using actual longitudinal vehicle speed and vertical load as input; wherein, the dynamic upper limit of yaw rate... Dynamic upper limit of centroid sideslip angle In the formula, For low adhesion attenuation coefficient, This refers to the actual longitudinal vehicle speed. The road surface adhesion coefficient, It is the acceleration due to gravity;

[0020] S204. The total expected longitudinal force demand of the vehicle is calculated using a PID controller based on the longitudinal speed deviation; wherein, the actual longitudinal speed deviation is the difference between the ideal longitudinal speed and the actual longitudinal speed; the ideal longitudinal speed is obtained by converting the throttle opening and brake pedal travel collected by the driver intention perception module through the vehicle inverse longitudinal dynamics model; the actual longitudinal speed is obtained after being collected, preprocessed, and verified through S101-S103.

[0021] S205. Based on the yaw rate deviation and the center of gravity sideslip angle deviation, the sliding mode control algorithm is used to calculate the additional expected yaw moment and roll moment required to maintain vehicle stability.

[0022] Wherein, the yaw rate deviation is the difference between the ideal yaw rate and the actual yaw rate. The ideal yaw rate is calculated by the two-degree-of-freedom vehicle model based on the steering wheel angle and the actual longitudinal vehicle speed. The actual yaw rate is the yaw rate after being collected, preprocessed and verified by S101-S103. Furthermore, both the ideal yaw rate and the actual yaw rate are corrected by the upper limit of yaw rate obtained in S203.

[0023] The centroid sideslip angle deviation is the difference between the ideal centroid sideslip angle and the estimated centroid sideslip angle; the ideal centroid sideslip angle is set to 0; the estimated centroid sideslip angle is obtained through S201; and the estimated centroid sideslip angle is obtained through the dynamic upper limit of the centroid sideslip angle obtained in S203. Perform amplitude limiting correction;

[0024] S206. The additional desired yaw moment obtained in S205 is corrected using the feedforward compensation yaw moment obtained in S202.

[0025] Furthermore, S5 includes:

[0026] S501. Construct a multi-objective optimization function;

[0027] The multi-objective optimization function includes a vehicle stability sub-objective, a tire adhesion utilization rate sub-objective, and an actuator energy consumption sub-objective. The vehicle stability sub-objective aims to minimize the yaw rate tracking error, the tire adhesion utilization rate sub-objective aims to minimize the sum of squares of the load rates of each tire, and the actuator energy consumption sub-objective aims to minimize the electrical power consumption of each actuator.

[0028] S502. The sub-objective functions in S501 are integrated into a comprehensive objective function by weighting coefficients. The basic weighting coefficients are dynamically adjusted according to driving conditions: the stability weight is increased in extreme conditions and the energy consumption weight is increased in normal conditions.

[0029] S503. Under the condition of satisfying the physical constraints of each execution subsystem, the comprehensive objective function is transformed into the standard quadratic programming (QP) form to obtain the optimal control allocation model based on the quadratic programming (QP) solution, and the effective set method is used to solve it.

[0030] S504. Obtain the optimal cooperative control allocation scheme based on the solution obtained by the effective set method, and generate cooperative control commands based on the cooperative control allocation scheme; wherein, the cooperative control allocation scheme includes the target driving torque, target braking torque, target steering angle and target suspension damping force of each wheel.

[0031] Furthermore, the optimal control allocation model is as follows:

[0032] ;

[0033] In the formula, All are dynamically weighted coefficients. These are the weighting coefficients for the stability sub-objective. The weighting coefficients for the tire adhesion utilization sub-objective are as follows: The weighting coefficients for the energy consumption sub-objectives; This is the sensitivity matrix of the yaw rate to the control quantity; The decision variable vector includes the driving torque, braking torque, steering angle, and suspension damping force of each wheel; for Transpose of; Ideal yaw rate; The weight matrix for tire adhesion utilization is a diagonal matrix. The actuator energy consumption weight matrix is ​​a diagonal matrix. For equality constraint matrices / vectors; Inequality constraint matrix / vector; These are the calibration basis weighting coefficients for the stability sub-objective. The calibration basis weighting coefficient for the tire adhesion utilization sub-target; The calibration constant is used to dynamically adjust the gain; The attachment margin index; As a stability risk index; This is the actual yaw rate; This is the upper limit of the yaw rate; To estimate the centroid sideslip angle; This is the dynamic upper limit of the centroid sideslip angle; For the longitudinal force of each tire; The lateral forces of each tire; For the vertical force of each tire; It has a low adhesion attenuation coefficient; The road surface adhesion coefficient; It is the acceleration due to gravity; This refers to the actual longitudinal speed; For road curvature; For vehicle mass parameters; This refers to the vehicle's wheelbase parameter. This is the equivalent lateral stiffness of the vehicle's front axle. This is the equivalent lateral stiffness of the rear axle of the vehicle.

[0034] Furthermore, the calibration base weighting coefficients in S502 are dynamically adjusted based on driving conditions, including: determining whether the vehicle status information meets the priority conditions of economy or stability, determining the driving conditions and vehicle driving mode, and determining the calibration base weighting coefficients, specifically including:

[0035] The absolute value of the lateral acceleration is compared with a first threshold and a second threshold, wherein the second threshold is greater than the first threshold;

[0036] The absolute value of the yaw rate deviation is compared with the third threshold and the fourth threshold, wherein the fourth threshold is greater than the third threshold;

[0037] Compare the estimated absolute value of the centroid sideslip angle with the fifth threshold;

[0038] If the absolute value of the lateral acceleration is not greater than the first threshold and the absolute value of the yaw rate deviation is not greater than the third threshold, then the vehicle status information is determined to meet the economy priority condition, which is the normal operating condition, and the vehicle driving mode is determined to be the economy priority control mode.

[0039] If the absolute value of the lateral acceleration is greater than the second threshold, or the absolute value of the yaw rate deviation is greater than the fourth threshold, or the absolute value of the estimated center of gravity sideslip angle is greater than the fifth threshold, then the vehicle state information is determined to meet the stability priority condition, which is the extreme condition, and the vehicle driving mode is determined to be the stability priority control mode.

[0040] The economy-first control mode and the stability-first control mode are used to adjust the weight coefficients of the sub-objectives in the optimal control allocation model; the energy consumption sub-objective of the economy-first control mode is used as the calibration basis weighting coefficient. The calibration basis weighting coefficients for the stability sub-objective The calibration basis weighting coefficient for the tire adhesion utilization rate sub-target Decrease sequentially; the calibration basis weighting coefficients of the stability sub-objectives in the stability priority control mode. The calibration basis weighting coefficient for the tire adhesion utilization rate sub-target The calibration basis weighting coefficients for energy consumption sub-targets Decrease sequentially;

[0041] If the vehicle status information does not meet the priority conditions of economy and stability, the vehicle driving mode will remain unchanged.

[0042] Furthermore, it also includes: S7, detecting whether each actuator is in a fault state; if any actuator is in a fault state, then determining that the vehicle driving mode is a fault-degradation control mode.

[0043] The detection of whether each actuator is in a faulty state is performed through a fault diagnosis circuit, including:

[0044] When no drive signal is output to the corresponding execution module, the voltage value of the drive end of each actuator is collected in real time through the fault diagnosis circuit, and the type of actuator fault is determined based on the collected voltage.

[0045] The fault diagnosis circuit includes a voltage divider resistor network, a current-limiting protection resistor, a clamping diode, an analog-to-digital converter sampling port, and a reference voltage. The voltage divider resistor network includes a first resistor and a second resistor connected in series. One end of the first resistor is connected to the reference voltage, and the other end is connected to the sampling node P of the corresponding actuator's drive end. One end of the second resistor is connected to the sampling node, and the other end is grounded. The sampling node is connected to the analog-to-digital converter sampling port via the current-limiting protection resistor, and a clamping diode is provided at the sampling node.

[0046] The present invention provides a chassis intelligent electronic control system, comprising a perception layer, a decision-making layer, and an execution layer;

[0047] The perception layer is configured to collect environmental information, vehicle status information, driver intention information, and the actual output status of each actuator, and send them to the decision layer.

[0048] The decision layer is configured to: receive signals collected by the perception layer, perform preprocessing and rationality verification; calculate feedforward compensation yaw moment based on the environmental information and vehicle state information, determine boundary constraints for correcting vehicle stability, and calculate the generalized target force and moment required to maintain vehicle stability, wherein the generalized target force and moment include total expected longitudinal force, additional expected yaw moment, and roll moment; obtain the current available execution capability range of each actuator, and determine whether the generalized target force and moment exceed the current available execution capability range of each actuator; if the generalized target force and moment do not exceed the current available execution capability range of each actuator... Using the execution capability range, the generalized target force and torque are converted into fixed allocation control commands for each wheel according to a preset fixed allocation rule; if the generalized target force and torque exceed the current available execution capability range of a certain actuator, a multi-objective quadratic programming (QP) model is constructed, and based on the rolling time-domain optimization of model predictive control, the target driving torque, target braking torque, target steering angle, and target suspension damping force of each wheel are redistributed and optimized to generate cooperative control commands; the fixed allocation control commands or the cooperative control commands are sent to the execution layer; the execution status of the commands is confirmed based on the actual output state of each actuator;

[0049] The execution layer is configured to receive the fixed allocation control command or the collaborative control command issued by the decision layer, and control the operation of each actuator.

[0050] Furthermore, the decision-making layer is a chassis domain controller, which includes a signal processing unit, a logic verification unit, a vehicle state estimation unit, a model solving unit, a cooperative control unit, and a communication unit.

[0051] The signal processing unit is configured to receive signals collected by the sensing layer and perform preprocessing.

[0052] The logic verification unit is configured to perform a rationality check on the preprocessed signal;

[0053] The vehicle state estimation unit is configured to: establish a state equation based on a two-degree-of-freedom vehicle model, take the steering wheel angle and actual longitudinal vehicle speed as inputs, take the yaw rate and lateral acceleration as observations, and use an unscented Kalman filter to perform optimal estimation of the centroid sideslip angle to obtain the estimated centroid sideslip angle.

[0054] The model solving unit is configured to: calculate the feedforward compensation yaw moment based on the environmental information and vehicle state information, determine the boundary constraints for correcting vehicle stability, and calculate the generalized target force and moment required to maintain vehicle stability, wherein the generalized target force and moment include the total desired longitudinal force, the additional desired yaw moment, and the roll moment.

[0055] The collaborative control unit is configured to: acquire the current available execution capability range of each actuator, and determine whether the generalized target force and torque exceed the current available execution capability range of each actuator; if the generalized target force and torque do not exceed the current available execution capability range of each actuator, convert the generalized target force and torque into fixed allocation control commands for each wheel according to a preset fixed allocation rule; if the generalized target force and torque exceed the current available execution capability range of a certain actuator, construct a multi-objective quadratic programming (QP) model, and based on the rolling time-domain optimization of model predictive control, redistribute and optimize the target driving torque, target braking torque, target steering angle, and target suspension damping force of each wheel to generate collaborative control commands; and confirm the command execution status based on the actual output state of each actuator.

[0056] The communication unit is configured to send the fixed allocation control command or the cooperative control command to the execution layer.

[0057] One vehicle of the present invention includes the aforementioned intelligent electronic control system for the chassis.

[0058] The beneficial effects of this invention are:

[0059] I. The chassis domain controller of this invention coordinates the execution commands of drive, braking, steering and suspension in a unified manner, which can prevent command conflicts.

[0060] Second, this invention has an environmental perception feedforward function, which introduces the road curvature and road surface adhesion coefficient from the environmental information, calculates the feedforward compensation yaw moment and the boundary constraints for correcting vehicle stability, and corrects the control target in real time before the vehicle enters a complex road (curve / low-adhesion road surface), improving the accuracy of pre-aiming control and changing "post-correction" to "pre-control".

[0061] III. When the target generalized force exceeds the capability of a single actuator, this invention constructs a multi-objective quadratic programming (QP) model, including: (a) yaw rate tracking deviation; (b) tire adhesion utilization rate (minimizing the sum of squares of load rates to prevent tire slippage); and (c) actuator energy consumption. Based on model predictive control and rolling time-domain optimization, different modes are switched through dynamic weighting, achieving a balance between stability, adhesion limit, and energy economy.

[0062] Fourth, this invention improves the fault diagnosis circuit. During the intervals when the domain controller is not outputting a drive signal (high impedance state) (idle periods of the control cycle), it reads the node voltage through a single ADC port. Different faults can be distinguished using only a single port and a single sample. Each actuator diagnosis in this invention requires only one ADC port, improving upon the traditional solution that requires two ADC ports. Furthermore, the fault diagnosis circuit in this invention does not require forced drive of the solenoid valve, achieving "non-destructive" diagnosis and significantly improving system safety. Attached Figure Description

[0063] To make the objectives, technical solutions, and beneficial effects of this invention clearer, the following figures are provided for illustration:

[0064] Figure 1 This is a flowchart illustrating the intelligent electronic control method for the chassis of the present invention.

[0065] Figure 2 This is a flowchart illustrating step S5 of the intelligent electronic control method for chassis of the present invention;

[0066] Figure 3 This is a schematic diagram of the control mode switching state machine of the present invention;

[0067] Figure 4 This is an architecture diagram of the intelligent electronic control system for the chassis of the present invention;

[0068] Figure 5 This is an architecture diagram of the decision-making layer of the intelligent electronic control system for the chassis of the present invention;

[0069] Figure 6 This is a schematic diagram of the fault diagnosis circuit of the intelligent electronic control system for chassis according to the present invention.

[0070] The following labels are shown in the attached diagram:

[0071] 100-Perception layer, 110-Environmental perception module, 111-Millimeter-wave radar, 112-LiDAR, 113-Camera, 120-Vehicle status perception module, 121-Inertial measurement unit, 122-Wheel speed sensor, 123-Steering wheel angle sensor, 124-Yaw rate sensor, 130-Driver intention perception module, 131-Accelerator pedal position sensor, 132-Brake pedal position sensor, 133-Steering wheel torque sensor, 140-Actuator status feedback module;

[0072] 200-Decision layer, 210-Chassis domain controller, 211-Signal processing unit, 212-Logic verification unit, 213-Vehicle state estimation unit, 214-Model solving unit, 215-Cooperative control unit, 2151-Execution capability assessment subunit, 2152-Command arbitration subunit, 2153-Failure analysis subunit, 216-Communication unit;

[0073] 300 - Execution layer, 310 - Drive-by-wire module, 320 - Brake-by-wire module, 330 - Steering-by-wire module, 340 - Active suspension-by-wire module. Detailed Implementation

[0074] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0075] Example 1:

[0076] like Figures 1-6 As shown, this embodiment provides a chassis intelligent electronic control method, including the following steps:

[0077] S1. Acquire environmental information, vehicle status information, driver intention information, and the actual output status of each actuator.

[0078] This step specifically includes: S101, collecting environmental information through the environmental perception module 110, the environmental information including road curvature and road surface adhesion coefficient; collecting vehicle status information through the vehicle status perception module 120, the vehicle status information including actual longitudinal vehicle speed, lateral vehicle speed, yaw rate, vehicle three-axis acceleration, vehicle three-axis angular velocity, wheel speed of each wheel, and steering wheel angle; collecting driver intention information through the driver intention perception module 130, the driver intention information including throttle opening, brake pedal travel, and steering wheel torque; and collecting the actual output status of each actuator through the execution status feedback module, the actual output status of each actuator including actual drive motor torque, actual brake cylinder pressure, actual steering angle, and actual suspension damping force.

[0079] The vehicle's three-axis acceleration includes longitudinal acceleration, lateral acceleration, and vertical acceleration. The vehicle's three-axis angular velocities include roll rate, pitch rate, and yaw rate.

[0080] The environmental perception module 110 includes a millimeter-wave radar 111, a lidar 112, and a camera 113, used to collect environmental information such as road type, obstacle information, road curvature, and road surface adhesion coefficient in front of the vehicle. The millimeter-wave radar 111 detects the distance and relative speed of obstacles ahead, the lidar 112 constructs a 3D point cloud map of the vehicle's surroundings, and the camera 113 identifies lane lines, traffic signs, and pedestrians. The environmental perception module 110 communicates with the chassis domain controller 210 via an in-vehicle Ethernet (100BASE-T1) to meet the large data transmission requirements of the camera 113 and lidar 112.

[0081] The vehicle state perception module 120 includes an inertial measurement unit (IMU) 121, wheel speed sensors 122, a steering wheel angle sensor 123, and a yaw rate sensor 124. The vehicle state perception module 120 acquires the vehicle's three-axis acceleration and three-axis angular velocity through the IMU 121, acquires the wheel speeds of the four wheels through the wheel speed sensors 122, acquires the steering wheel angle through the steering wheel angle sensor 123, and acquires the yaw rate through the yaw rate sensor 124. The three-axis angular velocities acquired by the IMU 121 include the yaw rate, which is the same as the yaw rate acquired by the yaw rate sensor 124.

[0082] The driver intention perception module 130 includes an accelerator pedal position sensor 131, a brake pedal position sensor 132, and a steering wheel torque sensor 133. The driver intention perception module 130 collects the accelerator opening through the accelerator pedal position sensor 131, collects the brake pedal travel through the brake pedal position sensor 132, and collects the steering wheel torque applied by the driver through the steering wheel torque sensor 133.

[0083] The actuator status feedback module 140 includes a drive motor actual torque feedback unit, a brake wheel cylinder actual pressure feedback unit, an actual steering angle feedback unit, and a suspension actual damping force feedback unit. In this embodiment, the drive motor controller of the drive-by-wire module 310 feeds back the actual torque value to the chassis domain controller 210 in real time via the CAN FD bus; a pressure sensor is installed at the brake wheel cylinder of the brake-by-wire module 320 to feed back the actual braking pressure to the chassis domain controller 210; an absolute encoder is installed at the output shaft of the steering-by-wire module 330 to feed back the actual steering angle to the chassis domain controller 210; and the CDC damper of the active suspension-by-wire module 340 integrates a displacement sensor and a force sensor to feed back the actual damping force to the chassis domain controller 210. The above feedback signals are updated at a frequency of not less than 100Hz to ensure the real-time performance of the closed-loop control.

[0084] The environmental perception module 110, vehicle status perception module 120, driver intention perception module 130 and actuator status feedback module 140 all communicate with the chassis domain controller 210 via CAN FD bus (500kbps).

[0085] S102. Preprocess the signal acquired in S101. The preprocessing includes filtering, noise reduction and calibration.

[0086] The signal processing unit 211 performs filtering, denoising, and calibration processing on the received various sensor signals to eliminate noise interference and convert the original signals into standardized physical quantity values. Filtering can employ methods such as low-pass filters, Kalman filters, or mean filters to remove high-frequency noise from the signal. Calibration processing converts the original voltage or digital signals output by the sensors into physically meaningful engineering unit values ​​(such as vehicle speed in m / s, acceleration in m / s², angle in rad, etc.).

[0087] S103. Perform a rationality check on the preprocessed signals in S102 and compare each signal with the preset normal working range. If all signals are normal signals that do not exceed the normal range, proceed directly to S2. If an abnormal signal that exceeds the normal range appears, mark the sensor corresponding to the abnormal signal as faulty and replace each abnormal signal with backup sensor data or estimated data from the vehicle state estimation unit 213 before proceeding to S2.

[0088] The logic verification unit 212 performs rationality verification and fault identification on the processed signals, determining whether each sensor signal is within the preset normal operating range. If a signal exceeds the normal range, the corresponding sensor is marked as faulty. For example, the vehicle speed signal should be within the range of 0 to the vehicle's maximum speed, the yaw rate signal should be within a reasonable range (e.g., -0.8 rad / s to 0.8 rad / s), and there should not be excessive differences between the wheel speed signals. If a signal exceeds the normal range, the corresponding sensor is marked as faulty, and in subsequent control, backup sensor data or estimation by the vehicle state estimation unit 213 is used.

[0089] S2. Based on the environmental information and vehicle state information, calculate the feedforward compensation yaw moment, determine the boundary constraints for correcting vehicle stability, and calculate the generalized target force and moment required to maintain vehicle stability. The generalized target force and moment include the total desired longitudinal force, the additional desired yaw moment, and the roll moment.

[0090] This step specifically includes: S201, establishing a state equation based on a two-degree-of-freedom vehicle model, using the steering wheel angle and actual longitudinal vehicle speed as inputs, and the yaw rate and lateral acceleration as observations, and using an unscented Kalman filter through the vehicle state estimation unit 213 to optimally estimate the centroid sideslip angle, thereby obtaining the estimated centroid sideslip angle.

[0091] S202. Using model solver 214, calculate the feedforward compensation yaw moment based on road curvature and actual longitudinal vehicle speed, using the following formula: ;

[0092] In the formula, To compensate for yaw moment; For road curvature; The feedforward gain related to the actual longitudinal vehicle speed is pre-calibrated into a MAP table for lookup. The MAP table is a table showing the relationship between the actual longitudinal vehicle speed and... The mapping table is calculated during calibration using the following formula: ,in, For vehicle mass parameters, For vehicle wheelbase parameters, This represents the equivalent lateral stiffness of the vehicle's front axle. This is the equivalent lateral stiffness of the vehicle's rear axle. This represents the actual longitudinal speed.

[0093] S203. Boundary constraints for vehicle stability correction based on road surface adhesion coefficient, including: dynamically correcting the upper limit of yaw rate and the upper limit of center-of-gravity sideslip angle based on tire friction circle theory and phase plane stability domain analysis method, using actual longitudinal vehicle speed and vertical load as input; wherein, the dynamic upper limit of yaw rate... Dynamic upper limit of centroid sideslip angle In the formula, For low adhesion attenuation coefficient, This refers to the actual longitudinal vehicle speed. The road surface adhesion coefficient, This represents gravitational acceleration. Vertical load is a parameter specific to the vehicle's configuration.

[0094] S204. The model solving unit 214 calculates the total expected longitudinal force demand of the vehicle based on the longitudinal speed deviation using a PID controller. The actual longitudinal speed deviation is the difference between the ideal longitudinal speed and the actual longitudinal speed. The ideal longitudinal speed is obtained by converting the throttle opening and brake pedal travel collected by the driver intention perception module 130 using the vehicle's inverse longitudinal dynamics model. The actual longitudinal speed is obtained after being collected, preprocessed, and verified through steps S101-S103.

[0095] S205. Using the model solver 214, based on the yaw rate deviation and the center of gravity sideslip angle deviation, the sliding mode control algorithm is used to calculate the additional expected yaw moment and roll moment required to maintain vehicle stability.

[0096] The yaw rate deviation is the difference between the ideal yaw rate and the actual yaw rate. The ideal yaw rate is calculated by the two-degree-of-freedom vehicle model based on the steering wheel angle and the actual longitudinal vehicle speed. The actual yaw rate is the yaw rate among the three-axis angular velocities of the vehicle after being collected, preprocessed, and verified by S101-S103. Both the ideal yaw rate and the actual yaw rate are corrected by the upper limit of the yaw rate obtained in S203.

[0097] The centroid sideslip angle deviation is the difference between the ideal centroid sideslip angle and the estimated centroid sideslip angle; the ideal centroid sideslip angle is set to 0; the estimated centroid sideslip angle is obtained through S201; and the estimated centroid sideslip angle is obtained through the dynamic upper limit of the centroid sideslip angle obtained in S203. Amplitude limiting correction was performed.

[0098] S206. The additional desired yaw moment obtained in S205 is corrected using the feedforward compensation yaw moment obtained in S202.

[0099] S3. Obtain the current available execution capability range of each actuator, and determine whether the generalized target force and torque exceed the current available execution capability range of each actuator; if the generalized target force and torque do not exceed the current available execution capability range of each actuator, proceed to S4; if the generalized target force and torque exceed the current available execution capability range of a certain actuator, proceed to S5.

[0100] Specifically, the execution capability assessment subunit 2151 of the cooperative control unit 215 obtains the current available execution capability range of each execution subsystem, including: the maximum available drive torque of the steer-by-wire module 310, the maximum available braking torque of the brake-by-wire module 320, the maximum available steering angle and steering angular velocity of the steering-by-wire module 330, and the available damping adjustment range of the active suspension-by-wire module 340. The cooperative control unit 215 determines whether the generalized target force / torque calculated in step S2 exceeds the execution capability of each execution subsystem.

[0101] S4. If the generalized target force and torque do not exceed the current available execution capability range of each actuator, the generalized target force and torque are converted into fixed allocation control commands for each wheel according to the preset fixed allocation rules.

[0102] The preset fixed allocation rules are vehicle preset parameters, such as allocating according to a fixed ratio of the front and rear axles of the vehicle.

[0103] S5. When the generalized target force and torque exceed the current available execution capability range of a certain actuator, a multi-objective quadratic programming (QP) model is constructed. Based on the rolling time-domain optimization of model predictive control, the target driving torque, target braking torque, target steering angle, and target suspension damping force of each wheel are redistributed and optimized to generate cooperative control commands.

[0104] This step S5 specifically includes: S501, constructing a multi-objective optimization function;

[0105] The multi-objective optimization function includes a vehicle stability sub-objective, a tire adhesion utilization sub-objective, and an actuator energy consumption sub-objective. The vehicle stability sub-objective aims to minimize the yaw rate tracking error, the tire adhesion utilization sub-objective aims to minimize the sum of squares of the load rates of each tire, and the actuator energy consumption sub-objective aims to minimize the electrical power consumption of each actuator.

[0106] S502. The sub-objective functions in S501 are integrated into a comprehensive objective function by weighting coefficients. The calibrated basic weighting coefficients are dynamically adjusted according to driving conditions: the stability weight is increased under extreme conditions, and the energy consumption weight is increased under normal conditions.

[0107] Optionally, the calibration base weighting coefficient in S502 is dynamically adjusted according to driving conditions, including: determining whether the vehicle status information meets the priority conditions of economy or stability, determining the driving conditions and vehicle driving mode, and determining the calibration base weighting coefficient, specifically including:

[0108] The absolute value of the lateral acceleration is compared with a first threshold and a second threshold, wherein the second threshold is greater than the first threshold;

[0109] The absolute value of the yaw rate deviation is compared with the third threshold and the fourth threshold, wherein the fourth threshold is greater than the third threshold;

[0110] Compare the estimated absolute value of the centroid sideslip angle with the fifth threshold;

[0111] If the absolute value of the lateral acceleration is not greater than the first threshold and the absolute value of the yaw rate deviation is not greater than the third threshold, then the vehicle status information is determined to meet the economy priority condition, which is the normal operating condition, and the vehicle driving mode is determined to be the economy priority control mode.

[0112] If the absolute value of the lateral acceleration is greater than the second threshold, or the absolute value of the yaw rate deviation is greater than the fourth threshold, or the absolute value of the estimated center of gravity sideslip angle is greater than the fifth threshold, then the vehicle state information is determined to meet the stability priority condition, which is the extreme condition, and the vehicle driving mode is determined to be the stability priority control mode.

[0113] The economy-first control mode and the stability-first control mode are used to adjust the weight coefficients of the sub-objectives in the optimal control allocation model; the energy consumption sub-objective of the economy-first control mode is used as the calibration basis weighting coefficient. The calibration basis weighting coefficients for the stability sub-objective The calibration basis weighting coefficient for the tire adhesion utilization rate sub-target Decrease sequentially; the calibration basis weighting coefficients of the stability sub-objectives in the stability priority control mode. The calibration basis weighting coefficient for the tire adhesion utilization rate sub-target The calibration basis weighting coefficients for energy consumption sub-targets Decrease sequentially;

[0114] If the vehicle status information does not meet the priority conditions of economy and stability, the vehicle driving mode will remain unchanged.

[0115] For example, when the absolute value of lateral acceleration |ay| is not greater than 0.3g and the absolute value of yaw rate deviation |Δγ| is not greater than 0.05 rad / s, the vehicle state information is determined to meet the economy priority condition, which is a normal operating condition, and the economy priority control mode is adopted. In this mode, the calibration basis weighting coefficient of the energy consumption sub-objective in the comprehensive objective function is... Set to 0.5, the baseline weighting coefficient for the stability sub-objective. Set to 0.3, tire adhesion utilization weight Set to 0.2 to prioritize ensuring that each actuator operates in the high-efficiency range and reduce the overall vehicle energy consumption.

[0116] When the absolute value of lateral acceleration |ay| is greater than 0.4g, the absolute value of yaw rate deviation |Δγ| is greater than 0.1 rad / s, or the absolute value of estimated center-of-gravity sideslip angle |β| is greater than 0.05 rad, the vehicle state information is determined to meet the stability priority condition, which is the extreme condition, and the stability priority control mode is adopted. In this mode, the calibration basis weighting coefficients of the stability sub-objective in the comprehensive objective function are... Set to 0.6, the calibration basis weighting coefficient for the tire adhesion utilization sub-target. Set to 0.3, the basic weighting coefficient for the energy consumption sub-target. Set to 0.1 to prioritize ensuring that the vehicle's yaw rate tracks the ideal value and prevent vehicle instability.

[0117] When the vehicle's driving state does not meet the conditions of any of the above modes (e.g., 0.3g < |ay| ≤ 0.4g), the current control mode is maintained to avoid frequent switching of the control mode switching state machine at the boundary.

[0118] S503. Under the condition of satisfying the physical constraints of each execution subsystem, the comprehensive objective function is transformed into the form of standard quadratic programming (QP) to obtain the optimal control allocation model based on the solution of quadratic programming (QP), and the effective set method is used to solve it.

[0119] In this embodiment, the solution speed meets the real-time control cycle requirements of the chassis electronic control system. An Infineon AURIX TC397 embedded microcontroller is used, with a single-step solution time of no more than 3ms, satisfying the real-time requirement of a 10ms control cycle. The solution result is the optimal cooperative control allocation scheme, including the target driving torque, target braking torque, target steering angle, and target suspension damping force for each wheel.

[0120] The optimal control allocation model is as follows:

[0121] ;

[0122] In the formula, All are dynamically weighted coefficients. These are the weighting coefficients for the stability sub-objective. The weighting coefficients for the tire adhesion utilization sub-objective are as follows: The weighting coefficients for the energy consumption sub-objectives; This is the sensitivity matrix of the yaw rate to the control quantity; The decision variable vector includes the driving torque, braking torque, steering angle, and suspension damping force of each wheel; for Transpose of; Ideal yaw rate; The weight matrix for tire adhesion utilization is a diagonal matrix. The actuator energy consumption weight matrix is ​​a diagonal matrix. For equality constraint matrices / vectors; Inequality constraint matrix / vector; These are the calibration basis weighting coefficients for the stability sub-objective. The calibration basis weighting coefficient for the tire adhesion utilization sub-target; The calibration constant is used to dynamically adjust the gain; The attachment margin index; As a stability risk index; This is the actual yaw rate; This is the upper limit of the yaw rate; To estimate the centroid sideslip angle; This is the dynamic upper limit of the centroid sideslip angle; For the longitudinal force of each tire; The lateral forces of each tire; For the vertical force of each tire; It has a low adhesion attenuation coefficient; The road surface adhesion coefficient; It is the acceleration due to gravity; This refers to the actual longitudinal speed; For road curvature; For vehicle mass parameters; This refers to the vehicle's wheelbase parameter. This is the equivalent lateral stiffness of the vehicle's front axle. This is the equivalent lateral stiffness of the rear axle of the vehicle.

[0123] S504. Obtain the optimal cooperative control allocation scheme based on the solution obtained by the effective set method, and generate cooperative control commands based on the cooperative control allocation scheme; wherein, the cooperative control allocation scheme includes the target driving torque, target braking torque, target steering angle and target suspension damping force of each wheel.

[0124] S6. Control each actuator based on the fixed allocation control command in S4 or the cooperative control command in S5, and confirm the command execution status based on the actual output status of each actuator.

[0125] Confirming the execution status of the command means: collecting the actual output status of each actuator through the actuator status feedback module 140, comparing each actual output status with the target command issued by S4 or S5, detecting whether the control command has been executed properly and whether there is any execution deviation, and then feeding back the execution status to the decision layer 200 to achieve closed-loop control.

[0126] Control commands are sent to each execution module of the execution layer 300 via communication unit 216. The actuator refers to the drive end corresponding to each execution module, such as a drive motor, brake actuator, or steering motor. Specifically, the drive-by-wire module 310 controls the drive motor to output a corresponding torque based on the target drive torque; the brake-by-wire module 320 controls the brake actuator to generate a corresponding braking force based on the target braking torque; the steering-by-wire module 330 controls the steering motor to rotate a corresponding angle based on the target steering angle; and the active suspension-by-wire module 340 adjusts the damping force of the CDC shock absorber based on the target damping force and adjusts the air spring stiffness as needed.

[0127] S7. Detect whether each actuator is in a fault state. If any actuator is in a fault state, determine that the vehicle driving mode is fault-degradation control mode. The priority of entering fault-degradation control mode is higher than the priority of switching between stability-priority control mode and economy-priority control mode.

[0128] During execution, the decision-making layer 200 also uses a fault diagnosis circuit: when no drive signal is output to the corresponding execution module, the fault diagnosis circuit collects the voltage value of the drive end of each actuator in real time, and determines the type of actuator fault based on the collected voltage. When an actuator fault is detected, it automatically switches to the fault degradation control mode, triggering the failure analysis subunit 2153 to formulate corresponding degradation strategies based on the fault type and severity. Degradation strategies include: shutting down the faulty actuator, having other normal actuators compensate for its function, reducing the vehicle's operating speed, or guiding the vehicle to a safe stop. For example, limiting the maximum vehicle speed to no more than 30 km / h and limiting the maximum steering wheel angle to no more than ±30°, while illuminating the instrument panel fault warning light to remind the driver to stop safely.

[0129] The fault diagnosis circuit of the decision layer 200 includes a voltage divider resistor network, a current-limiting protection resistor, a clamping diode, an analog-to-digital converter (ADC) sampling port, and a reference voltage. For example... Figure 6 As shown, the voltage divider resistor network consists of a first resistor R1 and a second resistor R2 connected in series. One end of the first resistor R1 is connected to the reference voltage Vref, and the other end of the first resistor R1 is connected to the sampling node P of the corresponding actuator drive end. One end of the second resistor R2 is connected to the sampling node P, and the other end of the second resistor R2 is grounded. The sampling node P is connected to the ADC sampling port through the current limiting protection resistor R3 (1kΩ). A clamping diode D1 is set at the sampling node P (cathode connected to Vref=5V, anode grounded) to clamp the sampling node voltage within the range of 0-(Vref+0.7V). Since Vref=5V, this range is 0-5.7V to prevent external high voltage (such as 12V / 48V power supply) from backflowing and damaging the ADC port.

[0130] In this embodiment, the reference voltage Vref is 5.0V, the first resistor R1 is 10kΩ, and the second resistor R2 is 10kΩ. Therefore, the theoretical voltage of the sampling node under normal conditions is 2.5V, and the normal threshold range is set to 2.0V-3.0V. When the chassis domain controller 210 does not output a drive signal to the corresponding module, it acquires the voltage value of the sampling node P through the ADC sampling port. When the sampling voltage is 0V±0.1V, it is determined to be a short circuit to ground; when the sampling voltage is 4.8V-5.7V (i.e., close to or reaching the Vref clamping value), it is determined to be a short circuit to the power supply (including the case of a short circuit to a 12V / 48V external power supply, where the voltage is clamped to Vref); when the sampling voltage is in the range of 0.5V-1.8V or 3.2V-4.5V, it is determined to be a poor contact or a partial short circuit in the coil; when the sampling voltage is in the range of 1.8V-2.0V or 3.0V-3.2V, it is determined to be actuator parameter drift, and a warning signal is issued. The above diagnostics are performed once during the no-output drive signal period of each control cycle (10ms). This fault diagnosis circuit can perform the diagnosis without driving the solenoid valve, and the diagnosis of each actuator only requires one ADC port resource.

[0131] As can be seen from the above, during the interval when the domain controller does not output a drive signal (high impedance state) (the idle period of the control cycle), it reads the node voltage through a single ADC port. With just a single port and a single sample, it can distinguish: (1) 0V → short circuit to ground; (2) 0.9Vref → short circuit to power supply (12V / 48V) (clamped protection); (3) voltage deviates from the normal threshold → poor contact or partial short circuit of the coil. In this embodiment, each actuator diagnosis only requires 1 ADC port, which improves the traditional technical solution that requires 2 ADC ports. Moreover, the fault diagnosis circuit in this embodiment does not require forced drive of the solenoid valve, realizing "non-destructive" diagnosis and significantly improving system safety.

[0132] It is worth noting that during the switching process between fault degradation control mode, stability priority control mode and economy priority control mode, the chassis domain controller 210 performs first-order low-pass filtering on the output control quantity (the filtering time constant is set to 0.1s) to avoid vehicle impact caused by control quantity jumps; at the same time, an anti-integral saturation strategy is adopted, which freezes the integral term when the actuator reaches the saturation boundary to prevent control overshoot caused by integral depth saturation.

[0133] Example 2:

[0134] like Figures 1-6 As shown, this embodiment provides a chassis intelligent electronic control system, including a perception layer 100, a decision layer 200, and an execution layer 300.

[0135] The perception layer 100 is configured to collect environmental information, vehicle status information, driver intention information, and the actual output status of each actuator, and send them to the decision layer 200.

[0136] Specifically, the perception layer 100 includes an environment perception module 110, a vehicle state perception module 120, a driver intention perception module 130, and an actuator state feedback module 140.

[0137] The environmental perception module 110 includes a millimeter-wave radar 111, a lidar 112, and a camera 113, used to collect environmental information such as road type, obstacle information, road curvature, and road surface adhesion coefficient in front of the vehicle. The millimeter-wave radar 111 detects the distance and relative speed of obstacles ahead, the lidar 112 constructs a 3D point cloud map of the vehicle's surroundings, and the camera 113 identifies lane lines, traffic signs, and pedestrians. The environmental perception module 110 communicates with the chassis domain controller 210 via an in-vehicle Ethernet (100BASE-T1) to meet the large data transmission requirements of the camera 113 and lidar 112.

[0138] The vehicle state perception module 120 includes an inertial measurement unit (IMU) 121, wheel speed sensors 122, a steering wheel angle sensor 123, and a yaw rate sensor 124. The vehicle state perception module 120 acquires the vehicle's three-axis acceleration and three-axis angular velocity through the IMU 121, acquires the wheel speeds of the four wheels through the wheel speed sensors 122, acquires the steering wheel angle through the steering wheel angle sensor 123, and acquires the yaw rate through the yaw rate sensor 124. The three-axis angular velocities acquired by the IMU 121 include the yaw rate, which is the same as the yaw rate acquired by the yaw rate sensor 124.

[0139] The driver intention perception module 130 includes an accelerator pedal position sensor 131, a brake pedal position sensor 132, and a steering wheel torque sensor 133. The driver intention perception module 130 collects the accelerator opening through the accelerator pedal position sensor 131, collects the brake pedal travel through the brake pedal position sensor 132, and collects the steering wheel torque applied by the driver through the steering wheel torque sensor 133.

[0140] The actuator status feedback module 140 includes a drive motor actual torque feedback unit, a brake wheel cylinder actual pressure feedback unit, an actual steering angle feedback unit, and a suspension actual damping force feedback unit. In this embodiment, the drive motor controller of the drive-by-wire module 310 feeds back the actual torque value to the chassis domain controller 210 in real time via the CAN FD bus; a pressure sensor is installed at the brake wheel cylinder of the brake-by-wire module 320 to feed back the actual braking pressure to the chassis domain controller 210; an absolute encoder is installed at the output shaft of the steering-by-wire module 330 to feed back the actual steering angle to the chassis domain controller 210; and the CDC damper of the active suspension-by-wire module 340 integrates a displacement sensor and a force sensor to feed back the actual damping force to the chassis domain controller 210. The above feedback signals are updated at a frequency of not less than 100Hz to ensure the real-time performance of the closed-loop control.

[0141] The environmental perception module 110, vehicle status perception module 120, driver intention perception module 130 and actuator status feedback module 140 all communicate with the chassis domain controller 210 via CAN FD bus (500kbps).

[0142] The decision layer 200 is configured to: receive signals collected by the perception layer 100, and perform preprocessing and rationality verification; calculate the feedforward compensation yaw moment based on the environmental information and vehicle state information, determine the boundary constraints for correcting vehicle stability, and calculate the generalized target force and moment required to maintain vehicle stability, wherein the generalized target force and moment include the total expected longitudinal force, the additional expected yaw moment, and the roll moment; obtain the current available execution capability range of each actuator, and determine whether the generalized target force and moment exceed the current available execution capability range of each actuator; if the generalized target force and moment do not exceed the current available execution capability range of each actuator... If the available execution capability range is within a certain range, the generalized target force and torque are converted into fixed allocation control commands for each wheel according to a preset fixed allocation rule. If the generalized target force and torque exceed the current available execution capability range of a certain actuator, a multi-objective quadratic programming (QP) model is constructed. Based on the rolling time-domain optimization of model predictive control, the target driving torque, target braking torque, target steering angle, and target suspension damping force of each wheel are redistributed and optimized to generate cooperative control commands. The fixed allocation control commands or the cooperative control commands are sent to the execution layer 300. The execution status of the commands is confirmed based on the actual output status of each actuator.

[0143] The decision layer 200 is a chassis domain controller 210. The chassis domain controller 210 communicates with the sensors of the perception layer 100 via a CAN FD bus or an in-vehicle Ethernet, receives various sensor signals, and performs comprehensive processing. The chassis domain controller 210 includes a signal processing unit 211, a logic verification unit 212, a vehicle state estimation unit 213, a model solving unit 214, a cooperative control unit 215, and a communication unit 216.

[0144] The signal processing unit 211 is configured to receive the signal collected by the sensing layer 100 and perform preprocessing.

[0145] Specifically, the signal processing unit 211 is used to filter, denoise, and calibrate the received sensor signals, eliminating noise interference in the signals and converting the original signals into standardized physical quantity values. Filtering can employ methods such as low-pass filters, Kalman filters, or mean filters to remove high-frequency noise from the signals. Calibration converts the original voltage or digital signals output by the sensors into physically meaningful engineering unit values ​​(such as vehicle speed in m / s, acceleration in m / s², angle in rad, etc.).

[0146] The logic verification unit 212 is configured to perform a rationality check on the preprocessed signal.

[0147] Specifically, the logic verification unit 212 is used to perform rationality verification and fault identification on the processed signal, and to determine whether the signals of each sensor are within the preset normal working range. If a signal exceeds the normal range, the corresponding sensor is marked as faulty.

[0148] The vehicle state estimation unit 213 is configured to: establish a state equation based on a two-degree-of-freedom vehicle model; use the steering wheel angle and actual longitudinal vehicle speed as inputs; use yaw rate and lateral acceleration as observations; and employ an unscented Kalman filter to optimally estimate the centroid sideslip angle, thus obtaining the estimated centroid sideslip angle. The algorithm runs in the chassis domain controller 210 with a period of 5ms. The estimation results are compared and verified with the GPS / INS integrated navigation system, and the steady-state error does not exceed 0.01 rad.

[0149] The model solving unit 214 is configured to: calculate the feedforward compensation yaw moment based on the environmental information and vehicle state information, determine the boundary constraints for correcting vehicle stability, and calculate the generalized target force and moment required to maintain vehicle stability, wherein the generalized target force and moment include the total desired longitudinal force, the additional desired yaw moment, and the roll moment.

[0150] Specifically, the model solving unit 214 incorporates vehicle dynamics models, including a two-DOF vehicle model, a fourteen-DOF vehicle model, and a vehicle inverse longitudinal dynamics model. Based on the vehicle's current state information (longitudinal speed, yaw rate, and estimated centroid sideslip angle), the model solving unit 214 calculates the generalized target forces / moments required to maintain vehicle stability, such as the total desired longitudinal force, additional desired yaw moment, and roll moment. Simultaneously, the model solving unit 214 also receives estimated values ​​of road curvature ρ and road surface adhesion coefficient μ from the environmental perception module 110. It calculates the feedforward compensation yaw moment based on the road curvature (e.g., calculating the steady-state yaw moment compensation required for cornering based on road curvature and vehicle speed), and corrects the boundary constraints of vehicle stability based on the road surface adhesion coefficient (e.g., lowering the target upper limit for yaw rate tracking on low-adhesion surfaces), enabling the control strategy to have look-ahead capabilities.

[0151] The collaborative control unit 215 is configured to: acquire the current available execution capability range of each actuator, and determine whether the generalized target force and torque exceed the current available execution capability range of each actuator; if the generalized target force and torque do not exceed the current available execution capability range of each actuator, convert the generalized target force and torque into fixed allocation control commands for each wheel according to a preset fixed allocation rule; if the generalized target force and torque exceed the current available execution capability range of a certain actuator, construct a multi-objective quadratic programming (QP) model, and based on the rolling time-domain optimization of model predictive control, redistribute and optimize the target driving torque, target braking torque, target steering angle, and target suspension damping force of each wheel to generate collaborative control commands; and confirm the command execution status based on the actual output state of each actuator.

[0152] Specifically, the model solving unit 214 can implement steps S3-S5 of the intelligent electronic control method for the chassis in Embodiment 1. The collaborative control unit 215 includes an execution capability assessment subunit 2151, a command arbitration subunit 2152, and a failure analysis subunit 2153. The execution capability assessment subunit 2151 is used to evaluate the current available execution capability boundaries of each execution subsystem (drive, braking, steering, suspension) in real time; the command arbitration subunit 2152 is used to make priority decisions when multiple control objectives (such as stability objectives and economic objectives) conflict; the failure analysis subunit 2153 is used to detect the fault status of each execution subsystem and formulate degradation strategies.

[0153] The communication unit 216 is configured to send the fixed allocation control command or the cooperative control command to the execution layer 300.

[0154] Specifically, the communication unit 216 is used to interact with the perception layer 100, the execution layer 300, and other domain controllers of the vehicle (such as the intelligent driving domain controller, the body domain controller, etc.).

[0155] The fault diagnosis circuit of the decision layer 200 includes a voltage divider resistor network, a current-limiting protection resistor, a clamping diode, an analog-to-digital converter (ADC) sampling port, and a reference voltage. For example... Figure 6 As shown, the voltage divider resistor network consists of a first resistor R1 and a second resistor R2 connected in series. One end of the first resistor R1 is connected to the reference voltage Vref, and the other end of the first resistor R1 is connected to the sampling node P of the corresponding actuator drive end. One end of the second resistor R2 is connected to the sampling node P, and the other end of the second resistor R2 is grounded. The sampling node P is connected to the ADC sampling port through the current limiting protection resistor R3 (1kΩ). A clamping diode D1 is set at the sampling node P (cathode connected to Vref=5V, anode grounded) to clamp the sampling node voltage within the range of 0-(Vref+0.7V). Since Vref=5V, this range is 0-5.7V to prevent external high voltage (such as 12V / 48V power supply) from backflowing and damaging the ADC port.

[0156] In this embodiment, the reference voltage Vref is 5.0V, the first resistor R1 is 10kΩ, and the second resistor R2 is 10kΩ. Therefore, the theoretical voltage of the sampling node under normal conditions is 2.5V, and the normal threshold range is set to 2.0V-3.0V. When the chassis domain controller 210 does not output a drive signal to the corresponding module, it acquires the voltage value of the sampling node P through the ADC sampling port. When the sampling voltage is 0V±0.1V, it is determined to be a short circuit to ground; when the sampling voltage is 4.8V-5.7V (i.e., close to or reaching the Vref clamping value), it is determined to be a short circuit to the power supply (including the case of a short circuit to a 12V / 48V external power supply, where the voltage is clamped to Vref); when the sampling voltage is in the range of 0.5V-1.8V or 3.2V-4.5V, it is determined to be a poor contact or a partial short circuit in the coil; when the sampling voltage is in the range of 1.8V-2.0V or 3.0V-3.2V, it is determined to be actuator parameter drift, and a warning signal is issued. The above diagnostics are performed once during the no-output drive signal period of each control cycle (10ms). This fault diagnosis circuit can perform the diagnosis without driving the solenoid valve, and the diagnosis of each actuator only requires one ADC port resource.

[0157] As can be seen from the above, during the interval when the domain controller does not output a drive signal (high impedance state) (the idle period of the control cycle), it reads the node voltage through a single ADC port. With just a single port and a single sample, it can distinguish: (1) 0V → short circuit to ground; (2) 0.9Vref → short circuit to power supply (12V / 48V) (clamped protection); (3) voltage deviates from the normal threshold → poor contact or partial short circuit of the coil. In this embodiment, each actuator diagnosis only requires 1 ADC port, which improves the traditional technical solution that requires 2 ADC ports. Moreover, the fault diagnosis circuit in this embodiment does not require forced drive of the solenoid valve, realizing "non-destructive" diagnosis and significantly improving system safety.

[0158] The execution layer 300 is configured to receive the fixed allocation control command or the collaborative control command issued by the decision layer 200, and control each actuator to work.

[0159] Specifically, the execution layer 300 includes a drive-by-wire module 310, a brake-by-wire module 320, a steering-by-wire module 330, and a steer-by-wire active suspension module 340. The steer-by-wire active suspension module 340 includes an air spring and a continuously damped controlled (CDC) shock absorber. The decision layer 200 and the execution layer 300 adopt a redundant dual-bus architecture of vehicle Ethernet and CAN FD, with the main communication link being vehicle Ethernet and the backup link being CAN FD. Each execution module communicates with the chassis domain controller 210, receiving and executing control commands issued by the chassis domain controller 210.

[0160] In this embodiment, the chassis domain controller 210 is also equipped with a communication redundancy scheme.

[0161] In addition to communicating with each actuator via the main drive-by communication link (vehicle Ethernet as primary and CAN FD as backup), the chassis domain controller 210 is also equipped with a backup wireless communication device. This backup wireless communication device adopts Dedicated Short Range Communication (DSRC) or Cellular Vehicle-to-Everything (C-V2X) technology.

[0162] The chassis domain controller 210 interacts with each actuator through the main drive-by communication link for routine data exchange, while simultaneously sending a heartbeat signal continuously at a 20ms cycle via a backup wireless communication device. Each actuator replies with an acknowledgment signal upon receiving the heartbeat signal. If the chassis domain controller 210 does not receive a heartbeat response from any actuator within a preset time threshold (e.g., 50ms, i.e., 2 to 3 consecutive heartbeat cycles), it determines that the main drive-by communication link has failed and automatically switches to the backup wireless communication device for command issuance and status feedback, ensuring reliable transmission of critical control commands.

[0163] Meanwhile, the backup wireless communication device can also be used to receive real-time road condition information (such as slippery roads ahead, construction, etc.) sent by roadside infrastructure, providing richer environmental perception data for chassis control and enabling forward-looking chassis pre-adjustment control.

[0164] Example 3:

[0165] This embodiment provides a vehicle, including the chassis intelligent electronic control system of Embodiment 2.

[0166] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for intelligent electronic control of a chassis, characterized in that, include: S1. Acquire environmental information, vehicle status information, driver intention information, and the actual output status of each actuator; S2. Based on the environmental information and vehicle state information, calculate the feedforward compensation yaw moment, determine the boundary constraints for correcting vehicle stability, and calculate the generalized target force and moment required to maintain vehicle stability. The generalized target force and moment include the total desired longitudinal force, the additional desired yaw moment, and the roll moment. S3. Obtain the current available execution capability range of each actuator, and determine whether the generalized target force and torque exceed the current available execution capability range of each actuator; if the generalized target force and torque do not exceed the current available execution capability range of each actuator, proceed to S4; if the generalized target force and torque exceed the current available execution capability range of a certain actuator, proceed to S5. S4. Convert the generalized target force and torque into fixed distribution control commands for each wheel according to the preset fixed distribution rules; S5. Construct a multi-objective quadratic programming (QP) model, and based on the rolling time-domain optimization of model predictive control, redistribute and optimize the target driving torque, target braking torque, target steering angle and target suspension damping force of each wheel to generate cooperative control commands. S6. Control each actuator based on the fixed allocation control command in S4 or the cooperative control command in S5, and confirm the command execution status based on the actual output status of each actuator.

2. The intelligent electronic control method for chassis according to claim 1, characterized in that, S1 includes: S101. Environmental information is collected through an environmental perception module, including road curvature and road surface adhesion coefficient; vehicle status information is collected through a vehicle status perception module, including actual longitudinal vehicle speed, lateral vehicle speed, yaw rate, vehicle three-axis acceleration, vehicle three-axis angular velocity, wheel speed of each wheel, and steering wheel angle; driver intention information is collected through a driver intention perception module, including throttle opening, brake pedal travel, and steering wheel torque; and the actual output status of each actuator is collected through an execution status feedback module, including actual torque of the drive motor, actual pressure of the brake wheel cylinder, actual steering angle, and actual damping force of the suspension. S102. Preprocess the signal acquired in S101. The preprocessing includes filtering, noise reduction and calibration. S103. Perform a rationality check on the preprocessed signals in S102 and compare each signal with the preset normal working range. If all signals are normal and do not exceed the normal range, proceed directly to S2. If an abnormal signal exceeds the normal range, mark the sensor corresponding to the abnormal signal as faulty and replace each abnormal signal with backup sensor data or estimated data from the vehicle state estimation unit before proceeding to S2.

3. The intelligent electronic control method for chassis according to claim 2, characterized in that, S2 includes: S201. Establish the state equation based on the two-degree-of-freedom vehicle model. Take the steering wheel angle and actual longitudinal vehicle speed as inputs, and the yaw rate and lateral acceleration as observations. Use an unscented Kalman filter to make the optimal estimate of the centroid sideslip angle and obtain the estimated centroid sideslip angle. S202. Calculate the feedforward yaw moment based on the road curvature and actual longitudinal vehicle speed using the following formula: ; In the formula, To compensate for yaw moment; For road curvature; The feedforward gain related to the actual longitudinal vehicle speed is pre-calibrated into a MAP table for lookup. The MAP table is a table showing the relationship between the actual longitudinal vehicle speed and... The mapping table is calculated during calibration using the following formula: ,in, For vehicle mass parameters, For vehicle wheelbase parameters, This represents the equivalent lateral stiffness of the vehicle's front axle. This is the equivalent lateral stiffness of the vehicle's rear axle. This refers to the actual longitudinal speed; S203. Boundary constraints for vehicle stability correction based on road surface adhesion coefficient, including: dynamically correcting the upper limit of yaw rate and the upper limit of center-of-gravity sideslip angle based on tire friction circle theory and phase plane stability domain analysis method, using actual longitudinal vehicle speed and vertical load as input; wherein, the dynamic upper limit of yaw rate... Dynamic upper limit of centroid sideslip angle In the formula, For low adhesion attenuation coefficient, This refers to the actual longitudinal vehicle speed. The road surface adhesion coefficient, It is the acceleration due to gravity; S204. The total expected longitudinal force demand of the vehicle is calculated using a PID controller based on the longitudinal speed deviation; wherein, the actual longitudinal speed deviation is the difference between the ideal longitudinal speed and the actual longitudinal speed; the ideal longitudinal speed is obtained by converting the throttle opening and brake pedal travel collected by the driver intention perception module through the vehicle inverse longitudinal dynamics model; the actual longitudinal speed is obtained after being collected, preprocessed, and verified through S101-S103. S205. Based on the yaw rate deviation and the center of gravity sideslip angle deviation, the sliding mode control algorithm is used to calculate the additional expected yaw moment and roll moment required to maintain vehicle stability. Wherein, the yaw rate deviation is the difference between the ideal yaw rate and the actual yaw rate. The ideal yaw rate is calculated by the two-degree-of-freedom vehicle model based on the steering wheel angle and the actual longitudinal vehicle speed. The actual yaw rate is the yaw rate after being collected, preprocessed and verified by S101-S103. Furthermore, both the ideal yaw rate and the actual yaw rate are corrected by the upper limit of yaw rate obtained in S203. The centroid sideslip angle deviation is the difference between the ideal centroid sideslip angle and the estimated centroid sideslip angle; the ideal centroid sideslip angle is set to 0; the estimated centroid sideslip angle is obtained through S201; and the estimated centroid sideslip angle is obtained through the dynamic upper limit of the centroid sideslip angle obtained in S203. Perform amplitude limiting correction; S206. The additional desired yaw moment obtained in S205 is corrected using the feedforward compensation yaw moment obtained in S202.

4. The intelligent electronic control method for chassis according to any one of claims 1-3, characterized in that, S5 includes: S501. Construct a multi-objective optimization function; The multi-objective optimization function includes a vehicle stability sub-objective, a tire adhesion utilization rate sub-objective, and an actuator energy consumption sub-objective. The vehicle stability sub-objective aims to minimize the yaw rate tracking error, the tire adhesion utilization rate sub-objective aims to minimize the sum of squares of the load rates of each tire, and the actuator energy consumption sub-objective aims to minimize the electrical power consumption of each actuator. S502. The sub-objective functions in S501 are integrated into a comprehensive objective function by weighting coefficients. The basic weighting coefficients are dynamically adjusted according to the working conditions: the stability weight is increased for extreme working conditions and the energy consumption weight is increased for normal working conditions. S503. Under the condition of satisfying the physical constraints of each execution subsystem, the comprehensive objective function is transformed into the standard quadratic programming (QP) form to obtain the optimal control allocation model based on the quadratic programming (QP) solution, and the effective set method is used to solve it. S504. Obtain the optimal cooperative control allocation scheme based on the solution obtained by the effective set method, and generate cooperative control commands based on the cooperative control allocation scheme; wherein, the cooperative control allocation scheme includes the target driving torque, target braking torque, target steering angle and target suspension damping force of each wheel.

5. The intelligent electronic control method for chassis according to claim 4, characterized in that, The optimal control allocation model is as follows: ； In the formula, All are dynamically weighted coefficients. These are the weighting coefficients for the stability sub-objective. The weighting coefficients for the tire adhesion utilization sub-objective are as follows: The weighting coefficients for the energy consumption sub-objectives; This is the sensitivity matrix of the yaw rate to the control quantity; The decision variable vector includes the driving torque, braking torque, steering angle, and suspension damping force of each wheel; for Transpose of; Ideal yaw rate; The weight matrix for tire adhesion utilization is a diagonal matrix. The actuator energy consumption weight matrix is ​​a diagonal matrix. For equality constraint matrices / vectors; Inequality constraint matrix / vector; These are the calibration basis weighting coefficients for the stability sub-objective. The calibration basis weighting coefficient for the tire adhesion utilization sub-target; The calibration constant is used to dynamically adjust the gain; The attachment margin index; As a stability risk index; This is the actual yaw rate; This is the upper limit of the yaw rate; To estimate the centroid sideslip angle; This is the dynamic upper limit of the centroid sideslip angle; For the longitudinal force of each tire; The lateral forces of each tire; For the vertical force of each tire; It has a low adhesion attenuation coefficient; The road surface adhesion coefficient; It is the acceleration due to gravity; This refers to the actual longitudinal speed; For road curvature; For vehicle mass parameters; This refers to the vehicle's wheelbase parameter. This is the equivalent lateral stiffness of the vehicle's front axle. This is the equivalent lateral stiffness of the rear axle of the vehicle.

6. The intelligent electronic control method for chassis according to claim 5, characterized in that, The calibration base weighting coefficients in S502 are dynamically adjusted based on driving conditions, including: determining whether the vehicle status information meets the priority conditions of economy or stability, determining the driving conditions and vehicle driving mode, and determining the calibration base weighting coefficients, specifically including: The absolute value of the lateral acceleration is compared with a first threshold and a second threshold, wherein the second threshold is greater than the first threshold; The absolute value of the yaw rate deviation is compared with the third threshold and the fourth threshold, wherein the fourth threshold is greater than the third threshold; Compare the estimated absolute value of the centroid sideslip angle with the fifth threshold; If the absolute value of the lateral acceleration is not greater than the first threshold and the absolute value of the yaw rate deviation is not greater than the third threshold, then the vehicle status information is determined to meet the economy priority condition, which is the normal operating condition, and the vehicle driving mode is determined to be the economy priority control mode. If the absolute value of the lateral acceleration is greater than the second threshold, or the absolute value of the yaw rate deviation is greater than the fourth threshold, or the absolute value of the estimated center of gravity sideslip angle is greater than the fifth threshold, then the vehicle state information is determined to meet the stability priority condition, which is the extreme condition, and the vehicle driving mode is determined to be the stability priority control mode. The economy-first control mode and the stability-first control mode are used to adjust the weight coefficients of the sub-objectives in the optimal control allocation model; the energy consumption sub-objective of the economy-first control mode is used as the calibration basis weighting coefficient. The calibration basis weighting coefficients for stability sub-objectives The calibration basis weighting coefficient for the tire adhesion utilization rate sub-target Decrease sequentially; the calibration basis weighting coefficients of the stability sub-objectives in the stability priority control mode. The calibration basis weighting coefficient for the tire adhesion utilization rate sub-target The calibration basis weighting coefficients for energy consumption sub-targets Decrease sequentially; If the vehicle status information does not meet the priority conditions of economy and stability, the vehicle driving mode will remain unchanged.

7. The intelligent electronic control method for chassis according to claim 1, characterized in that, Also includes: S7. Detect whether each actuator is in a fault state. If any actuator is in a fault state, determine that the vehicle driving mode is fault degraded control mode. The detection of whether each actuator is in a faulty state is performed through a fault diagnosis circuit, including: When no drive signal is output to the corresponding execution module, the voltage value of the drive end of each actuator is collected in real time through the fault diagnosis circuit, and the type of actuator fault is determined based on the collected voltage. The fault diagnosis circuit includes a voltage divider resistor network, a current-limiting protection resistor, a clamping diode, an analog-to-digital converter sampling port, and a reference voltage. The voltage divider resistor network includes a first resistor and a second resistor connected in series. One end of the first resistor is connected to the reference voltage, and the other end is connected to the sampling node P of the corresponding actuator's drive end. One end of the second resistor is connected to the sampling node, and the other end is grounded. The sampling node is connected to the analog-to-digital converter sampling port via the current-limiting protection resistor, and a clamping diode is provided at the sampling node.

8. A chassis intelligent electronic control system, characterized in that: It includes the perception layer, decision-making layer, and execution layer; The perception layer is configured to collect environmental information, vehicle status information, driver intention information, and the actual output status of each actuator, and send them to the decision layer. The decision-making layer is configured to receive signals collected by the perception layer and perform preprocessing and rationality verification. Based on the environmental information and vehicle status information, the feedforward compensation yaw moment is calculated, the boundary constraints for correcting vehicle stability are determined, and the generalized target force and moment required to maintain vehicle stability are calculated. The generalized target force and moment include the total desired longitudinal force, the additional desired yaw moment, and the roll moment. The current available execution capability range of each actuator is obtained, and it is determined whether the generalized target force and moment exceed the current available execution capability range of each actuator. If the generalized target force and moment do not exceed the current available execution capability range of each actuator, the generalized target force and moment are converted into fixed allocation control commands for each wheel according to a preset fixed allocation rule. If the generalized target force and torque exceed the current available execution capability of a certain actuator, a multi-objective quadratic programming (QP) model is constructed. Based on the rolling time-domain optimization of model predictive control, the target driving torque, target braking torque, target steering angle, and target suspension damping force of each wheel are redistributed and optimized to generate a cooperative control command. The fixed allocation control command or the cooperative control command is sent to the execution layer. The execution status of the command is confirmed based on the actual output state of each actuator. The execution layer is configured to receive the fixed allocation control command or the collaborative control command issued by the decision layer, and control the operation of each actuator.

9. The intelligent electronic control system for chassis according to claim 8, characterized in that: The decision-making layer is a chassis domain controller, which includes a signal processing unit, a logic verification unit, a vehicle state estimation unit, a model solving unit, a cooperative control unit, and a communication unit. The signal processing unit is configured to receive signals collected by the sensing layer and perform preprocessing. The logic verification unit is configured to perform a rationality check on the preprocessed signal; The vehicle state estimation unit is configured to: establish a state equation based on a two-degree-of-freedom vehicle model, take the steering wheel angle and actual longitudinal vehicle speed as inputs, take the yaw rate and lateral acceleration as observations, and use an unscented Kalman filter to perform optimal estimation of the centroid sideslip angle to obtain the estimated centroid sideslip angle. The model solving unit is configured to: calculate the feedforward compensation yaw moment based on the environmental information and vehicle state information, determine the boundary constraints for correcting vehicle stability, and calculate the generalized target force and moment required to maintain vehicle stability, wherein the generalized target force and moment include the total desired longitudinal force, the additional desired yaw moment, and the roll moment. The collaborative control unit is configured to: obtain the current available execution capability range of each actuator, and determine whether the generalized target force and torque exceed the current available execution capability range of each actuator; if the generalized target force and torque do not exceed the current available execution capability range of each actuator, then convert the generalized target force and torque into fixed allocation control commands for each wheel according to a preset fixed allocation rule; If the generalized target force and torque exceed the current available execution capability of a certain actuator, a multi-objective quadratic programming (QP) model is constructed. Based on the rolling time-domain optimization of model predictive control, the target driving torque, target braking torque, target steering angle, and target suspension damping force of each wheel are redistributed and optimized to generate cooperative control commands. The execution status of the commands is confirmed based on the actual output state of each actuator. The communication unit is configured to send the fixed allocation control command or the cooperative control command to the execution layer.

10. A vehicle, characterized in that: Includes the intelligent electronic control system for the chassis as described in claim 8 or 9.

Images