An electric vehicle intelligent driving steering control method and system

Abstract

The application discloses an intelligent driving steering control method and system for an electric vehicle, and belongs to the technical field of intelligent driving, which comprises the following steps: determining a main abnormal section and a suspected abnormal object; when verification is allowed, implementing limited driving torque verification on a driving object, determining an abnormal source according to response chain changes before and after the verification, otherwise generating an unverified abnormal identification and limiting the driving object; distributing a driving control amount according to the abnormal source or the unverified abnormal identification, a yaw trend and an execution capacity; after steering recovery, transferring the driving side yaw control amount to the steering side according to a pair of replacement steps; after the replacement, determining an arrearage recovery window and recovering a running state according to a response relationship, an intervention instruction arrearage, a target trajectory and a road boundary. The application improves the stability, trajectory control continuity and safe adaptive capacity of the electric vehicle under complex road adhesion states and actuator abnormal conditions as a whole.

Description

Technical Field

[0001] This invention relates to the field of intelligent driving technology, specifically to an intelligent driving steering control method and system for electric vehicles. Background Technology

[0002] With the development of autonomous driving technology and distributed drive technology, road vehicles typically adjust the steering actuators and driving torque of each drive wheel comprehensively based on the target trajectory, vehicle lateral position deviation, heading angle deviation, and yaw rate deviation to achieve trajectory tracking and vehicle stability control. When a vehicle is traveling on a wet or slippery road surface, a road surface with separate left and right adhesion, or a curve with high longitudinal driving force, various reasons such as steering actuator lag, limited steering ability under load, saturation of the combined longitudinal and lateral forces of the tires, decreased local road surface adhesion, and abnormal output of the drive actuators may all manifest as insufficient lateral acceleration, actual yaw rate deviating from the target yaw rate, or the vehicle veering outward from the target trajectory.

[0003] Existing control methods typically increase steering control input, adjust the difference in driving torque between the left and right wheels, or trigger braking stability control based on vehicle state deviation. It is difficult to locate the first abnormal link based on the timing transmission relationship between control commands, actuator responses, wheel responses, and vehicle motion responses. Therefore, it is easy to misjudge steering actuator abnormalities as tire adhesion abnormalities, or to misjudge tire force saturation as steering actuator failure.

[0004] Meanwhile, existing drive and steering coordinated control typically uses a preset ratio or a one-time switching method to allocate control authority. Prematurely exiting drive stability control before the lateral steering effect is fully established, or maintaining a large drive torque difference for an extended period after steering function has recovered, can both cause yaw fluctuations, increased wheel slippage, and amplified trajectory tracking errors. Furthermore, after stability control limits normal trajectory control commands, existing methods often fail to differentiate, record, and subsequently process the suppressed effective control demands, or only provide uniform compensation based on accumulated trajectory deviations. This makes it difficult to exclude changes caused by target trajectory replanning, easily leading to the vehicle reverting to ineffective historical control values ​​near road boundaries. Summary of the Invention

[0005] In view of the above-mentioned problems, the present invention provides a method for intelligent driving steering control of electric vehicles.

[0006] Therefore, the technical problem solved by the present invention is that it is difficult to accurately determine the source of abnormality in electric vehicles under complex adhesion conditions, high tire load, or abnormal steering and drive actuator response based solely on vehicle yaw deviation or trajectory deviation.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a method for intelligent driving steering control of electric vehicles, comprising,

[0008] Acquire driving information, response information, and constraint information; construct a response chain; compare it with the baseline response chain; and determine the main abnormal chain segment and suspected abnormal objects.

[0009] When verification is allowed, the restricted driving torque verification is performed on the driving object, and the source of the anomaly is determined based on the change in the response chain before and after verification; otherwise, an unverified anomaly flag is generated and the driving object is restricted.

[0010] Based on the source of the anomaly or unverified anomaly identifier, yaw trend, and execution capability, allocate drive control quantities; after steering is restored, transfer the drive-side yaw control quantities to the steering side according to the successive replacement steps;

[0011] After taking over, the system determines the recovery window based on the response relationship closure, the outstanding intervention instructions, the target trajectory, and the road boundaries, and then resumes operation.

[0012] As a preferred embodiment of the intelligent driving steering control method for electric vehicles according to the present invention, the construction of the response chain includes,

[0013] The validity of the driving information, response information, and constraint information is verified, and the time base is unified.

[0014] Identify the start time of a valid change in a control command or state variable, form a response event node, and associate the response event node with the corresponding control event;

[0015] Following the transmission sequence of control commands, execution responses, wheel-side responses, and body responses, adjacent response event nodes are connected into response chain segments, forming steering response chains and drive response chains respectively.

[0016] When multiple control events overlap within a related time period, a joint response record is established, and a corresponding response chain is formed based on subsequent independent control events or constrained driving torque verification to distinguish the response relationship corresponding to each control event;

[0017] When no effective control event occurs, perform steady-state relationship monitoring, establish steady-state response records or unverified steady-state anomaly indicators, and determine the credible status of each current response chain based on the data validity status, time synchronization status, and event attribution status.

[0018] As a preferred embodiment of the intelligent driving steering control method for electric vehicles according to the present invention, the step of determining the main abnormal chain segment and suspected abnormal objects includes,

[0019] Select the corresponding baseline response chain based on the current operating conditions;

[0020] Compare the response order, response interval, response direction, initial rate of change, and response missing status of adjacent response event nodes in the current response chain with those in the baseline response chain, and mark the abnormal response chain segments.

[0021] Determine the first abnormal chain segment from each current response chain along the direction of transmission of control commands to the vehicle body response;

[0022] Based on the order of occurrence, degree of abnormality, and duration of each initial abnormal chain segment, determine the main abnormal chain segment or concurrent abnormal chain segment;

[0023] During the maintenance period of the main anomaly chain segment, the main anomaly chain segment is updated only when an anomaly response chain segment with a higher degree of anomaly occurs; after the maintenance period ends, the main anomaly chain segment is updated according to the recovery status of the original main anomaly chain segment and the persistence status of other anomaly chain segments, and suspected anomaly objects are identified based on the updated main anomaly chain segment.

[0024] In a preferred embodiment of the intelligent driving steering control method for electric vehicles according to the present invention, wherein: during the verification process, a limited driving torque verification is performed on the driving object, and the source of the anomaly is determined based on the change in the response chain before and after verification; otherwise, an unverified anomaly identifier is generated and the driving object is restricted.

[0025] Determine whether the verification conditions are met based on the vehicle's yaw rate, wheel side conditions, road boundary conditions, target driving and execution capabilities.

[0026] If the verification conditions are not met, no verification drive torque adjustment will be performed, the drive control quantity of the associated drive object will be limited, and an unverified anomaly flag will be generated.

[0027] When the verification conditions are met, the driver object associated with the suspected abnormal object is identified as the verification object. Within the safety constraints, the driver control quantity of the verification object is reduced, and the limited compensation control quantity is allocated to other available driver objects according to the longitudinal operation maintenance requirements.

[0028] Reconstruct the response event nodes and restructure the relevant current response chain. Determine the source of the anomaly based on the disappearance, persistence, transfer, and addition of the main anomaly chain segment before and after verification.

[0029] When compensation control triggers a new anomaly, the corresponding compensation control is stopped and the verification-induced anomaly is recorded; if the verification conditions fail during the verification period, the verification is terminated and the drive control quantity is restored to the state before verification or the allowable state within the safety constraints.

[0030] As a preferred embodiment of the intelligent driving steering control method for electric vehicles according to the present invention, the allocation of drive control quantities includes,

[0031] Establish a set of available execution objects based on the vehicle's execution configuration and status, and record the incremental capability, decrement capability, and allowable output range of each execution object;

[0032] Based on the source of the anomaly or unverified anomaly identifier and the vehicle yaw trend, determine the execution objects that restrict the increase of drive control quantity, the execution objects that prioritize the reduction of drive control quantity, and the execution objects that undertake yaw control;

[0033] The target yaw control requirement is decomposed into the control change amount corresponding to each execution object responsible for yaw control, and the control change amount is limited according to the road boundary state, wheel side state and allowable output range.

[0034] Continuously monitor the response status of each execution object, update the available capabilities of the corresponding execution object based on new exceptions, and reallocate the remaining control change amount;

[0035] When the set of available executable objects cannot provide yaw control for the target direction or size, the allowed longitudinal control is reduced and the allowed steering control is limited, and the process enters a restricted operation phase.

[0036] As a preferred embodiment of the intelligent driving steering control method for electric vehicles according to the present invention, the step of transferring the yaw control amount from the drive side to the steering side according to the alternating steps includes,

[0037] Convert the drive-side control contribution and the steering-side control contribution into the same control representation;

[0038] The amount of control reduction on the drive side is determined based on the current yaw control requirements, and the amount of steering control recovery corresponding to it is determined based on the local response relationship on the steering side.

[0039] The drive-side control reduction and steering control recovery are executed synchronously, and a local steering response chain is formed based on the current steering control recovery.

[0040] When the local steering response chain is closed, and the vehicle yaw state, wheel side state, and road boundary state are within the allowable range, the current handover state is recorded as a successful handover state and the next handover step is executed.

[0041] If any condition is not met, stop increasing the steering-side control contribution and adjust the drive-side control contribution and steering-side control contribution to the previous successful takeover state; retain part of the drive-side control contribution or complete the takeover based on the available steering-side capacity.

[0042] As a preferred embodiment of the intelligent driving steering control method for electric vehicles according to the present invention, the step of determining the debt recovery window and restoring the operating state includes,

[0043] When a normal control command is first restricted by stable control, the difference between the cumulative normal control quantity to be executed and the actual allowable control quantity is recorded, forming an intervention command deficit. This intervention command deficit is then associated with the time of occurrence, the corresponding road section, the recovery direction, and the target trajectory information.

[0044] After the handover is completed, the window for recovery of the outstanding issues is determined when the steering response relationship meets the requirements of dynamic closure or steady-state closure, there are no new anomalies in the drive response chain, the confirmed anomalies have not recurred, the road boundary status is permissible, the outstanding intervention command is valid, the corresponding road section is in front of the vehicle and the current target trajectory is permissible for recovery.

[0045] Within the debt recovery window, additional allowable control quantities beyond the current target trajectory control will be used to gradually reduce the debt of intervention instructions;

[0046] When the target trajectory is continuously corrected, the outstanding intervention instructions can be retained, mapped, reduced, or canceled.

[0047] If the direction of travel of the target trajectory changes or the recovery conditions fail, the recovery will be terminated and the remaining outstanding intervention instructions will be frozen.

[0048] This invention provides an intelligent driving steering control system for electric vehicles.

[0049] To solve the above-mentioned technical problems, the present invention provides the following technical solution: an intelligent driving steering control system for electric vehicles, comprising: a response chain construction and anomaly localization module, a restricted verification and stability control module, and a restricted verification and stability control module;

[0050] The response chain construction and anomaly localization module is used to synchronously collect vehicle status, construct steering and drive response chains, and determine the first anomaly chain segment and suspected anomaly objects.

[0051] The constrained verification and stability control module is used to adjust the suspected abnormal wheel torque under safety constraints, confirm the source of the abnormality, and generate additional yaw control quantities.

[0052] The succession and overdue recovery module is used to complete the succession from drive control to steering control step by step, and to recover the overdue effective intervention commands based on the current trajectory and road boundaries.

[0053] The present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the intelligent driving steering control method for electric vehicles.

[0054] The present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the intelligent driving steering control method for electric vehicles.

[0055] The beneficial effects of the present invention are as follows: The present invention constructs a steering actuator drive response chain, a steering wheel side response chain, and a drive response chain corresponding to each drive wheel. It determines the first abnormal chain segment based on the occurrence sequence of response event nodes, the interval between adjacent responses, the direction of change, and the initial rate of change. This transforms the abnormal judgment from simply relying on the deviation of the vehicle body state to a segment-by-segment analysis of the control action transmission process. This can narrow down the range of suspected abnormal objects and improve the ability to distinguish between steering side abnormalities, drive side abnormalities, and tire adhesion abnormalities.

[0056] By implementing restricted torque reduction verification on suspected abnormal wheels or axles under independent safety constraint layers, and observing the disappearance, persistence, transfer, and addition of abnormal chain segments, the source of the anomaly can be further confirmed without significantly disrupting the vehicle's stability. This ensures that subsequent drive torque distribution is directly constrained by the source of the anomaly, the location of the abnormal chain segment, the vehicle's yaw trend, and the remaining capacity of the actuators. This reduces the risk of further torque increase on low-adhesion wheels or actuators with limited capacity. Furthermore, by setting up complete and partial replacements, and combining the gradual reduction of the additional yaw control on the drive side with the corresponding recovery of the control on the steering side into replacement step lengths, and then verifying each replacement step length using a local steering response chain, abrupt switching between drive control and steering control can be avoided. In the event of a replacement failure, the previous successful control state can be synchronously restored, improving the continuity and reversibility of yaw control.

[0057] Secondly, by establishing an intervention command deficit that includes curvature deficit, yaw control deficit, longitudinal driving force deficit, and lateral position deficit projection, and associating the deficit with road sections, deficit direction, anomaly source, target trajectory identifier, and validity period, it is possible to eliminate changes directly caused by trajectory replanning, prevent duplicate compensation or restoration of invalid historical control requirements, and gradually repay, remap, reduce, or cancel the deficit according to the current target trajectory and road boundary conditions. This improves the stability, trajectory control continuity, and safety adaptability of electric vehicles under complex road adhesion conditions and actuator abnormalities. Attached Figure Description

[0058] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0059] Figure 1 This is a general flowchart of an intelligent driving steering control method for electric vehicles provided in one embodiment of the present invention.

[0060] Figure 2The diagram below shows step S1 of an intelligent driving steering control method for electric vehicles according to an embodiment of the present invention. Detailed Implementation

[0061] To make the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0062] Example 1, referring to Figure 1 This is one embodiment of the present invention, which provides a method for intelligent driving steering control of an electric vehicle, comprising:

[0063] S1. Obtain driving information, response information, and constraint information, construct a response chain, compare it with the baseline response chain, and determine the main abnormal chain segment and suspected abnormal objects.

[0064] S2. When verification is allowed, perform restricted driving torque verification on the driving object and determine the source of the anomaly based on the change in the response chain before and after verification; otherwise, generate an unverified anomaly flag and restrict the driving object.

[0065] S3. Allocate drive control quantities based on the source of the anomaly or unverified anomaly identifier, yaw trend, and execution capability; after steering is restored, transfer the drive-side yaw control quantities to the steering side according to the established replacement steps.

[0066] S4. After taking over, determine the debt recovery window and restore the operation status based on the response relationship closure, intervention command backlog, target trajectory and road boundary.

[0067] It should be noted that this invention provides a drive-steering control method capable of identifying the first abnormal chain segment during the transmission of control commands to actuators, wheels, and vehicle body motion states, and distinguishing between abnormal steering actuator operation, limited steering capability under load, saturation of tire longitudinal and lateral force, decreased road surface adhesion, left and right separation adhesion, and abnormal drive actuator operation by verifying the limited driving torque that meets safety constraints. Simultaneously, it solves the problem of abrupt changes in yaw control caused by direct switching between drive stability control and steering control, enabling the vehicle to smoothly transfer yaw control tasks from the drive side to the steering side through a step-by-step verifiable and synchronously reversible succession process based on the actual established state of lateral steering action, while retaining necessary residual drive control even when steering capability is only partially restored.

[0068] In addition, it addresses the issues of lack of effective records of unexecuted control requirements after stability control restricts normal planning control instructions, unclear historical compensation objects after trajectory updates, and the potential increase in road boundary risks due to backlog recovery. It enables effective control requirements suppressed by stability control to be distinguished from replanning changes and ordinary trajectory errors, and to be retained, mapped, reduced, canceled, and gradually restored based on the current target trajectory, road section, abnormal state, and road boundary conditions.

[0069] Example 2, refer to Figure 2 As one embodiment of the present invention, based on the previous embodiment, an intelligent driving steering control method for electric vehicles is provided, comprising:

[0070] The present invention provides a detailed description of an intelligent driving steering control method for electric vehicles. It should be understood that the following specific embodiments are used to illustrate the technical solution of the present invention and are not intended to limit the scope of protection of the present invention. Those skilled in the art can adaptively adjust some detection parameters, judgment thresholds, control cycles, and execution methods according to the vehicle drive configuration, steering configuration, sensor configuration, actuator capabilities, and road operating environment without departing from the technical concept of the present invention.

[0071] This invention applies to road vehicles capable of autonomously adjusting their direction and speed according to a target trajectory. These road vehicles can be autonomous passenger vehicles, autonomous commercial vehicles, unmanned transport vehicles, unmanned delivery vehicles, or other vehicles with autonomous road driving capabilities. In this invention, the road vehicle is an electric vehicle. This invention constructs a steering actuator drive response chain, a steering wheel-side response chain, and a drive response chain, respectively, identifying abnormal chain segments in the transmission of control commands to the actuators, wheels, and vehicle body. When the vehicle's safety condition allows, it distinguishes between steering actuator malfunctions, limited steering capability, tire longitudinal and lateral force saturation, decreased road surface adhesion, and drive actuator malfunctions through restricted drive torque verification. Drive torque is allocated to each drive wheel based on the source of the anomaly, the vehicle's yaw tendency, and the remaining capacity of the actuators. After the lateral steering action is re-established, the yaw control task is gradually transferred from the drive side to the steering side through sequential replacement steps. After stable control is completed, the vehicle's operating state is restored based on the intervention command deficit, the current target trajectory, and road boundary conditions.

[0072] The vehicle described in this embodiment includes a trajectory planning module, a vehicle motion controller, a steering actuator, a drive actuator, a vehicle state detection unit, and a road environment detection unit, and has at least two drive wheels capable of independently adjusting the drive torque. When the vehicle is equipped with an independently controllable brake actuator, the brake actuator can serve as an auxiliary yaw control actuator when the drive execution capability is insufficient.

[0073] The trajectory planning module generates target driving information based on road structure, obstacle information, traffic rules, and the vehicle's current position. This target driving information includes the target trajectory, target lateral position, target heading angle, planned target curvature, target longitudinal velocity, target trajectory identifier, and target trajectory update time. The trajectory planning module can also directly output the target yaw rate. When the trajectory planning module does not directly output the target yaw rate, the vehicle motion controller determines the basic target yaw rate based on the planned target curvature and target longitudinal velocity. It then limits and smooths the basic target yaw rate based on vehicle stability constraints, steering actuator availability, tire adhesion status, and road boundary margins to obtain the target yaw rate used for vehicle control.

[0074] The target trajectory identifier is used to distinguish the target trajectory generated by the trajectory planning module at different times. The target trajectory update time is used to determine whether the target trajectory update occurs during abnormal positioning, limited driving torque verification, drive steering replacement, or intervention command backlog recovery.

[0075] The vehicle motion controller communicates with the trajectory planning module, steering actuator, each drive actuator, vehicle state detection unit, and road environment detection unit. Based on the target trajectory, planned target curvature, target longitudinal velocity, target yaw rate, and the actual motion state of the vehicle, the vehicle motion controller determines the steering control command under normal conditions, the target yaw control requirements, and the target driving torque of each drive wheel.

[0076] The target yaw control requirement is determined by the vehicle motion controller based on the deviation between the target yaw rate and the actual yaw rate, the trend of the yaw rate deviation, the vehicle's lateral position deviation, and the heading angle deviation. The target yaw control requirement can be expressed as the target additional yaw moment or as a uniform normalized yaw control contribution.

[0077] Before control allocation or replacement comparison, the drive-side control contribution and the steering-side control contribution are converted into the same physical quantity or the same normalized control quantity to avoid directly comparing the difference in drive torque and the increment of steering angle.

[0078] The steering actuator adjusts the steering angle of the vehicle's steering wheels according to the steering control command. The vehicle motion controller acquires the target steering angle, actual steering angle, steering angle change rate, steering actuator position deviation, steering motor control current, steering motor target output torque, steering motor actual output torque, steering actuator limit status, power supply status, communication status, and fault status.

[0079] The drive actuator is used to adjust the driving torque of the corresponding drive wheel. The vehicle motion controller acquires the target driving torque, actual driving torque, driving torque change rate, drive motor speed, drive motor control current, allowable driving torque range, drive actuator torque limit status, temperature status, power supply status, communication status, and fault status for each drive wheel.

[0080] The vehicle status detection unit is used to acquire the wheel speed, wheel angular acceleration, wheel steering angle, vehicle longitudinal speed, vehicle lateral speed, vehicle longitudinal acceleration, vehicle lateral acceleration, vehicle yaw rate, yaw rate change rate, vehicle lateral position deviation, and vehicle heading angle deviation of each wheel, and determine the slip state of each wheel based on the above status.

[0081] The road environment detection unit is used to acquire the distances from the vehicle to the left and right boundaries of the road, as well as the passable area of ​​the road. Road boundaries can be determined based on onboard cameras, LiDAR, millimeter-wave radar, high-precision maps, or a combination of these. The distance from the vehicle to the road boundary is determined by the minimum distance from the vehicle's safety envelope boundary to the road boundary. The vehicle's safety envelope includes the vehicle's actual outer contour and a lateral safety margin set based on vehicle speed, yaw rate, and positioning error.

[0082] Before distributing drive torque, the vehicle motion controller establishes a set of available actuators based on the vehicle's hardware configuration and the real-time status of the actuators. This set of available actuators records whether each drive wheel can independently increase or decrease torque, whether each axle can perform left-right differential drive, the current permissible drive torque range of each drive actuator, and whether the vehicle is equipped with independently controllable brake actuators.

[0083] For four-wheel independent drive vehicles, the vehicle motion controller can distribute drive torque among the four drive wheels. For vehicles where only the left and right drive wheels of the rear axle can be adjusted independently, additional yaw torque is mainly generated by the difference in drive torque between the left and right sides of the rear axle. For vehicles where only the left and right drive wheels of the front axle can be adjusted independently, additional yaw torque is generated by the difference in drive torque between the left and right sides of the front axle, while limiting the increase in drive torque of the front wheels suspected of having low traction.

[0084] For vehicles that can only adjust the total drive torque of the front and rear axles but not the independent left and right drive torque of the same axle, the longitudinal tire force borne by the axle suspected of being saturated can be released by transferring the drive torque of the front and rear axles. This vehicle only uses single-wheel braking torque to supplement additional yaw control when equipped with independently controllable brake actuators. When the vehicle is not equipped with independent brake actuators, single-wheel braking control is not performed.

[0085] When the set of available actuators cannot generate the additional yaw moment in the required direction or magnitude, the vehicle motion controller reduces the target longitudinal velocity, limits the planned target curvature, and enters a degraded operation state.

[0086] This implementation defines leftward turning, leftward yaw, and leftward lateral acceleration as positive directions, and rightward turning, rightward yaw, and rightward lateral acceleration as negative directions; an increase in driving torque is considered a positive change, and a decrease in driving torque is considered a negative change. The additional driving torque difference is calculated according to a preset fixed wheel arrangement order and remains consistent throughout the entire control process.

[0087] Additional driving torque difference represents the difference in driving torque between different drive wheels, while additional yaw moment represents the vehicle yaw effect caused by changes in driving torque through the longitudinal and lateral forces of the wheels. These are not the same physical quantity. The planned target curvature command represents the curvature command determined based on the current target trajectory and not subject to stability control restrictions. The stability-limited curvature command represents the curvature command allowed to be executed after stability control restrictions. The actual driving curvature represents the vehicle curvature determined based on the vehicle's actual motion state.

[0088] When the vehicle's longitudinal speed is higher than the minimum speed for calculating curvature, the actual driving curvature can be determined based on the ratio of the actual yaw rate to the vehicle's longitudinal speed. When the vehicle's longitudinal speed is lower than the minimum speed for calculating curvature, the actual driving curvature is not determined solely by the ratio, but rather by combining the actual steering angle, lateral acceleration, yaw rate, lateral position deviation, and heading angle deviation to determine the vehicle's lateral motion state.

[0089] A response event node refers to the starting moment when a control command or vehicle state variable undergoes a valid change. A response chain is a sequence of events formed by control command event nodes, actuator response event nodes, wheel response event nodes, and vehicle body motion response event nodes according to their response propagation relationships. An abnormal chain segment consists of two adjacent response event nodes in the response chain and the response relationship between them; it does not directly consist of non-adjacent response event nodes.

[0090] The first abnormal chain segment refers to the segment that first exhibits an anomaly when checking sequentially from the control command side to the vehicle motion response side in a single response chain. The main abnormal chain segment is the primary abnormal chain segment identified among the steering first abnormal chain segment and one or more drive first abnormal chain segments, based on the time of anomaly occurrence, anomaly level, and anomaly duration. Concurrent first abnormal chain segments refer to a concurrent abnormal result formed when multiple first abnormal chain segments occur within a concurrent abnormality time threshold, and a single main abnormal chain segment cannot be determined based on the anomaly level.

[0091] Intervention command deficit refers to the cumulative difference between the planned or normal control commands that should have been executed when not subject to stability control restrictions and the control commands that were actually allowed to be executed after stability control restrictions were imposed. Intervention command deficit differs from ordinary vehicle trajectory error and does not directly accumulate all state deviations caused by vehicle positioning errors, external disturbances, or target trajectory replanning.

[0092] The vehicle motion controller has a safety constraint layer independent of normal monitoring, limited drive torque verification, drive stability control, succession, and deficit recovery states. This safety constraint layer operates continuously under all control states, limiting actual yaw rate deviation, yaw rate of change, wheel slippage, the minimum distance from the vehicle's safety envelope to the road boundary, the rate of change of drive torque, the rate of change of steering angle, and the output range of each actuator.

[0093] The verification torque reduction and compensation torque increase generated by the limited drive torque verification are implemented only within the remaining actuator capabilities allowed by the safety constraints. Limited drive torque verification must not override or weaken existing emergency stability control, conservative stability control, or road boundary protection control. When verification actions may increase vehicle yaw rate, wheel slip, or road boundary risk, the safety constraints reduce or prohibit the execution of the corresponding verification action.

[0094] The vehicle motion controller first checks the validity and synchronizes the time of the information output by the trajectory planning module, steering actuator, drive actuator, vehicle state detection unit, and road environment detection unit. If the input information is valid, the vehicle motion controller detects the response event nodes of the control commands and corresponding state variables, and constructs the steering actuator drive response chain, the steering wheel side response chain, and the drive response chain corresponding to each drive wheel.

[0095] The vehicle motion controller compares the current response chain with the reference response chain corresponding to the current operating condition, determines the first abnormal chain segment of steering and the first abnormal chain segment of drive respectively, and determines the main abnormal chain segment or the concurrent first abnormal chain segment according to the time of occurrence of the abnormality, the level of the abnormality and the order of the corresponding control commands.

[0096] After identifying the main anomaly chain, the vehicle motion controller determines whether the vehicle is in a state of emergency yaw, severe wheel slip, emergency braking, emergency obstacle avoidance, insufficient road boundary margin, or complete failure of critical actuators. If the vehicle is in any of these states, the limited drive torque verification is not performed, and the vehicle directly enters emergency stability control, conservative stability control, or degraded operation.

[0097] When the vehicle is not in the aforementioned emergency state and the verification conditions are met, the vehicle motion controller performs a limited driving torque verification on the drive wheels or axles associated with the suspected abnormal object, and confirms the source of the abnormality based on the disappearance, continuation, transfer, or addition of the abnormal chain segment before and after the verification.

[0098] After confirming the source of the anomaly, the vehicle motion controller determines the drive torque distribution method based on the anomaly source, vehicle yaw trend, road boundary information, vehicle drive configuration, and actuator availability. After the steering lateral action is re-established, the additional yaw control on the drive side is gradually reduced using sequential alternating steps, while the yaw control on the steering side is simultaneously restored.

[0099] After the drive steering handover is completed, the vehicle motion controller determines whether a backlog recovery window has been created and processes valid intervention commands within that window. If a new high-level anomaly occurs at any stage of limited drive torque verification, drive stability control, handover completion, or backlog recovery, the vehicle motion controller interrupts the current control stage, freezes the current handover or backlog recovery amount, and re-enters the anomaly positioning state. If the new anomaly meets the emergency control conditions, the vehicle motion controller directly enters emergency stability control or degraded operation state.

[0100] In this implementation, the control priorities, from highest to lowest, are: emergency safety control, degraded operation and conservative stability control, vehicle stability control, limited drive torque verification, succession replacement control, intervention command backlog recovery control, and normal monitoring control. When a high-priority control condition occurs during a low-priority control process, the high-priority control directly preempts the current control process.

[0101] S1. Obtain driving information, response information, and constraint information, construct a response chain, compare it with the baseline response chain, and determine the main abnormal chain segment and suspected abnormal objects.

[0102] like Figure 2 As shown, the vehicle motion controller acquires target driving information, steering control information, drive control information, wheel motion information, vehicle body motion information and road boundary information according to a preset control cycle, and checks the continuity of timestamps, data update cycle, numerical range, rate of change and sensor or actuator status indicators of each input signal.

[0103] When a sensor data is missing, the retention time exceeds the data validity period, or exceeds the physical allowable range, the vehicle motion controller first generates a sensor data invalidation flag, and does not directly determine the missing data as an abnormal vehicle response.

[0104] When a vehicle is equipped with redundant sensors or alternative estimation information, redundant information is used to replace invalid signals. The actual steering angle can be cross-validated based on the positions of multiple steering position sensors or steering actuator motors; the vehicle's reference longitudinal speed can be determined jointly based on the wheel speeds of non-slip wheels, inertial measurement unit information, and positioning speed; the steering reaction load can be cross-validated based on the steering column torque, steering rack force, and the compensated steering motor output torque.

[0105] When there is no valid alternative source for the critical data necessary to construct a response chain, the corresponding response chain is marked as unavailable, and the source of the anomaly is not identified solely based on that response chain. The vehicle motion controller generates an unverified anomaly flag based on other valid response chains and the vehicle's stable state, and then enters a conservative stability control or degraded operation state.

[0106] Since the sampling frequencies of the trajectory planning module, steering actuator, drive actuator, and vehicle state detection unit may differ, the vehicle motion controller uniformly converts all types of information to its own time base and adds a unified timestamp to each piece of information. For information with different sampling frequencies, resampling can be performed by preserving the previous valid value, linear interpolation, or local interpolation based on the state change trend.

[0107] For the fixed communication delay between the vehicle motion controller and the steering actuator controller and drive actuator controller, the delay compensation value can be determined during the vehicle calibration phase. When calculating the response interval, the vehicle motion controller subtracts the corresponding fixed communication delay and fixed processing delay from the original response interval.

[0108] When communication latency varies with communication load, the current communication latency can be estimated based on the control message sending timestamp, control message receiving timestamp, and actuator feedback timestamp. If the fluctuation of the communication latency estimation result exceeds the latency confidence threshold, the confidence level of the corresponding response interval is reduced, and this response interval is not used as the sole basis for confirming the source of the anomaly.

[0109] The vehicle motion controller can perform low-pass filtering, moving average, median filtering, and outlier removal on the input data. The filtering method used should limit the phase delay caused by filtering, ensuring that the offset of the filtered response event node relative to the original signal event node is less than the allowable time deviation, and preventing data processing from changing the order of the response events.

[0110] In one implementation, the control cycle of the vehicle motion controller can be set to 5 milliseconds to 20 milliseconds, and the response event node is confirmed after the state change is continuously maintained for 2 to 5 control cycles. The above values ​​are calibrated based on the actuator response speed, sensor sampling frequency, and vehicle operating speed.

[0111] The vehicle motion controller assigns a unique control event identifier to each valid steering command change event and each valid target driving torque change event. In addition to recording the time of occurrence, direction of change, and initial rate of change, each response event node also records one or more control event identifiers associated with that response event node.

[0112] When the difference between the occurrence time of a steering command change event and a driving torque change event is greater than a preset decoupling time threshold, subsequent response events are preferentially associated with control events that occur closer in time, have a response direction consistent with the previous response event, and should trigger first according to the corresponding baseline response chain. The decoupling time threshold is calibrated based on the steering actuator response time, the drive actuator response time, and the vehicle yaw response time.

[0113] When the time difference between a steering command change event and at least one driving torque change event is not greater than the decoupling time threshold, the corresponding control event is marked as a coupled control event. For coupled control events, the lateral acceleration or yaw rate response is not individually attributed to a single control command, but rather a coupled response record containing multiple control event identifiers is established.

[0114] For coupled response records, the vehicle motion controller can wait for subsequent independent steering command change events or driving torque change events to reconstruct the corresponding response chain. When the verification conditions are met, it can also adjust the driving force rectangle of the suspected abnormal wheel or axle to form an independent driving verification event, and determine the original abnormality mainly from the steering side, driving side or tire adhesion side based on the change in the response chain after the independent event.

[0115] When the target driving torque of multiple drive wheels changes simultaneously within the same control window, the control event identifier corresponding to each drive wheel is retained. The effectiveness of each drive response chain is determined based on the actual driving torque change of each drive wheel, the wheel angular acceleration change, and the direction of the resulting additional yaw moment. If the contribution of each drive wheel to the vehicle body response cannot be distinguished, multiple drive control events are recorded as a group of joint drive events, and the source of the anomaly is not independently identified based on the response chain of a single wheel.

[0116] When the absolute value of the rate of change of any state variable exceeds the corresponding event trigger threshold and is maintained for a preset number of sampling periods, the vehicle motion controller determines the moment when the state variable first exceeds the event trigger threshold as the response event node.

[0117] When determining response event nodes, the occurrence time, direction of state variable change, initial rate of change after the event, corresponding control event identifier, vehicle speed, planned target curvature, and target trajectory identifier of the response event node are recorded simultaneously. The generation of response event nodes does not require the direction of state variable change to be consistent with the target direction. Whether the state variable changes along the target direction, changes in the opposite direction to the target, or remains unchanged within the preset observation time limit, all are used as criteria for comparing response chains.

[0118] The initial rate of change refers to the average rate of change of the corresponding state variable within a preset initial response window, starting from the moment the response event occurs. The initial response window is longer than the duration of a single sampled noise, but shorter than the time required for the corresponding state variable to enter a stable change phase. Different initial response windows can be used for different state variables.

[0119] When the state variable reverses direction within the initial response window, record the first change direction, the time of direction reversal, and the rate of change after reversal, and mark the direction reversal as an abnormal response direction or an unstable response characteristic.

[0120] When the absolute value of the change in the target steering angle or the rate of change of the target steering angle exceeds the steering command event threshold, a steering command change event node is determined. When the absolute value of the rate of change of at least one of the steering motor control current, the target output torque, or the actual output torque exceeds the corresponding drive event threshold, a steering actuator drive event node is determined.

[0121] When the absolute value of the actual steering angle change rate exceeds the actual steering event threshold and continues for a preset number of sampling periods, an actual steering angle change event node is determined. When the direction of the actual steering angle change is opposite to the direction of the target steering angle change, it is marked as a reverse response event node.

[0122] The steering reaction load used to determine wheel-side response refers to the reaction load transmitted to the steering mechanism by the tire contact with the ground, the mechanical resistance of the steering mechanism, and the road surface after the steering wheel rotates. The steering reaction load can be characterized by steering column torque, steering rack force, or wheel-side reaction torque.

[0123] When the vehicle is not equipped with a steering column torque sensor or a steering rack force sensor, the equivalent steering reaction load can be determined by subtracting the steering mechanism's inertial torque, friction torque, and calibrated transmission losses from the actual output torque of the steering motor. When the absolute value of the steering reaction load change rate exceeds the load event threshold, the steering reaction load change event node is determined.

[0124] The steering motor control current, target output torque, and actual output torque are used to determine whether the steering actuator has received a control command, whether it has established driving force as expected, whether it has reached the current or torque limit, and whether there is insufficient power supply, communication abnormality, or actuator limit. They do not directly replace the steering reaction load to build the wheel-side response chain.

[0125] When the absolute value of the change in target driving torque or the rate of change of target driving torque of any drive wheel exceeds the driving torque command event threshold, the target driving torque change event node for the corresponding drive wheel is determined. When the absolute value of the rate of change of actual driving torque of the corresponding drive wheel exceeds the actual driving torque event threshold, the actual driving torque change event node is determined.

[0126] The vehicle motion controller can determine the wheel angular acceleration based on the wheel speed changes in adjacent control cycles, or it can directly obtain the motor angular acceleration output by the drive motor controller and convert it into wheel angular acceleration according to the transmission ratio. When the absolute value of the wheel angular acceleration exceeds the wheel response threshold, a wheel angular acceleration change event node is determined.

[0127] When the absolute value of the vehicle's lateral acceleration, yaw rate, longitudinal acceleration, or yaw rate of change exceeds the corresponding vehicle body response event threshold, the corresponding vehicle body motion response event node is determined, and the state of the response changing along the target direction, changing in the opposite direction to the target, the response time exceeding the allowable range, or the initial rate of change exceeding the reference range is recorded.

[0128] If a subsequent response event node fails to appear within the corresponding observation period, the corresponding chain segment is identified as a response missing anomaly, and its response interval is not simply set to an infinite value. The current response interval between adjacent response event nodes is obtained by subtracting the occurrence time of the previous response event node from the occurrence time of the subsequent response event node, and deducting the corresponding communication delay and processing delay.

[0129] The steering-side response process includes a steering actuator drive sub-chain and a steering wheel-side response sub-chain. The steering actuator drive sub-chain sequentially includes a steering command change event node, a steering actuator drive event node, and an actual steering angle change event node. A steering command-actuator drive chain segment is formed between the steering command change event node and the steering actuator drive event node, and an actuator drive-actual steering angle chain segment is formed between the steering actuator drive event node and the actual steering angle change event node.

[0130] The steering wheel-side response subchain sequentially includes the actual steering angle change event node, the steering reaction load change event node, the lateral acceleration change event node, and the yaw rate change event node. These adjacent event nodes respectively form the actual steering angle-steering reaction load chain segment, the steering reaction load-lateral acceleration chain segment, and the lateral acceleration-yaw rate chain segment.

[0131] When the steering reaction load cannot be reliably obtained, the actual steering angle-steering reaction load chain segment and the steering reaction load-lateral acceleration chain segment are set as medium-reliability chain segments or unavailable chain segments, and tire adhesion abnormalities are not confirmed solely based on these chain segments. In this case, the lateral action of the steering wheel is determined by combining the actual steering angle change, front wheel slip state, lateral acceleration change, and yaw rate change.

[0132] Each drive wheel with independently adjustable drive torque corresponds to a drive response chain. The drive response chain includes, in sequence, the target drive torque change event node, the actual drive torque change event node, the wheel angular acceleration change event node, and the longitudinal acceleration change event node and / or yaw rate change event node.

[0133] A target driving torque change event node and an actual driving torque change event node form a target driving torque-actual driving torque chain segment; an actual driving torque change event node and a wheel angular acceleration change event node form an actual driving torque-wheel angular acceleration chain segment; and a wheel angular acceleration change event node and a vehicle motion response event node form a wheel angular acceleration-vehicle motion response chain segment.

[0134] When multiple drive wheels simultaneously adjust their drive torque, a drive response chain is established for each drive wheel, and the vehicle motion response corresponding to the change in drive torque of each drive wheel is determined by combining the longitudinal acceleration and yaw rate of change of the vehicle body.

[0135] Response chain closure includes dynamic response chain closure and steady-state response relationship closure. Dynamic response chain closure means that the response sequence, response interval, response direction, and initial rate of change of each adjacent response event node triggered by a change in effective control command are all within the allowable range of the corresponding baseline, and no new abnormal chain segments appear.

[0136] Steady-state response closure means that, in the absence of new effective control command changes, the relationships between the actual steering angle and steering reaction load, the planned target curvature and the actual driving curvature, the target yaw rate and the actual yaw rate, as well as the relationships between the target driving torque, the actual driving torque and the wheel slip state are all within the corresponding steady-state allowable range.

[0137] When the most recent dynamic response chain closure time is less than the current time, the most recent dynamic closure result and the current steady-state response relationship can be used together to determine the continued closure of the response chain. When the most recent dynamic closure result has exceeded the effective time of the dynamic chain, the steady-state response relationship closure is used as the basis for determining the stability maintenance, debt waiting, and debt recovery windows.

[0138] Each succession step in the succession process contains a new steering command recovery amount. Therefore, the local dynamic steering response chain triggered by this steering command recovery amount is used to determine whether the current succession step is successful.

[0139] The waiting and recovery phases of debt do not necessarily generate new complete dynamic response chains. Therefore, a combination of dynamic closure results and steady-state closure results is used to determine whether the vehicle remains stable.

[0140] The vehicle motion controller sets the response chain credibility for each current response chain. The response chain credibility is determined based on the valid state of the sensor signals used to construct the response chain, the timestamp synchronization accuracy, the communication delay credibility, the signal noise level, the duration of the response event, and the attribution of the control event.

[0141] A response chain is designated as a high-confidence response chain when all critical signals it depends on are valid, the time synchronization error is less than the synchronization error threshold, and there are no control event attribution conflicts. A response chain is designated as a medium-confidence response chain when at least one auxiliary signal is invalid, but the critical signals have valid alternative sources.

[0142] When a critical response event cannot be determined, the time synchronization error is too large, or the attribution of the response event cannot be distinguished, the response chain is set as a low-confidence response chain or an unavailable response chain. The vehicle motion controller can independently determine the main anomaly chain segment based on the high-confidence response chain; for the medium-confidence response chain, it is necessary to combine other response chains, wheel slip state, or limited drive torque verification results to confirm the source of the anomaly; for the low-confidence response chain or the unavailable response chain, the source of the anomaly is not confirmed based on it alone, but a data-limited flag is generated, and the system enters an unverified conservative control or degraded operation state.

[0143] The vehicle may enter a low-traction surface when the target steering angle and target driving torque remain essentially constant, or be affected by crosswinds, road cross slopes, localized water accumulation, and road impacts. To avoid the dynamic response chain only triggering after a change in control command, the vehicle motion controller performs steady-state relationship monitoring while the changes in the target steering angle and each target driving torque remain continuously below the corresponding event trigger threshold.

[0144] Steady-state relationship monitoring includes the deviation between the target yaw rate and the actual yaw rate, the deviation between the planned target curvature and the actual driving curvature, the relationship between the actual steering angle and the steering reaction load, the slip state of each wheel, the trend of lateral position deviation, and the trend of heading angle deviation.

[0145] A steady-state anomaly trigger event is generated when the deviation between the target yaw rate and the actual yaw rate, the deviation between the planned target curvature and the actual driving curvature, or the slip state of at least one wheel continuously exceeds the corresponding steady-state anomaly threshold.

[0146] If the time interval between the most recent effective control event and the current steady-state anomaly triggering event does not exceed the event association validity time, then the most recent effective control event is used as the reference event for the steady-state anomaly, and a hold-state response record is established based on the state changes after that event. If the most recent effective control event has exceeded the event association validity time from the current moment, then a complete dynamic response chain is not forced to be constructed, and an unverified steady-state anomaly flag is generated.

[0147] For unverified steady-state anomalies, the vehicle motion controller first restricts the torque increase authority of suspected abnormal wheels, and then determines whether to allow the restricted drive torque verification based on the vehicle yaw state, road boundary margin, and actuator capability.

[0148] The baseline response chain characterizes the normal transmission relationship between control commands, actuator responses, wheel responses, and vehicle motion responses when the vehicle is operating under corresponding conditions and without abnormalities. The baseline response chain includes the baseline response sequence of each response event node, the baseline response interval range between adjacent event nodes, the baseline response direction of each state variable, and the baseline initial rate of change range of each state variable.

[0149] The baseline response chain includes the basic safety baseline response chain obtained from the vehicle's factory calibration and the operational baseline response chain established during vehicle operation. The operational baseline response chain does not directly cover the basic safety baseline response chain.

[0150] The steering reference response chain can be stored according to at least two of the following categories: vehicle speed range, vehicle load range, planned target curvature range, target steering angle change range, and target steering angle change rate range. The drive reference response chain can be stored according to at least two of the following categories: vehicle speed range, vehicle load range, target drive torque change range, target drive torque change rate range, drive motor speed range, and current actual drive torque range.

[0151] The historical vehicle state sequence is allowed to be used to update the operating baseline response chain only when the vehicle is in a preset stable driving state. The preset stable driving state must at least meet the following conditions: the deviation between the actual yaw rate and the target yaw rate is less than the yaw deviation threshold, the slip state of each wheel is within the normal range, the position deviation of the steering actuator is less than the position deviation threshold, there are no abnormal chain segments in the current response chain, and the vehicle is not in a limited drive torque verification, drive stability control, succession replacement, or deficit recovery state.

[0152] The operating baseline response chain is only allowed to be updated when a high-reliability normal response chain with a cumulative number of samples not less than the preset minimum number is obtained under the same operating conditions, and the response interval, response direction and initial change rate of the normal response chain are not close to the corresponding abnormal boundary.

[0153] The amount of a single update to the operating baseline response chain shall not exceed the maximum update range of the baseline, and the cumulative change over multiple consecutive update cycles shall not exceed the maximum allowable offset range of the basic safety baseline. If the updated operating baseline response chain results in a significant reduction in anomaly detection capability, or if the vehicle subsequently displays related actuator fault indicators, the operating baseline response chain or the basic safety baseline response chain before the update shall be reverted to be used.

[0154] When a vehicle is subjected to prolonged conditions of deviation in the same direction, high load on the same actuator, slight slippage of the same wheel, or abnormal road surface adhesion, the corresponding data will not be used to update the baseline response chain. After the vehicle undergoes maintenance, tire replacement, or steering system recalibration, the operational baseline response chain can be re-established, but the basic safety baseline is still retained as the boundary for anomaly detection.

[0155] When there is no reference response chain that perfectly matches the current operating condition, the reference response chain of the adjacent operating condition can be selected and the corresponding reference allowable range can be expanded. Alternatively, the reference response interval and the reference initial change rate range of the adjacent operating conditions can be interpolated.

[0156] The vehicle motion controller compares the current response chain with the corresponding reference response chain. If any chain segment exhibits at least one of the following conditions: the current response order is inconsistent with the reference response order, the current response interval exceeds the reference allowable range, the current response direction is inconsistent with the reference response direction, the current initial rate of change exceeds the reference range, or the subsequent response event node does not occur within the observation time limit, the chain segment is marked as an abnormal chain segment.

[0157] For a response chain, each segment is checked sequentially from the control command side to the vehicle motion response side, and the segment that first shows an anomaly is identified as the first abnormal segment of the response chain.

[0158] To prevent multiple minor anomalies from causing frequent changes in the main anomaly chain, the vehicle motion controller sets anomaly levels according to the anomaly type and the degree of deviation. Missing response anomalies and reversed response directions are set as the first anomaly level; abnormal response sequence anomalies are set as the second anomaly level; and response interval exceeding limits and initial rate of change anomalies are set as the third or fourth anomaly level depending on the degree to which they exceed the baseline allowable range. The first anomaly level is higher than the second anomaly level, the second anomaly level is higher than the third anomaly level, and the third anomaly level is higher than the fourth anomaly level.

[0159] When the difference between the occurrence time of the first abnormal chain segment in the steering and the first abnormal chain segment in the driving sequence is greater than the concurrent abnormal time threshold, the first abnormal chain segment that occurred earlier is preferentially identified as the main abnormal chain segment. When the difference between the occurrence times of the two is not greater than the concurrent abnormal time threshold, the corresponding abnormal levels are compared, and the first abnormal chain segment with the higher abnormal level is identified as the main abnormal chain segment, and the other first abnormal chain segment is identified as the associated abnormal chain segment.

[0160] When two initial abnormal chain segments have the same abnormality level, or the difference in their abnormality levels is less than a preset difference, they are identified as concurrent initial abnormal chain segments, and the candidate type of the abnormality source is set as a composite abnormality. When multiple drive wheels have initial abnormal chain segments at the same time, the main drive abnormal chain segment is determined sequentially based on the abnormality level, the time of abnormality occurrence, the degree of wheel slippage, and the degree of deviation from the reference range.

[0161] Once the primary exception chain is determined, a minimum retention time is set. During this minimum retention time, the primary exception chain will not be switched to another chain unless a higher-level exception chain appears. After the minimum retention time expires, if the original primary exception chain has recovered and another exception chain persists, the primary exception chain can be updated.

[0162] When the main abnormal chain segment is the steering command-actuator drive chain segment, the steering control link, the power supply status of the steering actuator, or the steering actuator control unit are identified as suspected abnormal objects. When the main abnormal chain segment is the actuator drive-actual steering angle chain segment, the steering motor, the steering transmission mechanism, the limit status of the steering actuator, or the output capability of the steering actuator are identified as suspected abnormal objects.

[0163] When the main abnormal chain segment is the actual steering angle-steering reaction load chain segment, the steering transmission mechanism, steering wheel grounding action, or steering reaction load detection link are identified as suspected abnormal objects. When the main abnormal chain segment is the steering reaction load-lateral acceleration chain segment, the steering axle tire lateral action, steering wheel adhesion status, or tire longitudinal and lateral force saturation are identified as suspected abnormal objects.

[0164] When the main anomaly chain segment is the lateral acceleration-yaw rate chain segment, the front and rear axle lateral force distribution status, the non-steering axle tire status, or external disturbances to the vehicle body are identified as suspected anomalies. When the main anomaly chain segment is the target driving torque-actual driving torque chain segment, the corresponding drive actuator, drive control link, or driving torque limiting status are identified as suspected anomalies.

[0165] When the main anomaly chain segment is the actual driving torque-wheel angular acceleration chain segment, the corresponding drive wheel adhesion state, tire state, or transmission state is identified as a suspected anomaly. When the main anomaly chain segment is the wheel angular acceleration-vehicle motion response chain segment, the tire force transmission state to the vehicle body, the local road adhesion state, or the state in which multiple wheel forces cancel each other out is identified as a suspected anomaly. The above correspondence is used to generate suspected anomaly objects and is not directly used as the final anomaly source conclusion.

[0166] When the steering actuator drive sub-chain, actual steering angle-steering reaction load chain segment, and each drive response chain are all within the baseline allowable range, but the lateral acceleration-yaw rate chain segment is abnormal, and no corresponding abnormality is found in the slip state of each wheel, the candidate type of the abnormality source is set to external disturbance or unclassified body response abnormality.

[0167] When a vehicle is equipped with a crosswind detection device, road cross slope detection information, or vehicle body external force estimation function, the source of external disturbance can be further confirmed based on the crosswind direction and intensity, road cross slope direction and gradient, and vehicle load status. If the vehicle is not equipped with the corresponding detection capabilities, the anomaly is classified as a verification uncertainty anomaly and is not directly identified as a road surface adhesion anomaly.

[0168] In the event of external disturbances or unclassified abnormal vehicle body response, the vehicle motion controller generates a limited additional yaw moment based on the actual yaw rate deviation and lateral position deviation, while limiting the target longitudinal velocity and the planned target curvature, and continuously judging whether the steering response chain and drive response chain have recovered.

[0169] The wheel slip state is determined based on the difference between the wheel's circumferential velocity and the reference velocity of the wheel center along the wheel's rolling direction. For the i-th wheel, the vehicle motion controller determines the reference velocity of the wheel center along the wheel's rolling direction based on the vehicle's longitudinal velocity, lateral velocity, yaw rate, the wheel's position relative to the vehicle's center of gravity, and the wheel's steering angle.

[0170] Under driving conditions, the driving slip state is determined by the degree to which the wheel circumferential speed is higher than the wheel rolling direction reference speed. Under braking conditions, the braking slip state is determined by the degree to which the wheel circumferential speed is lower than the wheel rolling direction reference speed.

[0171] When the reference speed in the rolling direction of the wheel is lower than the preset minimum speed, the slip ratio value is not used directly. Instead, the wheel slip state is divided into normal, mild slip, moderate slip and severe slip levels based on the correspondence between the wheel circumferential speed difference, wheel angular acceleration, driving torque change and vehicle longitudinal acceleration.

[0172] When determining the vehicle reference speed, the vehicle motion controller discards wheels that have been marked as severely slipping or whose sensors are ineffective. If the number of remaining valid wheels is insufficient, it prioritizes inertial measurement unit information and vehicle positioning speed, and lowers the confidence level of the slip condition judgment result. When the confidence level of the slip condition is lower than the minimum confidence level for slip judgment, suspected low-adhesion wheels are not identified separately based on this slip result.

[0173] When the time difference between the changes in angular acceleration of the left and right corresponding wheels exceeds the wheel response time difference threshold, the actual driving torque change direction, wheel angular acceleration change direction, and wheel slip state of the left and right wheels are further compared.

[0174] When the actual driving torque of a wheel does not increase but the wheel angular acceleration changes along the direction of increasing wheel speed, or the change in angular acceleration corresponding to the change in unit driving torque of the wheel is greater than that of the corresponding wheel on the other side, or the degree of slippage of the wheel is higher than that of the corresponding wheel on the other side, or the actual driving torque-wheel angular acceleration chain segment of the wheel becomes abnormal before the corresponding chain segment on the other side, the side where the wheel is located is identified as the suspected low adhesion side.

[0175] When calculating the change in wheel angular acceleration corresponding to a unit change in driving torque, this calculation is performed only when the actual change in driving torque is greater than the minimum effective change in driving torque. If the actual change in driving torque is not greater than the minimum effective change in driving torque, this ratio is not used to identify the suspected low-adhesion side.

[0176] It should be noted that, in this embodiment, driving information refers to the target trajectory, target lateral position, target heading angle, planned target curvature, target longitudinal speed, target yaw rate, and target trajectory identifier, while response information refers to steering control and feedback information, driving torque control and feedback information, wheel side motion information, and vehicle body motion information.

[0177] It should be noted that constraint information can be defined as road boundary, vehicle safety envelope, wheel slip state, actuator allowable output range, and emergency control state. The current response chain can be defined as the currently constructed steering actuator drive response chain, steering wheel side response chain, and drive response chain of each drive wheel. The baseline response chain refers to the basic safety baseline response chain and the operational baseline response chain. The main abnormal chain segment refers to the main abnormal chain segment determined from the steering first abnormal chain segment and each drive first abnormal chain segment. The suspected abnormal object refers to the steering control link, steering actuator, steering transmission mechanism, tire, drive actuator, drive wheel, or road adhesion area. The drive object refers to the drive wheel with adjustable drive torque, multiple associated drive wheels, drive axle, or total drive torque of the axle. The restricted drive torque verification refers to verifying torque reduction, restricted compensation torque increase, response chain reconstruction, and comparison of abnormal chain segments before and after verification.

[0178] It should be noted that the sources of anomalies refer to abnormal steering actuator operation, limited steering performance, saturated tire force, decreased road surface adhesion, abnormal drive actuator operation, and combined anomalies. Unverified anomaly indicators refer to anomaly indicators generated when verification conditions are not met, critical response chains are unavailable, or anomalies cannot be verified. Execution capability refers to the torque amplification capability, torque reduction capability, differential control capability, permissible output range, and auxiliary braking capability of the driven object. Drive control quantities include single-wheel drive torque, axle drive torque, additional drive torque difference, and auxiliary braking torque.

[0179] It should be noted that the succession steps include corresponding drive-side control reduction, steering control recovery, succession confirmation, and succession reversal. Yaw control quantity refers to the additional yaw moment or the yaw control contribution after normalization. Response relationship closure refers to the closure of the dynamic response chain and the steady-state response relationship. Intervention command deficit refers to the projection components of curvature deficit, yaw control deficit, longitudinal drive force deficit, and lateral position deficit. Target trajectory and target trajectory information can be defined as the target trajectory body, target trajectory identifier, update time, trajectory position, and trajectory direction information. Road boundary refers to the distance from the vehicle safety envelope to the left and right boundaries of the road and the passable area of ​​the road.

[0180] It should be noted that the debt recovery window refers to a state where the response relationship is closed, there are no new anomalies, confirmed anomalies have not recurred, boundary conditions are allowed, the debt is valid, the road section is in front of the vehicle, and the target trajectory can be recovered. The response event node refers to the starting moment when the control command, actuator state, wheel side state, or vehicle body state undergoes a valid change. The control event refers to the steering command change event and the target driving torque change event. The response chain segment refers to two adjacent response event nodes and the response relationship between them. The steering response chain refers to the steering actuator drive response chain and the steering wheel side response chain. The drive response chain can be defined as a response chain composed of the target driving torque, the actual driving torque, the wheel motion response, and the vehicle body motion response.

[0181] It should be noted that joint response records refer to coupled control event records and joint drive event records. Steady-state relationship monitoring can be defined as continuous monitoring of the target and actual yaw state, planned and actual curvature, steering state, wheel slip state, and trajectory deviation trend. Steady-state response records refer to the hold-state response records formed after the most recent effective control event. Unverified steady-state anomaly identifiers are identifiers generated when the current steady-state anomaly cannot be reliably correlated with an effective control event. Trustworthy states include high trustworthiness, medium trustworthiness, low trustworthiness, and unavailable response chain states. Current operating conditions include vehicle speed, vehicle load, target curvature, control change state, and actuator operating state. Anomaly response chain segment refers to a chain segment with anomalies in response sequence, response interval, response direction, initial change rate, or response missing state. The first anomaly chain segment refers to the chain segment that first shows anomalies along the control command side towards the vehicle body response side. Concurrent anomaly chain segments refer to multiple first anomaly chain segments that occur within the associated time and for which a single main anomaly chain segment cannot be determined.

[0182] S2. When verification is allowed, perform restricted driving torque verification on the driving object and determine the source of the anomaly based on the change in the response chain before and after verification; otherwise, generate an unverified anomaly flag and restrict the driving object.

[0183] After identifying a suspected abnormal object, the vehicle motion controller determines whether the current vehicle state allows for the execution of the restricted drive torque verification. Verification conditions include: the distance from the vehicle's safety envelope to both the left and right boundaries of the road is greater than the corresponding verification distance threshold; the deviation between the actual yaw rate and the target yaw rate is less than the verification yaw threshold; the degree of slippage of each wheel is lower than the severe slippage threshold; the vehicle's longitudinal speed is greater than the minimum verification speed; at least one relevant drive actuator has torque reduction margin; the vehicle is not in an emergency braking state; the trajectory planning module has not output emergency obstacle avoidance or emergency stop commands; and the steering actuator does not have a complete failure indicator.

[0184] If all the above verification conditions are met, the system enters the restricted driving torque verification state. If at least one verification condition is not met, the verification driving torque adjustment is not performed; instead, an unverified anomaly flag is generated, and the system enters the conservative stable control state.

[0185] When the suspected abnormal object is a drive wheel, that drive wheel is designated as the verification object. When the suspected abnormal object is an axle, the drive wheel with the highest degree of slippage within that axle, or the total driving torque of that axle, is designated as the verification object. When the suspected abnormal object is a steering actuator or steering transmission mechanism, the steering wheel or steering axle currently bearing a large longitudinal force is selected as the verification object. By releasing part of the longitudinal tire force on that steering wheel or steering axle, it is determined whether the steering response abnormality still exists.

[0186] The vehicle motion controller reduces the driving torque of the test object within a preset verification period. The verification torque reduction can be a pre-calibrated fixed torque reduction, a preset proportion of the current driving torque of the test object, or a graded torque reduction corresponding to the wheel slip level, and is limited based on the distance from the vehicle to the road boundary and the current yaw rate deviation. The verification torque reduction does not exceed the current available torque reduction margin of the test object.

[0187] During the verification process, the driving torque of other non-suspected abnormal drive wheels can be increased to limit changes in the vehicle's longitudinal speed. When the vehicle has non-suspected abnormal drive wheels that are symmetrically arranged, the compensation driving torque should be distributed symmetrically to ensure that the net additional yaw moment generated by the compensation action is less than the verification allowable yaw moment.

[0188] If the vehicle does not have symmetrical compensation capabilities, the compensation torque can be reduced or eliminated, allowing the total longitudinal driving force of the vehicle to temporarily decrease during the verification period. The longitudinal speed change of the vehicle during the verification process shall not exceed the allowable speed change during verification.

[0189] The vehicle motion controller determines the change in driving force along the rolling direction of the drive wheels based on the actual driving torque, effective rolling radius, and transmission efficiency of the drive wheels. For drive wheels with a steering angle, the change in driving force along the rolling direction is transformed to the vehicle coordinate system to obtain the force change components in the longitudinal and lateral directions of the vehicle.

[0190] In the vehicle coordinate system, the longitudinal direction of the vehicle is defined as forward. In the positive direction of the axis, the left side of the vehicle is... In the positive direction of the axis, above the vehicle is In the positive direction of the axle, the additional yaw moment change caused by the change in the force of each wheel is:

[0191]

[0192] in, This represents the change in the vehicle's additional yaw moment caused by control actions such as verification torque reduction, compensation torque increase, differential drive, or additional braking, expressed in Newton-meters (N·m). . It represents the change in the corresponding physical quantity relative to the state before the control action was implemented or the state of the previous control cycle. Indicates torque; subscript This indicates that the torque is about the vehicle coordinate system. The axle generates torque, which is the torque in the yaw direction of the vehicle. This indicates a summation operation, which accumulates the changes in yaw moment generated by each wheel involved in the current verification control, stability control, or drive steering takeover control. This represents the wheel serial number, which is an index variable used to distinguish different wheels. This indicates the number of wheels involved in the current control action, and its value is a positive integer. For example, when the vehicle only adjusts the torque of the left and right drive wheels, Option 2 can be chosen; when the vehicle adjusts its four wheels independently, 4 is an acceptable value. Indicates the first The coordinates of the point of application of each wheel relative to the vehicle's center of gravity in the longitudinal direction of the vehicle are expressed in meters. (The text then continues with a description of the vehicle's longitudinal direction, which is not directly related to the coordinates of the wheels.) The convention for the positive axis direction is that the wheel located in front of the vehicle's center of gravity corresponds to... Typically a positive value, corresponding to the wheel located behind the vehicle's center of gravity. It is usually a negative value.

[0193] Indicates the first The coordinates of the point of action of each wheel relative to the vehicle's center of gravity in the lateral direction of the vehicle are shown in meters, with the left side of the vehicle as the reference point. The convention of the positive axis direction corresponds to the wheel on the left side of the vehicle. A positive value corresponds to the wheel on the right side of the vehicle. It is a negative value. Indicates the first The forces acting on each wheel in the vehicle coordinate system The component of change along the axial direction, i.e., the change in longitudinal force of the vehicle, measured in Newtons (N). Indicates the longitudinal direction of the vehicle, in the subscript Indicates the first For each wheel, this change can be determined by the changes in driving torque, braking torque, and wheel steering angle. Indicates the first The forces acting on each wheel in the vehicle coordinate system The component of change along the axial direction, i.e., the change in lateral force of the vehicle, measured in Newtons (N). Indicates the lateral direction of the vehicle, in the subscript. Indicates the first Each wheel. For a drive wheel with a steering angle, the force generated along the rolling direction of the wheel needs to be converted to the vehicle coordinate system to obtain... and .

[0194] The product in the formula Indicates the first The change in lateral force of each wheel is the change in yaw moment formed by its longitudinal lever arm relative to the vehicle's center of gravity; product Indicates the first The change in longitudinal force of each wheel is the change in yaw moment formed by its lateral lever arm relative to the vehicle's center of gravity.

[0195] For non-steering drive wheels with a steering angle close to zero, the above relationship can be simplified to:

[0196]

[0197] The vehicle motion controller determines the net additional yaw moment generated by the verification torque reduction, compensation torque increase, and stability control based on the above relationships, and ensures that the net additional yaw moment during the limited drive torque verification period does not exceed the verification allowable yaw moment. For the additional braking torque, it is converted into the longitudinal force change component of the corresponding wheel and then included in the additional yaw moment calculation.

[0198] The verification period is determined based on vehicle speed, drive actuator response time, and the longest response interval in the baseline response chain. In one implementation, the verification period is set to 50 milliseconds to 300 milliseconds.

[0199] During the verification period, the response event nodes are re-detected, and the steering actuator drive response chain, steering wheel side response chain, and related drive wheel drive response chain are reconstructed. The main abnormal chain segment before and after verification is compared to see if it disappears, remains unchanged, is transferred to other chain segments, or if a new abnormal chain segment appears. At the same time, it is determined whether the degree of suspected abnormal wheel slippage, actual steering angle response, steering reaction load response, lateral acceleration response, and yaw rate response have recovered.

[0200] When the compensation torque increase results in a new target driving torque—an abnormal actual driving torque, abnormal wheel slippage, or a torque-limiting state of the drive actuator—the corresponding compensation torque increase is stopped, and the newly emerging abnormal chain segment is recorded as a verification-induced abnormal chain segment. The verification-induced abnormal chain segment is not directly used as the original source of the abnormality, but is used to determine the available capability of the corresponding drive actuator.

[0201] If any of the following occurs during the verification process: the yaw rate deviation exceeds the verification termination threshold, the distance from the vehicle to any road boundary is less than the verification termination distance, any wheel slippage reaches the severe slippage level, the trajectory planning module outputs an emergency control command, the drive actuator generates a fault indicator, or the steering actuator generates a complete failure indicator, the verification will be terminated, the drive torque state before verification or after passing the safety limit will be restored, and the system will enter a conservative stable control state.

[0202] When the main abnormal chain segment is located in the steering reaction load-lateral acceleration chain segment or the actual driving torque-wheel angular acceleration chain segment, and after verification that the original main abnormal chain segment disappears after torque reduction, the degree of suspected abnormal wheel slippage decreases, the corresponding driving response chain recovers to the baseline allowable range, and the lateral acceleration or yaw rate response recovers, the source of the abnormality is determined to be a change in road surface adhesion state or saturation of the longitudinal and lateral combined force of the tires.

[0203] When attachment-related anomalies mainly occur when the vehicle's longitudinal driving force is large, and the response chain recovers after reducing the longitudinal driving force of the suspected abnormal wheel or axle, and the corresponding response relationship remains normal when the vehicle passes through the same road area under a lower longitudinal driving force, then the source of the anomaly is preferentially identified as tire longitudinal and lateral force saturation.

[0204] When a vehicle still experiences wheel slippage under low longitudinal driving force, a significant increase in the change of wheel angular acceleration corresponding to a change in unit driving torque, or repeated abnormal driving response of adjacent wheels in the same road area, the source of the abnormality should be identified as a decrease in road surface adhesion.

[0205] When the existing response information is insufficient to reliably distinguish between a decrease in road surface adhesion and saturation of the combined longitudinal and lateral force of the tire, the combined anomaly type of "change in road surface adhesion or saturation of the combined longitudinal and lateral force of the tire" is retained, and the torque limit is implemented according to the lower available adhesion.

[0206] If there is an anomaly in the steering command-actual steering angle response before verification, and the response returns to the baseline allowable range after reducing the longitudinal driving force of the steering axle, and the steering motor control current or actual output torque is close to the upper limit of output before verification, the source of the anomaly is determined to be that the steering execution capability is limited by load.

[0207] A load-limited steering capability indicates that the steering actuator still has the ability to move, but under the current conditions of the combined longitudinal and lateral forces of the tires, the mechanical resistance of the steering mechanism, or the power supply capacity, it cannot complete the response according to the original target steering command. For this anomaly, the vehicle motion controller limits the available target steering angle and the rate of change of the target steering angle, reduces the longitudinal driving force of the steering axle, and retains a portion of the additional driving torque difference to jointly bear the yaw control.

[0208] When the steering command-actual steering angle response remains abnormal after torque reduction verification, the steering actuator position deviation does not decrease, and at least one of the following: steering actuator control current, output torque, power supply voltage, limit status, communication status, or fault status exceeds the corresponding normal range, the source of the abnormality is determined to be an abnormal steering actuator operation.

[0209] The above-normal range includes control current reaching the upper limit, control current being significantly lower than the range required for normal operation, insufficient power supply voltage, actuator reaching mechanical limit, control communication interruption, or actuator fault indication being valid.

[0210] When the steering command—actual steering angle response is normal, and the steering reaction load—lateral acceleration chain segment recovers after the longitudinal driving force is reduced, the abnormal source is preferentially identified as saturation of the combined longitudinal and lateral forces of the tires or a decrease in road surface adhesion.

[0211] When the main abnormal chain segment is the target driving torque - actual driving torque chain segment, and the actual driving torque fails to reach the target driving torque within the allowable time, and the drive actuator outputs torque limit, over-temperature, over-current, power supply abnormality, or communication abnormality, the abnormality source is determined to be a drive actuator operation abnormality.

[0212] When there is a significant difference in the drive response chains of the corresponding left and right wheels, and the wheel slippage on the suspected low-adhesion side decreases after torque reduction, while the drive response chain of the other wheel remains normal, the source of the anomaly is determined to be left and right separation adhesion.

[0213] When the abnormal response of the steering actuator persists, and the attached chain segment recovers after the torque reduction is verified, the source of the abnormality is determined to be a composite abnormality of steering actuator action abnormality or steering execution capability being limited by load and road surface adhesion state.

[0214] When the original main exception chain disappears but a new exception chain appears in another response chain, the new exception chain is identified as the updated main exception chain.

[0215] When the original main abnormal chain segment is transferred from the steering response chain to the target driving torque - actual driving torque chain segment, and the new abnormal chain segment is triggered by the compensation torque increase, a driving execution capability limitation flag is generated, and the driving torque is stopped from being increased through the corresponding driving wheel.

[0216] When the constrained driving torque verification has been performed, but none of the response chains have recovered to the corresponding baseline allowable range and the explicit anomaly classification conditions are not met, a verification uncertainty anomaly flag is generated.

[0217] For unverified anomalies and unverified anomalies, the vehicle motion controller prohibits torque increase of suspected abnormal wheels, reduces the total driving torque of suspected abnormal axles, limits the rate of change of driving torque of each drive wheel, and limits the planned target curvature or reduces the target longitudinal speed according to the degree of anomaly.

[0218] In this embodiment, the verification conditions refer to the verification admission conditions composed of yaw state, wheel side state, road boundary state, vehicle operating state, and actuator capability. The verification object refers to the drive wheel, drive axle, or total drive torque of the axle associated with the suspected abnormal object. The safety constraint range can be defined as the allowable range for yaw state, wheel side slip, road boundary, control change rate, and actuator output settings. The limited compensation control quantity refers to the compensation drive torque allocated to other available drive objects to maintain the longitudinal operating state. The verification induced abnormality refers to the abnormal drive response, wheel side abnormality, or limited execution capability state caused by the compensation control.

[0219] S3. Allocate drive control quantities based on the source of the anomaly or unverified anomaly identifier, yaw trend, and execution capability; after steering is restored, transfer the drive-side yaw control quantities to the steering side according to the established replacement steps.

[0220] The vehicle motion controller determines the vehicle yaw trend based on the deviation between the target yaw rate and the actual yaw rate, the direction of change of the yaw rate deviation, the lateral offset direction of the vehicle relative to the current target trajectory, and the slip state of the front and rear axle wheels.

[0221] When the actual yaw rate is less than the target yaw rate, the actual curvature is less than the planned target curvature, and the vehicle deviates towards the outside of the curve, the vehicle is determined to be in an understeer tendency. When the actual yaw rate is greater than the target yaw rate, the yaw rate deviation continues to increase, or the slippage of the non-steering axle wheels increases significantly, the vehicle is determined to be in an oversteer tendency.

[0222] When the target yaw rate is opposite to the actual yaw rate, the vehicle is identified as being in an emergency yaw anomaly state. Without performing verification actions, the vehicle directly enters an emergency stability control or degraded operation state.

[0223] The vehicle motion controller determines the drive wheels that should not increase torque, the drive wheels that should reduce torque first, the drive wheels that should undertake additional yaw control, and the direction, upper limit and rate of change of the driving torque of each drive wheel based on the source of the anomaly, the main anomaly chain, the vehicle yaw trend, the set of available actuators and the allowable driving torque range of each drive wheel.

[0224] Drive torque distribution prioritizes prohibiting torque increase from wheels suspected of having low adhesion, prioritizes reducing the drive torque of wheels with saturated tire force, and prioritizes selecting wheels not suspected of abnormality to undertake additional yaw control. For wheels requiring torque increase, both permissible drive torque and wheel slip limit are applied simultaneously. When stability control requirements conflict with longitudinal speed maintenance requirements, longitudinal drive force is prioritized for limitation.

[0225] When the lateral action of the steering axle is abnormal and the vehicle is in a tendency to understeer, the torque increase of the suspected abnormal steering wheel is restricted, the driving torque of the steering axle is reduced first, and the left and right independently controllable drive wheels of the non-suspected abnormal axle are selected to form an additional driving torque difference, so that the direction of the additional yaw torque is the same as the target yaw direction.

[0226] When the non-steering axle drive response is abnormal and the vehicle is in an oversteering tendency, the drive torque of the suspected abnormal wheel is reduced first, the total drive torque of the corresponding axle is limited, and other available drive wheels are selected to form an additional yaw torque opposite to the actual excessive yaw direction.

[0227] When there is left and right separation adhesion, it is prohibited to increase the driving force of the low-adhesion side drive wheel to form an additional yaw moment. Instead, the driving torque of the low-adhesion side drive wheel should be reduced first, and the amplitude-limited torque increase should be implemented within the allowable range of the high-adhesion side drive wheel.

[0228] When the abnormality source includes abnormal steering actuator operation or limited steering execution capability, the available target steering angle is determined according to the actual available steering angle range of the steering actuator, and the yaw control amount undertaken by the additional driving torque difference is determined according to the difference between the target yaw control requirement and the yaw control effect that the available target steering angle can provide.

[0229] During the output of the additional driving torque difference, the vehicle motion controller continuously determines whether the lateral tire action generated by the steering control has been established. Before the steering lateral action reaches the replacement condition, at least the following necessary safety conditions should be met: the original main abnormal chain segment has disappeared or decreased to the permissible abnormal level, there is no new high-level abnormal chain segment in the steering response chain, the deviation between the actual yaw rate and the target yaw rate has not continued to increase, the distance from the vehicle safety envelope to the road boundary is greater than the replacement safety distance, and the slip state of the suspected abnormal wheel has not worsened.

[0230] Based on meeting the above-mentioned necessary safety conditions, when at least two of the following conditions are met consecutively for a preset number of control cycles: the actual steering angle reaches a preset ratio of the available target steering angle, the steering reaction load enters the corresponding reference range, and the lateral acceleration changes along the target yaw direction, the steering lateral action is determined to have reached the replacement condition.

[0231] When the steering response chain transitions from an open state to a closed state, each response interval must fall within the first allowable range, each response direction must conform to the reference direction, and the first confirmation time must be maintained continuously. When the steering response chain exits the closed state, it exits the closed state only if any response interval exceeds the second allowable range which is wider than the first allowable range, an anomaly of reversed response direction occurs, or an anomaly of missing response occurs.

[0232] The drive response chain closure, the limited drive torque verification state, the succession replacement state, and the debt recovery state all use different entry thresholds, exit thresholds, and minimum hold times to prevent frequent switching of control states.

[0233] Drive steering takeover includes full takeover and partial takeover. Full takeover is performed when the steering actuator and steering response chain return to their normal allowable range, and the steering side is able to independently handle the current target yaw control requirements. During full takeover, the additional drive torque difference is gradually reduced according to multiple complete takeover steps until the drive torque exit threshold is reached.

[0234] When the steering actuator exhibits limited steering capability, a limited usable steering angle range, or sluggish steering response, but still retains some usable capability, partial takeover is performed. During partial takeover, only the yaw control quantity that the steering side can stably handle is transferred from the drive side to the steering side, while retaining an additional drive torque difference not lower than the residual drive control threshold.

[0235] After partial handover is completed, the vehicle motion controller continues to limit the target longitudinal speed and plan the target curvature, and continuously monitors the steering actuator drive sub-chain and the steering wheel side response sub-chain. When the steering actuator returns to normal and the steering response chain closes again, the system transitions from the partial handover state to the full handover state.

[0236] When the steering actuator fails completely or the local steering response chain cannot be closed continuously, the switch from drive control to steering control is not executed. Instead, the currently available drive stability control is maintained, and the system enters a degraded operation or safe stop state based on the road boundary and the remaining capacity of the actuator.

[0237] Each successive replacement step includes an additional reduction in the difference in driving torque and a corresponding steering command recovery amount. The vehicle motion controller first determines the additional driving yaw torque to be released in the current replacement step based on the current driving torque of each drive wheel and the wheel force it generates; then, based on the current vehicle longitudinal speed, the planned target curvature, the vehicle load, the local response gain of the steering system, and the lateral action state of the tires, it determines the corresponding steering command recovery amount.

[0238] The correspondence can be determined by the calibrated "driving additional yaw moment - steering command increment" correspondence table, or by the local response gain between the change in steering command and the change in yaw rate under the current working conditions.

[0239] In each successive replacement step, the sum of the reduction in additional yaw control on the drive side and the increase in yaw control on the steering side should be within the allowable deviation range of the current target yaw control requirement. If the sum of the two is higher than the upper limit of the allowable range, the steering command recovery amount should be reduced; if the sum of the two is lower than the lower limit of the allowable range, the release amount of the additional drive torque difference should be reduced or the current replacement step should be paused.

[0240] The steering command recovery amount used to construct the local steering response chain should be greater than the minimum trigger amount of the local steering event. When the steering command recovery amount determined according to the yaw control requirements is less than the minimum trigger amount, adjacent successive step sizes can be merged or the successive observation window can be extended, and the local steering response chain should not be constructed based on steering command changes within the sensor noise range.

[0241] When executing each succession step, the amount of steering command recovery in that step is taken as a local steering command change event. The steering actuator drive increment, actual steering angle increment, steering reaction load increment, lateral acceleration increment, and yaw rate increment are detected sequentially in the succession observation window, thus forming a local steering response chain.

[0242] When the local steering response chain is closed, the deviation between the actual yaw rate and the target yaw rate does not increase, the wheel slip condition does not worsen, the distance between the vehicle and the road boundary does not decrease to the boundary distance threshold, and no new main abnormal chain segment appears, the current replacement step is determined to be successful, and the next replacement step is executed.

[0243] If the local steering response chain is not closed, a new abnormal chain segment appears, the yaw rate deviation increases, the wheel slip condition deteriorates, the road boundary margin is insufficient, or the steering actuator condition exceeds the allowable range, the current replacement step is determined to have failed.

[0244] When a handover fails, a synchronous rollback method is prioritized to simultaneously restore the additional driving torque difference and steering command recovery amount to the state corresponding to the previous successful handover step length. If a sudden rollback of the steering command may cause a shock to the steering actuator, the current steering command is kept from increasing further, and the additional driving torque difference corresponding to the previous successful handover step length is restored according to a limited rate of change. After the vehicle's condition stabilizes, the increment of steering commands exceeding the previous successful state is gradually reduced.

[0245] When the number of consecutive failed replacements reaches a preset number, the replacement state is exited, drive stability control is maintained, and the target longitudinal speed is reduced and / or the planned target curvature is limited.

[0246] In this embodiment, the set of available execution objects refers to the set of execution objects with available torque increase, torque decrease, differential control or auxiliary braking capabilities. The control change amount refers to the increase, decrease, transfer amount or auxiliary braking change amount of the driving torque allocated to each execution object. The restricted operation process can be defined as conservative stable control, degraded operation or safe stopping process.

[0247] It should be noted that the same control representation refers to the additional yaw moment representation or the unified normalized yaw control contribution representation. The local steering response chain refers to the execution response, steering response, wheel side response, and body response chain triggered by the current steering control recovery amount. The successful succession state can be defined as the state corresponding to the previous succession step size that has satisfied the local response chain closure and vehicle state constraints. The corresponding road segment can be defined as the target trajectory segment determined by the location, effective distance, or effective time of the debt generation. The recovery direction can be defined as the direction in which the intervention command debt is allowed to be repaid in the tangential-normal coordinate relationship of the target trajectory.

[0248] S4. After taking over, determine the debt recovery window and restore the operation status based on the response relationship closure, intervention command backlog, target trajectory and road boundary.

[0249] The vehicle motion controller establishes an intervention command deficit starting from the control cycle in which the planned control command is first constrained by stable control. The intervention command deficit includes curvature deficit components, yaw control deficit components, longitudinal driving force deficit components, and lateral position deficit projection components.

[0250] Curvature deficit is formed based on the difference between the planned target curvature command and the stabilized limit curvature command. Yaw control deficit is formed based on the difference between the planned target yaw control requirement without stabilized control and the actual allowable yaw control amount after stabilized control. Longitudinal drive force deficit is formed based on the difference between the planned target longitudinal drive force and the actual allowable longitudinal drive force after stabilized control.

[0251] For any instruction-level debt component, its update considers both the debt newly added in the current control cycle and the debt actually repaid. The update relationship is expressed as follows:

[0252]

[0253] in, Indicates the first At the end of the control cycle or the first The first control cycle corresponding to the first Accumulated value of outstanding intervention instructions. Indicates the cumulative amount of debt owed under intervention instructions, subscript This represents the debt category index, used to distinguish different types of debt components.

[0254] In this embodiment, These can be respectively addressed as curvature deficit, yaw control deficit, or longitudinal driving force deficit. To further avoid ambiguity, the three types of deficits can also be written as... and ,in Indicates curvature. This indicates the yaw control amount or the additional yaw moment. This indicates the longitudinal driving force.

[0255] This represents the discrete control cycle number of the vehicle motion controller, and is a non-negative integer. This indicates the outstanding debt value for the next control period after the calculation of new outstanding debts and debt repayments for the current control period is completed. The +1 in 1 represents relative to the first... The next control cycle after the previous control cycle does not mean that the outstanding debt value itself increases by 1. The symbol 'sat' represents a limiting function or saturation function, used to limit the outstanding debt value calculated within the parentheses to between a pre-set lower and upper limit to prevent the outstanding debt from accumulating indefinitely. Its processing relationship can be understood as follows: when the calculated result is higher than the upper limit, output the upper limit; when the calculated result is lower than the lower limit, output the lower limit; when the calculated result is between the upper and lower limits, output the original calculated result.

[0256] Indicates the first Within the first control cycle, the newly added [number] due to stability control limitations The class did not execute the control variable. Indicates a control command or control variable; subscript This indicates the category of outstanding debt to which the control quantity belongs; the parentheses This indicates that the control quantity corresponds to the first One control cycle. In Indicates the difference in control quantity before and after the stability control limit, indicated by the superscript. This indicates that the difference represents the newly added debt in the current control cycle. "new" is merely a superscript used to describe the nature of the variable and does not represent a mathematical exponentiation operation.

[0257] Specifically, when When indicating a curvature deficit, This represents the difference between the planned target curvature command in the reference normal control output and the executed curvature command output after stabilization and limiting; when When indicating a yaw control deficit, it represents the difference between the reference normal yaw control amount and the actual yaw control amount executed after stabilizing the amplitude; when When indicating a shortfall in longitudinal driving force, it represents the difference between the reference normal longitudinal driving force and the actual longitudinal driving force after stabilization and limiting.

[0258] Indicates the first Within each control cycle, in addition to fulfilling the normal control requirements for the current target trajectory, extra actions are performed to repay the [previous] [cycle]. The control quantity for historical debts, with the superscript "pay" indicating the amount of debt repaid, is also a textual superscript that describes the purpose of the variable and does not indicate exponentiation.

[0259] When the vehicle is in a stable intervention phase and historical debts cannot be restored. Zeroing. When the vehicle enters the debt recovery window, the vehicle motion controller determines the recovery time based on the current debt value, road boundary margin, actuator remaining capacity, and expected recovery duration. .

[0260] This indicates the control cycle or sampling cycle of the vehicle motion controller, in seconds, with the subscript... This indicates sampling, which refers to the sampling or control periodicity attribute. Multiply by the newly added unexecuted control quantity to convert the control quantity difference in the current period into the debt increment; multiply by the debt repayment control quantity to convert the repayment control quantity in the current period into the debt deduction quantity.

[0261] During the stabilization intervention phase, the amount of debt repayment is typically zero, and the debt accumulates based on the amount of unexecuted control caused by stabilization constraints. During the debt recovery phase, the vehicle motion controller determines the amount of debt repayment for the current control cycle based on the current debt value, expected recovery time, road boundary margin, wheel attachment status, and actuator availability.

[0262] The amount of control work performed beyond the normal control requirements for completing the current target trajectory is used to reduce historical debt. Control commands executed only for the current target trajectory are not counted as historical debt repayment, to avoid double-counting normal trajectory tracking control as debt repayment.

[0263] When both new stability constraints and partial repayment of historical debts exist within the same control cycle, the newly added debt amount and the debt repayment amount are calculated separately, and the debts are updated according to the above update relationship. When the debt direction is opposite to the currently allowed recovery direction, the corresponding debt repayment amount is set to zero, and the debt is canceled, reduced, or remapped according to the target trajectory update result.

[0264] The dimension of curvature deficit is the product of curvature command difference and time; the dimension of yaw control deficit is the product of yaw control demand difference and time; and the dimension of longitudinal drive force deficit is the product of longitudinal drive force difference and time. When generating recovery control commands, the accumulated deficit is converted into the recovery control quantity of the current control cycle based on the deficit recovery time, and the accumulated deficit value is not directly output as curvature command, yaw moment command, or longitudinal drive force command.

[0265] When the current planning control instruction is in the opposite direction to an existing debt, the existing debt is first offset using the difference in the current instruction; after the existing debt is reduced to zero, a new debt in the opposite direction is then established. If the corresponding planning control requirement has been cancelled or replaced by a new target trajectory, the debt is not accumulated.

[0266] When a new high-level anomaly occurs during the accumulated period of the intervention command's outstanding debt, the existing outstanding debt component is frozen, and the source of the new anomaly, the main anomaly chain identifier, and the target trajectory identifier are recorded. After the vehicle stabilizes again, the validity of the original outstanding debt is determined based on the current target trajectory and road boundaries.

[0267] The lateral position deficit projection component is not the total lateral position error that is independently accumulated, but rather the projection result of curvature deficit and yaw control deficit onto the vehicle's motion state.

[0268] Within a preset short-time prediction range, the vehicle motion controller determines the reference vehicle trajectory based on the planned target curvature and the planned target yaw control quantity when not subject to stable control limitations, and determines the restricted vehicle trajectory based on the stable amplitude limiting curvature command and the actual permissible yaw control quantity.

[0269] The difference in position increment between the reference vehicle trajectory and the restricted vehicle trajectory in the normal direction of the current target trajectory is determined as the lateral position deficit projection component. The short-term prediction can be achieved using a vehicle kinematics model, a simplified vehicle dynamics model, or a pre-calibrated correspondence between curvature change and lateral displacement increment.

[0270] The lateral position deficit projection component is used to determine the deficit recovery direction, road boundary margin, temporary recovery trajectory, and deficit validity. It is not directly superimposed with curvature deficit and yaw control deficit to generate steering control commands. Deficit recovery control is primarily generated based on the command-level deficit component, while the lateral position deficit projection component is used to limit the recovery direction and recovery magnitude.

[0271] Each debt component is associated with the time of its occurrence, the corresponding road segment, the direction of the debt, the boundary of adjacent roads, the validity period, the source of the anomaly, the main anomaly chain segment identifier, the suspected anomaly object, the target trajectory identifier, and the target trajectory update time.

[0272] Each control event, response event, and intervention command deficit component is associated with a target trajectory identifier and a target trajectory update time. When the target trajectory identifier changes, the vehicle motion controller determines the degree of change between the old and new target trajectories.

[0273] When the target trajectory update is caused by emergency obstacle avoidance, emergency stop, or a significant change in the target's driving direction, the current limited driving torque verification, succession replacement, or backlog recovery is immediately terminated, the incomplete response chain and incomplete replacement step length are abolished, and the system re-enters normal monitoring, abnormal positioning, or emergency control state according to the new target trajectory.

[0274] When the target trajectory update is only a small, continuous correction, and the target yaw direction, the planned target curvature direction, and the road segment where the vehicle is located remain unchanged, the current stable control state can be retained, but the control events before and after the update are not directly connected into the same response chain. The vehicle motion controller reconstructs the corresponding response chain starting from the first valid control command change event after the trajectory update.

[0275] Changes in curvature, lateral position, and heading directly resulting from target trajectory updates are not counted in the intervention command backlog. Backlogs incurred before the target trajectory update are only mapped to the new target trajectory and retained if the corresponding road segment is still in front of the vehicle, the direction of the backlog is consistent with the direction that the new target trajectory allows for recovery, and restoring the backlog will not increase the current road risk.

[0276] The road segment corresponding to the outstanding debt is represented by the target trajectory arc length coordinates. The vehicle motion controller uses the trajectory arc length coordinates corresponding to the location where the outstanding debt occurs as the starting point of the road segment, and determines the ending point of the road segment based on the effective distance of the outstanding debt, the effective time of the outstanding debt, or the expected distance the vehicle will travel.

[0277] When updating the target trajectory, the vehicle's current position, the start point of the outstanding road segment, and the end point of the outstanding road segment are projected onto the updated target trajectory. If the entire outstanding road segment is located behind the vehicle's current position, the outstanding amount is cancelled.

[0278] The mapping of the outstanding debt direction adopts the tangential-normal coordinate system of the updated target trajectory. The spatial direction corresponding to the original outstanding debt is projected onto the normal direction of the updated target trajectory. When the projection direction is consistent with the recoverable direction of the updated target trajectory, the corresponding projection component is retained; when the projection direction is opposite, the corresponding outstanding debt component is canceled; when the projection component is less than the outstanding debt retention threshold, the corresponding outstanding debt is considered to have been eliminated by the new target trajectory.

[0279] The vehicle motion controller does not immediately restore all outstanding balances after the drive steering handover is completed; instead, it first determines whether an outstanding balance restoration window has been created.

[0280] The debt recovery window requires at least the following conditions: the steering response chain meets the dynamic closure or steady-state closure conditions; no new abnormal chain segments appear in the drive response chain; the abnormality confirmed by the restricted drive torque verification does not reappear; the distance from the vehicle to the road boundary on the side of the debt recovery direction is greater than the recovery distance threshold; the corresponding debt component has not exceeded the validity period; the road segment corresponding to the debt is still in front of the vehicle; and the target trajectory currently output by the trajectory planning module allows recovery along the debt direction.

[0281] During the recovery process, curvature deficit, yaw control deficit, and longitudinal drive force deficit are all gradually reduced according to their corresponding recovery rates, without recovering all deficits at once. When recovering curvature or yaw control deficits, the rate of change of the target steering angle, the rate of change of the target yaw rate, and the lateral velocity of the vehicle approaching the target trajectory are limited. When recovering longitudinal drive force deficits, the rate of change of the target drive torque, the vehicle's longitudinal acceleration, and the slip state of each wheel are limited.

[0282] When the recovery direction is far from the road boundary that is close to the vehicle, the yaw control deficit and curvature deficit can be recovered first. After the vehicle's heading state enters the allowable range, the vehicle is gradually brought closer to the current target trajectory according to the projection component of the lateral position deficit, and finally the longitudinal driving force deficit is recovered.

[0283] When the recovery direction points to the road boundary that is close to the vehicle, instead of directly recovering all curvature deficit and yaw control deficit, a temporary driving trajectory is generated within the passable area of ​​the road, maintaining a preset distance from the current target trajectory. The vehicle is controlled to drive along the temporary driving trajectory, and the preset distance is gradually reduced according to the road boundary margin.

[0284] If restoring the longitudinal drive force deficit would increase the driving torque of a suspected abnormal wheel or axle, the restoration of the longitudinal drive force deficit shall be delayed until the corresponding drive response relationship is closed and the wheel slippage is within the allowable range.

[0285] If the absolute value of the outstanding debt is lower than the corresponding outstanding debt elimination threshold and the outstanding debt elimination confirmation time is maintained continuously, the corresponding outstanding debt will be cleared to zero.

[0286] If, during the debt recovery process, the actual yaw rate deviation increases, the steering dynamic response chain or steady-state response relationship is not closed, the drive response chain becomes abnormal, the road boundary margin decreases to the recovery termination threshold, or the wheel slippage condition worsens, the debt recovery is suspended, the current debt value is frozen, and the system re-enters the abnormal positioning state or the debt waiting state.

[0287] When the same unpaid component fails to recover for a preset number of consecutive times, the unpaid component will be reduced or canceled based on the current road boundary margin and target trajectory requirements. The vehicle is not required to fully recover all historical control differences.

[0288] In one implementation, the vehicle motion controller is configured with the following states: normal monitoring state, abnormal positioning state, verification condition judgment state, limited verification state, unverified conservative control state, drive stability control state, steering action confirmation state, partial replacement state, complete replacement state, replacement rollback state, outstanding waiting state, outstanding recovery state, and degraded operation state.

[0289] Under normal monitoring conditions, the vehicle performs normal driving and steering control according to the target driving information currently output by the trajectory planning module, while continuously building a dynamic response chain and performing steady-state relationship monitoring. Upon detecting an abnormal chain segment or a steady-state abnormality triggering event, the vehicle enters an abnormal positioning state.

[0290] In the anomaly localization state, the first anomaly chain segment of the steering and the first anomaly chain segment of the driving are determined, and the main anomaly chain segment, the first concurrent anomaly chain segment, the anomaly level, and the suspected anomaly object are also determined. After anomaly localization is completed, the verification condition judgment state is entered.

[0291] When the restricted driving torque verification conditions are met, the vehicle enters the restricted verification state; if the verification conditions are not met, an unverified anomaly flag is generated and the vehicle enters the unverified conservative control state. In the restricted verification state, restricted driving torque adjustments are applied to suspected abnormal wheels or axles to reconstruct the response chain and confirm the source of the anomaly.

[0292] After confirming the source of the anomaly, the system enters the drive stability control state. In the drive stability control state, an additional drive torque difference is formed based on the source of the anomaly, the main anomaly chain segment, the vehicle yaw trend, and the set of available actuators, and the system continuously judges whether the steering lateral action is established.

[0293] When the steering side has full yaw control capability, it enters a full takeover state; when the steering side only has partial yaw control capability, it enters a partial takeover state. If the takeover step fails, it enters a takeover rollback state.

[0294] After the takeover is completed, the system enters a waiting state for outstanding debts, continuously monitoring for the formation of an outstanding debt recovery window. Once an outstanding debt recovery window is formed, the system enters the outstanding debt recovery state, performing actions such as recovery, reduction, cancellation, or remapping on the outstanding debts. After all valid outstanding debt components are reduced to the corresponding outstanding debt elimination threshold, the system returns to normal monitoring status.

[0295] When an anomaly cannot be confirmed, actuator capacity is insufficient, road boundary margin is insufficient, critical response chain is unavailable, or successive replacements fail, the vehicle enters a degraded operation state, reducing the target longitudinal speed, limiting the planned target curvature, and performing a safe stop based on the vehicle status.

[0296] In one embodiment where steering actuation capability is limited by load, the vehicle enters a curve with a large longitudinal driving force on a surface with high adhesion. The vehicle motion controller detects steering command change events and steering actuator drive events, but the actual steering angle response is sluggish, and the steering motor control current is close to its output limit. Therefore, the actuator drive-actual steering angle chain segment is identified as the first abnormal chain segment.

[0297] When the vehicle meets the verification conditions, the vehicle motion controller reduces the longitudinal driving force of the steering axle. After the torque reduction, the actual steering angle response recovers, and the actuator drive-actual steering angle chain segment enters the reference allowable range.

[0298] The vehicle motion controller thus identifies the source of the anomaly as a load-limited steering capability, rather than a complete failure of the steering actuator. The vehicle motion controller limits the rate of change of the target steering angle, maintains torque reduction on the steering axle, and generates additional yaw moment through other available drive wheels.

[0299] When the steering side is able to handle some of the yaw control requirements, partial takeover is performed, maintaining an additional drive torque difference no less than the residual drive control threshold. Full takeover then commences once the steering actuator returns to normal and the steering response chain is closed.

[0300] In one implementation of a wet, slippery curve, the vehicle enters a left-turn section along a target trajectory. The trajectory planning module outputs the target left-turn steering angle, the planned target curvature, and the target longitudinal velocity. The vehicle motion controller determines the target yaw rate based on the planned target curvature and the target longitudinal velocity.

[0301] The vehicle motion controller detects that the actual steering angle and steering reaction load are in normal response, but does not detect the expected lateral acceleration change within the corresponding reference response interval, or the initial rate of change of lateral acceleration is lower than the reference allowable range. Therefore, the steering reaction load-lateral acceleration chain segment is identified as the first abnormal steering chain segment.

[0302] The front wheel slippage increased, while the rear wheel drive response chain showed no abnormalities. Therefore, the lateral action of the front axle tires and the front wheel adhesion status were identified as suspected abnormalities.

[0303] When the vehicle meets the verification conditions, the vehicle motion controller reduces the front axle drive torque and performs symmetrical compensation for the rear axle drive torque according to the allowable capacity of the rear axle drive wheels. After verifying the torque reduction, the front wheel slippage decreases, the steering reaction load—lateral acceleration chain segment returns to the reference allowable range, the lateral acceleration changes along the target left turn direction, and the yaw rate deviation begins to decrease.

[0304] The vehicle motion controller thus identifies the source of the anomaly as saturation of the longitudinal and lateral combined force of the front axle tires or a decrease in the road surface adhesion of the front axle, and prohibits the front axle drive wheels from increasing torque, thereby generating an additional yaw torque through the left and right drive wheels of the rear axle in the same direction as the target left turn yaw.

[0305] Once the original abnormal chain segment disappears, the steering response chain closes, and the front wheel slippage condition no longer worsens, the vehicle motion controller gradually reduces the difference in additional driving torque of the rear axle by forming a series of replacement steps and restores the steering control function.

[0306] In one embodiment of left-right separation adhesion, the vehicle travels to a road section where the left and right road surfaces have different adhesion states. Without an increase in actual driving torque, the left drive wheel's wheel angular acceleration changes along the direction of increasing wheel speed, causing an anomaly in the actual driving torque-wheel angular acceleration chain segment of the left drive wheel, while the corresponding drive response chain of the right drive wheel remains normal.

[0307] If the time difference between the changes in angular acceleration of the left and right drive wheels exceeds the wheel response time difference threshold, and the change in wheel angular acceleration corresponding to the unit driving torque change of the left drive wheel is greater than that of the right drive wheel, the vehicle motion controller will identify the left side as the suspected low adhesion side.

[0308] After verifying the reduction of the driving torque of the left drive wheel, the slippage of the left drive wheel decreased, and the actual driving torque-wheel angular acceleration chain segment returned to the baseline allowable range, thus confirming that the source of the anomaly was the separation and attachment of the left and right wheels.

[0309] The vehicle motion controller prohibits torque amplification on the left drive wheel and controls vehicle yaw by reducing torque on the left and limiting torque amplification on the right. If right-side compensation torque amplification results in limited drive performance or abnormal right wheel slippage, the controller stops increasing right-side drive torque and reduces the target longitudinal speed.

[0310] In one embodiment of abnormal steering actuator operation, after the trajectory planning module outputs a steering command, the vehicle motion controller detects a steering command change event node, but the steering actuator drive event does not occur normally, or although the steering actuator drive event occurs, the actual steering angle does not change within the reference response interval.

[0311] The vehicle motion controller identifies the steering command-actuator drive chain segment or the actuator drive-actual steering angle chain segment as the main abnormal chain segment, and identifies the steering control link, steering actuator, and steering transmission mechanism as suspected abnormal objects.

[0312] When the vehicle meets the verification conditions, the vehicle motion controller selects the steering axle currently bearing the largest longitudinal force to perform verification torque reduction. If the abnormal steering response still exists after verification torque reduction, the steering actuator position deviation has not decreased, and at least one of the following is outside the normal range: control current, supply voltage, limit status, communication status, or fault status, then the steering actuator is confirmed to be malfunctioning.

[0313] The vehicle motion controller determines the available target steering angle based on the current available steering angle range of the steering actuator, and determines the portion of the target yaw control requirement that cannot be handled by the steering actuator, and forms an additional driving torque difference based on this portion of the yaw control requirement.

[0314] The steering actuator performs partial takeover only when there is hysteresis and it still has some capability; when the steering actuator completely loses its ability to move, it does not perform takeover to the steering side, but reduces the target speed, limits the planned target curvature, and controls the vehicle to enter a safe area or performs a safe stop.

[0315] In one implementation of a dynamic response transitioning to a steady-state response, the vehicle enters a curve with substantially constant curvature after completing one effective steering control response. The most recent steering dynamic response chain is in a closed state, but the target steering angle remains unchanged thereafter, and no new complete dynamic response chain is generated.

[0316] During the effective holding time of the dynamic chain, the vehicle motion controller combines the most recent dynamic closure result, actual steering angle, steering reaction load, actual driving curvature, and actual yaw rate to determine the vehicle's continued stability.

[0317] After the effective holding time of the dynamic chain is exceeded, the vehicle motion controller determines whether to allow continued recovery based on whether the steady-state response relationship is closed. When the actual driving curvature deviates from the planned target curvature, the deviation between the target yaw rate and the actual yaw rate increases, or the wheel slip condition deteriorates, the steady-state closed state is exited and recovery is suspended.

[0318] In one implementation of an emergency target trajectory update, the vehicle is performing a drive steering transition when the trajectory planning module generates a new target trajectory with a significantly different direction due to the sudden appearance of an obstacle ahead.

[0319] The vehicle motion controller detects a change in the target trajectory identifier and a change in the target yaw direction or planned target curvature direction of the old and new target trajectories. As a result, it immediately terminates the current succession step, freezes the outstanding intervention commands, and redetermines the target yaw control requirements based on the new target trajectory.

[0320] The steering command recovery amount and the reduction amount of the additional driving torque difference in the original replacement step size will no longer be executed, and the control events before the trajectory update will not form the same response chain with the response events under the new trajectory.

[0321] The vehicle motion controller re-checks input information, detects events, and locates anomalies. Delays incurred under the original trajectory are only mapped to the new target trajectory and retained if the corresponding road segment is still in front of the vehicle, the direction of the delay is consistent with the direction allowed for recovery under the new target trajectory, and recovery will not increase the current road risk.

[0322] The event trigger threshold, initial response window, response observation time limit, decoupling time threshold, event association effective time, dynamic chain effective holding time, concurrent anomaly time threshold, anomaly level threshold, verification distance threshold, verification yaw threshold, verification torque reduction, verification period, succession step size, succession observation window, residual drive control threshold, debt effective period, debt recovery rate, and debt elimination threshold in the above embodiments can be determined through vehicle calibration tests, road tests, or vehicle dynamics simulations.

[0323] Each parameter can be set in zones based on vehicle speed, vehicle load, planned target curvature, vehicle drive configuration, tire condition, sensor accuracy, and actuator capability, and can also be corrected online within a preset safety range. Without changing the technical chain of response chain anomaly localization, restricted drive torque verification, drive torque distribution based on anomaly chain segments, complete or partial replacement, and intervention command backlog recovery, the response state quantities, verification objects, and drive torque adjustment methods used can be replaced or combined according to the vehicle hardware configuration.

[0324] Example 3 is an embodiment of the present invention, which provides an intelligent driving steering control system for electric vehicles, including a response chain construction and anomaly localization module, a restricted verification and stability control module, and a restricted verification and stability control module;

[0325] The response chain construction and anomaly localization module is used to synchronously collect vehicle status, construct steering and drive response chains, and determine the first anomaly chain segment and suspected anomaly objects.

[0326] The constrained verification and stability control module is used to adjust the suspected abnormal wheel torque under safety constraints, confirm the source of the abnormality, and generate additional yaw control quantities.

[0327] The succession and overdue recovery module is used to complete the succession from drive control to steering control step by step, and to recover the overdue effective intervention commands based on the current trajectory and road boundaries.

[0328] This embodiment also provides an electronic device applicable to an intelligent driving steering control method for electric vehicles, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the intelligent driving steering control method for electric vehicles as proposed in the above embodiment.

[0329] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements an intelligent driving steering control method for electric vehicles as proposed in the above embodiments.

[0330] The storage medium proposed in this embodiment belongs to the same inventive concept as the method for implementing intelligent driving steering control of electric vehicles proposed in the above embodiments. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0331] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0332] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not 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 driving steering control of electric vehicles, characterized in that: include, Acquire driving information, response information, and constraint information; construct a response chain; compare it with the baseline response chain; and determine the main abnormal chain segment and suspected abnormal objects. When verification is allowed, the restricted driving torque verification is performed on the driving object, and the source of the anomaly is determined based on the change in the response chain before and after verification; otherwise, an unverified anomaly flag is generated and the driving object is restricted. Based on the source of the anomaly or unverified anomaly identifier, yaw trend, and execution capability, allocate drive control quantities; after steering is restored, transfer the drive-side yaw control quantities to the steering side according to the successive replacement steps; After taking over, the system determines the recovery window based on the response relationship closure, the outstanding intervention instructions, the target trajectory, and the road boundaries, and then resumes operation.

2. The electric vehicle intelligent driving steering control method as described in claim 1, characterized in that: The construction of the response chain includes, The validity of the driving information, response information, and constraint information is verified, and the time base is unified. Identify the start time of a valid change in a control command or state variable, form a response event node, and associate the response event node with the corresponding control event; Following the transmission sequence of control commands, execution responses, wheel-side responses, and body responses, adjacent response event nodes are connected into response chain segments, forming steering response chains and drive response chains respectively. When multiple control events overlap within a related time period, a joint response record is established, and a corresponding response chain is formed based on subsequent independent control events or constrained driving torque verification to distinguish the response relationship corresponding to each control event; When no effective control event occurs, perform steady-state relationship monitoring, establish steady-state response records or unverified steady-state anomaly indicators, and determine the credible status of each current response chain based on the data validity status, time synchronization status, and event attribution status.

3. The electric vehicle intelligent driving steering control method as described in claim 2, characterized in that: The process of determining the main abnormal chain segment and suspected abnormal objects includes selecting the corresponding baseline response chain based on the current operating conditions; comparing the response order, response interval, response direction, initial change rate, and response missing state of adjacent response event nodes in the current response chain with those in the baseline response chain, and marking the abnormal response chain segment. Along the direction of transmission of control commands to the vehicle body response, the first abnormal chain segment is determined from each current response chain; based on the order of occurrence, degree of abnormality, and duration of each first abnormal chain segment, the main abnormal chain segment or concurrent abnormal chain segment is determined; during the period of maintaining the main abnormal chain segment, the main abnormal chain segment is updated only when an abnormal response chain segment with a higher degree of abnormality occurs; after the period of maintenance ends, the main abnormal chain segment is updated according to the recovery status of the original main abnormal chain segment and the duration of other abnormal chain segments, and suspected abnormal objects are determined based on the updated main abnormal chain segment.

4. The electric vehicle intelligent driving steering control method as described in claim 3, characterized in that: During the verification process, a restricted driving torque verification is performed on the driving object. The source of the anomaly is determined based on the changes in the response chain before and after verification. Otherwise, an unverified anomaly flag is generated, and the driving object is restricted. Determine whether the verification conditions are met based on the vehicle's yaw rate, wheel side conditions, road boundary conditions, target driving and execution capabilities. If the verification conditions are not met, no verification drive torque adjustment will be performed, the drive control quantity of the associated drive object will be limited, and an unverified anomaly flag will be generated. When the verification conditions are met, the driver object associated with the suspected abnormal object is identified as the verification object. Within the safety constraints, the driver control quantity of the verification object is reduced, and the limited compensation control quantity is allocated to other available driver objects according to the longitudinal operation maintenance requirements. Reconstruct the response event nodes and restructure the relevant current response chain. Determine the source of the anomaly based on the disappearance, persistence, transfer, and addition of the main anomaly chain segment before and after verification. When compensation control triggers a new anomaly, the corresponding compensation control is stopped and the verification-induced anomaly is recorded; if the verification conditions fail during the verification period, the verification is terminated and the drive control quantity is restored to the state before verification or the allowable state within the safety constraints.

5. The electric vehicle intelligent driving steering control method as described in claim 4, characterized in that: The allocation drive control quantity includes Establish a set of available execution objects based on the vehicle's execution configuration and status, and record the incremental capability, decrement capability, and allowable output range of each execution object; Based on the source of the anomaly or unverified anomaly identifier and the vehicle yaw trend, determine the execution objects that restrict the increase of drive control quantity, the execution objects that prioritize the reduction of drive control quantity, and the execution objects that undertake yaw control; The target yaw control requirement is decomposed into the control change amount corresponding to each execution object responsible for yaw control, and the control change amount is limited according to the road boundary state, wheel side state and allowable output range. Continuously monitor the response status of each execution object, update the available capabilities of the corresponding execution object based on new exceptions, and reallocate the remaining control change amount; When the set of available executable objects cannot provide yaw control for the target direction or size, the allowed longitudinal control is reduced and the allowed steering control is limited, and the process enters a restricted operation phase.

6. The electric vehicle intelligent driving steering control method as described in claim 5, characterized in that: The step of transferring the drive-side yaw control to the steering-side according to the assembly replacement includes: Convert the drive-side control contribution and the steering-side control contribution into the same control representation; The amount of control reduction on the drive side is determined based on the current yaw control requirements, and the amount of steering control recovery corresponding to it is determined based on the local response relationship on the steering side. The drive-side control reduction and steering control recovery are executed synchronously, and a local steering response chain is formed based on the current steering control recovery. When the local steering response chain is closed, and the vehicle yaw state, wheel side state, and road boundary state are within the allowable range, the current handover state is recorded as a successful handover state and the next handover step is executed. If any condition is not met, stop increasing the steering-side control contribution and adjust the drive-side control contribution and steering-side control contribution to the previous successful takeover state; retain part of the drive-side control contribution or complete the takeover based on the available steering-side capacity.

7. The electric vehicle intelligent driving steering control method as described in claim 6, characterized in that: The process of determining the debt recovery window and restoring the running status includes, When a normal control command is first restricted by stable control, the difference between the cumulative normal control quantity to be executed and the actual allowable control quantity is recorded, forming an intervention command deficit. This intervention command deficit is then associated with the time of occurrence, the corresponding road section, the recovery direction, and the target trajectory information. After the handover is completed, the window for recovery of the outstanding issues is determined when the steering response relationship meets the requirements of dynamic closure or steady-state closure, there are no new anomalies in the drive response chain, the confirmed anomalies have not recurred, the road boundary status is permissible, the outstanding intervention command is valid, the corresponding road section is in front of the vehicle and the current target trajectory is permissible for recovery. Within the debt recovery window, additional allowable control quantities beyond the current target trajectory control will be used to gradually reduce the debt of intervention instructions; When the target trajectory is continuously corrected, the outstanding intervention instructions can be retained, mapped, reduced, or canceled. If the direction of travel of the target trajectory changes or the recovery conditions fail, the recovery will be terminated and the remaining outstanding intervention instructions will be frozen.

8. An intelligent driving steering control system for electric vehicles, employing the intelligent driving steering control method for electric vehicles as described in any one of claims 1 to 7, characterized in that, include: Response chain construction and anomaly localization module, restricted verification and stability control module; The response chain construction and anomaly localization module is used to synchronously collect vehicle status, construct steering and drive response chains, and identify the first anomaly segment and suspected anomaly objects. The constrained verification and stability control module is used to adjust the suspected abnormal wheel torque under safety constraints, confirm the source of the anomaly, and generate additional yaw control quantities. The succession and backlog recovery module is used to complete the succession from drive control to steering control step by step, and recover the backlog of effective intervention commands based on the current trajectory and road boundaries.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the electric vehicle intelligent driving steering control method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the intelligent driving steering control method for electric vehicles according to any one of claims 1 to 7.

Images