Electric vehicle on-board charging system based on data analysis

Abstract

The invention provides an electric vehicle on-board charging system based on data analysis, and relates to the technical field of on-board charging system circuit diagnosis, the charging system comprises a multi-dimensional parameter acquisition module, a self-adaptive analysis module, a distributed response optimization module and a feedback learning module; a multi-dimensional parameter acquisition module is adopted, multi-dimensional data including current, voltage, power, temperature, humidity, charging interface resistance, electromagnetic interference intensity, battery internal voltage distribution, vehicle running state, geographic position, external climate condition, driver behavior pattern and the like are acquired through a distributed sensor network, and fine-grained monitoring of the charging process is realized. Through comprehensive and real-time data acquisition, the system can more accurately understand various working conditions in the charging process, the stability and safety of the charging process under different environmental conditions are ensured, and the reliability of the system is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of on-board charging system circuit diagnosis, and in particular to an on-board charging system for electric vehicles based on data analysis. Background Art

[0002] According to the electric vehicle charging system and method disclosed in Chinese patent publication number "CN103715725B", it includes: a travel plan creation unit, which is configured to create a travel plan based on travel time and travel distance; a charging plan calculation unit, which is configured to calculate the charging plan based on the travel plan; a monitoring unit, which is configured to monitor whether there is a deviation between the actual travel and the travel plan; and an analysis unit, which is configured to: if there is a deviation between the actual travel and the travel plan, analyze whether the deviation affects the nearest charging point in the charging plan; if the deviation affects the nearest charging point in the charging plan, enable the travel plan creation unit to recreate the travel plan, and enable the charging plan calculation unit to recalculate the charging plan based on the recreated travel plan.

[0003] The electric vehicle charger and intelligent charging system disclosed in Chinese Patent Publication No. CN112776626B includes a storage module and a control module. The storage module is used to store at least two charging curves. The control module is in communication with the battery to be charged and retrieves the corresponding charging curve based on the battery's information to charge the battery. This solution uses the control module to identify and determine the battery's information and retrieve the corresponding charging curve to charge the battery.

[0004] The above patent documents and prior art have the following technical problems when used:

[0005] 1. Traditional electric vehicle onboard charging systems can usually only monitor a limited number of parameters and lack the ability to monitor and analyze the complex multi-dimensional parameters involved in the charging process in real time. This results in the system being unable to detect and respond to abnormal situations in a timely manner, thus affecting charging efficiency and battery life.

[0006] 2. In traditional charging systems, the detection of abnormal signals and the prediction of faults are often delayed, resulting in potential problems not being dealt with in a timely manner, causing serious system failures and even safety accidents. Summary of the Invention

[0007] Technical problems solved

[0008] In view of the shortcomings of the existing technology, the present invention provides an electric vehicle on-board charging system based on data analysis, which solves the following problems:

[0009] 1. Traditional electric vehicle onboard charging systems can usually only monitor a limited number of parameters and lack the ability to monitor and analyze the complex multi-dimensional parameters involved in the charging process in real time. This results in the system being unable to detect and respond to abnormal situations in a timely manner, thus affecting charging efficiency and battery life.

[0010] 2. In traditional charging systems, the detection of abnormal signals and the prediction of faults are often delayed, resulting in potential problems not being dealt with in a timely manner, causing serious system failures and even safety accidents.

[0011] Technical Solution

[0012] To achieve the above objectives, the present invention is implemented through the following technical solutions: an electric vehicle onboard charging system based on data analysis, the charging system comprising:

[0013] A multi-dimensional parameter acquisition module, which is used to collect data from the vehicle charging system and its surrounding environment. The data includes current, voltage, power, temperature, humidity, charging interface resistance, electromagnetic interference intensity, internal battery voltage distribution, vehicle operating status, geographic location, external climate conditions, and driver behavior patterns. This enables fine-grained monitoring of the entire charging process through a distributed sensor network.

[0014] An adaptive analysis module, electrically connected to the multidimensional parameter acquisition module. Based on a fusion of an improved multi-scale convolutional neural network and an enhanced Bayesian inference model, the module processes data from the multidimensional parameter acquisition module in real time to identify potential failure modes and abnormal signals. The module also leverages big data analysis features to automatically build and update diagnostic models, accurately predicting and detecting abnormalities caused by electrical faults, hardware aging, and environmental factors in the charging system.

[0015] A distributed response optimization module, which is connected to the adaptive analysis module and uses an adaptive parameter optimization algorithm to adjust the current, voltage, and charging time parameters of the charging system in real time based on detected abnormal signals. It also dynamically optimizes the charging strategy by combining multi-source data to prevent fault propagation and improve charging efficiency and battery life.

[0016] The feedback learning module is electrically connected to the multi-dimensional parameter acquisition module and the adaptive analysis module. The feedback learning module integrates user operation feedback, historical fault records, charging scenario data and long-term operation data of the vehicle system, and inputs the big data feedback and learning module to enhance the self-learning ability of the algorithm, forming a dynamic update mechanism based on big data, and continuously optimizing the diagnostic accuracy and system response speed.

[0017] Preferably, the multi-dimensional parameter acquisition module also includes an environmental dynamic perception unit, which dynamically perceives changes in the external environment based on vehicle-mounted sensors, including temperature, humidity, air quality and vibration, and adjusts the working state of the charging system in real time in combination with environmental data.

[0018] Preferably, the adaptive analysis module further includes an abnormal data isolation and marking unit, which is used to isolate abnormal data and mark abnormal data points to prevent them from having a negative impact on the overall data analysis results, and to correlate and analyze all relevant data when the abnormal event occurs through a backtracking analysis algorithm to improve the accuracy of fault cause locating.

[0019] Preferably, the distributed response optimization module includes a self-healing circuit control unit, which activates a redundant circuit or switches to a backup charging channel based on a detected fault condition, so as to improve the robustness and fault tolerance of the system.

[0020] Preferably, the feedback learning module further includes a scenario adaptive tuning unit, which adjusts and optimizes the diagnostic model of the charging system according to the driving scenario and charging habits.

[0021] Preferably, the multi-dimensional parameter acquisition module includes a vehicle vibration detection unit, which analyzes the vibration frequency and amplitude of the vehicle under different road conditions and determines the potential impact of the vibration on the charging system circuit.

[0022] Preferably, the adaptive analysis module includes a charging process behavior analysis unit, which is used to analyze the driver's charging habits and vehicle trajectory, identify bad charging behaviors and provide personalized suggestions.

[0023] Preferably, the distributed response optimization module includes a remote control interface, which is used for users to monitor and control the operating status of the charging system.

[0024] Preferably, the feedback learning module further includes a data sharing unit, which is used to upload diagnostic data to a cloud database for improving future versions of the charging system and developing a more forward-looking fault diagnosis model.

[0025] Preferably, the multi-dimensional parameter acquisition module further integrates an electromagnetic interference shielding unit. The high-precision electromagnetic interference shielding unit is used to monitor and shield external electromagnetic interference signals in real time to prevent them from affecting the normal operation of the charging system circuit.

[0026] Beneficial effects

[0027] The present invention provides an electric vehicle onboard charging system based on data analysis. It has the following beneficial effects:

[0028] 1. This invention uses a multi-dimensional parameter acquisition module. Through a distributed sensor network, it collects multi-dimensional data including current, voltage, power, temperature, humidity, charging interface resistance, electromagnetic interference intensity, internal battery voltage distribution, vehicle operating status, geographic location, external climate conditions, and driver behavior patterns. This enables fine-grained monitoring of the charging process. Through comprehensive and real-time data acquisition, the system can more accurately understand the various operating conditions during the charging process, ensuring the stability and safety of the charging process under different environmental conditions, thereby improving system reliability.

[0029] 2. The present invention uses an adaptive analysis module, a fusion of an improved multi-scale convolutional neural network and an enhanced Bayesian inference model, to process multidimensional data in real time and identify potential fault modes and abnormal signals. At the same time, the abnormal data isolation and marking unit isolates and marks abnormal data to prevent it from affecting the overall analysis results, and improves the accuracy of fault location through retrospective analysis. The system can detect abnormal signals in real time and predict occurring faults, thereby achieving preventive maintenance, reducing unplanned downtime of the system, and improving the safety and stability of the charging system.

[0030] 3. The present invention adopts a distributed intelligent response and optimization module to adjust the current, voltage and charging time parameters of the charging system in real time based on the detected abnormal signals through an adaptive parameter optimization algorithm, and dynamically optimizes the charging strategy in combination with multi-source data. In addition, the self-healing circuit control unit can activate redundant circuits or switch to backup charging channels when a fault is detected to improve the robustness and fault tolerance of the system. The system can intelligently adjust the charging strategy according to the real-time operating conditions, optimize the charging process, improve charging efficiency, and extend battery life. It can also take quick measures when a fault occurs to prevent the fault from spreading and ensure the continuity and stability of the charging process. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a unit composition diagram of the present invention;

[0032] Figure 2 It is an operation flow chart of the present invention. DETAILED DESCRIPTION

[0033] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention. Specific embodiment one:

[0035] As shown in the figure, the electric vehicle on-board charging system based on data analysis includes: a multi-dimensional parameter acquisition module, which is used to collect data from the on-board charging system and its surrounding environment. The data includes current, voltage, power, temperature, humidity, charging interface resistance, electromagnetic interference intensity, battery internal voltage distribution, vehicle operating status, geographic location, external climate conditions, and driver behavior patterns. Fine-grained monitoring of the entire charging process is achieved through a distributed sensor network. The multi-dimensional parameter acquisition module also includes an environmental dynamic perception unit. The environmental dynamic perception unit dynamically perceives changes in the external environment based on on-board sensors, including temperature, humidity, air quality, and vibration, and adjusts the working status of the charging system in real time based on environmental data. The multi-dimensional parameter acquisition module further integrates an electromagnetic interference shielding unit. The high-precision electromagnetic interference shielding unit is used to monitor and shield external electromagnetic interference signals in real time to prevent them from affecting the normal operation of the charging system circuit; the hardware composition of the multi-dimensional parameter acquisition module is mainly sensors, including: current sensor: used to monitor the charging current in real time, voltage sensor: used to monitor the charging voltage in real time, power sensor: used to calculate the charging power, temperature sensor: monitors the temperature of the charging system and battery, humidity sensor: monitors the ambient humidity, resistance sensor: measures the resistance of the charging interface, electromagnetic interference sensor: monitors the intensity of electromagnetic interference, battery internal voltage distribution sensor: monitors the voltage of each unit inside the battery, vibration sensor: monitors vehicle vibration, position sensor: obtains the vehicle's geographic location, environmental sensor: monitors external climate conditions, behavior monitoring equipment: records the driver's behavior patterns.

[0036] The process is as follows:

[0037] Data acquisition: Sensors send data to a central data processing unit via wireless networks or wired interfaces.

[0038] Data preprocessing: After data collection, it is preprocessed through steps such as denoising, normalization, and feature extraction to ensure the accuracy and consistency of the data.

[0039] Data storage: Preprocessed data is stored in a local data cache, ready for further analysis;

[0040] Real-time monitoring: The multi-dimensional parameter acquisition module enables real-time monitoring of the charging system and its surrounding environment, ensuring timely acquisition of various operating status data.

[0041] Data integration: All collected data are integrated through the sensor network to provide a comprehensive information basis for subsequent analysis.

[0042] The multi-dimensional parameter acquisition module includes a vehicle vibration detection unit, which analyzes the vehicle's vibration frequency and amplitude under different road conditions to determine the potential impact of vibration on the charging system circuit;

[0043] Adaptive analysis modules include:

[0044] Processor: A high-performance multi-core processor or GPU for running complex data analysis algorithms.

[0045] Storage devices: Large-capacity storage devices (such as SSDs) are used to store analytical models and intermediate data results.

[0046] Data interface: interface for data transmission with the multi-dimensional parameter acquisition module (including but not limited to USB, Ethernet, I / O interface, etc.).

[0047] The process is as follows:

[0048] Data reception: Receive pre-processed data from the multi-dimensional parameter acquisition module.

[0049] Multi-scale convolutional neural network extracts multi-scale features from data and identifies fault modes.

[0050] The enhanced Bayesian reasoning model combines fault modes and environmental factors to identify abnormal signals and predict faults.

[0051] Model update: Based on the results of big data analysis, the diagnostic model is automatically built and updated to adapt to different working environments and failure modes.

[0052] Fault identification: Through fusion algorithm analysis, potential fault modes in the system can be identified in real time.

[0053] Improved model accuracy: Continuously updated diagnostic models improve the detection accuracy of electrical faults, hardware aging, and environmental anomalies.

[0054] The adaptive analysis module is electrically connected to the multi-dimensional parameter acquisition module, and the adaptive analysis module is based on a fusion of improved multi-scale convolutional neural networks and enhanced Bayesian inference models. The multi-scale convolutional neural network extracts features from input data at different scales to capture different levels of information in the data. The multi-scale convolutional neural network extracts features from the charging system data through multi-layer convolution and pooling operations, and uses different receptive fields on different convolution layers to cope with the multi-scale information in the data. The multi-scale convolutional neural network is used to process data from the multi-dimensional parameter acquisition module, including current, voltage, power, temperature, humidity, charging interface resistance, electromagnetic interference intensity, battery internal voltage distribution, vehicle operating status, geographic location, external climate conditions, and driver behavior patterns. These data show different change patterns at different time scales. Through the design of the multi-scale convolution kernel, the multi-scale convolutional neural network can capture short-term rapid fluctuations and long-term trend changes, thereby improving the accuracy of fault pattern recognition. The algorithm formula is as follows:

[0055]

[0056] Indicates the The output of the layer, is the convolution kernel weight, represents the convolution operation, is the bias, is the activation function;

[0057] The Bayesian inference model infers posterior probabilities by utilizing prior probabilities and likelihood functions. In this solution, the enhanced Bayesian inference model introduces a dynamic prior update mechanism, combined with a real-time learning process driven by big data, enabling the model to continuously update its prior knowledge to adapt to emerging fault modes in the charging system and infer the most likely fault type from the detected feature data. The model combines multi-source data and dynamically adjusts its prior probabilities through historical fault records, thereby improving diagnostic accuracy. The algorithm formula is as follows:

[0058]

[0059] is the posterior probability of hypothesis H given data D, is the likelihood function of observing data D when hypothesis H holds, is the prior probability of hypothesis H, is the marginal probability of observing data D.

[0060] By processing data from the multi-dimensional parameter acquisition module in real time, potential failure modes and abnormal signals are identified. The adaptive analysis module leverages big data analysis capabilities to automatically build and update diagnostic models, accurately predicting and detecting abnormalities caused by electrical faults, hardware aging, and environmental factors in the charging system. The adaptive analysis module further includes an abnormal data isolation and marking unit, which isolates and marks abnormal data points to prevent them from negatively impacting the overall data analysis results. A retrospective analysis algorithm is used to correlate and analyze all relevant data at the time of the abnormal event, improving the accuracy of fault location. The multi-dimensional parameter acquisition module collects multi-dimensional data from the charging system and its surrounding environment, including current, voltage, power, temperature, humidity, and more. All data is collected and transmitted in real time via a distributed sensor network.

[0061] Data preprocessing: Before entering the retrospective analysis algorithm, the data undergoes denoising, normalization, and feature extraction to ensure data quality. The input data for retrospective analysis includes key parameters such as current waveform, voltage curve, and charging interface resistance.

[0062] The adaptive analysis module uses a fusion of an improved multi-scale convolutional neural network (CNN) and an enhanced Bayesian inference model to process data in real time and identify potential failure modes and abnormal signals. The convolutional neural network extracts features from the data, while the Bayesian inference model performs probabilistic analysis.

[0063] Abnormal data isolation and labeling: The retrospective analysis algorithm first isolates and labels abnormal points in the data. These abnormal data points are not directly used for model training to avoid negatively affecting the overall diagnosis results.

[0064] Backtracking analysis steps:

[0065] Fault detection and marking: When a fault event is detected, the system records the relevant timestamp, parameter values ​​and environmental status.

[0066] Fault tree construction: Build a fault tree model based on the fault detection results. The root node of the fault tree is the final detected fault phenomenon, and the other nodes represent the fault source and intermediate states.

[0067] Backtracking path calculation: Based on the fault tree model, a backtracking analysis algorithm is used to calculate all paths from the fault phenomenon to the fault source. This path calculation combines features extracted by a multi-scale convolutional neural network with the inference results of a Bayesian model to calculate the failure probability of each path.

[0068] Path evaluation: Evaluate the comprehensive failure probability of each path and select the most promising path for in-depth analysis.

[0069] Root Cause Analysis: Identify the root cause of the failure and generate a detailed fault report including the root cause and recommended repair measures.

[0070] The feedback learning module integrates user feedback, historical fault records, and long-term operating data to continuously optimize the retrospective analysis model. Through the self-learning mechanism, the accuracy and response speed of the retrospective analysis algorithm are improved.

[0071] Scenario-adaptive tuning: Adjust the parameter settings of the retrospective analysis model according to different charging scenarios and driving habits to adapt to the actual operating environment.

[0072] After receiving the results of the backtracking analysis, the distributed response optimization module adjusts the current, voltage and charging time parameters of the charging system in real time through an adaptive parameter optimization algorithm.

[0073] Self-healing circuit: Based on the results of backtracking analysis, when the system detects a potential fault, the self-healing circuit can automatically switch to the backup circuit to prevent the fault from spreading.

[0074] Failure probability calculation of the backtracking analysis algorithm:

[0075]

[0076] : The posterior probability of the i-th fault source F_i, :Data D is at the fault source The likelihood function when exists, :Fault source The prior probability of : The marginal probability of data D.

[0077] Path probability calculation:

[0078]

[0079] :path The probability of :path The i-th node in .

[0080] The backtracking analysis algorithm, combined with the onboard intelligent charging system's circuit charging system, provides accurate fault location and repair recommendations through multi-dimensional data collection, anomaly detection, and path analysis. The algorithm optimizes fault diagnosis accuracy through big data analysis, ensuring the stability and safety of the charging system.

[0081] The adaptive analysis module includes a charging process behavior analysis unit, which is used to analyze the driver's charging habits and vehicle trajectory, identify bad charging behavior and provide personalized suggestions;

[0082] The distributed response optimization module includes:

[0083] Intelligent controller: used to adjust the current, voltage and other parameters of the charging system in real time.

[0084] Self-healing circuit control unit: manages redundant circuits and switches.

[0085] Remote control interface: Interface for remote monitoring and operation (Wi-Fi module or Bluetooth).

[0086] The process is as follows:

[0087] Abnormal response: Receive abnormal signals from the adaptive analysis module and adjust charging parameters in real time based on the analysis results.

[0088] Real-time adjustment: Ability to quickly respond to abnormal situations and adjust charging parameters to prevent faults from spreading.

[0089] System robustness: Improve the system's fault tolerance and reliability through self-healing circuits.

[0090] The distributed response optimization module is connected to the adaptive analysis module. Based on detected abnormal signals, the distributed response optimization module uses an adaptive parameter optimization algorithm to adjust the charging system's current, voltage, and charging time parameters in real time. It also dynamically optimizes the charging strategy by combining multi-source data to prevent fault propagation and improve charging efficiency and battery life. The adaptive parameter optimization algorithm automatically adjusts system parameters to optimize system performance by monitoring system status in real time. Using feedback control theory and online learning methods, the system can quickly respond to external environmental changes and internal faults. Based on abnormal signals from the adaptive analysis module, the operating parameters of the charging system are adjusted to prevent fault expansion and improve charging efficiency. The algorithm formula is as follows:

[0091]

[0092] is the current parameter, is the learning rate, is the parameter gradient, which is used to guide parameter updates.

[0093] The distributed response optimization module includes a self-healing circuit control unit. Based on detected fault conditions, the self-healing circuit control unit activates redundant circuits or switches to backup charging channels to improve the system's robustness and fault tolerance. A self-healing circuit introduces redundancy into the circuit. When a fault is detected in the main circuit, the self-healing circuit automatically switches to a backup circuit or activates a self-healing mechanism to ensure continued normal operation of the system. The self-healing circuit includes the following components:

[0094] Detection unit: includes current detector, voltage detector and temperature detector, used to monitor circuit status in real time.

[0095] Control unit: implemented by a microcontroller or FPGA, responsible for initiating the self-healing mechanism when an anomaly is detected.

[0096] Backup power supply and redundant circuit: The backup power supply module is combined with the redundant circuit module to provide power support and circuit replacement during fault switching.

[0097] If the detection unit detects an anomaly in the main charging circuit, the control unit triggers the backup power module and redundant circuit module through a signal, switching to the backup channel. At this point, the redundant circuit takes over the main charging function, ensuring uninterrupted system operation.

[0098] The control unit will determine the severity of the fault based on the results of the backtracking analysis algorithm and decide whether manual intervention is required or to continue using the backup channel.

[0099] The redundant switching circuit is used to transfer the load to the redundant circuit through a fast switching mechanism when the main circuit fails. This circuit can complete the switch in milliseconds, minimizing system downtime.

[0100] The redundant circuit is composed as follows:

[0101] Fast transfer switch: used to quickly switch the power path after a fault is detected.

[0102] Redundant circuit: A backup circuit parallel to the main circuit, including a backup power supply, backup charging module, etc.

[0103] Control logic circuit: The logic circuit that controls the switching switch and controls the switching operation by detecting signals.

[0104] Detection: When a main circuit fault is detected, the control logic circuit activates the fast transfer switch.

[0105] Switching: The switch disconnects the main power path and connects the redundant power module at the same time to ensure that the load is transferred to the redundant circuit.

[0106] Recovery: After the main circuit fault is eliminated, the switch is switched back to the main circuit and the redundant circuit returns to the standby state.

[0107] The distributed response optimization module includes a remote control interface, which is used by users to monitor and control the operating status of the charging system;

[0108] The Feedback Learning Module includes:

[0109] Data processing unit: responsible for integrating user feedback, historical records and other data.

[0110] Cloud interface: used to upload data to the cloud database (via API or FTP) and encrypt the data in advance to prevent data leakage.

[0111] The process is as follows:

[0112] Data integration: Input user operation feedback, historical fault records, charging scenario data, etc. into the feedback learning module.

[0113] Dynamic update: Use incremental learning algorithms to dynamically update the diagnostic model and continuously optimize model accuracy.

[0114] Scenario tuning: Adapt the model according to different charging scenarios and driving habits.

[0115] Learning and optimization: Continuously optimize diagnostic accuracy and improve system response speed through big data feedback.

[0116] Dynamic adaptation: Adjust algorithm parameters according to actual usage to ensure the accuracy and practicality of the model.

[0117] The feedback learning module is electrically connected to the multi-dimensional parameter acquisition module and the adaptive analysis module. It integrates user operational feedback, historical fault records, charging scenario data, and long-term vehicle system operation data. This data is fed into the big data feedback and learning module to enhance the algorithm's self-learning capabilities. This dynamic update mechanism, based on big data, continuously optimizes diagnostic accuracy and system response speed. The feedback learning module further includes a scenario-adaptive tuning unit, which adjusts and optimizes the charging system's diagnostic model based on driving scenarios and charging habits.

[0118] The feedback learning module also includes a data sharing unit that is used to upload diagnostic data to a cloud database for improving future versions of the charging system and developing more forward-looking fault diagnosis models. Specific embodiment two:

[0120] As shown in the figure, the operation process of the electric vehicle on-board charging system based on data analysis:

[0121] Sp1: Multi-dimensional parameter acquisition and data monitoring:

[0122] Sp1.1: The system first starts the multi-dimensional parameter acquisition module, which is composed of a distributed sensor network that covers every corner of the vehicle charging system and its surrounding environment.

[0123] Sp1.2: Collect data: including current, voltage, power, temperature, humidity, charging interface resistance, electromagnetic interference intensity, battery internal voltage distribution, vehicle operating status, geographic location, external climate conditions and driver behavior patterns.

[0124] Sp1.3: The environmental dynamic perception unit monitors changes in the external environment in real time, including temperature, humidity, air quality and vibration, and adjusts the working status of the charging system in real time based on the environmental data.

[0125] Sp1.4: The vehicle vibration detection unit analyzes the vibration frequency and amplitude under different road conditions to determine the potential impact of vibration on the charging system circuit.

[0126] Sp2: Adaptive analysis and fault identification:

[0127] Sp2.1: Multidimensional data is transmitted to the adaptive analysis module via an electrical connection. This module uses a fusion of an improved multi-scale convolutional neural network and an enhanced Bayesian inference model to perform data analysis.

[0128] Sp2.2: The adaptive analysis module processes data from the acquisition module in real time to identify potential failure modes and abnormal signals;

[0129] Sp2.3: The abnormal data isolation and marking unit isolates and marks the identified abnormal data to prevent it from negatively impacting the overall data analysis results, and uses a backtracking analysis algorithm to correlate and analyze related data to locate the cause of the fault;

[0130] Sp2.4: The charging process behavior analysis unit analyzes the driver's charging habits and vehicle trajectory, identifies bad charging behaviors, and provides personalized suggestions to optimize charging habits.

[0131] Sp3: Distributed intelligent response and optimization:

[0132] Sp3.1: The detected abnormal signal is transmitted to the distributed response optimization module, which adjusts the current, voltage and charging time parameters of the charging system in real time based on the adaptive parameter optimization algorithm.

[0133] Sp3.2: The self-healing circuit control unit activates the redundant circuit or switches to the backup charging channel according to the fault situation to improve the robustness and fault tolerance of the system.

[0134] Sp3.3: The system dynamically optimizes the charging strategy by combining multi-source data to prevent fault spread and improve charging efficiency and battery life.

[0135] Sp3.4: The remote control interface allows users to monitor and control the operating status of the charging system, enabling remote adjustment and optimization of the system.

[0136] Sp4: Feedback learning and continuous optimization:

[0137] Sp4.1: Collect and integrate data: The feedback learning module collects user operation feedback, historical fault records, charging scenario data and long-term operation data of the vehicle system, integrates them and inputs them into the big data feedback and learning module.

[0138] Sp4.2: The scenario-adaptive tuning unit dynamically adjusts and optimizes the diagnostic model based on driving scenarios and charging habits to improve the system's self-learning ability and response speed. Specific embodiment three:

[0140] This system combines multi-dimensional data acquisition, adaptive analysis, distributed intelligent response, and feedback learning modules to create a comprehensive, intelligent onboard charging solution for electric vehicles. Through real-time data collection and analysis, the system dynamically adjusts charging strategies to ensure stable and efficient operation under various operating conditions, and continuously improves long-term performance through learning and optimization.

[0141] After the system is started, the multi-dimensional parameter acquisition module is first activated. This module obtains real-time data from the on-board charging system and its surrounding environment through a distributed sensor network. Specific parameters collected include current, voltage, power, temperature, humidity, charging interface resistance, electromagnetic interference intensity, internal battery voltage distribution, vehicle operating status, geographic location, external climate conditions, and driver behavior patterns. These data provide the basis for subsequent analysis of the system, ensuring fine-grained monitoring of the charging process. In particular, the environmental dynamic perception unit monitors changes in the external environment, such as temperature, humidity, air quality, and vibration, in real time, and combines this environmental data with the internal state of the system to make real-time adjustments to the operating conditions during the charging process. In addition, the vehicle vibration detection unit determines the potential impact of vibration on the charging system circuit by analyzing the vibration frequency and amplitude of the vehicle under different road conditions, ensuring that the system can still operate stably in harsh environments.

[0142] When data collection is completed, all collected data will be transmitted to the adaptive analysis module for in-depth processing. The adaptive analysis module relies on the fusion of improved multi-scale convolutional neural networks and enhanced Bayesian inference models to process and analyze these multi-dimensional data in real time to identify potential fault modes and abnormal signals. To ensure the accuracy of data analysis, the system introduces an abnormal data isolation and marking unit, which is specifically used to isolate and mark abnormal data to prevent its negative impact on the overall analysis results. Through a retrospective analysis algorithm, this unit correlates and analyzes all relevant data at the time of the abnormal event, further improving the accuracy of fault cause location. In addition, the adaptive analysis module also includes a charging process behavior analysis unit, which identifies poor charging behavior by analyzing the driver's charging habits and vehicle trajectory, and provides personalized optimization suggestions to users based on this, helping users improve their charging behavior, thereby extending battery life and improving the overall efficiency of the system.

[0143] After completing the analysis, the adaptive analysis module passes the identified abnormal signals to the distributed response optimization module. Based on the adaptive parameter optimization algorithm, this module adjusts the current, voltage, and charging time parameters of the charging system in real time to ensure the optimal operating state of the system under different operating conditions. If the system detects a serious fault, the distributed intelligent response module will start the self-healing circuit control unit, activate the redundant circuit or switch to the backup charging channel to avoid system interruption and ensure the continuity and stability of the charging process. This dynamic optimization mechanism of multi-source data can effectively prevent the spread of faults, improve charging efficiency, and extend the service life of the battery. At the same time, the distributed response optimization module is also equipped with a remote control interface, through which users can monitor and adjust the operating status of the charging system in real time to ensure that the charging process always operates under optimal conditions.

[0144] Throughout the system's operation, the feedback learning module will continuously collect user operation feedback, historical fault records, charging scenario data, and long-term operating data of the vehicle system. These data will be integrated and input into the big data feedback and learning module to enhance the system's self-learning capabilities. Through the scenario-adaptive tuning unit, the system can dynamically adjust and optimize the diagnostic model based on actual driving scenarios and user charging habits, making it more adaptable to different usage scenarios. At the same time, the data sharing unit will upload the diagnostic data collected by the system to the cloud database for subsequent system improvements and the development of a new generation of fault diagnosis models. In this way, the system can not only achieve rapid optimization in the short term, but also provide solid data support for long-term use, ensuring that the system continues to evolve and improve in future versions.

[0145] In summary, the entire operational logic includes the following aspects: First, comprehensive environmental and system status monitoring is achieved through the multi-dimensional data acquisition module. Next, the adaptive analysis module performs in-depth data processing and anomaly detection to ensure that every piece of data is fully utilized. Finally, the distributed intelligent response module adjusts the charging strategy in real time to ensure system stability and efficiency. Finally, the feedback learning module continuously optimizes system performance and lays a data foundation for future upgrades. The close coordination of these modules throughout the entire process ensures stable and efficient operation of the charging system under various complex operating conditions.

[0146] It should be noted that, in this document, relational terms such as first and second, etc., are used only 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 terms "comprises," "comprising," or any other variants thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further restrictions, an element defined by the statement "comprising a reference structure" does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0147] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. An electric vehicle on-board charging system based on data analysis, characterized in that: The charging system comprises: A multi-dimensional parameter acquisition module, which is used to collect data from the vehicle charging system and its surrounding environment. The data includes current, voltage, power, temperature, humidity, charging interface resistance, electromagnetic interference intensity, battery internal voltage distribution, vehicle operating status, geographic location, external climate conditions, and driver behavior patterns, and achieves fine-grained monitoring of the entire charging process through a distributed sensor network; An adaptive analysis module, the adaptive analysis module is electrically connected to the multi-dimensional parameter acquisition module, and the adaptive analysis module is based on a fusion of an improved multi-scale convolutional neural network and an enhanced Bayesian reasoning model, and processes data from the multi-dimensional parameter acquisition module in real time to identify potential fault modes and abnormal signals, and the adaptive analysis module uses big data analysis features to automatically build and update a diagnostic model, and accurately predict and detect electrical faults in the charging system, hardware aging, and abnormal conditions caused by environmental factors; A distributed response optimization module, wherein the distributed response optimization module is connected to the adaptive analysis module, and the distributed response optimization module adjusts the current, voltage, and charging time parameters of the charging system in real time based on the detected abnormal signal through an adaptive parameter optimization algorithm, and dynamically optimizes the charging strategy in combination with multi-source data to prevent the spread of faults and improve charging efficiency and battery life; A feedback learning module is electrically connected to the multi-dimensional parameter acquisition module and the adaptive analysis module, and the feedback learning module integrates the user's operation feedback, historical fault records, charging scenario data and long-term operation data of the vehicle system, and inputs the big data feedback and learning module to enhance the self-learning ability of the algorithm, forming a dynamic update mechanism based on big data, and continuously optimizing the diagnostic accuracy and system response speed.

2. The electric vehicle on-board charging system based on data analysis according to claim 1 is characterized in that: The multi-dimensional parameter acquisition module also includes an environmental dynamic perception unit, which dynamically perceives changes in the external environment, including temperature, humidity, air quality and vibration, based on vehicle-mounted sensors, and adjusts the working state of the charging system in real time in combination with environmental data.

3. The electric vehicle on-board charging system based on data analysis according to claim 1 is characterized in that: The adaptive analysis module further includes an abnormal data isolation and marking unit, which is used to isolate abnormal data and mark abnormal data points to prevent their negative impact on the overall data analysis results, and to correlate and analyze all relevant data when the abnormal event occurs through a backtracking analysis algorithm to improve the accuracy of fault cause locating.

4. The electric vehicle on-board charging system based on data analysis according to claim 1 is characterized in that: The distributed response optimization module includes a self-healing circuit control unit, which activates a redundant circuit or switches to a backup charging channel based on a detected fault condition to improve the robustness and fault tolerance of the system.

5. The electric vehicle on-board charging system based on data analysis according to claim 1 is characterized in that: The feedback learning module further includes a scenario adaptive tuning unit, which adjusts and optimizes the diagnostic model of the charging system according to the driving scenario and charging habits.

6. The electric vehicle on-board charging system based on data analysis according to claim 1 is characterized in that: The multi-dimensional parameter acquisition module includes a vehicle vibration detection unit, which analyzes the vibration frequency and amplitude of the vehicle under different road conditions and determines the potential impact of the vibration on the charging system circuit.

7. The electric vehicle on-board charging system based on data analysis according to claim 1 is characterized in that: The adaptive analysis module includes a charging process behavior analysis unit, which is used to analyze the driver's charging habits and vehicle trajectory, identify bad charging behaviors and provide personalized suggestions.

8. The electric vehicle on-board charging system based on data analysis according to claim 1 is characterized in that: The distributed response optimization module includes a remote control interface, which is used for a user to monitor and control the operating status of the charging system.

9. The electric vehicle on-board charging system based on data analysis according to claim 1 is characterized in that: The feedback learning module also includes a data sharing unit, which is used to upload diagnostic data to a cloud database for improving future versions of the charging system and developing a more forward-looking fault diagnosis model.

10. The electric vehicle on-board charging system based on data analysis according to claim 1, characterized in that: The multi-dimensional parameter acquisition module further integrates an electromagnetic interference shielding unit. The high-precision electromagnetic interference shielding unit is used to monitor and shield external electromagnetic interference signals in real time to prevent them from affecting the normal operation of the charging system circuit.

Images