Method for scheduling and communication delay compensation of multiple controllers of vehicle chassis electronic control system

Abstract

The present application belongs to the technical field of chassis electric control, and belongs to a vehicle chassis electric control system multi-controller task scheduling and communication delay compensation method, which comprises obtaining chassis multi-source data.Analyzing the influence degree of task attribute data on scheduling certainty to obtain scheduling priority, adjusting the scheduling priority to obtain task scheduling priority by analyzing the jitter influence data caused by different scheduling types and bus communication data during switching and transmission.Analyzing the linkage relationship between task scheduling and communication delay compensation to obtain scheduling delay correlation data.Establishing a scheduling compensation coordination model, updating according to the task scheduling priority and scheduling delay correlation data, inputting real-time chassis operating condition data and real-time hardware state data, and outputting to obtain scheduling and delay compensation strategy;The present application can improve the certainty and real-time of task scheduling, enhance the accuracy and stability of communication delay compensation, and improve the operating condition adaptability and robustness of the system.

Description

Technical Field

[0001] This invention belongs to the field of chassis electronic control technology, specifically a method for multi-controller task scheduling and communication delay compensation in vehicle chassis electronic control systems. Background Technology

[0002] With the increasing electrification and intelligence of automobiles, vehicle chassis electronic control systems integrate multiple controllers such as Electronic Stability Control (ESC), Electric Power Steering (EPS), and Brake-by-Wire (BBW), making multi-task parallel execution and high-frequency data interaction commonplace. However, in existing technologies, multi-controller task scheduling and communication delay compensation mostly adopt independent design patterns, resulting in significant technical shortcomings.

[0003] Current task scheduling strategies often prioritize tasks based on a crude approach, failing to dynamically adjust according to the impact of task attributes on scheduling determinism. This leads to resource contention between high-security control tasks and auxiliary diagnostic tasks, causing scheduling conflicts and making it difficult to guarantee the worst-case response time (WCRT) for critical tasks. Furthermore, communication latency compensation is optimized independently only for bus transmission characteristics, disconnected from the scheduling strategy. When scheduling types change or priorities are adjusted, compensation parameters are not synchronously adapted, exacerbating message jitter and timing misalignment issues. In addition, traditional strategies do not consider real-time chassis operating conditions. Fixed scheduling types cannot balance the real-time requirements of extreme scenarios such as emergency braking and high-speed steering with resource utilization in conventional scenarios, resulting in unreasonable system resource allocation and hindering the control accuracy and functional safety level of the chassis electronic control system.

[0004] Therefore, there is an urgent need for a multi-controller task scheduling and communication delay compensation method for vehicle chassis electronic control systems to solve the problems of scheduling conflicts, compensation disconnection and resource waste in the existing technology. Summary of the Invention

[0005] To overcome the shortcomings of existing technologies, this invention proposes a method for multi-controller task scheduling and communication delay compensation in vehicle chassis electronic control systems. This invention primarily addresses the problems of task scheduling conflicts among multiple controllers in vehicle chassis electronic control systems, the disconnect between scheduling and communication delay compensation, and unreasonable allocation of system resources.

[0006] The technical solution adopted by this invention to solve its technical problem is: the method for multi-controller task scheduling and communication delay compensation of a vehicle chassis electronic control system provided by this invention includes: It acquires bus communication data and task attribute data and hardware status data of the chassis electronic control system and multiple controllers, and collects chassis operating condition data of the vehicle in real time.

[0007] The scheduling priority is obtained by analyzing the impact of task attribute data on scheduling determinism. Demand scenarios are divided based on chassis operating condition data, and scheduling types are determined based on the demand scenarios.

[0008] By analyzing the jitter impact data caused by switching and transmitting different scheduling types and bus communication data, the scheduling priority is adjusted to obtain the task scheduling priority.

[0009] The delay scheduling analysis method is used to analyze the linkage between task scheduling and communication delay compensation to obtain scheduling delay correlation data.

[0010] Establish a scheduling compensation and coordination model, update it according to task scheduling priority and scheduling delay correlation data, input real-time chassis operating condition data and real-time hardware status data, and output scheduling and delay compensation strategies.

[0011] The method for multi-controller task scheduling and communication delay compensation in a vehicle chassis electronic control system provided by this invention includes the following steps for obtaining task attribute data: Based on the tasks undertaken by each controller, they are divided into control controllers and auxiliary diagnostic controllers. A task list is made for each controller, and the control tasks are obtained by classifying them according to the characteristics of the tasks.

[0012] Resource consumption characteristics during the execution of control tasks are collected from CPU utilization, memory requirements, and data interaction features.

[0013] By combining preset functional safety standards, task attribute data is obtained by analyzing the safety-related attributes of resource consumption characteristics from the perspectives of functional safety level, fault impact degree, and redundancy strategy.

[0014] The method for multi-controller task scheduling and communication delay compensation in a vehicle chassis electronic control system provided by this invention includes the following steps for obtaining scheduling priority: The scheduling evaluation index is defined from three aspects: worst-case response time, deadline satisfaction rate, and scheduling conflict probability.

[0015] By analyzing the impact of task attribute data on scheduling evaluation indicators, the influence of control tasks and auxiliary tasks can be determined.

[0016] The analytic hierarchy process (AHP) is used to assign weights to the degree of influence of different task attributes, and the weighted sum is used to obtain the scheduling deterministic influence coefficient for each task.

[0017] The scheduling priority is obtained by weighting and sorting each task based on the deterministic influence coefficient of the comprehensive scheduling, the task type, and the execution cycle.

[0018] The method for multi-controller task scheduling and communication delay compensation in a vehicle chassis electronic control system provided by this invention includes the following steps for classifying demand scenarios: From the chassis operating condition data, vehicle motion state characteristics, chassis system load characteristics, and external environmental disturbance characteristics are selected as operating condition characteristics.

[0019] Based on the real-time requirements of chassis control tasks, a quantitative threshold is set for each working condition feature, and the classification criteria for different scenarios are determined to obtain the preset boundary threshold.

[0020] The real-time chassis operating condition data is compared with preset boundary thresholds to divide different scenarios. The required scenarios are obtained by adjusting the data based on the dynamic changes in operating condition characteristics.

[0021] The method for multi-controller task scheduling and communication delay compensation in a vehicle chassis electronic control system provided by this invention includes the following steps for determining the scheduling type: For each defined demand scenario, the characteristics of the scenario are analyzed based on task priority distribution, resource load level, and deterministic requirements to determine the scheduling target.

[0022] Based on the scheduling objective, select the appropriate type from static priority, dynamic priority, time-triggered, and hybrid scheduling.

[0023] For the selected adaptation type, the scheduling parameters are configured in combination with the characteristics of the scenario, and the linkage switching logic between the requirement scenario and the adaptation type is defined to obtain the scheduling type.

[0024] The method for multi-controller task scheduling and communication delay compensation in a vehicle chassis electronic control system provided by this invention includes the following steps for obtaining jitter impact data: When switching scheduling types, task switching jitter, priority preemption jitter, and scheduling cycle jitter are extracted as scheduling jitter indicators. When communicating, message transmission jitter, bus load jitter, and clock synchronization jitter are extracted as communication jitter indicators.

[0025] For each scheduling type, comparative tests were conducted in conjunction with different bus communication load scenarios. The changes in scheduling jitter and communication jitter indicators during the tests were collected to obtain jitter test data.

[0026] By performing hierarchical analysis on jitter test data, the correlation between scheduling type, bus communication and jitter is obtained to obtain jitter impact data.

[0027] The method for multi-controller task scheduling and communication delay compensation in a vehicle chassis electronic control system provided by this invention includes the following steps for adjusting task scheduling priority: The jitter impact data is standardized to obtain the jitter value corresponding to the jitter index. Based on the real-time requirements of the task, a jitter risk threshold is set to divide multiple jitter risk levels.

[0028] Based on the jitter risk level, adjustment coefficients are formulated in combination with the security and resource characteristics of the task, and priority adjustment rules are formulated in combination with jitter risk scenarios.

[0029] For different chassis requirements, the scheduling priority is adjusted according to the priority adjustment rules to generate task scheduling priority.

[0030] The method for multi-controller task scheduling and communication delay compensation in a vehicle chassis electronic control system provided by this invention includes the following steps for obtaining scheduling delay correlation data: Based on the jitter impact data, communication delay compensation data is extracted from the bus communication data and time-aligned with task scheduling related data.

[0031] A two-way linkage impact table is defined based on the positive impact of task scheduling data on communication delay compensation data and the negative impact of communication delay compensation effect on task scheduling strategy.

[0032] Based on the two-way linkage impact table, the Pearson correlation coefficient between the scheduling side index and the compensation index is calculated, and the correlation analysis results are obtained by removing spurious correlations through conditional invariance detection.

[0033] From the correlation analysis results, select the combination of indicators that meet the preset correlation coefficient as strong correlation features, and transform them into engineering rules to obtain scheduling delay correlation data.

[0034] The method for multi-controller task scheduling and communication delay compensation in a vehicle chassis electronic control system provided by this invention includes the following steps for updating the scheduling compensation coordination model: A deep learning model was chosen as the scheduling compensation collaborative model. Task attribute data, bus communication data, and scheduling priority were used as the input dataset, and historical scheduling resolution events and communication delay compensation parameters were used as the label data.

[0035] The input dataset and label data are divided into training and test sets according to a preset ratio. Initial training is performed using the training set, and the model performance is evaluated using the test set.

[0036] Import scheduling delay correlation data, construct an incremental training set by combining it with the training set, use the engineering rules corresponding to the current task scheduling priority as real-time constraints, and use an online learning algorithm to update the parameters of the scheduling compensation collaborative model.

[0037] The method for multi-controller task scheduling and communication delay compensation in a vehicle chassis electronic control system provided by this invention includes the following steps for obtaining the scheduling and delay compensation strategy: The real-time chassis operating condition data and real-time hardware status data are preprocessed to extract features that match the input layer of the updated scheduling compensation coordination model, forming a real-time feature vector.

[0038] Determine the demand scenario in which the real-time feature vector is located, and retrieve the feature processing vector by retrieving the priority of the corresponding task according to the task scheduling priority.

[0039] The feature processing vector is input into the updated scheduling compensation collaborative model. When the preset conditions are met, the inference mechanism is triggered. The model is then weighted through the attention mechanism and combined with the preset target to calculate and output the scheduling optimization parameters and compensation optimization parameters.

[0040] Priorities are adjusted, resources are reallocated, and execution cycles are corrected based on scheduling optimization parameters. Algorithm parameters and network scheduling are configured based on compensation optimization parameters, and clock calibration is performed to obtain scheduling and delay compensation strategies.

[0041] The beneficial effects of this invention are as follows: 1. This invention reduces task execution jitter through jitter impact analysis and dynamic priority adjustment, and optimizes communication latency compensation and task scheduling in a coordinated manner to improve system response speed. By dynamically dividing demand scenarios based on chassis operating conditions, it adapts to different driving conditions, and the scheduling strategy is adjusted in real time according to hardware status, enhancing system fault tolerance. Scheduling priorities are quantified based on task attribute data, and computing resources are rationally allocated. Combined with functional safety levels and redundancy strategies, critical tasks are ensured to be executed first, avoiding resource waste and scheduling conflicts. By analyzing task safety-related attributes in conjunction with functional safety standards, and through jitter risk level classification and priority adjustment, the risk of system failure is reduced, achieving intelligent scheduling decisions. Attached Figure Description

[0042] The invention will now be further described with reference to the accompanying drawings.

[0043] Figure 1 This is one of the flowcharts illustrating the multi-controller task scheduling and communication delay compensation method for a vehicle chassis electronic control system provided in this embodiment of the invention. Figure 2 This is the second flowchart of the method for multi-controller task scheduling and communication delay compensation of vehicle chassis electronic control system provided in this embodiment of the invention.

[0044] Figure 3 This is the third flowchart of the method for multi-controller task scheduling and communication delay compensation of the vehicle chassis electronic control system provided in this embodiment of the invention. Detailed Implementation

[0045] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0046] like Figures 1 to 3 As shown in the embodiment of the present invention, the method for multi-controller task scheduling and communication delay compensation of a vehicle chassis electronic control system includes: It acquires bus communication data and task attribute data and hardware status data of the chassis electronic control system and multiple controllers, and collects chassis operating condition data of the vehicle in real time.

[0047] The steps to obtain task attribute data include: Based on the tasks undertaken by each controller, they are divided into control controllers and auxiliary diagnostic controllers. A task list is made for each controller, and the control tasks are obtained by classifying them according to the characteristics of the tasks.

[0048] Control tasks can be divided into critical control tasks: tasks that directly affect vehicle driving safety and need to meet high real-time requirements, such as emergency braking command execution, dynamic adjustment of steering assist, vehicle attitude stability control, and real-time processing of wheel speed signals. The execution cycle of these tasks is usually ≤10ms, and the tolerance for deadline deviation is <5%.

[0049] Auxiliary diagnostic tasks: Non-real-time tasks that ensure the normal operation of the system, such as sensor fault diagnosis, bus communication health monitoring, controller parameter calibration, historical data storage, etc., with an execution cycle of ≥100ms and a deadline deviation tolerance of ≥20%.

[0050] Resource consumption characteristics during the execution of control tasks are collected from CPU utilization, memory requirements, and data interaction features.

[0051] CPU utilization: Statistically measure the CPU usage time of a single task during a single execution, calculate its proportion of the task cycle time, record the average utilization rate and peak utilization rate, and also record whether interrupt preemption is triggered during task execution.

[0052] Memory requirements: Measure the static memory capacity required for task execution, such as the space occupied by code segments and data segments, and the peak dynamic memory usage, such as the space occupied by stack and buffer, to determine the memory allocation method.

[0053] Data interaction characteristics: Statistically analyze the frequency of message interactions between the task and other controllers, sensors / actuators, record the frame format, data length, and transmission protocol of the interaction data, and clearly define the input / output data interface of the task.

[0054] By combining preset functional safety standards, task attribute data is obtained by analyzing the safety-related attributes of resource consumption characteristics from the perspectives of functional safety level, fault impact degree, and redundancy strategy.

[0055] Functional safety level: Determine the ASIL level corresponding to each task. For example, the ESC emergency braking task is the highest level ASIL-D, while auxiliary diagnostic tasks are mostly ASIL-QM (quality management item).

[0056] Failure impact level: Assess the consequences of task execution failure and classify them into four levels: fatal (affecting driving safety, serious), system degradation (minor), and functional limitation (no impact).

[0057] Redundancy strategy: Record whether the task has a redundancy backup mechanism, and clarify the triggering conditions, switching logic and switching time of the redundant task.

[0058] The scheduling priority is obtained by analyzing the impact of task attribute data on scheduling determinism. Demand scenarios are divided based on chassis operating condition data, and scheduling types are determined based on the demand scenarios.

[0059] The steps to obtain the scheduling priority include: The scheduling evaluation index is defined from three aspects: worst-case response time, deadline satisfaction rate, and scheduling conflict probability.

[0060] Worst Case Response Time (WCRT): The longest execution time of a task under the worst load and highest priority preemption, which is the core indicator of scheduling determinism.

[0061] Deadline fulfillment rate: The probability that the actual execution time of a task is less than or equal to the preset deadline, which directly reflects the reliability of scheduling.

[0062] Scheduling conflict probability: The probability of priority preemption conflict or resource deadlock when multiple tasks compete for CPU / bus resources. The higher the conflict probability, the lower the scheduling determinism.

[0063] By analyzing the impact of task attribute data on scheduling evaluation indicators, the influence of control tasks and auxiliary tasks can be determined.

[0064] For critical control tasks: the higher the functional safety level and the more fatal the impact of a fault, the stricter the requirements for WCRT and deadline compliance. The higher the peak CPU utilization and the more frequent the data interaction, the higher the probability of scheduling conflicts.

[0065] For auxiliary diagnostic tasks: resource consumption is low, fault impact is minimal, and the requirement for scheduling determinism is relatively low, allowing for a certain degree of delay or preemption. Based on these characteristics, a mathematical correlation model can be constructed, such as using linear weighting or machine learning methods, to quantify the mapping relationship between task attributes and scheduling determinism indicators.

[0066] The analytic hierarchy process (AHP) is used to assign weights to the degree of influence of different task attributes, and the weighted sum is used to obtain the scheduling deterministic influence coefficient for each task.

[0067] The steps for assigning weights may include: subjective weighting: combining functional safety requirements, assigning higher weights to functional safety levels and fault impact, such as 40%, because these attributes directly determine the safety priority of the task.

[0068] Objective weights: Based on measured data, calculate the contribution of CPU utilization and data interaction frequency to the probability of scheduling conflicts, such as 30%, and the matching degree between WCRT and deadline accounts for 30%.

[0069] The scheduling priority is obtained by weighting and sorting each task based on the deterministic influence coefficient of the comprehensive scheduling, the task type, and the execution cycle.

[0070] Step 1: Layer tasks by type, with critical control tasks at a higher level than auxiliary diagnostic tasks.

[0071] Step 2: Within the same level, sort by the working condition adaptability influence coefficient from high to low.

[0072] Step 3: For tasks with the same coefficient, sort them from shortest to longest execution cycle. The shorter the cycle, the higher the real-time requirement and the higher the priority.

[0073] The steps for segmenting demand scenarios include: From the chassis operating condition data, vehicle motion state characteristics, chassis system load characteristics, and external environmental disturbance characteristics are selected as operating condition characteristics.

[0074] Vehicle motion characteristics: driving speed, steering angle / angular velocity, brake pedal opening / braking intensity, accelerator pedal opening, and longitudinal / lateral acceleration of the vehicle body.

[0075] Chassis system load characteristics: bus load rate, such as CANFD / Ethernet, controller CPU utilization, critical task trigger frequency, such as ESC intervention frequency, EPS adjustment frequency.

[0076] External environmental interference characteristics: road surface adhesion coefficient, ambient temperature, and electromagnetic interference intensity are estimated by wheel speed difference and slip ratio.

[0077] Based on the real-time requirements of chassis control tasks, a quantitative threshold is set for each working condition feature, and the classification criteria for different scenarios are determined to obtain the preset boundary threshold.

[0078] Threshold comparison table:

[0079] The real-time chassis operating condition data is compared with preset boundary thresholds to divide different scenarios. The required scenarios are obtained by adjusting the data based on the dynamic changes in operating condition characteristics.

[0080] Determine if real-time operating data meets the core characteristic thresholds of a specific scenario, such as braking intensity ≥ 80% → initially classify as an emergency braking scenario. Further confirmation is made by combining other related characteristics, such as emergency braking scenarios requiring braking intensity ≥ 80% + ESC trigger frequency ≥ 10Hz, to avoid misjudgment.

[0081] If the operating data falls within the threshold boundary of two scenarios, such as a speed of 75 km / h and a steering angle of 12°, a weighted scoring method is used to determine the classification. For example, the weight of high-speed steering features is 60%, and the weight of normal driving is 40%. The one with the higher total score is classified as the corresponding scenario.

[0082] Set a lag threshold for scene switching to avoid frequent scene switching caused by small fluctuations in operating conditions. For example, after the emergency braking scene is released, the braking intensity must be less than 30% for 1 second before switching to the normal driving scene.

[0083] The steps to determine the scheduling type include: For each defined demand scenario, the characteristics of the scenario are analyzed based on task priority distribution, resource load level, and deterministic requirements to determine the scheduling target.

[0084] For example, in an emergency braking scenario, the characteristics include dense triggering of high-priority tasks (BBW, ESC), a sudden surge in bus load, and millisecond-level real-time requirements. The scheduling goal is to ensure hard real-time performance, allowing high-priority tasks to preempt all resources at the expense of auxiliary task execution efficiency.

[0085] High-speed steering scenario: Characterized by high-frequency execution of EPS and vehicle stability control tasks, and sensitivity to communication latency. The scheduling objective is low-jitter deterministic scheduling, strict control of task switching overhead, and avoidance of communication and scheduling timing misalignment.

[0086] Typical driving scenarios: Characterized by balanced load between control and auxiliary tasks, with no extreme real-time requirements. The scheduling objective is to optimize resource utilization while balancing the determinism of control tasks and the completion rate of auxiliary tasks.

[0087] Low-speed diagnostic scenarios: Characterized by auxiliary diagnostic tasks dominating and control tasks running infrequently. The scheduling goal is to prioritize non-critical tasks, allow control tasks to execute at a lower frequency, and make full use of idle resources.

[0088] Based on the scheduling objective, select the appropriate type from static priority, dynamic priority, time-triggered, and hybrid scheduling.

[0089]

[0090] For the selected adaptation type, the scheduling parameters are configured in combination with the characteristics of the scenario, and the linkage switching logic between the requirement scenario and the adaptation type is defined to obtain the scheduling type.

[0091] Static preemptive scheduling parameters: Configure priority level, preemption threshold, such as P0 task can preempt all low priority tasks, task core binding policy, such as P0 task is bound to CPU core 0, exclusively occupying hardware resources.

[0092] Time-triggered + static priority hybrid scheduling parameters: plan the time slot length, such as 5ms for EPS task time slot, time slot period, static priority mapping table, and clarify the correspondence between the priority of each task and the time slot.

[0093] Dynamic priority scheduling parameters: Set deadline calculation rules, dynamically adjust based on task cycle and working conditions, and load threshold. When the load is >70%, switch to static priority to avoid scheduling jitter.

[0094] Same-priority round-robin scheduling parameters: Configure time slice length, such as 10ms for diagnostic tasks; round-robin trigger conditions, start when CPU load < 30%; task suspension threshold, suspend unnecessary diagnostic tasks when the load is high.

[0095] The linkage switching logic may include: triggering conditions: cross-determination based on operating condition characteristic thresholds, such as braking intensity ≥80% and ESC trigger frequency ≥10Hz → triggering static preemptive scheduling of emergency braking scenarios.

[0096] Switching lag mechanism: Set the delay time for scene switching, such as 50ms, to avoid frequent switching of scheduling types caused by instantaneous fluctuations in working conditions, which could lead to system jitter.

[0097] Degradation protection strategy: Under extreme operating conditions, force a switch to safety-priority scheduling mode, retain only P0 / P1 level critical tasks, and suspend all auxiliary tasks.

[0098] By analyzing the jitter impact data caused by switching and transmitting different scheduling types and bus communication data, the scheduling priority is adjusted to obtain the task scheduling priority.

[0099] The steps to obtain jitter impact data include: When switching scheduling types, task switching jitter, priority preemption jitter, and scheduling cycle jitter are extracted as scheduling jitter indicators. When communicating, message transmission jitter, bus load jitter, and clock synchronization jitter are extracted as communication jitter indicators.

[0100] For each scheduling type, comparative tests were conducted in conjunction with different bus communication load scenarios. The changes in scheduling jitter and communication jitter indicators during the tests were collected to obtain jitter test data.

[0101] With a fixed scheduling type, run 1000 task cycles continuously under different bus loads, collect task execution timing and message transmission delay data, and calculate the steady-state values ​​of various jitter indicators.

[0102] Simulate demand scenario switching, such as normal driving → emergency braking, triggering a scheduling type switch, from dynamic priority to static preemption. Record jitter data for the first 100 task cycles during the switching process, focusing on collecting the jitter peak at the moment of switching. Focus on core control tasks such as ESC and EPS, and test their jitter characteristics separately under different scheduling types and communication loads to eliminate interference from other tasks.

[0103] By performing hierarchical analysis on jitter test data, the correlation between scheduling type, bus communication and jitter is obtained to obtain jitter impact data.

[0104] The contribution rate of different scheduling types to task switching jitter was statistically analyzed. The impact of different bus loads on message transmission jitter was also investigated; for example, transmission jitter increased by 200% when the load was >70%. For instance, time-triggered scheduling had minimal timing jitter under low load, but time slot conflicts under high load led to a sharp increase in jitter. Static priority scheduling was less sensitive to communication jitter under high load.

[0105] By comparing data, we can distinguish the source of jitter: whether it is the task preemption overhead caused by scheduling type switching, the message delay fluctuation caused by bus communication congestion, or the coupled jitter caused by scheduling-communication timing misalignment.

[0106] The steps to adjust and obtain task scheduling priorities include: The jitter impact data is standardized to obtain the jitter value corresponding to the jitter index. Based on the real-time requirements of the task, a jitter risk threshold is set to divide multiple jitter risk levels.

[0107] Risk Level Table:

[0108] Based on the jitter risk level, adjustment coefficients are formulated in combination with the security and resource characteristics of the task, and priority adjustment rules are formulated in combination with jitter risk scenarios.

[0109] Priority adjustment rules may include: if the task jitter risk is high and the adjustment coefficient is ≥0.8, the priority is directly increased by 1 level.

[0110] If the task jitter risk is medium risk and 0.5 ≤ adjustment coefficient < 0.8, the priority is increased by half a level.

[0111] The priority of low-risk jitter tasks remains unchanged.

[0112] Priority downgrade restriction: Priority downgrade is allowed only when the task is an auxiliary diagnostic task, has a low risk of jitter, and system resources are scarce. Priority downgrade is prohibited for critical control tasks.

[0113] For high-jitter-risk tasks upgraded to P0 level, a dedicated CPU core and bus time slot are bound to avoid competing for resources with other tasks, thereby further reducing jitter.

[0114] For different chassis requirements, the scheduling priority is adjusted according to the priority adjustment rules to generate task scheduling priority.

[0115] For example, in emergency braking scenarios, the BBW / ESC task is prioritized to P0 due to the high risk of vibration. In high-speed steering scenarios, the EPS task is prioritized to P0, and the priority of other tasks is adjusted according to the coefficient.

[0116] When the operating condition characteristic data triggers a scene switch and the change in jitter risk level is greater than 0.3, priority adjustment is immediately triggered.

[0117] Set the lag time for priority adjustment, such as 20ms, to avoid frequent priority switching caused by instantaneous fluctuations in operating conditions and jitter data.

[0118] The delay scheduling analysis method is used to analyze the linkage between task scheduling and communication delay compensation to obtain scheduling delay correlation data.

[0119] The steps to obtain scheduling delay correlation data include: Based on the jitter impact data, communication delay compensation data is extracted from the bus communication data and time-aligned with task scheduling related data.

[0120] Communication delay compensation data may include: message transmission delay, bus load rate, clock synchronization deviation, Mann filter gain, clock calibration coefficient, post-compensation delay deviation, jitter suppression rate, etc. Task scheduling related data may include task scheduling priority, task execution timing data for different scheduling types under different demand scenarios, and CPU / memory resource usage data, etc.

[0121] A two-way linkage impact table is defined based on the positive impact of task scheduling data on communication delay compensation data and the negative impact of communication delay compensation effect on task scheduling strategy.

[0122] Two-way linkage impact table:

[0123] Based on the two-way linkage impact table, the Pearson correlation coefficient between the scheduling side index and the compensation index is calculated, and the correlation analysis results are obtained by removing spurious correlations through conditional invariance detection.

[0124] The steps for calculating the Pearson correlation coefficient between scheduling-side indicators and compensation-side indicators may include: The correlation coefficient between task priority and the compensated delay deviation is negative, indicating that the higher the priority, the smaller the compensated delay deviation.

[0125] The correlation coefficient between the matching degree of scheduling time slots and communication time slots and the jitter suppression rate is positive, indicating that the higher the matching degree, the better the jitter suppression effect. A correlation coefficient with an absolute value ≥ 0.6 is defined as strong correlation, 0.3-0.6 as moderate correlation, and < 0.3 as weak correlation.

[0126] False associations are eliminated through conditional invariance testing. For example, when the bus load rate is kept constant, it is verified whether increasing the task priority does indeed lead to a reduction in the computational overhead of the compensation algorithm. This eliminates the interference of the confounding factor of reduced bus load.

[0127] From the correlation analysis results, select the combination of indicators that meet the preset correlation coefficient as strong correlation features, and transform them into engineering rules to obtain scheduling delay correlation data.

[0128] Feature combinations with a correlation coefficient ≥ 0.6 can be retained, such as: P0 level task + CANFD high priority message + Kalman filter gain 0.8, time-triggered scheduling slot + TSN gated window + clock synchronization deviation ≤ 50ns.

[0129] Engineering rules may include: when the task priority is increased from P1 to P0, the Kalman filter gain of the communication delay compensation algorithm needs to be increased by 0.2 simultaneously to improve the compensation response speed.

[0130] When the bus load rate is greater than 70%, if the jitter suppression rate after compensation is less than 80%, the scheduling cycle of the corresponding task needs to be shortened by 20% to reduce the message exchange frequency.

[0131] When the scheduling-communication time slot matching degree is less than 0.5, a forced switch to static priority scheduling is made to avoid timing misalignment leading to compensation failure.

[0132] Establish a scheduling compensation and coordination model, update it according to task scheduling priority and scheduling delay correlation data, input real-time chassis operating condition data and real-time hardware status data, and output scheduling and delay compensation strategies.

[0133] The steps for updating the scheduling compensation coordination model include: A deep learning model was chosen as the scheduling compensation collaborative model. Task attribute data, bus communication data, and scheduling priority were used as the input dataset, and historical scheduling resolution events and communication delay compensation parameters were used as the label data.

[0134] The scheduling compensation and coordination model includes: Model input layer: Scheduling side features: Task scheduling priority, scheduling type, task execution cycle, CPU / memory resource utilization.

[0135] Communication-side characteristics: bus transmission delay, bus load rate, clock synchronization deviation, and initial compensation parameters.

[0136] Correlation characteristics: correlation impact coefficient and linkage rule label in scheduling delay correlation data, such as priority upgrade → compensation gain upgrade.

[0137] Model hidden layers: adopt a hybrid structure of convolutional layers and fully connected layers.

[0138] Convolutional layers: Extract the temporal coupling features between scheduling priority and communication delay, such as the alignment features between task triggering time and message transmission time.

[0139] Fully Connected Layer: Integrates scheduling side, communication side, and related features to learn the nonlinear mapping relationship between scheduling priority changes, communication delay fluctuations, and compensation parameter adjustments.

[0140] Introducing an attention mechanism: assigning higher weights to high-priority tasks and strongly correlated features to improve the model's sensitivity to key variables.

[0141] Model output layer: Scheduling optimization parameters: task core binding scheme, priority dynamic adjustment threshold, scheduling cycle correction value.

[0142] Compensation optimization parameters: Kalman filter gain, clock synchronization calibration frequency, and TSN time slot planning scheme.

[0143] The input dataset and label data are divided into training and test sets according to a preset ratio. Initial training is performed using the training set, and the model performance is evaluated using the test set.

[0144] The Adam optimizer is selected, with the objective function being to maximize the overall system performance score.

[0145] Iterative training continues until the model converges, with validation set score fluctuations <0.5%.

[0146] Use the test set to evaluate the model's performance. If the core metrics, such as the P0 task deadline satisfaction rate, are not met, adjust the number of hidden layer neurons or the attention weights, and retrain.

[0147] Import scheduling delay correlation data, construct an incremental training set by combining it with the training set, use the engineering rules corresponding to the current task scheduling priority as real-time constraints, and use an online learning algorithm to update the parameters of the scheduling compensation collaborative model.

[0148] Import structured scheduling delay correlation data, including scenario type, task priority, correlation impact coefficient, linkage rules, and compensation effect indicators. Use the priority-correlation coefficient-compensation parameter mapping relationship from the correlation data as new training samples to supplement the original training set.

[0149] Freeze the weights of the model's convolutional layers and update only the weights of the fully connected layers and the attention mechanism to avoid disrupting the learned temporal coupling features. Compare the deviation between the compensated parameters output by the model before and after the update and the actual optimal parameters. If the deviation is >5%, increase the amount of associated data samples and readjust.

[0150] Real-time collection of operating data, triggering scene switching, and obtaining the current task scheduling priority.

[0151] The corresponding priority engineering rules are matched from the scheduling delay correlation data and used as real-time constraints for the model.

[0152] Online learning algorithms, such as the Ftrl algorithm, are employed: based on newly acquired real-time data on priority, communication latency, and compensation effect, the model parameters are incrementally updated, and the learning rate is set to dynamically decay to avoid model oscillation.

[0153] Set update trigger threshold: Online update is triggered only when the task scheduling priority is adjusted by ≥1 level, or when the communication latency jitter exceeds the tolerance threshold, so as to reduce resource consumption.

[0154] The steps to obtain the scheduling and delay compensation strategy include: The real-time chassis operating condition data and real-time hardware status data are preprocessed to extract features that match the input layer of the updated scheduling compensation coordination model, forming a real-time feature vector.

[0155] Hardware status data may include CPU core load rate, memory usage rate, hardware timestamp synchronization deviation, message transmission error rate, actuator response latency, etc.

[0156] Determine the demand scenario in which the real-time feature vector is located, and retrieve the feature processing vector by retrieving the priority of the corresponding task according to the task scheduling priority.

[0157] The real-time operating condition characteristics are compared with the boundary thresholds of the preset demand scenarios to determine the current scenario, such as braking intensity ≥80% → emergency braking scenario.

[0158] Based on the current requirements, retrieve the baseline priority of the corresponding task from the task scheduling priority list.

[0159] Verify whether the real-time hardware status data meets the model safety constraints, for example: CPU core load rate ≤ 80% (if it exceeds this, a degradation strategy will be triggered).

[0160] Clock synchronization deviation ≤ 50ns (if it exceeds this, clock calibration compensation will be triggered).

[0161] If the constraints are not met, a hardware fault preprocessing strategy is first triggered, such as task core migration or redundant controller switching, and then the processed data is input into the model.

[0162] The feature processing vector is input into the updated scheduling compensation collaborative model. When the preset conditions are met, the inference mechanism is triggered. The model is then weighted through the attention mechanism and combined with the preset target to calculate and output the scheduling optimization parameters and compensation optimization parameters.

[0163] Preset conditions may include: reaching a preset inference cycle, such as 10ms / inference, and being synchronized with the scheduling cycle.

[0164] Sudden changes in operating conditions, such as a jump in braking intensity from 10% to 90%.

[0165] Hardware status exceeds warning thresholds, such as CPU load rate rising from 50% to 85%.

[0166] The preset goals are: maximizing the task deadline fulfillment rate, minimizing communication latency jitter, and optimizing resource utilization.

[0167] Scheduling optimization parameters may include: dynamic adjustment values ​​for task priorities, core binding schemes, and execution cycle correction values.

[0168] Compensation optimization parameters may include: Kalman filter gain, TSN slot length, clock synchronization calibration frequency, etc.

[0169] Priorities are adjusted, resources are reallocated, and execution cycles are corrected based on scheduling optimization parameters. Algorithm parameters and network scheduling are configured based on compensation optimization parameters, and clock calibration is performed to obtain scheduling and delay compensation strategies.

[0170] Priority Adjustment: Based on the priority adjustment values ​​output by the model, the baseline priority is fine-tuned to generate a real-time task priority list.

[0171] Resource allocation: Based on the binding scheme, P0 / P1 level critical tasks are bound to idle CPU cores to avoid resource contention.

[0172] Cycle Correction: Adjust the task execution frequency according to the execution cycle correction value. For example, under high load, the cycle of auxiliary diagnostic tasks will be extended from 100ms to 200ms.

[0173] Algorithm parameter configuration: Write the Kalman filter gain output by the model into the communication compensation module to optimize the state prediction accuracy.

[0174] Network scheduling configuration: Update the bus gating scheduling table according to the TSN timeslot length to reserve timeslots for high-priority packets.

[0175] Clock calibration: Based on the synchronization calibration frequency, trigger PTP clock synchronization between controller nodes to compensate for transmission delay.

[0176] The dispatch instruction set is sent to the chassis domain controller via the vehicle Ethernet, and the compensation instruction set is sent to the communication gateway, with the sending delay controlled within ≤1ms.

[0177] In summary, the multi-controller task scheduling and communication delay compensation method for the vehicle chassis electronic control system provided in this embodiment dynamically adjusts priorities by combining task attributes, operating scenarios, and jitter impacts. This ensures that the worst-case response time of critical tasks meets deadline requirements, reduces the probability of scheduling conflicts, and improves the determinism and real-time performance of task scheduling. Optimizing compensation parameters based on a two-way linkage rule between scheduling and compensation effectively suppresses bus transmission jitter and clock synchronization deviations, enhancing the accuracy and stability of communication delay compensation. Through linkage analysis and a collaborative model, scheduling strategy adjustments and compensation parameter optimization are performed synchronously, avoiding timing misalignment issues. This improves the chassis control response speed under extreme conditions such as emergency braking and high-speed steering, achieving deep collaborative optimization of scheduling and compensation. The collaborative model supports online incremental updates and can dynamically optimize strategies based on real-time operating data and hardware status, improving the system's adaptability and robustness. Matching scheduling types to different scenarios improves the execution efficiency of auxiliary diagnostic tasks while ensuring critical tasks are completed, optimizing system resource utilization.

[0178] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence 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 ROM / RAM, magnetic disk, 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 various embodiments or some parts of embodiments.

[0179] Finally, 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 the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. These modifications or substitutions do not cause the essence of the corresponding technical solutions to depart from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for multi-controller task scheduling and communication delay compensation in a vehicle chassis electronic control system, comprising: Acquire bus communication data and task attribute data and hardware status data of the chassis electronic control system and multiple controllers, and collect chassis operating condition data of the vehicle in real time; The scheduling priority is obtained by analyzing the degree of influence of the task attribute data on scheduling determinism. Demand scenarios are divided according to the chassis operating condition data, and scheduling types are determined according to the demand scenarios. By analyzing the jitter impact data caused by different scheduling types and the switching and transmission of bus communication data, the scheduling priority is adjusted to obtain the task scheduling priority; The delay scheduling analysis method is used to analyze the linkage between task scheduling and communication delay compensation to obtain scheduling delay correlation data; A scheduling compensation and coordination model is established, which is updated according to the task scheduling priority and the scheduling delay correlation data. Real-time chassis operating condition data and real-time hardware status data are input, and the scheduling and delay compensation strategy is output.

2. The method for multi-controller task scheduling and communication delay compensation of a vehicle chassis electronic control system according to claim 1, characterized in that: The steps to obtain the task attribute data include: Based on the tasks undertaken by each controller, they are divided into control controllers and auxiliary diagnostic controllers. A task list is made for each controller, and the control tasks are obtained by classifying them according to their characteristics. Resource consumption characteristics during the execution of the control task are collected from CPU utilization, memory requirements, and data interaction characteristics. By combining preset functional safety standards, the task attribute data is obtained by analyzing the safety-related attributes of the resource consumption characteristics from the perspectives of functional safety level, fault impact degree, and redundancy strategy.

3. The method for multi-controller task scheduling and communication delay compensation of a vehicle chassis electronic control system according to claim 1, characterized in that: The steps to obtain the scheduling priority include: The scheduling evaluation index is defined from three aspects: worst-case response time, deadline satisfaction rate, and scheduling conflict probability. The influence of the task attribute data on the scheduling evaluation index is analyzed to determine the degree of influence of the control task and the degree of influence of the auxiliary task. The analytic hierarchy process is used to assign weights to the degree of influence of different task attributes, and the weighted sum is used to obtain the scheduling deterministic influence coefficient for each task. The scheduling priority is obtained by weighting and sorting each task based on the scheduling deterministic impact coefficient, task type, and execution cycle.

4. The method for multi-controller task scheduling and communication delay compensation of a vehicle chassis electronic control system according to claim 1, characterized in that: The steps for classifying the aforementioned demand scenarios include: From the chassis operating condition data, vehicle motion state characteristics, chassis system load characteristics, and external environmental interference characteristics are selected as operating condition characteristics. In combination with the real-time requirements of chassis control tasks, a quantitative threshold is set for each working condition feature, and the classification criteria for different scenarios are determined to obtain the preset boundary threshold. The real-time chassis operating condition data is compared with the preset boundary threshold to divide different scenarios. The required scenarios are then adjusted based on the dynamic changes in the operating condition characteristics.

5. The method for multi-controller task scheduling and communication delay compensation of a vehicle chassis electronic control system according to claim 1, characterized in that: The steps for determining the scheduling type include: For each defined demand scenario, the characteristics of the scenario are analyzed based on task priority distribution, resource load level, and deterministic requirements to determine the scheduling objective; Based on the scheduling objective, an adaptation type is selected from static priority, dynamic priority, time-triggered, and hybrid scheduling. For the selected adaptation type, the scheduling parameters are configured in conjunction with the characteristics of the scenario, and the linkage switching logic between the required scenario and the adaptation type is defined to obtain the scheduling type.

6. The method for multi-controller task scheduling and communication delay compensation of a vehicle chassis electronic control system according to claim 1, characterized in that: The steps for obtaining the jitter impact data include: When switching scheduling types, task switching jitter, priority preemption jitter, and scheduling cycle jitter are extracted as scheduling jitter indicators. When communicating, message transmission jitter, bus load jitter, and clock synchronization jitter are extracted as communication jitter indicators. For each type of scheduling, comparative tests were conducted in conjunction with different bus communication load scenarios. The change amplitude of the scheduling jitter index and the communication jitter index during the test was collected to obtain jitter test data. The jitter impact data is obtained by performing hierarchical analysis on the jitter test data to determine the correlation between scheduling type, bus communication and jitter.

7. The method for multi-controller task scheduling and communication delay compensation of a vehicle chassis electronic control system according to claim 1, characterized in that: The steps for adjusting the task scheduling priority include: The jitter impact data is standardized to obtain the jitter value corresponding to the jitter index. Based on the real-time requirements of the task, a jitter risk threshold is set to divide multiple jitter risk levels. Based on the jitter risk level, adjustment coefficients are formulated in combination with the security and resource characteristics of the task, and priority adjustment rules are formulated in combination with jitter risk scenarios; For different chassis requirements, the scheduling priority is adjusted according to the priority adjustment rules to generate the task scheduling priority.

8. The method for multi-controller task scheduling and communication delay compensation of a vehicle chassis electronic control system according to claim 1, characterized in that: The steps for obtaining the scheduling delay correlation data include: Based on the jitter impact data, communication delay compensation data is extracted from the bus communication data and time-aligned with task scheduling related data. A two-way linkage impact table is defined based on the positive impact of the task scheduling-related data on the communication delay compensation data and the negative impact of the communication delay compensation effect on the task scheduling strategy. Based on the bidirectional linkage impact table, the Pearson correlation coefficient between the scheduling side index and the compensation index is calculated, and the correlation analysis results are obtained by removing false correlations through conditional invariance detection. From the correlation analysis results, select the index combination that meets the preset correlation coefficient as strong correlation features, and transform it into engineering rules to obtain the scheduling delay correlation data.

9. The method for multi-controller task scheduling and communication delay compensation of a vehicle chassis electronic control system according to claim 1, characterized in that: The steps for updating the scheduling compensation coordination model include: A deep learning model is selected as the scheduling compensation collaborative model, and the task attribute data, the bus communication data and the scheduling priority are used as the input dataset, and historical scheduling resolution events and communication delay compensation parameters are used as label data. The input dataset and the label data are divided into a training set and a test set according to a preset ratio. Initial training is performed based on the training set, and the model performance is evaluated using the test set. Import the scheduling delay correlation data, construct an incremental training set by combining it with the training set, use the engineering rules corresponding to the current task scheduling priority as real-time constraints, and update the parameters of the scheduling compensation collaborative model using an online learning algorithm.

10. The method for multi-controller task scheduling and communication delay compensation of a vehicle chassis electronic control system according to claim 1, characterized in that: The steps to obtain the scheduling and delay compensation strategy include: The real-time chassis operating condition data and real-time hardware status data are preprocessed to extract features that match the input layer of the updated scheduling compensation coordination model and form a real-time feature vector. Determine the demand scenario in which the real-time feature vector is located, and retrieve the priority of the corresponding task according to the task scheduling priority to obtain the feature processing vector; The feature processing vector is input into the updated scheduling compensation collaborative model. After the preset conditions are met, the inference mechanism is triggered. Weighted processing is performed through the attention mechanism, and the scheduling optimization parameters and compensation optimization parameters are calculated and output in combination with the preset target. The scheduling optimization parameters are used to adjust priorities, reallocate resources, and correct execution cycles. The compensation optimization parameters are used to configure algorithm parameters and network scheduling. At the same time, clock calibration is performed to obtain the scheduling and delay compensation strategy.

Images