A method for monitoring the health status of automotive fuel cells based on multi-dimensional sensing

Abstract

The present application provides a method for monitoring the health status of a vehicle fuel cell based on multi-dimensional sensing, and relates to the technical field of vehicle fuel cells. The method includes: configuring a cellular network to digitally communicate with a multi-dimensional sensor group; performing time alignment of the multi-dimensional sensor group, obtaining monitoring data of the multi-dimensional sensor group, generating a time-series monitoring data set, and sending it to a data processing center; establishing a prediction network based on basic battery information, which is built into the data processing center; predicting the health status of the time-series monitoring data set, and generating a health status prediction result; configuring a dynamic health index model, performing health index decay analysis, generating a real-time health index, and managing the health status of vehicle fuel cells. The present application can solve the technical problem in the prior art that most of the battery performance is reduced due to the lack of effective monitoring based on multi-dimensional sensors, achieve the technical goal of effectively monitoring fuel cells with multi-dimensional sensors, and achieve the technical effect of improving battery performance.

Description

Technical Field

[0001] The present application relates to the technical field of vehicle fuel cells, and in particular to a method for monitoring the health status of vehicle fuel cells based on multi-dimensional sensing. Background Art

[0002] A vehicle fuel cell is a device that converts chemical energy directly into electrical energy. It uses hydrogen as fuel and reacts with oxygen to produce electricity and water to drive the vehicle's electric motor, thus achieving zero-emission clean energy transportation.

[0003] At present, in the existing battery health status monitoring process, most of them do not adjust the fuel cell in time according to the battery status monitored by multi-dimensional sensors, resulting in battery performance degradation, including reduced power output, reduced efficiency and slow response time. The lack of effective health status monitoring leads to potential problems and failures not being discovered, thereby increasing the risk of sudden failure of the battery system. Therefore, a method is needed to solve the above problems.

[0004] In summary, the prior art has a technical problem that most of the fuel cells are not effectively monitored based on multi-dimensional sensors, resulting in failure to adjust the fuel cell in a timely manner, causing degradation of battery performance and further causing the risk of battery failure. Summary of the invention

[0005] The purpose of this application is to provide a method for monitoring the health status of automotive fuel cells based on multi-dimensional sensing, in order to solve the technical problem in the prior art that most of the fuel cells are not effectively monitored based on multi-dimensional sensors, resulting in failure to adjust the fuel cells in a timely manner, causing battery performance degradation and further causing the risk of battery failure.

[0006] In view of the above problems, the present application provides a method for monitoring the health status of a vehicle fuel cell based on multi-dimensional sensing.

[0007] The present application provides a method for monitoring the health status of a vehicle fuel cell based on multidimensional sensing, wherein the method includes: configuring a cellular network, the cellular network being used for digital communication with a multidimensional sensor group, the multidimensional sensor group including a voltage sensor, a temperature sensor, an impedance sensor and a hydrogen flow sensor; performing timing alignment of the multidimensional sensor group based on the cellular network, and acquiring monitoring data of the multidimensional sensor group after timing alignment, generating a timing monitoring data set, and sending the timing monitoring data set to a data processing center; acquiring basic battery information of the vehicle fuel cell, establishing a prediction network based on the basic battery information, the establishment data of the prediction network including internal data and encrypted external data, and the prediction network is built in the data processing center; predicting the health status of the timing monitoring data set according to the prediction network, and generating a health status prediction result; configuring a dynamic health index model according to the health status prediction result, performing health index decay analysis through the dynamic health index model, generating a real-time health index, and managing the health status of the vehicle fuel cell according to the real-time health index.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] By configuring a cellular network, the cellular network is used for digital communication with a multidimensional sensor group, and the multidimensional sensor group includes a voltage sensor, a temperature sensor, an impedance sensor and a hydrogen flow sensor; based on the cellular network, the timing alignment of the multidimensional sensor group is performed, and the monitoring data of the multidimensional sensor group after the timing alignment is obtained, a timing monitoring data set is generated, and the timing monitoring data set is sent to a data processing center; basic battery information of a vehicle fuel cell is obtained, and a prediction network is established based on the basic battery information, the establishment data of the prediction network includes internal data and encrypted external data, and the prediction network is built in the data processing center; the health status of the timing monitoring data set is predicted according to the prediction network, and a health status prediction result is generated; a dynamic health index model is configured according to the health status prediction result, and a health index attenuation analysis is performed through the dynamic health index model to generate a real-time health index, and the health status of the vehicle fuel cell is managed according to the real-time health index, so as to achieve the technical goal of multidimensional sensors to effectively monitor the fuel cell and achieve the technical effect of improving battery performance.

[0010] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0012] Figure 1 This is a flow chart of a method for monitoring the health status of a fuel cell for a vehicle based on multi-dimensional sensing in this application;

[0013] Figure 2 This is a flow chart of establishing a prediction network based on the basic battery information in the vehicle fuel cell health status monitoring method based on multi-dimensional sensing in this application. DETAILED DESCRIPTION

[0014] This application provides a method for monitoring the health status of a vehicle fuel cell based on multi-dimensional sensing, which solves the technical problem in the prior art that most of the fuel cells are not effectively monitored based on multi-dimensional sensors, resulting in failure to adjust the fuel cell in time, causing battery performance degradation and further causing the risk of battery failure. The technical goal of effectively monitoring the fuel cell with a multi-dimensional sensor is achieved, and the technical effect of improving battery performance is achieved.

[0015] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0016] Please see attached Figure 1 The present application provides a vehicle fuel cell health status monitoring method based on multi-dimensional sensing, wherein the vehicle fuel cell health status monitoring method based on multi-dimensional sensing specifically includes the following steps:

[0017] Step 1: configuring a cellular network, wherein the cellular network is used for digital communication with a multi-dimensional sensor group, wherein the multi-dimensional sensor group includes a voltage sensor, a temperature sensor, an impedance sensor, and a hydrogen flow sensor.

[0018] Specifically, a cellular network is a wireless communication network that can provide wide coverage, stable and reliable communication services. In the vehicle fuel cell health status monitoring system, the cellular network is used to realize digital communication between the multidimensional sensor group and the monitoring center. The multidimensional sensor group can monitor the operating status of the fuel cell in real time to ensure safe, stable and efficient operation. The multidimensional sensor group usually includes a voltage sensor, a temperature sensor, an impedance sensor and a hydrogen flow sensor, which can monitor the parameters of the fuel cell, such as the voltage, temperature distribution, internal impedance and hydrogen consumption rate of the battery stack. The multidimensional sensor group communicates digitally with the monitoring center through the cellular network, and transmits the collected fuel cell operating status data to the monitoring center in real time. The communication method can be based on the TCP / IP protocol, such as MQTT, HTTP, etc.

[0019] Step 2: Performing time alignment of the multidimensional sensor group based on the cellular network, acquiring monitoring data of the multidimensional sensor group after time alignment, generating a time series monitoring data set, and sending the time series monitoring data set to a data processing center.

[0020] Specifically, a common reference time point is determined. For example, the starting point of data collection or the time when a specific event occurs is extracted. Then, according to the timestamp information of the sensor, the data of each sensor is adjusted to correspond to the reference time point. When the sensor data have completed the timing alignment, the aligned data set is obtained, and a timing monitoring data set is generated, which contains the measurement values ​​of all sensors at the same time point, providing a basis for subsequent data analysis. The timing monitoring data set after timing alignment is sent to the data processing center via the cellular network. The data processing center can be a data analysis workstation of a cloud server or a local server, which further processes and analyzes the data to evaluate the health status of the fuel cell.

[0021] Step three: Obtain basic battery information of the vehicle fuel cell, and establish a prediction network based on the basic battery information. The establishment data of the prediction network includes internal data and encrypted external data, and the prediction network is built in the data processing center.

[0022] Specifically, the basic information of the battery of the vehicle fuel cell is obtained, including the type, manufacturer, model, production date, installation date, expected life, design specifications, operating conditions, historical performance data, etc. of the battery, and the behavior and performance characteristics of the battery are obtained. Internal data refers to sensor data from the fuel cell itself, such as voltage, temperature, impedance and hydrogen flow, which reflects the real-time operating status and health status of the battery. External data includes environmental conditions (such as temperature, humidity, air pressure), vehicle usage patterns (such as driving distance, speed, acceleration), driving behavior, etc., which affect the performance and life of the fuel cell. When processing external data, encryption is performed to protect privacy and data confidentiality. The prediction network is trained using internal data and encrypted external data. The prediction network can be a machine learning model, such as a neural network, support vector machine, decision tree, etc., in order to predict the health status and potential failures of the fuel cell. The prediction network is built into the data processing center to directly access and process real-time data from the fuel cell. The data processing center has sufficient computing resources to process large-scale data and complex models.

[0023] Step 4: Perform health status prediction on the time series monitoring data set according to the prediction network to generate a health status prediction result.

[0024] Specifically, the collected time-series monitoring data set is input into the prediction network, which contains the measured values ​​of fuel cell voltage, temperature, impedance, hydrogen flow and other parameters. The prediction network uses the input data set to generate health status prediction results, including data preprocessing, feature extraction, model training and prediction generation, and generates prediction results, including the evaluation of the current health status of the fuel cell, the prediction of the future health status, the early warning of potential failures, and the estimation of the remaining service life.

[0025] Step 5: Configure a dynamic health index model according to the health status prediction result, perform health index decay analysis through the dynamic health index model, generate a real-time health index, and manage the health status of the vehicle fuel cell according to the real-time health index.

[0026] Specifically, the dynamic health index model is configured based on the operating characteristics, historical performance, environmental factors, and maintenance activities of the fuel cell. The dynamic health index model is a health status indicator that is adjusted based on the real-time data and prediction results of the fuel cell. The health index decay analysis is performed based on factors such as battery aging, performance degradation, failure history, and maintenance effects. The health index decay analysis refers to analyzing the trend of the health index over time to predict possible future state changes. Based on the dynamic health index model and decay analysis, a real-time health index is generated. It reflects the current health status of the fuel cell and can be adjusted based on real-time data and historical trends. Based on the real-time health index, the health status of automotive fuel cells is managed, including adjusting the operating mode, planning preventive maintenance, replacing parts that are about to fail, optimizing energy use, and making retirement decisions.

[0027] The vehicle fuel cell health status monitoring method based on multi-dimensional sensing can achieve the technical goal of multi-dimensional sensors to effectively monitor fuel cells and achieve the technical effect of improving battery performance.

[0028] Further, if Figure 2 As shown, this application also includes:

[0029] Obtain the data sample size of the internal data, and generate a first influence coefficient according to the data sample size; perform data richness analysis on the internal data, and generate a second influence coefficient according to the data richness analysis result; perform adaptation verification on the external data, and establish a third influence coefficient based on the adaptation verification result; obtain the data credibility of the external data, and establish a fourth influence coefficient according to the data credibility; establish a prediction network through the first influence coefficient, the second influence coefficient, the third influence coefficient, and the fourth influence coefficient.

[0030] Specifically, internal data refers to sensor data inside the fuel cell system, such as voltage, temperature, impedance, and hydrogen flow. The data sample size refers to the number of data points collected within a certain time range. The size of the sample size affects the accuracy of subsequent analysis and modeling. If the sample size is insufficient, the dynamic characteristics and abnormal conditions of the system cannot be accurately captured. Based on the internal data sample size, the first impact coefficient is generated, which reflects the impact of the data sample size on the accuracy of the prediction model. The larger the sample size, the higher the first impact coefficient, indicating that the model is more representative of the system, and vice versa.

[0031] Then, data richness refers to the diversity and comprehensiveness of data. Perform richness analysis on internal data to evaluate whether the data contains enough information to describe the operating status of the fuel cell. For example, whether the data covers different working conditions, whether it contains enough abnormal situations, etc. Based on the results of data richness analysis, a second impact coefficient is generated, which reflects the impact of data diversity on the performance of the prediction model. The richer the data, the higher the second impact coefficient, indicating that the model can better capture the different states of the system, and vice versa.

[0032] Next, external data includes environmental conditions, vehicle usage patterns, driving behaviors, etc. The external data is adapted and verified to ensure compatibility with the internal data and to reflect the actual operating environment of the fuel cell. Based on the adaptation verification results, a third influence coefficient is established to reflect the adaptive impact of external data on the prediction model. If the external data is highly adapted to the internal data, the third influence coefficient is higher, indicating that the model can better consider the impact of external factors on the health status of the fuel cell, otherwise, it is worse.

[0033] Next, the credibility of external data refers to the accuracy and reliability of the data. For example, data from different sources may have different precision and errors. Evaluate the credibility of external data and determine its weight in the prediction model. According to the credibility of external data, establish the fourth impact coefficient, which reflects the impact of data quality on the performance of the prediction model. The more credible the data, the higher the fourth impact coefficient, indicating that the model relies more on this data, and vice versa.

[0034] Then, by combining the first influence coefficient, the second influence coefficient, the third influence coefficient, and the fourth influence coefficient, a comprehensive prediction network is established to improve the prediction accuracy of the health status of the fuel cell according to the sample size and richness of the internal data and the adaptability and credibility of the external data.

[0035] By establishing a predictive model to monitor the health status of automotive fuel cells, the safety and reliability of the vehicle can be improved.

[0036] Furthermore, the present application also includes:

[0037] Among them, ML_Predict(t) is the prediction result of the prediction network, f ML Representation prediction network, Characterizes the moving average value of voltage at time t, Characterizes the moving average temperature at time t, Characterizes the moving average value of impedance at time t, Characterizes the moving average value of hydrogen flow at time t, I VT(t) is the interaction term between voltage and temperature, representing the joint effect of voltage and temperature, I ZQH (t) is the interaction term between impedance and hydrogen flow rate, representing the joint effect of impedance and hydrogen flow rate.

[0038] Specifically, the moving average of the voltage It can smooth short-term fluctuations and help analyze long-term trends. Moving average of temperature Helps identify temperature stability and trends. Moving average of impedance Helps analyze long-term changes in impedance. Moving average of hydrogen flow It helps to understand the consumption pattern of hydrogen and the operating status of the fuel cell. There is an interaction between voltage and temperature. The interaction term I VT (t) jointly affect the performance of the fuel cell. Impedance and hydrogen flow interact with each other, and the interaction term I ZQH (t) Affect the overall performance of the fuel cell.

[0039] According to the prediction network f ML , taking multiple parameters and interaction terms as input features, and using machine learning algorithms to train the prediction network f ML . Prediction network f ML The output is the prediction result ML_Predict(t), which represents the health status of the fuel cell at time t.

[0040] Through the interaction of multiple parameters and interaction terms, the prediction network can more accurately simulate the behavior of the fuel cell and predict its future health status, which helps to identify potential problems in advance, so as to perform preventive maintenance or adjust the operation strategy to extend the service life of the fuel cell and ensure safe operation.

[0041] Furthermore, the present application also includes:

[0042] DHI(t)=γ·DHI(t-1)+(1-γ)·ML_Predict(t); wherein DHI(t) represents the real-time health index, DHI(t-1) is the health index at the previous time, γ is the attenuation factor, representing the impact of the health index at the previous time on the current evaluation, and ML_Predict(t) is the prediction result of the prediction network.

[0043] Specifically, the real-time health index DHI(t) characterizes the value of the health status of the fuel cell at time t, which is obtained based on the influence of multiple parameters such as voltage, temperature, impedance and hydrogen flow rate. The health index DHI(t-1) at the previous time is the real-time health index value at time t-1, reflecting the health status of the fuel cell at the previous time point. The attenuation factor γ is a value between 0 and 1, which is used to characterize the degree of influence of the health index of the previous time period on the current evaluation. The size of the attenuation factor γ can be determined according to the characteristics and application scenarios of the fuel cell. A larger attenuation factor means that the historical data has a greater impact on the current health status, while a smaller attenuation factor means that more attention is paid to the current data. The prediction result ML_Predict(t) of the prediction network is the prediction result obtained by the prediction network based on the sensor data at the current time point, reflecting the health status of the fuel cell predicted based on the historical data and the learning algorithm. The real-time health index is obtained by weighted average of the health index at the previous time and the further prediction results at the current time. The attenuation factor determines the relative importance of the two factors in calculating the current health index.

[0044] The real-time health index can be continuously updated to reflect the latest health status of the fuel cell, helping to promptly understand the status of the fuel cell and take action when necessary to maintain performance and extend life.

[0045] Furthermore, the present application also includes:

[0046] The time series monitoring data set is input into the impact verification network of long and short windows, and the impact verification of long and short windows is performed, including: S1: calling the short time series window of the impact verification network to perform abnormal event impact verification on the time series monitoring data set, and establish a transient health impact result; S2: calling the long time series window of the impact verification network to perform abnormal event impact verification on the time series monitoring data set, and establish a long-term health impact result; S3: calling the fusion verification subnetwork of the impact verification network to perform fusion smoothing verification of transient health impact results and long-term health impact results, and generating impact verification results of long and short windows according to the fusion smoothing verification results; compensating the real-time health index according to the impact verification result, and managing the health status of the vehicle fuel cell according to the compensated real-time health index.

[0047] Specifically, in the vehicle fuel cell health status monitoring method, the impact verification network of long and short windows is used to analyze the time series monitoring data set and evaluate the impact of abnormal events on the health status of the fuel cell. First, the time series monitoring data set is analyzed using the short time series window of the impact verification network. The short time series window focuses on the changes in data in a short period of time, which is used to detect and verify the transient response of the fuel cell and abnormal events in the short term, and establish the transient health impact results, that is, to evaluate the immediate impact of abnormal events in the short term on the health status of the fuel cell. Next, the same time series monitoring data set is analyzed using the long time series window of the impact verification network. The long time series window focuses on the change trend of the data over a long period of time, which is used to detect and verify the long-term performance changes and chronic abnormal events of the fuel cell, and establish the long-term health impact results, that is, to evaluate the cumulative impact of abnormal events in the long term on the health status of the fuel cell. Finally, the fusion verification subnetwork of the impact verification network is called to fuse the transient health impact results and the long-term health impact results, combining the short-term and long-term impacts to obtain a more comprehensive and accurate health status assessment. According to the results of the fusion smoothing verification, the impact verification results of the long and short windows are generated, and the immediate and long-term health impacts of the fuel cell are generated.

[0048] The real-time health index is then compensated based on the impact verification results. If the impact verification results indicate that the health of the fuel cell has been negatively impacted by the abnormal event, the real-time health index will be adjusted accordingly to reflect the impact.

[0049] Finally, the health status of the vehicle fuel cell is managed based on the compensated real-time health index, including adjusting operating parameters, planning maintenance activities or replacing problematic parts to ensure the performance and reliability of the fuel cell system.

[0050] Through the impact verification method of long and short windows, the health status of automotive fuel cells can be more accurately monitored and managed, thereby improving the overall performance and safety of the vehicle.

[0051] Furthermore, the present application also includes:

[0052] A transient breakthrough record is made for the transient health impact result, and a health penalty impact factor is established according to the transient breakthrough record result; the impact verification result is corrected by the health penalty impact factor, and real-time health index compensation is completed according to the corrected impact verification result.

[0053] Specifically, transient breakthrough recording refers to recording and analyzing abnormal events that occur in a fuel cell in a short period of time, including sudden voltage fluctuations, rapid changes in temperature, abnormal increases in impedance, or sharp drops in hydrogen flow, etc., to capture short-term transient events that may have a significant impact on the health of the fuel cell. Based on the results of transient breakthrough recording, a health penalty impact factor is established. The health penalty impact factor refers to a quantitative indicator of the decline in the health of a fuel cell when encountering an abnormal event. The size of the health penalty impact factor depends on the severity and duration of the abnormal event, as well as the sensitivity of the fuel cell to these events.

[0054] Then, the impact verification results are corrected by the health penalty impact factor to more accurately reflect the short-term and long-term impact of abnormal events on the health status of the fuel cell. The corrected impact verification results will more realistically reflect the actual health status of the fuel cell. Based on the corrected impact verification results, the real-time health index is compensated to ensure that the real-time health index can accurately reflect the health status of the fuel cell after the abnormal event, thereby providing a reliable basis for the health status management of the fuel cell.

[0055] By effectively monitoring and managing the health status of automotive fuel cells, it helps to improve the reliability of the fuel cell and the safety of the vehicle, while also helping to extend the service life of the fuel cell.

[0056] Furthermore, the present application also includes:

[0057] Establish a life cycle of a vehicle fuel cell according to the basic battery information; configure an additional correlation factor of the vehicle fuel cell based on the life cycle; adjust the real-time health index with the additional correlation factor, and manage the health status of the vehicle fuel cell according to the adjustment result.

[0058] Specifically, the life cycle refers to the entire use process of a fuel cell from manufacturing, installation, operation to final retirement. The life cycle can be established based on basic information about the battery, including the battery type, manufacturing date, installation date, expected life, maintenance history, operating conditions, etc. By analyzing the life cycle, the behavior and performance changes of the battery at different stages of use can be predicted.

[0059] Then, additional correlation factors refer to other factors that affect the health status of the fuel cell, which change with the different stages of the battery life cycle. For example, battery aging, usage frequency, changes in the operating environment, maintenance status, etc. can be used as additional correlation factors to help more accurately evaluate the health status of the fuel cell.

[0060] Next, based on the additional correlation factors, the real-time health index is adjusted so that it can better reflect the true health status of the fuel cell at a specific stage of its life cycle. For example, if the battery is approaching the end of its expected life, then even if the real-time health index is high, the impact of aging effects still needs to be considered. Based on the adjusted real-time health index, the health status of automotive fuel cells is managed, including adjusting maintenance plans, replacing parts that are about to fail, optimizing operating conditions, or determining the retirement time of the battery, to ensure that the fuel cell maintains optimal performance and reliability throughout its life cycle.

[0061] By combining the life cycle with the real-time health index, the health status of automotive fuel cells can be managed more effectively, thereby improving the overall performance and safety of the vehicle, while also helping to extend the service life of the fuel cell.

[0062] Furthermore, the present application also includes:

[0063] Perform preset periodic maintenance of the vehicle fuel cell and create a periodic maintenance report; generate network correction data of the prediction network through the periodic maintenance report, and optimize the prediction network based on the network correction data.

[0064] Specifically, the periodic maintenance of automotive fuel cells includes scheduled inspections and repairs, which are carried out according to the maintenance plan. Maintenance activities include checking the integrity of the battery stack, cleaning sensors, replacing damaged parts, checking the cooling system, confirming the electrical connections of the battery stack, etc., to prevent failures and repair potential problems in a timely manner.

[0065] Then, after each maintenance is performed, a periodic maintenance report is created, recording information such as problems found during maintenance activities, measures taken, replaced parts, maintenance date and time, and the signature of the maintenance personnel. The information contained in the periodic maintenance report is used to generate network correction data for the prediction network, which reflects the problems found in actual maintenance activities and the actual condition of the battery. The comparison with the output of the prediction network reveals the accuracy and deviation of the prediction model. Based on the network correction data, the prediction network is optimized, including adjusting the parameters of the model, updating the learning algorithm, improving feature engineering, adding new training data, etc., to improve the accuracy and robustness of the prediction network, thereby more accurately predicting the health status of the fuel cell.

[0066] By performing periodic maintenance and using maintenance reports to optimize the predictive network, we ensure the continuous improvement and effectiveness of the vehicle fuel cell health status monitoring system, which helps to improve the reliability and operating efficiency of the fuel cell system while reducing maintenance costs.

[0067] In summary, the vehicle fuel cell health status monitoring method based on multi-dimensional sensing provided in this application has the following technical effects:

[0068] By configuring a cellular network, the cellular network is used for digital communication with a multidimensional sensor group, and the multidimensional sensor group includes a voltage sensor, a temperature sensor, an impedance sensor and a hydrogen flow sensor; based on the cellular network, the timing alignment of the multidimensional sensor group is performed, and the monitoring data of the multidimensional sensor group after the timing alignment is obtained, a timing monitoring data set is generated, and the timing monitoring data set is sent to a data processing center; basic battery information of a vehicle fuel cell is obtained, and a prediction network is established based on the basic battery information, the establishment data of the prediction network includes internal data and encrypted external data, and the prediction network is built in the data processing center; the health status of the timing monitoring data set is predicted according to the prediction network, and a health status prediction result is generated; a dynamic health index model is configured according to the health status prediction result, and a health index attenuation analysis is performed through the dynamic health index model to generate a real-time health index, and the health status of the vehicle fuel cell is managed according to the real-time health index, so as to achieve the technical goal of multidimensional sensors to effectively monitor the fuel cell and achieve the technical effect of improving battery performance.

[0069] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0070] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. A method for monitoring the health status of a fuel cell for a vehicle based on multi-dimensional sensing, characterized in that: The method comprises: configuring a cellular network, the cellular network being used for digital communication with a multi-dimensional sensor group, the multi-dimensional sensor group comprising a voltage sensor, a temperature sensor, an impedance sensor, and a hydrogen flow sensor; Performing time alignment of the multidimensional sensor group based on the cellular network, acquiring monitoring data of the multidimensional sensor group after time alignment, generating a time series monitoring data set, and sending the time series monitoring data set to a data processing center; Obtaining basic battery information of a vehicle fuel cell, and establishing a prediction network based on the basic battery information, wherein the establishment data of the prediction network includes internal data and encrypted external data, and the prediction network is built in the data processing center; Performing health status prediction on the time series monitoring data set according to the prediction network to generate a health status prediction result; configuring a dynamic health index model according to the health status prediction result, performing health index decay analysis through the dynamic health index model to generate a real-time health index, and performing health status management of the vehicle fuel cell according to the real-time health index; Wherein, the method further comprises: The time series monitoring data set is input into the impact verification network of the long and short windows to perform the impact verification of the long and short windows, including: S1: calling the short time series window of the impact verification network to verify the impact of abnormal events on the time series monitoring data set and establish a transient health impact result; S2: calling the long time series window of the impact verification network to verify the impact of abnormal events on the time series monitoring data set, and establishing long-term health impact results; S3: calling the fusion verification sub-network of the impact verification network to perform fusion smoothing verification of transient health impact results and long-term health impact results, and generating impact verification results of long and short windows according to the fusion smoothing verification results; The real-time health index is compensated according to the impact verification result, and the health status of the vehicle fuel cell is managed according to the compensated real-time health index.

2. The vehicle fuel cell health status monitoring method based on multi-dimensional sensing as claimed in claim 1, characterized in that: The establishing of a prediction network based on the battery basic information further includes: Acquire a data sample size of the internal data, and generate a first influence coefficient according to the data sample size; Performing data richness analysis on the internal data, and generating a second influence coefficient according to the data richness analysis result; Performing adaptation verification on the external data, and establishing a third influence coefficient based on the adaptation verification result; Acquiring data credibility of the external data, and establishing a fourth influence coefficient according to the data credibility; A prediction network is established through the first influence coefficient, the second influence coefficient, the third influence coefficient, and the fourth influence coefficient.

3. The vehicle fuel cell health status monitoring method based on multi-dimensional sensing as claimed in claim 2, characterized in that: The prediction network is as follows: ； ； ； in, To predict the prediction results of the network, Representation prediction network, Characterization time The voltage moving average under Characterization time The moving average temperature under Characterization time The impedance moving average under Characterization time The moving average value of hydrogen flow rate under is the interaction term between voltage and temperature, representing the joint effect of voltage and temperature, is the interaction term between impedance and hydrogen flow rate, representing the joint effect of impedance and hydrogen flow rate.

4. The vehicle fuel cell health status monitoring method based on multi-dimensional sensing as claimed in claim 3, characterized in that: The dynamic health index model is configured according to the health status prediction result as follows: ； in, Characterize real-time health index, is the health index of the previous time, is the attenuation factor, which represents the impact of the previous health index on the current assessment. is the prediction result of the prediction network.

5. The vehicle fuel cell health status monitoring method based on multi-dimensional sensing as claimed in claim 1, characterized in that: The method further comprises: Recording transient breakthroughs of the transient health impact results, and establishing health penalty impact factors according to the transient breakthrough record results; The impact verification result is corrected by the health penalty impact factor, and real-time health index compensation is completed according to the corrected impact verification result.

6. The vehicle fuel cell health status monitoring method based on multi-dimensional sensing as claimed in claim 1, characterized in that: The health status management of the vehicle fuel cell according to the real-time health index also includes: Establishing a life cycle of a vehicle fuel cell according to the basic battery information; configuring additional correlation factors for the vehicle fuel cell based on the life cycle; The real-time health index is adjusted using the additional correlation factor, and the health status of the vehicle fuel cell is managed according to the adjustment result.

7. The vehicle fuel cell health status monitoring method based on multi-dimensional sensing as claimed in claim 1, characterized in that: The method further comprises: Perform scheduled periodic maintenance of vehicle fuel cells and create periodic maintenance reports; The network correction data of the predicted network is generated through the periodic maintenance report, and the predicted network is optimized based on the network correction data.

Images