Remote fault monitoring method and system for vehicle-mounted mobile meteorological monitoring equipment

Abstract

The invention discloses a remote fault monitoring method and system for vehicle-mounted mobile meteorological monitoring equipment, and belongs to the technical field of meteorological monitoring, and the method specifically comprises the steps: collecting meteorological data in real time through a sensor, carrying out the self-adaptive filtering preprocessing, and transmitting the data to a remote monitoring center; the remote monitoring center uses a deep learning model, time sequence analysis and an anomaly detection algorithm to evaluate the operation state of the equipment, predicts the fault possibility and the emergency degree, and triggers multi-level early warning; in combination with historical fault data and current meteorological conditions, fault types, positions and reasons are judged by using an integrated learning method, an optimal fault removal scheme is automatically generated, and simulation verification is carried out; the block chain is adopted to record fault information, an intelligent feedback system optimization deep learning model is established, a preventive maintenance plan is made, and the accuracy and efficiency of fault monitoring are improved.

Description

Technical Field

[0001] The invention belongs to the technical field of meteorological monitoring, in particular to a remote fault monitoring method and system for vehicle-mounted mobile meteorological monitoring equipment. Background Art

[0002] With the growing demand for meteorological monitoring, vehicle-mounted mobile meteorological monitoring equipment has been widely used in agriculture, environment, urban planning and other fields due to its high flexibility and wide coverage. However, when the equipment is operating in a complex and changeable field environment, it is susceptible to environmental factors and malfunctions, such as sensor damage and data transmission interruption, which affects the continuity and accuracy of the data. The traditional manual inspection method is not only inefficient, but also difficult to detect and solve problems in a timely manner. Therefore, it is particularly important to develop a system that can remotely monitor equipment status, intelligently warn and assist in remote maintenance.

[0003] For example, the Chinese patent application with publication number CN112766372A discloses a method and system for monitoring and warning the status of power equipment based on micro-meteorological perception, including: using a binocular camera to form a deep image to recognize the external appearance of the power equipment, and combining micro-meteorological perception to analyze the evolution process of power equipment faults, and to issue early warnings for power equipment discharge, external damage, etc., to provide technical support for improving the operating safety margin of my country's power equipment in weather conditions such as strong winds, high temperatures, severe cold, wind and snow.

[0004] For example, the patent with the authorization announcement number CN221595311U discloses an automatic weather station fault monitoring device, including: a central processing unit; a sensor module and a comparison module, wherein the sensor module is connected to the comparison module, and the comparison module is also connected to the central processing unit, wherein the sensor module is used to monitor meteorological characteristics in real time, and the comparison module is used to compare the data information monitored by two identical sensors; an alarm module, which is connected to the central processing unit; a storage module, which is connected to the central processing unit; a display module, which is connected to the central processing unit; a positioning module, which is connected to the central processing unit; and a host computer, which is connected to the central processing unit. This technical solution can quickly and accurately detect sensor faults by comparing the data information monitored by two identical sensors, so as to achieve timely maintenance and reduce the consumption of time and manpower.

[0005] The above existing technologies all have the following problems: the fault detection means are relatively single, and lack deep learning and adaptive capabilities, and lack automatic generation and optimization of fault handling solutions; the accuracy of fault warning is limited, and the scope of application is limited. Summary of the invention

[0006] In view of the shortcomings of the prior art, the present invention proposes a remote fault monitoring method and system for vehicle-mounted mobile meteorological monitoring equipment. The meteorological data is collected in real time by sensors and sent to a remote monitoring center after adaptive filtering preprocessing. The remote monitoring center uses deep learning models, time series analysis and anomaly detection algorithms to evaluate the operating status of the equipment, predict the possibility and urgency of faults, and trigger multi-level warnings. Combined with historical fault data and current meteorological conditions, an integrated learning method is used to determine the type, location and cause of the fault, automatically generate the optimal troubleshooting plan and perform simulation verification. Blockchain is used to record fault information, an intelligent feedback system is established to optimize the deep learning model, and a preventive maintenance plan is formulated. The present invention improves the accuracy and efficiency of fault monitoring.

[0007] To achieve the above object, the present invention provides the following technical solutions: A remote fault monitoring method for vehicle-mounted mobile meteorological monitoring equipment, comprising: Step S1: The vehicle-mounted equipment collects meteorological data in real time through sensors, pre-processes the collected meteorological data using an adaptive filtering strategy, and sends the pre-processed meteorological data to a remote monitoring center; Step S2: The remote monitoring center uses a deep learning model to identify potential fault modes in the preprocessed meteorological data received, combines time series analysis and anomaly detection algorithms, and evaluates the operating status of the on-board equipment in real time. Based on the evaluation results, the possibility and urgency of the fault are predicted, and a multi-level early warning mechanism is triggered based on the urgency of the fault prediction. Step S3: Combine historical fault data and current meteorological conditions, use ensemble learning methods to comprehensively judge the fault type, specific location and potential causes, and introduce optimization strategies to automatically generate the optimal troubleshooting solution based on the diagnosis results, and perform simulation verification on the optimal troubleshooting solution; Step S4: Use blockchain method to record the time, location, cause, handling method and result information of the fault. At the same time, establish an intelligent feedback system, integrate the feedback and fault handling results into the deep learning model, and formulate a preventive maintenance plan based on the optimized deep learning model.

[0008] Specifically, the specific steps of the adaptive filtering strategy in step S1 include: S1.1: Obtain meteorological data and store it as time series meteorological data ,in, represents the mth time series meteorological data, and m represents the number of time series meteorological data; S1.2: Set the initial parameters of the filter according to the current filter output and the expected output , calculate the error signal ,in, represents the current filter coefficient vector, represents the meteorological data vector input to the filter, and T represents transpose; S1.3: According to the error signal and the input signal, the coefficient of the filter is updated using the adaptive filtering algorithm, and the updated filter is used to filter the meteorological data. The formula is: ; in, represents the updated filter coefficient vector, represents the dynamic learning rate adjustment factor, represents the step size factor, Indicates the error signal The variance of Represents a positive number, represents the momentum term coefficient, Displays weather data The amount of change, represents the regularization term coefficient, and n represents the current time step.

[0009] Specifically, the specific steps of step S2 include: S2.1: The remote monitoring center receives meteorological data from vehicle-mounted equipment ,in, represents the meteorological data received by the mth remote monitoring center, and m represents the number of meteorological data received by the remote monitoring center; S2.2: Load the pre-built deep learning model, collect historical meteorological data and vehicle equipment failure records, and use the historical meteorological data and vehicle equipment failure records to train the pre-built deep learning model to generate a meteorological failure prediction model; S2.3: Input the meteorological fault prediction model to perform meteorological fault prediction and obtain the meteorological fault prediction result; S2.4: Yes Perform time series analysis and use anomaly detection algorithms to detect Combine the abnormal values ​​or abnormal patterns in the data with the meteorological data and the real-time monitoring data of the on-board equipment to evaluate the operating status of the equipment and analyze the correlation between the failure of the on-board equipment and the meteorological conditions; S2.5: Use meteorological fault prediction results and time series analysis results to predict the likelihood and urgency of vehicle equipment failures, and determine the severity of the failure, possible impact range, and resources required for repair based on the fault prediction results; S2.6: According to the urgency of the fault prediction, a multi-level warning mechanism is triggered and warning information is sent.

[0010] Specifically, the specific steps of S2.4 include: S2.41: Use time series analysis algorithms to Analyze and obtain the changing trend of meteorological data. The formula is: ; in, represents the weighted moving average at time t, represents the time weight, represents a constant, and , k represents the window size of the moving average; S2.42: Detection using anomaly detection algorithms The anomaly detection algorithm formula is: ; Among them, M represents The mean value in express and The weight of represents the time decay function, represents the attenuation factor, represents a Boolean function; S2.43: Set the abnormal threshold h, like , then The corresponding data points in are taken as outliers; S2.44: Combine meteorological data with real-time monitoring data from onboard equipment and use machine learning models to analyze the relationship between meteorological data and equipment operating status; S2.45: Use the correlation analysis method to analyze the correlation between vehicle equipment failure and meteorological conditions, and identify the meteorological conditions that cause equipment failure. The formula is: ; Where g represents the correlation between vehicle equipment failure and meteorological conditions, and express Two different data points.

[0011] Specifically, the specific steps of step S3 include: S3.1: Collect historical fault data, obtain real-time meteorological data, and pre-process the historical fault data and real-time meteorological condition data, wherein the historical fault data includes the fault type, location, cause, and meteorological conditions at the time; S3.2: Extract the fault type, location and cause characteristics from the historical fault data, and extract the abnormal meteorological characteristics from the meteorological data, and form the meteorological fault characteristics through integration. The abnormal meteorological characteristics include extreme temperature, humidity change, and abnormal wind speed; S3.3: Load the pre-built ensemble learning model, and train the ensemble learning model using the meteorological fault features to obtain a trained ensemble learning model; S3.4: Match the real-time meteorological data with the extracted meteorological fault features, and use the trained ensemble learning model to perform fault diagnosis on the matching results to generate a preliminary judgment of the fault type, location and potential cause; S3.5: Based on the fault diagnosis results, according to the fault type and historical experience, an optimization strategy is selected from a preset optimization strategy library to automatically generate an optimal troubleshooting solution; S3.6: Perform simulation verification on the generated troubleshooting solution, and adjust and optimize the solution according to the simulation results to form a final optimal troubleshooting solution.

[0012] Specifically, the meteorological data in step S1 includes visibility, ozone concentration, ultraviolet radiation intensity, temperature, humidity, wind speed, wind direction, air pressure, and precipitation.

[0013] A remote fault monitoring system for vehicle-mounted mobile meteorological monitoring equipment, including: a data processing module, a fault identification module, a fault diagnosis module, and a blockchain recording module; The data processing module is used to collect meteorological data from the vehicle-mounted equipment in real time and pre-process the collected meteorological data; The fault identification module is used to receive pre-processed meteorological data, identify potential fault modes through deep learning models, evaluate the operating status of on-board equipment in real time, predict the possibility and urgency of faults, and trigger corresponding early warning mechanisms; The fault diagnosis module is used to combine historical fault data and current meteorological conditions to comprehensively determine the fault type, specific location and potential cause, automatically generate the optimal fault elimination plan, and perform simulation verification; The blockchain recording module is used to record fault information using blockchain technology, establish an intelligent feedback system, integrate feedback and fault handling results into a deep learning model, optimize model performance, and formulate a preventive maintenance plan.

[0014] Specifically, the fault identification module includes: a deep learning model unit, a time series analysis unit, an anomaly detection unit, and an early warning mechanism unit; The deep learning model unit is used to perform in-depth analysis on the preprocessed meteorological data using the deep learning model to identify potential fault modes in the data; The time series analysis unit is used to perform time series analysis on meteorological data to reveal the laws and trends of data changes over time; The anomaly detection unit is used to combine the output of time series analysis and deep learning models to identify abnormal data, evaluate the operating status of the equipment, and predict the possibility and urgency of the occurrence of failures; The early warning mechanism unit is used to trigger a multi-level early warning mechanism according to the urgency of the fault prediction.

[0015] Specifically, the fault diagnosis module includes: a fault diagnosis unit, a solution optimization unit, and a simulation verification unit; The fault diagnosis unit is used to comprehensively determine the fault type, specific location and potential cause by using an integrated learning method; The solution optimization unit is used to introduce an optimization strategy and automatically generate an optimal troubleshooting solution based on the fault diagnosis result; The simulation verification unit is used to perform simulation verification on the generated troubleshooting solution, and evaluate the feasibility and effectiveness of the solution by simulating the actual operating environment and fault scenarios.

[0016] Compared with the prior art, the present invention has the following beneficial effects: The present invention proposes a remote fault monitoring method for vehicle-mounted mobile meteorological monitoring equipment. By real-time collection and preprocessing of meteorological data, combined with deep learning models, time series analysis and anomaly detection algorithms, it can accurately identify potential fault modes and evaluate the operating status of vehicle-mounted equipment in real time. This not only improves the accuracy and timeliness of fault prediction, but also effectively reduces the risk and loss of fault occurrence through a multi-level early warning mechanism. At the same time, the integrated learning method is used to comprehensively judge the fault type, specific location and potential causes, providing strong support for the rapid location and resolution of faults.

[0017] The present invention proposes a remote fault monitoring method for vehicle-mounted mobile meteorological monitoring equipment. By adopting the blockchain method to record fault information, the data is ensured to be tamper-proof and traceable, providing reliable data support for fault analysis and processing. At the same time, an intelligent feedback system is established to integrate the feedback and fault processing results into a deep learning model, thereby realizing continuous optimization and improvement of the model. This not only improves the efficiency and accuracy of fault processing, but also helps to extend the service life of the vehicle-mounted equipment and reduce maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A schematic diagram of a remote fault monitoring method for a vehicle-mounted mobile meteorological monitoring device according to the present invention; Figure 2 This is a principle flow chart of a remote fault monitoring method for vehicle-mounted mobile meteorological monitoring equipment according to the present invention; Figure 3 It is a process flow chart of triggering a multi-level early warning mechanism by a remote fault monitoring method for vehicle-mounted mobile meteorological monitoring equipment of the present invention; Figure 4 This is an architecture diagram of the remote fault monitoring system used for vehicle-mounted mobile meteorological monitoring equipment in the present invention. DETAILED DESCRIPTION

[0019] See also Figure 1-Figure 3 , an embodiment provided by the present invention: a remote fault monitoring method for a vehicle-mounted mobile meteorological monitoring device, comprising the following steps: Step S1: The vehicle-mounted equipment collects meteorological data in real time through sensors, pre-processes the collected meteorological data using an adaptive filtering strategy, and sends the pre-processed meteorological data to a remote monitoring center; The meteorological data in step S1 include visibility, ozone concentration, ultraviolet radiation intensity, temperature, humidity, wind speed, wind direction, air pressure, and precipitation.

[0020] Step S2: The remote monitoring center uses a deep learning model to identify potential fault modes in the preprocessed meteorological data received, combines time series analysis and anomaly detection algorithms, and evaluates the operating status of the on-board equipment in real time. Based on the evaluation results, the possibility and urgency of the fault are predicted, and a multi-level early warning mechanism is triggered based on the urgency of the fault prediction. Step S3: Combine historical fault data and current meteorological conditions, use ensemble learning methods to comprehensively judge the fault type, specific location and potential causes, and introduce optimization strategies to automatically generate the optimal troubleshooting solution based on the diagnosis results, and perform simulation verification on the optimal troubleshooting solution; Step S4: Use blockchain method to record the time, location, cause, handling method and result information of the fault. At the same time, establish an intelligent feedback system, integrate the feedback and fault handling results into the deep learning model, and formulate a preventive maintenance plan based on the optimized deep learning model.

[0021] Furthermore, the specific steps of step S4 include: (1) Blockchain system construction and configuration Select Hyperledger Fabric and configure the blockchain network, including node settings and consensus mechanism selection. At the same time, design smart contracts to record fault information, including time, location, cause, handling method and result. (2) Fault information recording process When a fault occurs, fault-related information is collected through manual input or automatic sensors and submitted to the blockchain network for verification and recording through smart contracts, ensuring that once the fault information is recorded on the blockchain, it cannot be tampered with or deleted; (3) Establishment of intelligent feedback system Design the architecture of the intelligent feedback system, including data collection, processing, analysis and feedback. Use deep learning models to process and analyze fault data and extract useful information. Feedback the analysis results to relevant personnel or systems to improve the fault handling process or optimize equipment performance. (4) Deep learning model optimization Collect fault handling results and feedback data for training and optimizing deep learning models. Use ensemble learning methods or other optimization algorithms to improve the accuracy and generalization ability of the models. At the same time, continuously iterate and optimize the models to adapt to new fault types and handling methods. (5) Development of preventive maintenance plan Based on the optimized deep learning model, the types and probabilities of possible future failures are predicted. According to the prediction results, a preventive maintenance plan is formulated, including maintenance time, content, and required resources. The preventive maintenance plan is incorporated into the equipment management system or production plan to ensure that it is executed as planned.

[0022] Among them, the smart contract is used to record fault information, and its logic includes: (1) Verify the integrity and authenticity of fault information; (2) Store the fault information on the blockchain and generate a unique transaction ID; (3) Ensure that once fault information is recorded on the blockchain, it cannot be tampered with or deleted.

[0023] The preventive maintenance planning algorithm includes: (1) Determine the type and probability of future failures based on the prediction results of the deep learning model; (2) Calculate maintenance priority and sequence based on failure type and probability; (3) Develop a preventive maintenance plan based on maintenance priorities and sequence, including maintenance time, content, and required resources.

[0024] The specific steps of the adaptive filtering strategy in step S1 include: S1.1: Obtain meteorological data and store it as time series meteorological data ,in, represents the mth time series meteorological data, and m represents the number of time series meteorological data; Furthermore, the specific process of storing meteorological data as time series meteorological data includes: (1) Determine the required meteorological data type, obtain meteorological data, and clean the acquired meteorological data, including removing duplicate data, processing missing values, and processing outliers; (2) Store the preprocessed meteorological data in CSV; (3) Use Python to convert the date and time into a time series format and store the processed time series data in the database.

[0025] S1.2: Set the initial parameters of the filter according to the current filter output and the expected output , calculate the error signal ,in, represents the current filter coefficient vector, represents the meteorological data vector input to the filter, and T represents transpose; S1.3: According to the error signal and the input signal, the coefficient of the filter is updated using the adaptive filtering algorithm, and the updated filter is used to filter the meteorological data. The formula is: ; in, represents the updated filter coefficient vector, represents the dynamic learning rate adjustment factor, represents the step size factor, Indicates the error signal The variance of Represents a positive number, represents the momentum term coefficient, Displays weather data The amount of change, represents the regularization term coefficient, and n represents the current time step.

[0026] It should be noted that the formula in the present invention introduces the adjustment factor , so that the weight update step can be adjusted dynamically, thereby improving the adaptability and flexibility of the algorithm; secondly, the error term With the input meteorological data vector and its variation The product of not only considers the current input signal, but also takes into account the changing trend of the signal, which helps the algorithm to capture data features more accurately; finally, by introducing the regularization term , which effectively prevents the model from overfitting and improves the generalization ability of the algorithm.

[0027] The specific steps of step S2 include: S2.1: The remote monitoring center receives meteorological data from vehicle-mounted equipment ,in, represents the meteorological data received by the mth remote monitoring center, and m represents the number of meteorological data received by the remote monitoring center; S2.2: Load the pre-built deep learning model, collect historical meteorological data and vehicle equipment failure records, and use the historical meteorological data and vehicle equipment failure records to train the pre-built deep learning model to generate a meteorological failure prediction model; Furthermore, the specific steps of S2.2 include: (1) Collect historical meteorological data to ensure that the data is complete, accurate, and contains the required meteorological parameters. At the same time, collect vehicle equipment fault records, including the time, location, fault type, maintenance records, and other information of the fault, and pre-process the data, including data cleaning, missing value processing, data standardization, etc., to ensure data quality; (2) Pre-constructing a deep learning model, which should be suitable for processing meteorological data and predicting equipment failures. At the same time, loading the model architecture and weights and preparing for training. The deep learning model is the prior art content in this field and is not the inventive solution of this application, and will not be described in detail here; (3) Divide the historical meteorological data and vehicle equipment fault records into a training set and a validation set. The training set is used to train the deep learning model, and the validation set is used to evaluate the performance of the deep learning model. (4) Using the training set data to train the deep learning model, setting training parameters, such as learning rate, batch size, and number of training rounds. At the same time, during the training process, using a loss function to measure the difference between the model prediction results and the actual results, and adjusting the model parameters through an optimizer to reduce the loss. Among them, the loss function and optimizer in the deep learning model are the prior art content in this field, and are not the inventive solution of this application, and are not elaborated here; (5) Use the validation set data to evaluate the performance of the deep learning model and adjust the deep learning model parameters based on the evaluation results; (6) Based on the evaluation results, fine-tune the deep learning model to improve prediction accuracy; (7) When the performance of the deep learning model meets the requirements, it is saved as a meteorological fault prediction model.

[0028] S2.3: Input the meteorological fault prediction model to perform meteorological fault prediction and obtain the meteorological fault prediction result; Furthermore, the specific steps of S2.3 include: (1) Loading a trained meteorological fault prediction model, where the meteorological fault prediction model should be able to receive meteorological data as input and output meteorological fault prediction results; (2) inputting the preprocessed meteorological data into the meteorological fault prediction model, wherein the method is implemented through an interface provided by the meteorological fault prediction model; (3) After receiving meteorological data, the meteorological fault prediction model performs internal calculations, including data feature extraction, application of model parameters, etc., to generate meteorological fault prediction results and output the prediction results. This result may include information such as the type, probability, and severity of the meteorological fault.

[0029] S2.4: Yes Perform time series analysis and use anomaly detection algorithms to detect Abnormal values ​​or abnormal patterns in the data are combined with meteorological data and real-time monitoring data of on-board equipment to evaluate the operating status of the equipment. At the same time, the correlation between on-board equipment failures and meteorological conditions is analyzed; S2.5: Use meteorological fault prediction results and time series analysis results to predict the likelihood and urgency of vehicle equipment failures, and determine the severity of the failure, possible impact range, and resources required for repair based on the fault prediction results; Furthermore, the specific steps of S2.5 include: (1) Integrate meteorological fault prediction results and time series analysis results to form a comprehensive fault prediction data set; (2) Use a logistic regression model to analyze the integrated fault prediction data set to predict the possibility of vehicle-mounted equipment failure, that is, the probability of failure, and evaluate the urgency of the failure based on the probability of failure and the severity of the failure mode. The logistic regression model is a prior art in this field and is not an inventive solution of the present application, and is not described in detail here. (3) Analyze the historical records of vehicle-mounted equipment failures to understand the impact of failure modes on equipment performance, safety, and reliability. At the same time, combine the current operating status of the equipment and real-time monitoring data to predict the possible impact of the failure; (4) Based on the type and severity of the fault, evaluate the resources required to repair the fault, including human, material and financial resources, and output the prediction results and evaluation results in the form of visualization or report so that relevant personnel can understand the operating status of the equipment and the risk of failure.

[0030] S2.6: According to the urgency of the fault prediction, a multi-level warning mechanism is triggered and warning information is sent.

[0031] The specific steps of S2.4 include: S2.41: Use time series analysis algorithms to Analyze and obtain the changing trend of meteorological data. The formula is: ; in, represents the weighted moving average at time t, represents the time weight, represents a constant, and ,when When , the difference between weights is small and the smoothness is high. When , the difference between weights is large and the degree of smoothness is low. k represents the window size of the moving average, that is, how many past time points are considered to calculate the weighted moving average of the current time point; S2.42: Detection using anomaly detection algorithms The anomaly detection algorithm formula is: ; Among them, M represents The mean value in express and The weight of represents the time decay function, represents the attenuation factor, represents a Boolean function; S2.43: Set the abnormal threshold h, like , then The corresponding data points in are taken as outliers; S2.44: Combine meteorological data with real-time monitoring data from onboard equipment and use machine learning models to analyze the relationship between meteorological data and equipment operating status; S2.45: Use the correlation analysis method to analyze the correlation between vehicle equipment failure and meteorological conditions, and identify the meteorological conditions that cause equipment failure. The formula is: ; Where g represents the correlation between vehicle equipment failure and meteorological conditions, and express Two different data points.

[0032] It should be noted that the formula in the present invention introduces weights , so that different data points have different importance in the correlation calculation, which improves the flexibility and accuracy of the analysis. Secondly, the time decay function in the formula Taking into account the characteristics of time series data, the contribution of older data points to the correlation calculation gradually decreases, which helps to reflect the trend of data changes over time and makes the analysis results more in line with reality. At the same time, this time decay processing method also helps to reduce the deviation caused by outdated data.

[0033] The specific steps of step S3 include: S3.1: Collect historical fault data, obtain real-time meteorological data, and pre-process the historical fault data and real-time meteorological condition data, wherein the historical fault data includes the fault type, location, cause, and meteorological conditions at the time; S3.2: Extract the fault type, location and cause characteristics from the historical fault data, and extract the abnormal meteorological characteristics from the meteorological data, and form the meteorological fault characteristics through integration, wherein the abnormal meteorological characteristics include extreme temperature, humidity change, and abnormal wind speed, wherein a data mining algorithm is used in the feature extraction, and the data mining algorithm is the prior art content in this field, and is not the inventive solution of the present application, and will not be described in detail here; S3.3: Loading a pre-built integrated learning model, and using the meteorological fault characteristics to train the integrated learning model to obtain a trained integrated learning model, wherein the construction and training process of the integrated learning model is the prior art content in this field, and is not the inventive solution of the present application, and will not be described in detail here; S3.4: Match the real-time meteorological data with the extracted meteorological fault features, and use the trained ensemble learning model to perform fault diagnosis on the matching results to generate a preliminary judgment of the fault type, location and potential cause; Furthermore, the specific process of matching the real-time meteorological data with the extracted meteorological fault features includes: (1) Obtaining real-time meteorological data: Collecting meteorological data in real time through meteorological sensor networks and meteorological stations to ensure the accuracy, completeness and timeliness of real-time meteorological data; (2) Extract meteorological characteristics related to the fault type, location, and potential causes from historical meteorological data and fault records. These characteristics may include specific thresholds or changing trends of meteorological elements such as visibility, ozone concentration, ultraviolet radiation intensity, temperature, humidity, wind speed, wind direction, and air pressure; (3) Unify the data formats of real-time meteorological data and meteorological fault features, including data type, data precision, and timestamp format. If the dimension of real-time meteorological data is inconsistent with the dimension of meteorological fault features, the dimension needs to be adjusted. For example, if the meteorological fault features are based on hourly data and the real-time meteorological data is based on minute-level data, the real-time meteorological data needs to be aggregated or interpolated to match the dimension of the meteorological fault features. (4) Determine a time window based on the time characteristics of the meteorological fault characteristics, where the size of the time window depends on the timeliness of the meteorological fault characteristics and the update frequency of the real-time meteorological data; (5) Align the timestamp of the real-time meteorological data with the time window of the meteorological fault feature. If the timestamp of the real-time meteorological data is not within the time window of the meteorological fault feature, it may be necessary to perform time interpolation or discard the data point; (6) Convert real-time meteorological data into feature vectors, where each feature corresponds to the value of a meteorological element; (7) Calculating the similarity between the feature vector of the real-time meteorological data and the meteorological fault feature, wherein the similarity calculation method adopts the Euclidean distance method. The Euclidean distance method is the prior art content in this field and is not the inventive solution of the present application, and is not described in detail here; (8) Based on the similarity calculation results, determine whether the real-time meteorological data matches the meteorological fault characteristics. If the similarity exceeds the preset threshold, it is considered that the real-time meteorological data matches the meteorological fault characteristics successfully; (9) The real-time meteorological data of successful or failed matching and its corresponding meteorological fault characteristics shall be recorded, and subsequent meteorological fault diagnosis, early warning or decision support shall be carried out based on the matching results.

[0034] S3.5: Based on the fault diagnosis results, according to the fault type and historical experience, select the optimization strategy from the preset optimization strategy library to automatically generate the optimal troubleshooting plan, where the optimal troubleshooting plan should clearly specify the specific operation steps, required resources, time requirements and expected results; Furthermore, the specific process of selecting an optimization strategy from the preset optimization strategy library includes: (1) Identify the specific problems or fault types currently faced, as well as the goals or effects expected to be achieved by selecting optimization strategies; (2) Be familiar with the contents of the preset optimization strategy library, including the names, applicable scopes, advantages and disadvantages, and historical application of various strategies; Among them, the optimization strategy library is a database or knowledge base that brings together a variety of optimization methods and techniques, aiming to provide a series of verified and screened strategies for solving specific problems or achieving specific goals. These strategies are usually based on historical experience, professional knowledge, industry best practices, and data analysis, etc., and are intended to help decision makers quickly find the most suitable solution for the current situation. In the present invention, the preset optimization strategy library includes optimization algorithms such as genetic algorithms and simulated annealing algorithms; optimization strategies for business processes, production processes, service processes, etc.; including performance optimization strategies for computer systems, network systems, database systems, etc., and the preset optimization strategy library is a dynamically updated system. With the emergence of new technologies, new methods, and new experiences, new strategies will be continuously added and outdated strategies will be eliminated.

[0035] (3) Select suitable candidate strategies from the optimization strategy library based on the specific nature of the problem or fault and the target requirements; (4) Evaluate each of the selected candidate strategies one by one, considering factors such as their feasibility, effectiveness, cost-effectiveness, and potential risks in practical applications; (5) Based on the evaluation results, select the optimization strategy that best suits the current problem or fault. This may require comprehensive consideration of multiple factors, such as the strategy’s long-term effects, short-term benefits, and resource investment. (6) Based on the selected strategy, develop the optimal troubleshooting plan, including specific operation steps, time schedule, resource allocation, and expected results.

[0036] S3.6: Perform simulation verification on the generated troubleshooting solution, and adjust and optimize the solution according to the simulation results to form a final optimal troubleshooting solution.

[0037] See also Figure 4 Another embodiment provided by the present invention is a remote fault monitoring system for a vehicle-mounted mobile meteorological monitoring device, comprising: Data processing module, fault identification module, fault diagnosis module, blockchain recording module; The data processing module is used to collect meteorological data from the vehicle-mounted equipment in real time and pre-process the collected meteorological data for subsequent analysis and evaluation; The fault identification module is used to receive pre-processed meteorological data, identify potential fault modes through deep learning models, evaluate the operating status of on-board equipment in real time, predict the possibility and urgency of faults, and trigger corresponding early warning mechanisms; The fault diagnosis module is used to combine historical fault data and current meteorological conditions to comprehensively determine the fault type, specific location and potential causes, automatically generate the optimal fault elimination plan, and perform simulation verification; The blockchain recording module is used to record fault information using blockchain technology, establish an intelligent feedback system, integrate feedback and fault handling results into the deep learning model, optimize model performance, and formulate a preventive maintenance plan.

[0038] The data processing module includes: a sensor unit, an adaptive filtering unit, and a data transmission unit; Sensor unit, including high-precision, multi-parameter meteorological sensors for real-time collection of meteorological data such as visibility, ozone concentration, ultraviolet radiation intensity, temperature, humidity, wind speed, wind direction, air pressure, precipitation, etc.; An adaptive filtering unit is used to pre-process the collected meteorological data using an adaptive filtering strategy to remove noise and outliers, retain valid data, and improve data quality; The data transmission unit is used to send the pre-processed meteorological data to the remote monitoring center through a stable communication network to ensure the real-time and accuracy of the data.

[0039] The fault identification module includes: deep learning model unit, time series analysis unit, anomaly detection unit, and early warning mechanism unit; A deep learning model unit, which is used to perform in-depth analysis on the pre-processed meteorological data using a deep learning model to identify potential failure modes in the data; Time series analysis unit, used to perform time series analysis on meteorological data to reveal the laws and trends of data changes over time; Anomaly detection unit, which combines the output of time series analysis and deep learning models to identify abnormal data, evaluate the operating status of equipment, and predict the possibility and urgency of failures; The early warning mechanism unit is used to trigger a multi-level early warning mechanism according to the urgency of the fault prediction, including instant alarms, SMS notifications, email reminders, etc., to ensure that relevant personnel can respond quickly.

[0040] The fault diagnosis module includes: a fault diagnosis unit, a solution optimization unit, and a simulation verification unit; Fault diagnosis unit, used to comprehensively determine the fault type, specific location and potential cause using ensemble learning methods; Solution optimization unit, used to introduce optimization strategies and automatically generate the best troubleshooting solution based on the fault diagnosis results; The simulation verification unit is used to simulate and verify the generated troubleshooting solution, and evaluate the feasibility and effectiveness of the solution by simulating the actual operating environment and fault scenarios.

[0041] The blockchain recording module includes: blockchain recording unit, intelligent feedback unit, model optimization unit, and preventive maintenance unit; Blockchain recording unit, which uses blockchain technology to record the time, location, cause, treatment method and result information of the fault, ensuring the data cannot be tampered with and is traceable; Intelligent feedback unit, which is used to collect multi-dimensional feedback information such as on-site photos and videos, and use natural language processing technology to understand the semantics of the feedback information; Model optimization unit, which is used to integrate user feedback and fault handling results into the deep learning model, conduct continuous learning and model optimization, and improve the accuracy of fault diagnosis and the effectiveness of troubleshooting solutions; The preventive maintenance unit is used to predict future failure trends and potential risk points based on the optimized deep learning model and combined with big data analysis technology, and to formulate targeted preventive maintenance plans.

[0042] To summarize, the remote fault monitoring system of the present invention realizes comprehensive monitoring, fault diagnosis, solution optimization and preventive maintenance of vehicle-mounted mobile meteorological monitoring equipment through the collaborative work of the data processing module, fault identification module, fault diagnosis module, blockchain recording module and its subordinate units, thereby improving the stability and reliability of the equipment.

[0043] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0044] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0045] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in the field may also change, modify, replace and modify the above-mentioned embodiments without departing from the purpose and scope of protection of the present invention, and all of these are within the protection of the present invention.

Claims

1. A remote fault monitoring method for a vehicle-mounted mobile meteorological monitoring device, characterized in that: include: Step S1: The vehicle-mounted equipment collects meteorological data in real time through sensors, pre-processes the collected meteorological data using an adaptive filtering strategy, and sends the pre-processed meteorological data to a remote monitoring center; Step S2: The remote monitoring center uses a deep learning model to identify potential fault modes in the preprocessed meteorological data received, combines time series analysis and anomaly detection algorithms, and evaluates the operating status of the on-board equipment in real time. Based on the evaluation results, the possibility and urgency of the fault are predicted, and a multi-level early warning mechanism is triggered based on the urgency of the fault prediction. Step S3: Combine historical fault data and current meteorological conditions, use ensemble learning methods to comprehensively judge the fault type, specific location and potential causes, and introduce optimization strategies to automatically generate the optimal troubleshooting solution based on the diagnosis results, and perform simulation verification on the optimal troubleshooting solution; Step S4: Use blockchain method to record the time, location, cause, handling method and result information of the fault. At the same time, establish an intelligent feedback system, integrate the feedback and fault handling results into the deep learning model, and formulate a preventive maintenance plan based on the optimized deep learning model.

2. The remote fault monitoring method for vehicle-mounted mobile meteorological monitoring equipment according to claim 1, characterized in that: The specific steps of the adaptive filtering strategy in step S1 include: S1.1: Obtain meteorological data and store it as time series meteorological data ,in, represents the mth time series meteorological data, and m represents the number of time series meteorological data; S1.2: Set the initial parameters of the filter according to the current filter output and the expected output , calculate the error signal ,in, represents the current filter coefficient vector, represents the meteorological data vector input to the filter, and T represents transpose; S1.3: According to the error signal and the input signal, the coefficient of the filter is updated using the adaptive filtering algorithm, and the updated filter is used to filter the meteorological data. The formula is: ； in, represents the updated filter coefficient vector, represents the dynamic learning rate adjustment factor, represents the step size factor, Indicates the error signal The variance of Represents a positive number, represents the momentum term coefficient, Displays weather data The amount of change, represents the regularization term coefficient, and n represents the current time step.

3. The remote fault monitoring method for vehicle-mounted mobile meteorological monitoring equipment according to claim 2, characterized in that: The specific steps of step S2 include: S2.1: The remote monitoring center receives meteorological data from vehicle-mounted equipment ,in, represents the meteorological data received by the mth remote monitoring center, and m represents the number of meteorological data received by the remote monitoring center; S2.2: Load the pre-built deep learning model, collect historical meteorological data and vehicle equipment failure records, and use the historical meteorological data and vehicle equipment failure records to train the pre-built deep learning model to generate a meteorological failure prediction model; S2.3: Input the meteorological fault prediction model to perform meteorological fault prediction and obtain the meteorological fault prediction result; S2.4: Yes Perform time series analysis and use anomaly detection algorithms to detect Abnormal values ​​or abnormal patterns in the data are combined with meteorological data and real-time monitoring data of on-board equipment to evaluate the operating status of the equipment. At the same time, the correlation between on-board equipment failures and meteorological conditions is analyzed; S2.5: Use meteorological fault prediction results and time series analysis results to predict the likelihood and urgency of vehicle equipment failures, and determine the severity of the failure, possible impact range, and resources required for repair based on the fault prediction results; S2.6: According to the urgency of the fault prediction, a multi-level warning mechanism is triggered and warning information is sent.

4. The remote fault monitoring method for vehicle-mounted mobile meteorological monitoring equipment according to claim 3, characterized in that: The specific steps of S2.4 include: S2.41: Use time series analysis algorithms to Analyze and obtain the changing trend of meteorological data. The formula is: ； in, represents the weighted moving average at time t, represents the time weight, represents a constant, and , k represents the window size of the moving average; S2.42: Detection using anomaly detection algorithms The anomaly detection algorithm formula is: ； Among them, M represents The mean value in express and The weight of represents the time decay function, represents the attenuation factor, represents a Boolean function; S2.43: Set the abnormal threshold h, like , then The corresponding data points in are taken as outliers; S2.44: Combine meteorological data with real-time monitoring data from onboard equipment and use machine learning models to analyze the relationship between meteorological data and equipment operating status; S2.45: Use the correlation analysis method to analyze the correlation between vehicle equipment failure and meteorological conditions, and identify the meteorological conditions that cause equipment failure. The formula is: ； Where g represents the correlation between vehicle equipment failure and meteorological conditions, and express Two different data points.

5. The remote fault monitoring method for vehicle-mounted mobile meteorological monitoring equipment according to claim 4, characterized in that: The specific steps of step S3 include: S3.1: Collect historical fault data, obtain real-time meteorological data, and pre-process the historical fault data and real-time meteorological condition data, wherein the historical fault data includes the fault type, location, cause, and meteorological conditions at the time; S3.2: Extract the fault type, location and cause characteristics from the historical fault data, and extract the abnormal meteorological characteristics from the meteorological data, and form the meteorological fault characteristics through integration. The abnormal meteorological characteristics include extreme temperature, humidity change, and abnormal wind speed; S3.3: Load the pre-built ensemble learning model, and train the ensemble learning model using the meteorological fault features to obtain a trained ensemble learning model; S3.4: Match the real-time meteorological data with the extracted meteorological fault features, and use the trained ensemble learning model to perform fault diagnosis on the matching results to generate a preliminary judgment of the fault type, location and potential cause; S3.5: Based on the fault diagnosis results, according to the fault type and historical experience, an optimization strategy is selected from a preset optimization strategy library to automatically generate an optimal troubleshooting solution; S3.6: Perform simulation verification on the generated troubleshooting solution, and adjust and optimize the solution according to the simulation results to form a final optimal troubleshooting solution.

6. The remote fault monitoring method for vehicle-mounted mobile meteorological monitoring equipment according to claim 5, characterized in that: The meteorological data in step S1 include visibility, ozone concentration, ultraviolet radiation intensity, temperature, humidity, wind speed, wind direction, air pressure, and precipitation.

7. A remote fault monitoring system for vehicle-mounted mobile meteorological monitoring equipment, which is used to implement the remote fault monitoring method for vehicle-mounted mobile meteorological monitoring equipment according to any one of claims 1 to 6, characterized in that: include: Data processing module, fault identification module, fault diagnosis module, blockchain recording module; The data processing module is used to collect meteorological data from the vehicle-mounted equipment in real time and pre-process the collected meteorological data; The fault identification module is used to receive pre-processed meteorological data, identify potential fault modes through deep learning models, evaluate the operating status of on-board equipment in real time, predict the possibility and urgency of faults, and trigger corresponding early warning mechanisms; The fault diagnosis module is used to combine historical fault data and current meteorological conditions to comprehensively determine the fault type, specific location and potential causes, automatically generate the optimal fault elimination plan, and perform simulation verification; The blockchain recording module is used to record fault information using blockchain technology, establish an intelligent feedback system, integrate feedback and fault handling results into a deep learning model, optimize model performance, and formulate a preventive maintenance plan.

8. The remote fault monitoring system for vehicle-mounted mobile meteorological monitoring equipment according to claim 7, characterized in that: The fault identification module includes: a deep learning model unit, a time series analysis unit, an anomaly detection unit, and an early warning mechanism unit; The deep learning model unit is used to perform in-depth analysis on the preprocessed meteorological data using the deep learning model to identify potential fault modes in the data; The time series analysis unit is used to perform time series analysis on meteorological data to reveal the laws and trends of data changes over time; The anomaly detection unit is used to combine the output of time series analysis and deep learning models to identify abnormal data, evaluate the operating status of the equipment, and predict the possibility and urgency of the occurrence of failures; The early warning mechanism unit is used to trigger a multi-level early warning mechanism according to the urgency of the fault prediction.

9. The remote fault monitoring system for vehicle-mounted mobile meteorological monitoring equipment according to claim 8, characterized in that: The fault diagnosis module includes: a fault diagnosis unit, a solution optimization unit, and a simulation verification unit; The fault diagnosis unit is used to comprehensively determine the fault type, specific location and potential cause by using an integrated learning method; The solution optimization unit is used to introduce an optimization strategy and automatically generate an optimal troubleshooting solution based on the fault diagnosis result; The simulation verification unit is used to perform simulation verification on the generated troubleshooting solution, and evaluate the feasibility and effectiveness of the solution by simulating the actual operating environment and fault scenarios.

Images