New energy automobile calibration data processing cloud platform and method

Abstract

The invention discloses a new energy vehicle calibration data processing cloud platform and method, and the system comprises a front-end interaction module which constructs a visual operation interface based on a Vue framework, and is used for receiving a calibration data package uploaded by a user, and achieving the 1MB fragmentation cutting and transmission progress dynamic display; the front-end interaction module and the back-end service module form a bidirectional data channel; fragmented data uploaded by the front-end interaction module is transmitted to the back-end service module for verification through an HTTP (Hyper Text Transport Protocol); and the back-end service module adopts a Node.js architecture to process the core service flow, and is used for receiving the front-end fragmented data, checking the integrity through a CRC32 algorithm, calling MATLAB Runtime through a sub-process to execute calibration calculation, routing a generated. Mat file to the Alicloud OSS for storage, and establishing a task scheduling pipeline with the MATLAB calculation engine module. And an intelligent resource scheduling strategy can dynamically allocate computing nodes according to the data scale, so that the service stability in a high-concurrency scene is guaranteed, and resource idle waste is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of automotive data processing, and particularly to a cloud platform and method for processing calibration data of new energy vehicles. Background Art

[0002] An automotive calibration data processing cloud platform is a cloud service system that integrates data collection, storage, processing, analysis, and optimization, and is specifically designed to provide efficient and accurate vehicle performance calibration services for automotive manufacturers and R & D institutions. The platform collects real-time data of vehicles under various working conditions, and uses cloud computing and big data analysis technologies to quickly process and optimize calibration parameters, so as to improve the fuel economy, power performance, and emission performance of vehicles. At the same time, the cloud platform supports remote access, which is convenient for engineers to collaborate across regions, monitor the calibration progress in real time, ensure the accuracy and security of data processing, greatly shorten the automotive R & D cycle, and improve product quality and market competitiveness.

[0003] However, automotive calibration data processing on the market is all based on MATLAB scripts on local PC computers. If local data is lost, it will seriously affect the company's after-sales service and product testing. At the same time, the current calibration data processing of new energy vehicles highly depends on local tools (such as MATLAB scripts), and there are problems such as scattered data storage, strong environmental dependence, and low processing efficiency. Traditional solutions require testers to install a specific version of MATLAB software, and have high requirements for hardware performance, and it takes a significant amount of time for large-scale data processing. In addition, local storage has the risk of data loss and cannot achieve team collaboration and remote real-time verification. Summary of the Invention

[0004] The purpose of the present invention is to provide a cloud platform for processing calibration data of new energy vehicles in order to solve the above-mentioned problems.

[0005] The technical solution adopted by the present invention is as follows: A cloud platform for processing calibration data of new energy vehicles, characterized in that: the cloud platform includes: A front-end interaction module: Construct a visual operation interface based on the Vue framework, which is used to receive the calibration data compression package uploaded by the user, and realize 1MB sharding and dynamic display of the transmission progress. The front-end interaction module and the back-end service module form a two-way data channel: the sharded data uploaded by the front-end interaction module is transmitted to the back-end for verification via the HTTP protocol; A back-end service module: Adopt the Node.js architecture to process the core business flow, which is used to verify the integrity through the CRC32 algorithm after receiving the front-end sharded data, execute the calibration calculation by calling MATLAB Runtime through a child process, route the generated.mat file to Alibaba Cloud OSS for storage, and establish a task scheduling pipeline with the MATLAB calculation engine module.

[0006] MATLAB Computing Engine Module: Build a distributed computing cluster based on MATLAB Production Server, which is used to convert the original.fig file into a.mat structured file containing X / Y / Z coordinate matrices and color mapping data through the changeMat script; Utilize memory mapping technology to achieve high-speed reading and writing of TB-level calibration data, and generate three-dimensional rendering data streams in a WebGL-compatible format.

[0007] This module is deeply integrated with the cloud storage module: After receiving the computing tasks dispatched by Node.js, it reads the original data from OSS and performs the conversion, and directly writes the generated.mat file back to the user's exclusive storage path. At the same time, it forms a rendering closed-loop with the front end: The MATLAB Web App dynamically loads the.mat file from the specified OSS path according to the user ID requested by the front end, and after completing the three-dimensional graph rendering on the browser side, it synchronizes the interaction instructions back to the front-end interface in reverse, realizing the full-link collaboration of "computing-storage-presentation".

[0008] In a preferred embodiment, the cloud platform further includes: a cloud storage module, which builds a multi-layer data warehouse based on Alibaba Cloud OSS, encrypts and stores the user's original data and processing results using AES-256, and reduces the data reading latency to less than 50 ms; The cold storage archiving unit automatically migrates historical data to the low-cost storage layer according to the access frequency, while maintaining multi-region redundant backups. This module provides a secure data channel to the MATLAB Web App through the pre-signed URL mechanism, and synchronizes the file status to the task monitoring interface of the back-end service module in real time.

[0009] In a preferred embodiment, the back-end service module is further provided with a conditional trigger unit for automatically starting the MATLAB computing engine module under pre-designed computing requirement conditions; the pre-designed computing requirement conditions include: (1) Detect that the calibration data compression package uploaded by the user exceeds the processing capacity threshold of the local computing node (≥50 GB); (2) The MATLAB Runtime component is not installed or the version does not match in the user terminal environment; (3) When receiving a batch processing request, automatically activate the distributed computing cluster.

[0010] In a preferred embodiment, the conditional trigger unit of the back-end service module includes at least one of a computing resource evaluation unit, an environment detection unit, and a batch task recognition unit; The condition trigger unit is specifically configured to automatically activate the MATLAB computing engine module when the computing resource evaluation unit detects that the calibrated data volume exceeds the local node processing capacity threshold, or when the environment detection unit determines that the specified version of the MATLAB component is not installed on the user terminal, or when the batch task recognition unit receives a concurrent processing request; The data routing module of the backend service module includes at least one of a path generation unit, a cache synchronization unit, and a permission verification unit; The path generation unit is used to generate an OSS storage path rule containing a timestamp according to the user account identifier; The cache synchronization unit is used to store the.mat file path mapping table generated recently through Redis and establish a real-time data synchronization channel with the front-end MATLAB Web App component; The permission verification unit is used to verify the binding relationship between the token and the account identifier when a.mat file access request is triggered, and activate the Alibaba Cloud WAF protection interception mechanism for illegal access requests; The data routing module of the backend service module is specifically implemented as: (1) Generate an OSS storage path rule according to the user account identifier: / user / {accountID} / mat / {timestamp} / plot3d.mat, where timestamp is accurate to the millisecond level; (2) Cache the.mat file path mapping table generated within the last 24 hours through Redis and synchronize it in real time with the MATLAB Web App component of the front-end interaction module; (3) When a.mat file access request is detected, verify the binding relationship between the request token and the user account identifier, and the illegal request triggers the Alibaba Cloud WAF protection mechanism to intercept.

[0011] In a preferred embodiment, The sharding fault tolerance unit of the front-end interaction module includes at least one of a transmission protocol switching unit, a sharding cache management unit, and a task cleaning execution unit; The sharding fault tolerance unit is specifically configured to control the transmission protocol switching unit to perform a protocol switching operation and re-initiate sharding upload when the sharding transmission fails due to network fluctuations; when the sharding verification fails continuously for a preset number of times threshold, trigger the sharding cache management unit to store the incomplete sharding data and give priority to transmission after the network recovers; The task cleaning execution unit is used to perform an automatic cleaning operation on the sharding tasks that have timed out and push the task termination status information to the user interface through the WebSocket communication unit.

[0012] The sharding fault tolerance unit of the front-end interaction module is specifically implemented as: (1)When the sharded transmission fails due to network fluctuations, automatically switch to the backup transmission protocol (such as the QUIC protocol) to initiate the sharded upload again; (2)When the shard verification fails continuously for 3 times, trigger the local cache mechanism to temporarily store the unfinished shard data, and give priority to transmitting the unfinished shards after the network is restored; (3)Automatically clean up the shard tasks that have not been completed for more than 72 hours, and push the task termination notice to the user interface through WebSocket.

[0013] In a preferred embodiment, the MATLAB computing engine module is further provided with a dynamic scheduling unit for: (1)Dynamically allocate computing resources according to the complexity of the calibration data: When the data volume of a single file < 10GB, call the single-node MATLAB Runtime to execute; When 10GB ≤ data volume < 100GB, enable the local computing cluster for parallel processing; When the data volume ≥ 100GB, trigger the cross-cloud platform (AWS / Aliyun) hybrid computing resource pool; (2)Set a preemptive computing policy for high-priority tasks, pause the GPU resource occupancy of low-priority tasks, and ensure that critical tasks are completed within 15 minutes.

[0014] In a preferred embodiment, the method for processing new energy vehicle calibration data is characterized in that: the processing method includes: S1: Sharded upload of calibration data. The user uploads the new energy vehicle calibration data compression package (including parameters such as torque and speed) through the front-end interaction module. The front-end automatically cuts the data into 1MB shards, uses the Axios library to implement asynchronous sharded transmission, and displays the upload progress bar in real time. If network fluctuations are detected during the transmission process, automatically switch to the QUIC protocol to retransmit the failed shards, and cache the unfinished shards in the local browser for 72 hours to achieve resume from breakpoint.

[0015] S2: Shard verification and data merging. After the back-end service module receives the shard data, it verifies the hash value of each shard through the CRC32 algorithm. The shards with failed verification trigger the front-end retransmission mechanism. After all shards pass the verification, they are automatically merged into a complete compression package and decompressed to a temporary directory. At the same time, the user configuration parameters (such as the number of pole pairs and sampling frequency) are extracted to generate MATLAB execution environment variables.

[0016] S3: MATLAB distributed computing scheduling. Node.js calls the child_process module to start the MATLAB Runtime, dynamically generates an.m script file (such as changeMat.m) carrying environment variables, and distributes tasks to the MATLAB Production Server cluster through an SSH tunnel.

[0017] S4: Three-dimensional data conversion and storage The MATLAB engine performs calibration calculations, converts the original.fig file into a.mat structured file containing the X / Y / Z coordinate matrices and color mapping data C, and encrypts and stores it in Alibaba Cloud OSS according to the rule of "user ID / timestamp / plot3d.mat".

[0018] S5: Three-dimensional visual dynamic rendering. The front-end MATLAB Web App component obtains the.mat file from OSS according to the received file path and loads it into an iframe window, and uses WebGL technology to implement three-dimensional graph rendering on the browser side, supporting window rotation, zooming, and color parameter adjustment within 0.8 seconds.

[0019] S6: Security audit and resource recovery. The platform records the full-process operation logs (including shard MD5 values, computing node IDs, and storage paths), and abnormal operations trigger two-factor authentication interception. The calculation results are automatically migrated to the cold storage layer after 30 days, and at the same time, temporary computing resources are released to ensure that the cluster resource utilization rate is always higher than 85%.

[0020] In a preferred embodiment, in the step S3, the computing engine dynamically allocates resources according to the data scale, including: For data less than 10GB, a single node is called for processing; For 10 - 100GB of data, local GPU cluster parallel computing is enabled; For data exceeding 100GB, a cross-cloud (AWS / Aliyun) hybrid resource pool is triggered, and memory mapping technology is used to accelerate the reading and writing of TB-level data.

[0021] In a preferred embodiment, in the step S4, after storage, the backend pushes the.mat file path and access token to the front-end through WebSocket.

[0022] In a preferred embodiment, in the step S5, the rendering data stream is optimized by WebAssembly, reducing the GPU resource occupancy rate to less than 15%.

[0023] In summary, due to the adoption of the above technical solutions, the beneficial effects of the present invention are: 1. In the present invention, one-stop processing of calibration data is achieved through a cloud architecture. Users only need to upload data and configure parameters through a browser to automatically trigger a distributed computing cluster to complete complex operations. The three-dimensional charts generated by the calculation are rendered online in real time, supporting multi-terminal synchronous viewing and interactive operations, avoiding the cumbersome process of local file transfer. The test team can share data results based on a unified platform, quickly locate abnormal data points, and significantly shorten the cycle from data collection to analysis and decision-making.

[0024] 2. In the present invention, cloud encryption storage and multi-copy disaster recovery mechanisms are adopted to ensure traceability throughout the life cycle of calibration data. The strict isolation design of user accounts and data permissions prevents unauthorized data access. At the same time, intelligent resource scheduling strategies can dynamically allocate computing nodes according to the data scale, ensuring service stability in high-concurrency scenarios and avoiding resource idle waste. Even in the face of sudden large-scale data processing requirements, the system can still maintain efficient response, providing continuous and reliable technology for the new energy vehicle testing link. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is the overall system block diagram of the present invention; Figure 2 is the execution flow chart of the cloud platform in the present invention; Figure 3 is the operation flow chart of the new energy vehicle calibration data processing cloud platform in the present invention; Figure 4 is the three-dimensional diagram of the.mat file generated by the changeMat script in the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0026] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. Embodiment:

[0027] Refer to Figures 1-4 , a new energy vehicle calibration data processing cloud platform, the cloud platform includes: Front-end interaction module: A visualization operation interface is built based on the Vue framework, which is used to receive the calibration data compression package uploaded by the user, and realizes 1MB sharding and dynamic display of the transmission progress. The front-end interaction module and the back-end service module form a two-way data channel: the sharded data uploaded by the front-end interaction module is transmitted to the back-end for verification via the HTTP protocol; Backend service module: Adopts the Node.js architecture to process the core business flow. It is used to receive the sharded data from the front end, verify the integrity through the CRC32 algorithm, call MATLAB Runtime through a child process to perform calibration calculations, route the generated.mat files to Alibaba Cloud OSS for storage, and establish a task scheduling pipeline with the MATLAB computing engine module.

[0028] MATLAB computing engine module: Builds a distributed computing cluster based on MATLAB Production Server. It is used to convert the original.fig files into.mat structured files containing X / Y / Z coordinate matrices and color mapping data through the changeMat script; Utilizes memory mapping technology to achieve high-speed reading and writing of TB-level calibration data, and generates three-dimensional rendering data streams in a WebGL-compatible format.

[0029] This module is deeply integrated with the cloud storage module: After receiving the computing tasks dispatched by Node.js, it reads the original data from OSS and performs the conversion, and directly writes the generated.mat files back to the user's exclusive storage path. At the same time, it forms a rendering closed-loop with the front end: The MATLAB Web App dynamically loads the.mat files from the specified OSS path according to the user ID requested by the front end, and after completing the three-dimensional graph rendering on the browser side, it synchronizes the interaction instructions back to the front-end interface in reverse, realizing the full-link collaboration of "computing - storage - presentation".

[0030] The cloud platform also includes: A cloud storage module that builds a multi-layer data warehouse based on Alibaba Cloud OSS, encrypts and stores the user's original data and processing results using AES-256, reducing the data reading latency to less than 50ms; The cold storage archiving unit automatically migrates historical data to the low-cost storage layer according to the access frequency, while maintaining multi-region redundant backups. This module provides a secure data channel to the MATLAB Web App through the pre-signed URL mechanism, and synchronizes the file status to the task monitoring interface of the backend service module in real time.

[0031] The backend service module also has a conditional trigger unit, which is used to automatically start the MATLAB computing engine module under pre-designed computing requirement conditions; The pre-designed computing requirement conditions include: (1) Detect that the compressed package of the calibration data uploaded by the user exceeds the processing capacity threshold of the local computing node (≥50GB); (2) The MATLAB Runtime component is not installed or the version does not match in the user terminal environment; (3) When receiving a batch processing request, automatically activate the distributed computing cluster.

[0032] The data routing module of the backend service module is specifically implemented as: (1)Generate the OSS storage path rule based on the user account identifier: / user / {accountID} / mat / {timestamp} / plot3d.mat, where the timestamp is accurate to the millisecond level; (2)Cache the mapping table of.mat file paths generated within the last 24 hours through Redis and synchronize it in real time with the MATLAB Web App component of the front-end interaction module; (3)When a.mat file access request is detected, verify the binding relationship between the request token and the user account identifier. An illegal request triggers the Alibaba Cloud WAF protection mechanism to intercept.

[0033] The shard fault tolerance unit of the front-end interaction module is specifically implemented as follows: (1)When the shard transmission fails due to network fluctuations, automatically switch to the backup transmission protocol (such as the QUIC protocol) to initiate the shard upload again; (2)When the shard verification fails continuously three times, trigger the local cache mechanism to temporarily store the incomplete shard data, and give priority to transmitting the incomplete shards after the network resumes; (3)Automatically clean up the shard tasks that have not been completed for more than 72 hours, and push a task termination notice to the user interface through WebSocket.

[0034] The MATLAB computing engine module also has a dynamic scheduling unit for: (1)Dynamically allocate computing resources according to the complexity of the calibration data: When the data volume of a single file < 10GB, call the single-node MATLAB Runtime to execute; When 10GB ≤ data volume < 100GB, enable the local computing cluster for parallel processing; When the data volume ≥ 100GB, trigger the cross-cloud platform (AWS / Aliyun) hybrid computing resource pool; (2)Set a preemptive computing policy for high-priority tasks, pause the GPU resource occupancy of low-priority tasks, and ensure that critical tasks are completed within 15 minutes.

[0035] The processing method includes: S1: Upload the calibration data in shards. The user uploads the compressed package of new energy vehicle calibration data (including parameters such as torque and speed) through the front-end interaction module. The front-end automatically cuts the data into 1MB shards, uses the Axios library to implement asynchronous shard transmission, and displays the upload progress bar in real time. If network fluctuations are detected during the transmission process, automatically switch to the QUIC protocol to retransmit the failed shards, and cache the incomplete shards in the local browser for 72 hours to achieve breakpoint resumption.

[0036] S2: Piecewise Verification and Data Merging. After the backend service module receives the piecewise data, it verifies the hash value of each piece through the CRC32 algorithm. For pieces with failed verification, the front-end retransmission mechanism is triggered. After all pieces pass the verification, they are automatically merged into a complete compressed package and decompressed to a temporary directory. At the same time, user configuration parameters (such as the number of pole pairs and sampling frequency) are extracted to generate MATLAB execution environment variables.

[0037] S3: MATLAB Distributed Computing Scheduling. Node.js calls the child_process module to start MATLAB Runtime, dynamically generates a.m script file (such as changeMat.m) carrying environment variables, and distributes tasks to the MATLAB Production Server cluster through an SSH tunnel.

[0038] S4: 3D Data Conversion and Storage The MATLAB engine executes the calibration calculation, converts the original.fig file into a.mat structured file containing the X / Y / Z coordinate matrices and color mapping data C, and encrypts and stores it in Alibaba Cloud OSS according to the rule of "user ID / timestamp / plot3d.mat".

[0039] S5: 3D Visualization Dynamic Rendering. The front-end MATLAB Web App component obtains the.mat file from OSS according to the received file path and loads it into an iframe window, and uses WebGL technology to implement 3D graph rendering on the browser side, supporting window rotation, zooming, and color parameter adjustment within 0.8 seconds.

[0040] S6: Security Audit and Resource Recycling. The platform records the full-process operation logs (including piecewise MD5 values, computing node IDs, and storage paths), and abnormal operations trigger two-factor authentication interception. The calculation results are automatically migrated to the cold storage layer after 30 days, and at the same time, the temporary computing resources are released to ensure that the cluster resource utilization rate is always higher than 85%.

[0041] In step S3, the computing engine dynamically allocates resources according to the data scale, including: For data less than 10GB, a single node is called for processing; For 10 - 100GB of data, local GPU cluster parallel computing is enabled; For data exceeding 100GB, a cross-cloud (AWS / Aliyun) hybrid resource pool is triggered, and memory mapping technology is used to accelerate the reading and writing of TB-level data.

[0042] In step S4, after storage, the backend pushes the.mat file path and access token to the front-end through WebSocket.

[0043] In step S5, the rendering data stream is optimized by WebAssembly, reducing the GPU resource occupancy rate to below 15%.

[0044] As can be seen from the above: In the present invention, one-stop processing of calibration data is achieved through a cloud architecture. Users only need to upload data and configure parameters through a browser to automatically trigger a distributed computing cluster to complete complex operations. The three-dimensional charts generated by the calculation are rendered online in real time, supporting multi-terminal synchronous viewing and interactive operations, avoiding the cumbersome process of local file transfer. The test team can share data results based on a unified platform, quickly locate abnormal data points, and significantly shorten the cycle from data collection to analysis and decision-making.

[0045] In the present invention, cloud encryption storage and a multi-copy disaster tolerance mechanism are adopted to ensure traceability of calibration data throughout its life cycle. The strict isolation design of user accounts and data permissions prevents unauthorized data access. At the same time, an intelligent resource scheduling strategy can dynamically allocate computing nodes according to the data scale, ensuring service stability in high-concurrency scenarios and avoiding resource idle waste. Even in the face of sudden large-scale data processing requirements, the system can still maintain an efficient response, providing continuous and reliable technology for the new energy vehicle test link.

[0046] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or device. Without further limitation, an element defined by the statement "comprising a..." does not exclude the existence of additional identical elements in the process, method, article, or device comprising the element.

[0047] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A calibration data processing cloud platform for new energy vehicles, characterized in that: The cloud platform includes: Front-end interaction module: A visualization operation interface is built based on the Vue framework, which is used to receive the calibration data compression package uploaded by the user, and realizes 1MB sharding cutting and dynamic display of the transmission progress; A two-way data channel is formed between the front-end interaction module and the back-end service module: the sharded data uploaded by the front-end interaction module is transmitted to the back-end for verification via the HTTP protocol; Back-end service module: The Node.js architecture is adopted to process the core business flow. After receiving the front-end sharded data, it verifies the integrity through the CRC32 algorithm, calls MATLAB Runtime through a child process to perform calibration calculations, and routes the generated.mat files to Alibaba Cloud OSS for storage, and establishes a task scheduling pipeline with the MATLAB calculation engine module; MATLAB calculation engine module: A distributed computing cluster is built based on MATLAB Production Server, which is used to convert the original.fig file into a.mat structured file containing X / Y / Z coordinate matrices and color mapping data through the changeMat script.

2. The new energy vehicle calibration data processing cloud platform according to claim 1, characterized in that: The cloud platform also includes: Cloud storage module: A multi-layer data warehouse is built based on Alibaba Cloud OSS, and AES-256 encryption is used to store the user's original data and processing results; Cold storage archiving unit: Automatically migrates historical data to the low-cost storage layer according to the access frequency, while maintaining multi-region redundant backups, provides a secure data channel to the MATLAB Web App through the pre-signed URL mechanism, and synchronizes the file status to the task monitoring interface of the back-end service module in real time.

3. The new energy vehicle calibration data processing cloud platform according to claim 1, characterized in that: The back-end service module is also provided with a conditional trigger unit, which is used to automatically start the MATLAB calculation engine module under the pre-designed calculation requirement conditions.

4. The new energy vehicle calibration data processing cloud platform according to claim 1, characterized in that: The conditional trigger unit of the back-end service module includes at least one of a computing resource evaluation unit, an environment detection unit, and a batch task recognition unit; The conditional trigger unit is specifically used to automatically activate the MATLAB calculation engine module when the computing resource evaluation unit detects that the calibration data volume exceeds the local node processing capacity threshold, or when the environment detection unit determines that the specified version of the MATLAB component is not installed on the user terminal, or when the batch task recognition unit receives a concurrent processing request; The data routing module of the back-end service module includes at least one of a path generation unit, a cache synchronization unit, and a permission verification unit; The path generation unit is used to generate an OSS storage path rule containing a timestamp according to the user account identifier; The cache synchronization unit is used to store the path mapping table of the.mat files generated recently through Redis, and establish a real-time data synchronization channel with the front-end MATLAB Web App component; The permission verification unit is used to verify the binding relationship between the token and the account identifier when a.mat file access request is triggered, and start the Alibaba Cloud WAF protection interception mechanism for illegal access requests.

5. The new energy vehicle calibration data processing cloud platform according to claim 1, characterized in that: The sharding fault tolerance unit of the front-end interaction module includes at least one of a transmission protocol switching unit, a sharding cache management unit, and a task cleaning execution unit; The sharding fault tolerance unit is specifically used to control the transmission protocol switching unit to perform protocol switching operations and re-initiate sharding upload when sharding transmission fails due to network fluctuations; when the sharding verification fails continuously for a preset number of times threshold, trigger the sharding cache management unit to store the incomplete sharding data and give priority to transmission after the network recovers; The task cleaning execution unit is used to perform automatic cleaning operations on the sharding tasks that have not been completed within the timeout period, and push task termination status information to the user interface through the WebSocket communication unit.

6. The new energy vehicle calibration data processing cloud platform according to claim 1, wherein: The MATLAB computing engine module is also provided with a dynamic scheduling unit.

7. A method for processing calibration data of a new energy vehicle, characterized in that: The processing method includes: S1: Calibration data sharding upload. The user uploads the compressed package of new energy vehicle calibration data through the front-end interaction module. The front-end automatically cuts the data into 1MB shards, uses the Axios library to achieve asynchronous sharding transmission, and displays the upload progress bar in real time; if network fluctuations are detected during the transmission process, automatically switch to the QUIC protocol to retransmit the failed shards, and cache the incomplete shards in the local browser for 72 hours to achieve breakpoint resumption. S2: Sharding verification and data merging. After the back-end service module receives the sharding data, it verifies the hash value of each shard through the CRC32 algorithm. The shards with verification failures trigger the front-end retransmission mechanism; after all shards pass the verification, they are automatically merged into a complete compressed package and decompressed to a temporary directory, and at the same time, the user configuration parameters are extracted to generate MATLAB execution environment variables. S3: MATLAB distributed computing scheduling. Node.js calls the child_process module to start the MATLAB Runtime, dynamically generates a.m script file carrying environment variables, and distributes the tasks to the MATLAB ProductionServer cluster through the SSH tunnel. S4: 3D data conversion and storage. The MATLAB engine performs calibration calculations, converts the original.fig file into a.mat structured file containing the X / Y / Z coordinate matrix and color mapping data C, and encrypts and stores it in Alibaba Cloud OSS according to the rule of "user ID / timestamp / plot3d.mat". S5: 3D visualization dynamic rendering. The front-end MATLAB Web App component obtains the.mat file from OSS according to the received file path and loads it into the iframe window, and uses WebGL technology to implement browser-side 3D map rendering, supporting window rotation, zooming, and color parameter adjustment within 0.8 seconds. S6: Security auditing and resource recovery. The platform records the operation logs of the entire process, and abnormal operations trigger two-factor authentication interception; the calculation results are automatically migrated to the cold storage layer after 30 days, and at the same time, the temporary computing resources are released to ensure that the cluster resource utilization rate is always higher than 85%.

8. The new energy vehicle calibration data processing method according to claim 7, characterized in that: In step S3, the computing engine dynamically allocates resources according to the data scale, including: For data less than 10GB, single-node processing is called; For 10-100GB data, local GPU cluster parallel computing is enabled; Data exceeding 100GB triggers a cross-cloud hybrid resource pool, and memory mapping technology is used to accelerate the reading and writing of TB-level data.

9. The new energy vehicle calibration data processing method according to claim 7, wherein: In step S4, after storage is completed, the backend pushes the.mat file path and access token to the frontend via WebSocket.

10. The new energy vehicle calibration data processing method according to claim 1, wherein: In step S5, the rendering data stream is optimized by WebAssembly, reducing the GPU resource occupancy rate to below 15%.

Images