A new energy vehicle cooperative control fault processing method, device and medium

Abstract

The application discloses a new energy vehicle cooperative control fault processing method and device and medium, and relates to the technical field of cooperative control, which comprises the following steps: after receiving battery historical operation data and real-time planning path information, the digital twin cloud compares and analyzes the battery historical operation data with a group battery aging database, generates a health state quantitative value, and judges future driving load according to the real-time planning path information; when the health state quantitative value is in a healthy available interval, the battery attenuation time sequence is generated by fusing the health state quantitative value and the future driving load judgment result; when the health state quantitative value presents an abnormal trend, the battery historical operation data is subjected to multiple verifications and corrections, and the corrected battery historical operation data is recompared and analyzed until the health state quantitative value is stable. Through multi-modal fault-tolerant control, the application realizes dynamic response and safety intervention on the running state of the new energy vehicle.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of cooperative control, and in particular to a cooperative control fault processing method, device and medium for a new energy vehicle. BACKGROUND

[0002] With the rapid development of new energy vehicle technology, the intelligentization and networking degree of vehicle power are continuously improved, and the operation safety and reliability of the battery as the core energy carrier directly affect the vehicle performance. Under complex working conditions, the coupling relationship between the dynamic evolution of the battery state and the vehicle driving demand is increasingly close. The digital twin technology provides a new technical path for realizing the state monitoring and prediction of the vehicle throughout its life cycle by constructing a real-time data interaction mechanism between the physical entity and the virtual model. Under this background, the cooperative control method driven by the fusion of multi-source data gradually becomes an important research direction for improving the operation safety of new energy vehicles, and especially shows significant potential in the linkage optimization of state evaluation, load prediction and control decision.

[0003] The traditional control strategy based on local state feedback has been difficult to meet the requirements of forward-looking identification and response to vehicle-level fault risks. Although the existing technology has attempted to use a cloud platform for battery state monitoring, in the health state quantization process, it often relies on single-time or fragmented historical data, lacks the ability to continuously model long-term degradation trends, and leads to inaccurate state evaluation due to noise interference; at the same time, when facing future driving load changes, it is difficult to dynamically associate the battery degradation time sequence with path planning, limiting the spatiotemporal accuracy of fault prediction. These have limited adaptability in dealing with nonlinear aging processes and sudden working conditions, and there are problems of misjudgment or response lag. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a cooperative control fault processing method for a new energy vehicle to solve the problem of being difficult to meet the requirements of forward-looking identification and response to vehicle-level fault risks.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In a first aspect, the present application provides a cooperative control fault processing method for a new energy vehicle, which comprises collecting battery historical operation data and real-time planning path information of the new energy vehicle, and uploading the battery historical operation data and the real-time planning path information to a digital twin cloud after encryption.

[0008] The digital twin cloud receives the battery historical operation data and real-time planning path information, compares and analyzes the battery historical operation data with the group battery aging database, generates a health state quantitative value, and judges the future driving load according to the real-time planning path information; when the health state quantitative value is in the health available interval, the battery attenuation time sequence is generated by fusing the health state quantitative value and the future driving load judgment result; when the health state quantitative value presents an abnormal trend, the battery historical operation data is subjected to multiple verifications and corrections, and the health state quantitative value is stably obtained based on the corrected battery historical operation data through re-comparison and analysis;

[0009] The new energy vehicle requests the battery attenuation time sequence to the digital twin cloud during driving, and obtains the current state and future path information of the vehicle; the battery attenuation time sequence, the current state and the future path information of the vehicle are compared and analyzed, and it is judged whether there is a high-risk scene in space-time overlap;

[0010] When it is determined that there is a high-risk scene, the new energy vehicle starts the multi-modal fault-tolerant control until the high-risk scene is removed.

[0011] As a preferred scheme of the cooperative control fault handling method of the new energy vehicle, the battery historical operation data and the real-time planning path information are encrypted and uploaded to the digital twin cloud, and the specific steps are as follows,

[0012] The battery historical operation data and the real-time planning path information are integrated to generate a unified format of the fusion data set;

[0013] The fusion data set is encrypted by applying the SM4 encryption algorithm to generate an encrypted data packet;

[0014] The encrypted data packet is transmitted to the digital twin cloud through wireless communication to complete data uploading.

[0015] As a preferred scheme of the cooperative control fault handling method of the new energy vehicle, the health state quantitative value is generated, and the specific steps are as follows,

[0016] The reference battery sample set is retrieved from the group battery aging database;

[0017] The battery historical operation data and the reference battery sample set are aligned in the aging track, and the capacity attenuation deviation of the current battery under the same cumulative use time is calculated;

[0018] The capacity attenuation deviation is converted into a health state quantitative value and output.

[0019] As a preferred scheme of the cooperative control fault handling method of the new energy vehicle, the future driving load is judged according to the real-time planning path information, and the specific steps are as follows,

[0020] The real-time planning path information is input into a path load processor constructed based on a graph neural network, the path load processor encodes the vehicle path length, average slope, road speed limit, traffic flow and environmental temperature of the real-time planning path information as node attributes of a path topology graph;

[0021] The path load processor constructs an adjacency matrix based on the connection relationship between the trip road segments, and takes the node attributes as initial feature vectors, performs multi-layer graph convolution operation, each layer aggregates the feature information of adjacent trip road segments, and updates the load feature representation of each trip road segment layer by layer;

[0022] After completing the last layer of graph convolution, the updated load feature representation of each trip road segment is arranged in the order of the vehicle path to form a spatiotemporal load distribution sequence covering the entire vehicle path;

[0023] The health state quantitative value and the spatiotemporal load distribution sequence are subjected to multi-dimensional risk coupling analysis, and a future driving load judgment result is output.

[0024] As a preferred scheme of the cooperative control fault handling method of the new energy vehicle, wherein: the trip road segment is a continuous road segment divided based on the vehicle path length, average slope, road speed limit, traffic flow and environmental temperature of the real-time planning path information.

[0025] As a preferred scheme of the cooperative control fault handling method of the new energy vehicle, wherein: the abnormal trend is a trend that any one of the change rate, fluctuation amplitude and group aging trajectory deviation degree of the health state quantitative value exceeds a preset dynamic threshold, and the health state quantitative value is outside the health available interval.

[0026] As a preferred scheme of the cooperative control fault handling method of the new energy vehicle, wherein: the rolling comparison of the battery attenuation time sequence with the current state of the vehicle and the future path information is performed to determine whether there is a spatiotemporally overlapping high-risk scenario, specifically as follows,

[0027] A rolling comparison window is set with the current timestamp as the starting point;

[0028] In the rolling comparison window, the battery attenuation prediction value, the driving speed in the current state of the vehicle, the battery temperature in the current state of the vehicle and the road segment position in the future path information are aligned point by point at the corresponding timestamp;

[0029] The new energy vehicle detects the battery attenuation prediction value and the road segment position respectively based on the point-by-point alignment result;

[0030] If any one of the conditions that the battery attenuation prediction value is lower than a preset attenuation threshold and the corresponding road segment position belongs to a high-load road segment exists, it is determined that the new energy vehicle has a spatiotemporally overlapping high-risk scenario.

[0031] If the battery attenuation prediction value is detected to be higher than the preset attenuation threshold and the corresponding road segment position does not belong to a high-load road segment, it is determined that the new energy vehicle does not exist in the spatiotemporal overlapping high-risk scenario.

[0032] The rolling comparison window is slid forward by one time step, and the rolling operation is repeatedly performed until the new energy vehicle reaches the journey endpoint, and if a new path planning update is received, the rolling operation is stopped.

[0033] As a preferred scheme of the cooperative control fault handling method of the new energy vehicle, the multi-modal fault-tolerant control is a fault-tolerant control that dynamically switches between a power retention priority mode, a steering compensation priority mode and a safety braking cooperation mode by continuously monitoring a driving motor power change trend, a steering wheel angle change rate and a vehicle lateral acceleration.

[0034] In a second aspect, the present application provides a computer device comprising a memory and a processor, the memory storing a computer program, wherein the computer program, when executed by the processor, implements any step of the cooperative control fault handling method of the new energy vehicle according to the first aspect of the present application.

[0035] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements any step of the cooperative control fault handling method of the new energy vehicle according to the first aspect of the present application.

[0036] The present application has the following beneficial effects: through the multi-modal fault-tolerant control, dynamic response and safety intervention of the running state of the new energy vehicle are realized. The multi-modal fault-tolerant control mechanism autonomously selects a power retention priority mode, a steering compensation priority mode or a safety braking cooperation mode according to real-time monitoring results of a driving motor power change trend, a steering wheel angle change rate and a vehicle lateral acceleration, so that the vehicle can still maintain stable operation of key functions under the condition of limited battery performance. Seamless switching is performed between different control modes according to actual running parameters, sudden operating conditions in complex driving environments are effectively coped with, and response failure caused by limitations of a single control strategy is avoided. The running robustness of the vehicle under the condition of high battery attenuation is improved, potential instability risks are inhibited, and driving safety and driving continuity in high-risk scenarios are ensured. BRIEF DESCRIPTION OF DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0038] Fig. 1 A flowchart of a cooperative control fault processing method for a new energy vehicle;

[0039] Fig. 2 A flowchart of processing data in a digital twin cloud;

[0040] Fig. 3 A flowchart of rolling comparison to determine a high-risk scenario;

[0041] Fig. 4 A flowchart of multi-modal fault-tolerant control. DETAILED DESCRIPTION

[0042] In order to make the above objectives, features and advantages of the present application more apparent, specific embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0043] In the following description, a large number of specific details are set forth in order to facilitate a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the spirit of the present application, therefore the present application is not limited to the specific embodiments disclosed below.

[0044] Secondly, the "one embodiment" or "embodiment" referred to herein means that a specific feature, structure or characteristic can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an embodiment that is separate or alternative to other embodiments.

[0045] Reference Figs. 1-4 For one embodiment of the present application, the embodiment provides a cooperative control fault processing method for a new energy vehicle, comprising the following steps:

[0046] S1, collect the battery historical operation data and real-time planning path information of the new energy vehicle, and encrypt and upload the battery historical operation data and real-time planning path information to the digital twin cloud.

[0047] Start the new energy vehicle, activate the vehicle data recorder and navigation function;

[0048] Read the battery voltage, current, temperature, charge-discharge cycle number, remaining capacity and use time from the vehicle data recorder of the new energy vehicle, and aggregate to form the battery historical operation data;

[0049] Obtain the vehicle path length, average slope, road speed limit, traffic flow and environmental temperature from the navigation function of the new energy vehicle, and aggregate to form the real-time planning path information;

[0050] Integrate the battery historical operation data and the real-time planning path information to generate a unified format fusion data set;

[0051] Apply the SM4 encryption algorithm to the fusion data set for encryption to generate an encrypted data packet; specifically, the new energy vehicle initializes the SM4 encryption algorithm using a preset 128-bit key, divides the fusion data set into multiple 128-bit grouping blocks, and applies the SM4 encryption algorithm to each grouping block for four rounds of function conversion; each round of function conversion rearranges the bit positions of the grouping block through permutation and substitution; the conversion process is repeated for 32 rounds until all grouping blocks are processed, and the new energy vehicle concatenates all the converted grouping blocks to form a complete ciphertext; the ciphertext is used as the encrypted data packet; wherein the preset 128-bit key is based on the standard setting of the SM4 encryption algorithm, providing high security and efficient processing, and the 128-bit length balances the encryption strength and computational efficiency, making it more suitable for new energy vehicle data encryption than other bit lengths (such as 64-bit security is insufficient, and 256-bit computational overhead is too large);

[0052] The encrypted data packet is transmitted to the digital twin cloud through wireless communication to complete data uploading.

[0053] S2, after receiving the battery historical operation data and the real-time planning path information, the digital twin cloud compares and analyzes the battery historical operation data with the group battery aging database to generate a health state quantitative value, and judges the future driving load according to the real-time planning path information; when the health state quantitative value is in the healthy available interval, the health state quantitative value is fused with the future driving load judgment result to generate a battery degradation time sequence; when the health state quantitative value presents an abnormal trend, the battery historical operation data is subjected to multiple verifications and corrections, and based on the corrected battery historical operation data, the comparison and analysis is re-performed until the health state quantitative value is stable.

[0054] The digital twin cloud receives the encrypted data packet and performs SM4 decryption on the encrypted data packet to obtain the battery historical operation data and the real-time planning path information; specifically, the digital twin cloud initializes the SM4 decryption algorithm using a preset 128-bit key, divides the encrypted data packet into multiple 128-bit grouping blocks, applies the inverse round function of the SM4 decryption algorithm to each grouping block for four rounds of conversion, each round of inverse round function rearranges the bit positions of the grouping block through inverse permutation and inverse substitution, and the digital twin cloud repeats the inverse round function conversion process for 32 rounds until all grouping blocks are processed, and concatenates all the converted grouping blocks to form a complete plaintext; the plaintext is the decrypted battery historical operation data and real-time planning path information; wherein the conversion process adopts 32 rounds because 32 rounds provide sufficient security strength and decryption efficiency, compared with less rounds of security deficiency and more rounds of efficiency reduction, which is suitable for new energy vehicle data processing;

[0055] The process involves retrieving a set of reference batteries from a group battery aging database that are most similar to historical battery data in terms of usage environment, charge / discharge mode, and temperature distribution. This retrieval process in the digital twin cloud involves comparing historical battery data with records in the group battery aging database one by one to extract the set of reference batteries with the highest similarity. This comparison is based on the usage environment (determined by using time and temperature), charge / discharge mode (determined by the number of charge / discharge cycles and current), and temperature distribution characteristics (determined by temperature data).

[0056] Align the battery's historical operating data with a reference battery sample set to determine the aging trajectory, and calculate the capacity decay deviation of the current battery under the same cumulative usage time. The expression is as follows:

[0057] ;

[0058] in, This refers to the capacity decay deviation, which indicates the difference in capacity decay over the same cumulative usage time. Below is the capacity difference between the current battery and the reference battery sample set. This represents the current battery capacity value, indicating historical battery operating data over the same cumulative usage time. The remaining capacity is extracted from the battery's historical operating data and determined after alignment with the aging trajectory. The reference battery capacity value represents the reference battery sample set after the same cumulative usage time. The capacity value is obtained from the reference sample set of the group battery aging database and determined based on the aging trajectory alignment;

[0059] The capacity decay deviation is converted into a health status quantification value through normalization and then output. Normalization means comparing the capacity decay deviation with the maximum capacity value of the reference battery sample set and adjusting it to a health status quantification value range (SOH ratio) of 0 to 1.

[0060] The real-time planned route information is input into the route load processor built on graph neural network. The route load processor encodes the vehicle route length, average gradient, road speed limit, traffic flow and ambient temperature of the real-time planned route information into node attributes of the route topology graph.

[0061] Specifically, the digital twin cloud uses the vehicle path length, average gradient, road speed limit, traffic flow, and ambient temperature from real-time planned path information as path feature vectors. It then maps these parameters to each node in the path topology map, forming a node attribute vector containing five features. The digital twin cloud uses these node attribute vectors as node attributes in the path topology map.

[0062] The path load processor constructs an adjacency matrix based on the connection relationship between travel segments and uses node attributes as the initial feature vector to perform multi-layer graph convolution operations. Each layer aggregates the feature information of adjacent travel segments and updates the load feature representation of each travel segment layer by layer. Among them, the travel segment is a continuous segment divided based on the vehicle path length, average gradient, road speed limit, traffic flow and ambient temperature of real-time planned path information.

[0063] Perform multi-layer graph convolution operation, specifically: use the connection relationship between the travel segments as the edge of the path topology graph, the digital twin cloud creates an adjacency matrix for the path topology graph, the rows and columns of the adjacency matrix correspond to the number of travel segments, and the digital twin cloud sets the position of adjacent travel segments to 1 and the position of non-adjacent travel segments to 0 in the adjacency matrix;

[0064] Using node attributes as initial feature vectors, the digital twin cloud updates the load feature representation of the central travel segment by combining node attribute information from adjacent travel segments in the first layer of graph convolution; in the second layer of graph convolution, it updates the load feature representation of the central travel segment by combining the updated node attribute information from adjacent travel segments; in the third layer of graph convolution, it updates the load feature representation of the central travel segment by combining the further updated node attribute information from adjacent travel segments. The digital twin cloud repeats the process of updating the load feature representation until the three-layer graph convolution is completed, fusing the node attribute information from adjacent travel segments layer by layer to generate the final load feature representation. Among them, the multi-layer graph convolution operation is a hierarchical process of propagating node attribute information through the adjacency matrix and fusing the features of adjacent travel segments to update the load feature representation. It is set to three layers to capture the path dependency relationship from local to global. The advantage is that it balances expressiveness and processing efficiency. Three layers are sufficient to integrate features such as vehicle path length and average slope to form a spatiotemporal load distribution sequence. Compared with two layers, which have insufficient fusion, and four layers, which are prone to information overfitting and processing complexity, three layers are more suitable for new energy vehicle path load processing.

[0065] After completing the last layer of graph convolution, the updated load feature representations of each travel segment are arranged in the order of vehicle path to form a spatiotemporal load distribution sequence covering the entire vehicle path.

[0066] It should be noted that the training process of the path load processor built based on graph neural networks is as follows:

[0067] Historical path data is collected as a training set, including vehicle path length, average gradient, road speed limit, traffic flow, and ambient temperature. The path load processor's parameter weights are initialized using this historical path data. In each iteration, the path load processor updates the parameter weights to match the label load values ​​in the historical path data. The path load processor continuously iterates and updates the parameter weights until the error between the label load value and the predicted load value is minimized (e.g., the difference between the predicted load value and the label load value stabilizes within 0.01). The predicted load value is generated by the path load processor through multi-layer graph convolution operations, encoding the vehicle path length, average gradient, road speed limit, traffic flow, and ambient temperature from the historical path data into node attributes of the path topology graph. The feature information of adjacent travel segments is aggregated layer by layer to generate the final load feature representation.

[0068] A multidimensional risk coupling analysis is performed on the quantitative value of health status and the spatiotemporal load distribution sequence to output the judgment result of future driving load.

[0069] Specifically, the digital twin cloud maps the quantified health status value to each sequence element of the spatiotemporal load distribution sequence; it integrates the power demand element, energy consumption rate element, and environmental load coefficient element of the spatiotemporal load distribution sequence into the quantified health status value, and assesses the risk level corresponding to each of the power demand element, energy consumption rate element, and environmental load coefficient element. The digital twin cloud integrates all risk levels to generate a future driving load judgment result. The assessment process involves comparing the power demand element, energy consumption rate element, and environmental load coefficient element with the quantified health status value to determine whether each element exceeds a preset risk threshold, thus determining the corresponding risk level (i.e., low, medium, and high risk levels). Elements below the lower limit of the risk threshold are considered risk levels. Low risk is defined as the risk threshold, medium risk is defined as the risk threshold between the lower and upper limits, and high risk is defined as the risk threshold above the upper limit. The risk threshold is set based on the statistical distribution of the quantitative value of the health status, with a lower limit of 0.8 and an upper limit of 0.6 (SOH ratio). The values ​​of 0.8 and 0.6 are chosen to balance safety and availability, which has the advantage of reducing false alarms and improving response efficiency. If a higher value is used (such as the lower limit of 0.9), fault tolerance may be triggered more frequently, while if the value is too low (such as the upper limit of 0.4), risks may be missed and accidents may occur. The environmental load coefficient is a normalized coefficient generated from the spatiotemporal load distribution sequence using a graph neural network based on the ambient temperature. The value ranges from 0 to 1, and the value is based on the statistical impact of ambient temperature on battery aging.

[0070] When the quantified health status value is within the healthy and usable range, the quantified health status value is fused with the future driving load judgment result to generate a battery degradation time series including timestamps and vehicle path location markers. The healthy and usable range is set based on the standard range of the quantified health status value, with a value range of 0.8 to 1.0 (SOH ratio). The value of 0.8 to 1.0 is to ensure the safety and reliability of battery performance. The advantage is that it balances battery life and fault prevention. If a narrower range (such as 0.9 to 1.0) is used, the verification will be triggered too early and the efficiency will be reduced. If the range is too wide (such as 0.7 to 1.0), there may be a risk of missed detection.

[0071] When the quantitative value of health status shows an abnormal trend, the battery historical operation data is checked for time continuity, physical rationality and sensor consistency. Abnormal data that violates any of the verification rules in the battery historical operation data is identified and removed. The remaining valid data in the battery historical operation data is then re-integrated to generate a correction dataset.

[0072] It should be noted that the healthy usable range is the acceptable numerical range of the quantified health status value, while an abnormal trend is defined as any of the following exceeding the preset dynamic threshold: the rate of change, the amplitude of fluctuation, or the degree of deviation from the group aging trajectory of the quantified health status value, and the quantified health status value being outside the healthy usable range. The dynamic threshold is set based on historical statistical data from the group battery aging database. For example, the rate of change threshold is 0.01 / hour, the amplitude of fluctuation threshold is 0.05, and the degree of deviation threshold is 0.1 (SOH ratio). The value balances detection sensitivity and false alarm rate. The advantage is that it ensures timely detection of abnormal trends. If a higher dynamic threshold (such as 0.2) is used, there may be a risk of missed detection. If a lower dynamic threshold (such as 0.05) is used, it is easy to trigger verification frequently, reducing efficiency.

[0073] The time continuity verification, physical rationality verification, and sensor consistency verification are as follows:

[0074] Time continuity verification refers to the use of the battery's historical operating data time sequence by checking the digital twin cloud to determine whether the timestamps of battery voltage, current and temperature data points are continuously arranged without missing or skipping.

[0075] Physical rationality verification refers to the digital twin cloud verification of the battery's historical operating data, including battery voltage, current, and remaining capacity, to confirm whether the voltage, current, and remaining capacity values ​​conform to the battery's physical characteristics. Battery physical characteristics refer to the physical behavior patterns of the battery with voltage values ​​between 2.5V and 4.2V, current values ​​within the safe operating range, and remaining capacity values ​​between 0% and 100%.

[0076] Sensor consistency verification refers to the use of digital twin cloud analysis of battery voltage, current and temperature data from historical battery operation data to verify whether the same data values ​​recorded by multiple sensors at the same time stamp are consistent.

[0077] Use the corrected dataset as new battery history running data and re-execute the group battery aging database retrieval operation until the health status quantification value stabilizes.

[0078] S3. When a new energy vehicle is in motion, it requests the battery degradation timeline from the digital twin cloud and obtains the vehicle's current status and future path information. The battery degradation timeline is compared with the vehicle's current status and future path information to determine whether there are high-risk scenarios with spatiotemporal overlap.

[0079] New energy vehicles send requests to the digital twin cloud while in motion. The requests include the vehicle's identity identifier and the current trip identifier.

[0080] The digital twin cloud retrieves the corresponding battery degradation timeline based on the vehicle's identity identifier and the current trip identifier, and then returns the battery degradation timeline to the new energy vehicle.

[0081] New energy vehicles receive battery degradation time data and obtain the vehicle's current status, which includes the current timestamp, current vehicle location, current remaining battery power, current battery temperature, and current driving speed.

[0082] New energy vehicles acquire future route information, which is the complete navigation route planned by the vehicle navigation from the current vehicle location to the destination. It includes multiple travel segments arranged in the order of vehicle travel, and each travel segment corresponds to an estimated passing timestamp and a segment location range.

[0083] The battery degradation time series is parsed according to the timestamps and vehicle path location markers of each time point contained in the battery degradation time series to obtain the battery degradation prediction value and the path location corresponding to each time point. The parsing refers to the new energy vehicle extracting the timestamp sequence, the battery degradation prediction value sequence, and the vehicle path location marker sequence from the battery degradation time series, and pairing each time point of the timestamp sequence with the battery degradation prediction value of the battery degradation prediction value sequence and the vehicle path location marker of the vehicle path location marker sequence to generate the degradation prediction value and path location corresponding to each time point.

[0084] A scrolling comparison window is set starting from the current timestamp. The scrolling comparison window is a time interval of a preset duration that extends into the future from the current timestamp, and is used to cover part of the travel route in the future path information.

[0085] Within the scrolling comparison window, the predicted battery degradation value, the vehicle's current speed, the vehicle's current battery temperature, and the road segment location in the future path information are aligned point by point under the corresponding timestamp. Point-by-point alignment involves matching the predicted battery degradation value, driving speed, battery temperature, and road segment location one by one according to each time point in the timestamp sequence to form a unified data record for the corresponding time point.

[0086] Based on the point-by-point alignment results, new energy vehicles detect the predicted battery degradation value and the road segment location respectively;

[0087] If the detected battery degradation prediction value is lower than the preset degradation threshold and the corresponding road segment is a high-load road segment, then the new energy vehicle is determined to be in a high-risk scenario with spatiotemporal overlap.

[0088] If the predicted battery degradation value is higher than the preset degradation threshold and the corresponding road segment is not a high-load road segment, it is determined that the new energy vehicle does not have a high-risk scenario with spatiotemporal overlap.

[0089] It should be noted that the degradation threshold is set based on the capacity degradation statistics of the group battery aging database. For example, a value of 0.7 (SOH ratio) ensures that new energy vehicles can identify high-risk battery states in a timely manner. The advantage is that it balances detection sensitivity and false positive rate. If a higher value (such as 0.8) is used, it may miss risks, while a lower value (such as 0.6) may frequently trigger fault tolerance and reduce efficiency. High-load road segments are road segments in new energy vehicles where either the power demand or the energy consumption rate exceeds the comprehensive threshold. The comprehensive threshold is a power demand threshold and an energy consumption rate threshold set based on the vehicle path length, average gradient, road speed limit, traffic flow and ambient temperature based on the path load processor's analysis of real-time planned path information. For example, the power demand threshold is set to 80kW and the energy consumption rate threshold is set to 0.2kWh / km. These two values ​​are chosen because they can accurately identify high-load scenarios such as steep slopes or congested road segments. The advantage is that it improves the accuracy of path planning and reduces false positives. If a higher value (such as 90kW) is used, it may miss risks, while a lower value (such as 70kW) may be overly conservative.

[0090] The scroll comparison window is slid forward by one time step (i.e., a fixed interval, such as every second), and the scrolling operation (i.e., all operations from setting the scroll comparison window to determining whether there are high-risk scenarios of spatiotemporal overlap for new energy vehicles) is repeated until the new energy vehicle reaches the end of its journey. If a new path planning update is received, the process stops.

[0091] S4. When a high-risk scenario is determined to exist, the new energy vehicle will activate multimodal fault-tolerant control until the high-risk scenario is resolved.

[0092] Multimodal fault-tolerant control is activated to monitor the real-time trends of drive motor power change, steering wheel angle change rate, and vehicle lateral acceleration of new energy vehicles. These trends are detected by onboard sensors.

[0093] If the trend of the drive motor power change shows that there is a risk of continuous increase, then enter the power hold priority mode, keep the current output power of the drive motor unchanged, and do not respond to new power increase requests;

[0094] If the rate of change of steering wheel angle exceeds a preset safety threshold and lane departure warning is triggered, the vehicle enters steering compensation priority mode, activates the redundant torque output channel of electric power steering, and applies auxiliary steering torque based on the front wheel angle feedback and the deviation of the vehicle's yaw rate to suppress the vehicle from deviating from the expected driving trajectory. The safety threshold is set based on statistical analysis of vehicle dynamic stability, for example, a value of 5 degrees / second. A value of 5 degrees / second ensures timely detection of steering anomalies; a higher value (e.g., 7 degrees / second) avoids the risk of missed detections, while a lower value (e.g., 3 degrees / second) may trigger frequently, reducing driving efficiency. Lane departure warning is triggered when the new energy vehicle detects the lane line position using onboard cameras and lidar, and determines that the deviation of the vehicle's current trajectory from the lane line exceeds a lane departure distance threshold (e.g., 1.5 meters). The lane departure distance threshold is set based on lane width and vehicle dynamic statistics; for example, a value of 1.5 meters effectively identifies significant deviations without interfering with normal driving; a higher value (e.g., 2.0 meters) avoids the risk of missed detections, while a lower value (e.g., 0.8 meters) is prone to frequent triggering and reduces efficiency.

[0095] If the vehicle's lateral acceleration continues to increase and the collision risk is predicted to be high, it enters a safe braking coordination mode. This mode coordinates electric motor braking and hydraulic braking, applying differentiated braking forces to the wheels on both sides to assist the vehicle in returning to a safe driving path and reducing the overall vehicle speed. Collision risk prediction is achieved by the new energy vehicle using onboard radar and cameras to detect the distance and relative speed of obstacles ahead. If the collision time interval is lower than the collision time threshold (e.g., 2 seconds), the collision risk is determined to be high; if the collision time interval is higher than the collision time threshold, the collision risk is determined to be low. In the low-risk state, the current driving state is maintained and continuous monitoring continues. The collision time threshold is set based on vehicle dynamic response and driving safety statistics. For example, a value of 2 seconds ensures that the vehicle can identify high collision risks in a timely manner. Compared to higher values ​​(e.g., 3 seconds), it can avoid missed detections, while lower values ​​(e.g., 1 second) are prone to frequent triggering, reducing efficiency.

[0096] New energy vehicles dynamically switch between power maintenance priority mode, steering compensation priority mode and safety braking coordination mode based on their own operating dynamics to maintain vehicle operation;

[0097] If the test results of the new energy vehicle show that the trend of change in drive motor power does not exceed the normal fluctuation, the rate of change of steering wheel angle does not exceed the safe angle, and the lateral acceleration of the vehicle does not exceed the stable limit, then the current operating status of the new energy vehicle will be maintained, and the changes in vehicle status will be continuously monitored.

[0098] When no longer any of the following conditions are detected in the rolling comparison window: the predicted battery degradation value is lower than the degradation threshold and the corresponding road segment is a high-load road segment, the high-risk scenario is confirmed to be resolved and the multimodal fault-tolerant control is exited.

[0099] This embodiment also provides a computer device applicable to the collaborative control fault handling method for new energy vehicles, including: 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 collaborative control fault handling method for new energy vehicles as proposed in the above embodiment.

[0100] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0101] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the collaborative control fault handling method for new energy vehicles as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0102] In summary, this invention achieves dynamic response and safety intervention for the operating status of new energy vehicles through multimodal fault-tolerant control. The multimodal fault-tolerant control mechanism autonomously selects between power hold priority, steering compensation priority, or safe braking coordination mode based on real-time monitoring results of drive motor power change trends, steering wheel angle change rate, and vehicle lateral acceleration. This ensures stable operation of critical functions even under battery performance limitations. Seamless switching between different control modes based on actual operating parameters effectively addresses sudden conditions in complex driving environments, avoiding response failures due to limitations of a single control strategy. This improves the vehicle's robustness under high battery degradation, suppresses potential instability risks, and ensures driving safety and continuity in high-risk scenarios.

[0103] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for handling collaborative control faults in new energy vehicles, characterized in that: include, Collect historical battery operation data and real-time planned route information of new energy vehicles, and encrypt and upload the historical battery operation data and real-time planned route information to the digital twin cloud; After receiving historical battery operation data and real-time planned route information, the digital twin cloud compares and analyzes the historical battery operation data with the group battery aging database to generate a quantitative value of health status, and judges the future driving load based on the real-time planned route information; when the quantitative value of health status is in the healthy and usable range, it is merged with the judgment result of future driving load to generate a battery degradation time series. When the quantified health status value shows an abnormal trend, multiple verifications and corrections are performed on the battery's historical operating data, and the corrected battery historical operating data is re-compared and analyzed until the quantified health status value stabilizes. New energy vehicles request battery degradation timeline from the digital twin cloud while driving, and obtain information on the vehicle's current status and future path; the battery degradation timeline is compared with the vehicle's current status and future path information in a rolling manner to determine whether there are high-risk scenarios with spatiotemporal overlap. When a high-risk scenario is identified, the new energy vehicle will activate multimodal fault-tolerant control until the high-risk scenario is resolved. The multimodal fault-tolerant control is a fault-tolerant control that dynamically switches between power hold priority mode, steering compensation priority mode and safety braking coordination mode by continuously monitoring the trend of drive motor power change, steering wheel angle change rate and vehicle lateral acceleration. The trends in drive motor power change, steering wheel angle change rate, and vehicle lateral acceleration are detected by onboard sensors. If the trend of the drive motor power change shows that there is a risk of continuous increase, then enter the power hold priority mode, keep the current output power of the drive motor unchanged, and do not respond to new power increase requests; If the rate of change of steering wheel angle exceeds the preset safety threshold and lane departure warning is triggered, the steering compensation priority mode is entered, the electric power steering redundant torque output channel is activated, and auxiliary steering torque is applied according to the front wheel angle feedback and the deviation of the vehicle body yaw rate to suppress the vehicle from deviating from the expected driving trajectory. If the vehicle's lateral acceleration continues to increase and the collision risk is predicted to be high, it will enter the safety braking coordination mode, which coordinates the electric motor braking and hydraulic braking, and assists the vehicle to return to a safe driving path by applying differentiated braking forces to the wheels on both sides, and reduces the overall vehicle speed. New energy vehicles dynamically switch between power maintenance priority mode, steering compensation priority mode and safety braking coordination mode based on their own operating dynamics to maintain vehicle operation; If the test results of the new energy vehicle show that the trend of change in drive motor power does not exceed the normal fluctuation, the rate of change of steering wheel angle does not exceed the safe angle, and the lateral acceleration of the vehicle does not exceed the stable limit, then the current operating status of the new energy vehicle will be maintained, and the changes in vehicle status will be continuously monitored. When no longer any of the following conditions are detected in the rolling comparison window: the predicted battery degradation value is lower than the degradation threshold and the corresponding road segment is a high-load road segment, the high-risk scenario is confirmed to be resolved and the multimodal fault-tolerant control is exited.

2. The method for handling collaborative control faults in new energy vehicles according to claim 1, characterized in that: The process of encrypting and uploading historical battery operating data and real-time planned path information to the digital twin cloud is as follows: By integrating historical battery operation data with real-time route planning information, a unified fusion dataset is generated. The fused dataset is encrypted using the SM4 encryption algorithm to generate encrypted data packets. The encrypted data packet is transmitted wirelessly to the digital twin cloud to complete the data upload.

3. The method for handling collaborative control faults in new energy vehicles according to claim 1, characterized in that: The generation of quantitative health status values ​​is as follows: Retrieve a reference battery sample set from the group battery aging database; Align the battery's historical operating data with the reference battery sample set to determine the aging trajectory and calculate the capacity decay deviation of the current battery under the same cumulative usage time. Convert the capacity decay deviation into a health status quantification value and output it.

4. The method for handling collaborative control faults in new energy vehicles according to claim 1, characterized in that: The determination of future driving load based on real-time planned route information is as follows: The real-time planned route information is input into the route load processor built on graph neural network. The route load processor encodes the vehicle route length, average gradient, road speed limit, traffic flow and ambient temperature of the real-time planned route information into node attributes of the route topology graph. The path load processor constructs an adjacency matrix based on the connection relationship between the travel segments, and uses the node attributes as the initial feature vector to perform multi-layer graph convolution operation. Each layer aggregates the feature information of adjacent travel segments and updates the load feature representation of each travel segment layer by layer. After completing the last layer of graph convolution, the updated load feature representations of each travel segment are arranged in the order of vehicle path to form a spatiotemporal load distribution sequence covering the entire vehicle path. A multidimensional risk coupling analysis is performed on the quantitative value of health status and the spatiotemporal load distribution sequence to output the judgment result of future driving load.

5. The method for handling collaborative control faults in new energy vehicles according to claim 4, characterized in that: The travel segment is a continuous segment divided based on real-time planned route information, including vehicle route length, average gradient, road speed limit, traffic flow, and ambient temperature.

6. The method for handling collaborative control faults in new energy vehicles according to claim 1, characterized in that: The abnormal trend is when any of the rate of change, fluctuation amplitude, or deviation from the aging trajectory of the population in the quantitative value of health status exceeds a preset dynamic threshold, and the quantitative value of health status is outside the healthy usable range.

7. The method for handling collaborative control faults in new energy vehicles according to claim 1, characterized in that: The process of comparing the battery degradation timeline with the vehicle's current state and future path information to determine whether there are high-risk scenarios with spatiotemporal overlap is as follows. Set a scrolling comparison window starting from the current timestamp; Within the scrolling comparison window, the predicted battery degradation value, the vehicle's current speed, the vehicle's current battery temperature, and the road segment location in the future path information are aligned point by point under the corresponding timestamp. Based on the point-by-point alignment results, new energy vehicles detect the predicted battery degradation value and the road segment location respectively; If the detected battery degradation prediction value is lower than the preset degradation threshold and the corresponding road segment is a high-load road segment, then the new energy vehicle is determined to be in a high-risk scenario with spatiotemporal overlap. If the predicted battery degradation value is higher than the preset degradation threshold and the corresponding road segment is not a high-load road segment, it is determined that the new energy vehicle does not have a high-risk scenario with spatiotemporal overlap. The scrolling comparison window is moved forward by one time step, and the scrolling operation is repeated until the new energy vehicle reaches the end of the journey. If a new route planning update is received, the process stops.

8. 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 collaborative control fault handling method for new energy vehicles as described in any one of claims 1 to 7.

9. 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 collaborative control fault handling method for new energy vehicles as described in any one of claims 1 to 7.

Images