A monitoring and early warning system for electric tricycle controller

Abstract

The present application provides a monitoring and early warning system for an electric tricycle controller, which relates to the field of intelligent early warning technology, including: an alarm database generation module, which is used to determine the monitoring control points of the electric tricycle, perform abnormal parameter control threshold mining, and generate an alarm database; a data preprocessing module, which is used to retrieve sensor monitoring data, perform data preprocessing, and integrate to determine effective monitoring data; an early warning signal determination module, which is used to analyze and predict abnormal control decisions and control trends based on the alarm database, and determine early warning signals; an early warning response module, which is used to analyze early warning signals and perform response processing, and determine early warning response lists; an operation control early warning management module, which is used to execute early warning response lists and perform operation control early warning management on electric tricycles. The present application solves the technical problem that the data processing capacity of the prior art is insufficient, resulting in low accuracy of early warnings, and achieves the technical effect of intelligently mining and analyzing monitoring data and performing accurate early warnings.

Description

Technical Field

[0001] The present application relates to the field of intelligent early warning technology, and in particular to a monitoring and early warning system for an electric tricycle controller. Background Art

[0002] As a convenient and economical means of transportation, electric tricycles have been widely used in urban transportation and rural areas. With the increase in the number of electric tricycles, their operating safety and stability have received increasing attention. Monitoring and early warning technology has become the key to ensuring the safety of electric tricycles.

[0003] Existing monitoring and early warning technologies mostly rely on simple sensors and fixed thresholds to judge abnormalities. The early warning effect of this method depends on the accuracy of the sensor and the setting of the threshold. However, the operating status of electric tricycles will vary significantly under different operating environments, such as different road conditions, climate conditions and vehicle loads. This makes the fixed threshold method limited in practical applications and difficult to adapt to complex and changing operating environments. At the same time, the existing monitoring and early warning technology lacks the ability to deeply process and analyze real-time data, and cannot fully utilize the information contained in the data to identify and warn potential safety hazards early, which greatly reduces the effectiveness of the early warning.

[0004] In summary, the existing electric tricycle monitoring and early warning technology has obvious deficiencies in terms of intelligence, environmental adaptability, data processing, etc., and it is difficult to meet the monitoring and early warning needs in complex and changeable operating environments. Summary of the invention

[0005] The present application provides a monitoring and early warning system for an electric tricycle controller, which solves the technical problem that the existing electric tricycle monitoring and early warning technology has insufficient data processing capabilities, is unable to deeply analyze monitoring data and flexibly adjust thresholds, and is difficult to make accurate early warnings in complex and changeable operating environments. It achieves the technical effect of deeply mining monitoring data and performing data preprocessing to generate adaptive abnormal control decisions and accurate early warning signals, improves the accuracy and adaptability of electric tricycle monitoring and early warning, and at the same time improves the safety of use and life of electric tricycles.

[0006] In view of the above problems, the present application provides a monitoring and early warning system for an electric tricycle controller, the system comprising: an alarm database generation module, the alarm database generation module is used to determine the monitoring control points of the electric tricycle, perform abnormal parameter control threshold mining, and generate an alarm database, the alarm database includes an independent threshold sequence and a collaborative threshold sequence; a data preprocessing module, the data preprocessing module is used to retrieve sensor monitoring data, transmit it back to a data processing model, perform data preprocessing and integrate to determine effective monitoring data, the sensor monitoring data is marked with a timestamp, and the data processing model includes a three-layer fully connected data correction layer, a heterogeneous conversion layer and a de-redundancy layer; an early warning signal determination module, the early warning signal determination module is used to train an adaptive decision module based on the alarm database, perform abnormal control decisions and control trend prediction on the effective monitoring data, and determine an early warning signal, the early warning signal has an impact compensation based on the operating environment; an early warning response module, the early warning response module is used to perform signal analysis and response processing on the early warning signal, and determine an early warning response list; an operation control early warning management module, the operation control early warning management module is used to execute the early warning response list and perform operation control early warning management on the electric tricycle.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] The alarm database generation module sets monitoring control points, conducts abnormal parameter control threshold mining, generates an alarm database, accurately monitors electric tricycles, and collects sensor monitoring data with timestamps. The data preprocessing module uses a data processing model to preprocess data and integrate and determine effective monitoring data to correct, convert, and remove redundancy to ensure the accuracy and effectiveness of monitoring data. The early warning signal determination module considers the impact of the operating environment and compensates for it, thereby improving the accuracy and adaptability of the early warning. The early warning response module can analyze and process the early warning signal, determine the early warning response list, and execute it through the operation control early warning management module to achieve timely control and early warning management of electric tricycles.

[0009] Through the coordinated processing of the above modules, in-depth analysis and mining of monitoring data, the system can intelligently identify and predict potential abnormal situations, and can adapt to different operating environment conditions, provide more accurate early warning and control decisions, ensure the safe operation of electric tricycles, and extend their service life.

[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 in accordance with the contents of the specification. 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 listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 A schematic structural diagram of a monitoring and early warning system for an electric tricycle controller provided in an embodiment of the present application.

[0012] Figure 2 A schematic diagram of a process for generating an alarm database in a monitoring and early warning system for an electric tricycle controller provided in an embodiment of the present application.

[0013] Figure 3 A schematic diagram of a process for determining effective monitoring data in a monitoring and early warning system for an electric tricycle controller provided in an embodiment of the present application.

[0014] Figure 4 A schematic diagram of a flow chart for determining a warning response column in a monitoring and warning system of an electric tricycle controller provided in an embodiment of the present application.

[0015] Explanation of the reference numerals: alarm database generation module 10 , data preprocessing module 20 , warning signal determination module 30 , warning response module 40 , operation control warning management module 50 . DETAILED DESCRIPTION

[0016] The embodiment of the present application provides a monitoring and early warning system for an electric tricycle controller, deeply mines monitoring data and performs data preprocessing to generate adaptive abnormal control decisions and accurate early warning signals. The existing electric tricycle monitoring and early warning technology has insufficient data processing capabilities and is unable to deeply analyze monitoring data and flexibly adjust thresholds, resulting in technical problems such as difficulty in making accurate early warnings in complex and changeable operating environments. The accuracy and adaptability of electric tricycle monitoring and early warning are improved, while also improving the safety and lifespan of electric tricycles.

[0017] like Figure 1 As shown, the embodiment of the present application provides a monitoring and early warning system for an electric tricycle controller, the system comprising:

[0018] The alarm database generation module 10 is used to determine the monitoring control points of the electric tricycle, perform abnormal parameter control threshold mining, and generate an alarm database. The alarm database includes an independent threshold sequence and a collaborative threshold sequence.

[0019] Specifically, monitoring control points are key locations that need to be monitored during the operation of electric tricycles, such as battery voltage, motor temperature, vehicle speed, etc. Abnormal parameter control thresholds are parameter control thresholds between normal operation and abnormal conditions determined by mining and analyzing historical data. These thresholds are determined based on the differences between normal and abnormal conditions in historical data, and they help the system identify and predict possible abnormal conditions. The alarm database is a database containing abnormal parameter control thresholds, which provides data support for the safety monitoring system of electric tricycles. The alarm database includes two sequences: independent threshold sequence and collaborative threshold sequence. Independent threshold sequence refers to the abnormal control standard of a single monitoring point, which is used for anomaly detection of a single control point; collaborative threshold sequence is a control standard that is interrelated between multiple monitoring points, which is used for joint anomaly detection of multiple control points.

[0020] The alarm database generation module 10 first collects the operating data of each monitoring point of the electric tricycle through sensors and data acquisition systems. These sensors include speed sensors, temperature sensors, battery voltage sensors, GPS positioning sensors and acceleration sensors, etc., which are installed at the corresponding monitoring positions of the electric tricycle. The data acquisition system has a built-in electric tricycle control system, and uses a real-time data acquisition device or a wireless data transmission system to transmit the collected sensor data to the central control system for processing. After collecting the operating data, the data are analyzed using a data mining algorithm to determine the normal operating range and abnormal threshold of each monitoring point. The data mining algorithm can use a clustering algorithm (such as a K-means algorithm), a classification algorithm (such as a decision tree algorithm) and an anomaly detection algorithm (such as an isolated forest algorithm). For example, the K-means algorithm can be used to cluster sensor data under different operating states to find the data distribution under different states, thereby determining the parameter range of normal operation; or the isolated forest algorithm can be used to detect abnormal points in the operating data and identify the parameter threshold of abnormal operation. Then these thresholds are classified and sorted according to independent and collaborative relationships. The independent threshold refers to the abnormal control standard of a single monitoring point, such as a battery voltage that is too low or a motor temperature that is too high. The collaborative threshold refers to the control standard that is interrelated between multiple monitoring points. For example, when an electric tricycle is running at high speed, if the battery voltage drops and the temperature rises at the same time, it may indicate that there is a problem with the battery or the motor. These collaborative thresholds are determined by analyzing the data relationship between multiple monitoring points. An alarm database is constructed using the obtained independent threshold sequence and collaborative threshold sequence. This alarm database can be stored in the form of a relational database, including the identification of the monitoring point, independent threshold, collaborative threshold and related context information. For example, in a relational database, a table can be established, which contains fields such as monitoring point ID, sensor type, independent threshold upper limit, independent threshold lower limit, collaborative threshold condition, etc. In this way, in actual applications, the system can quickly retrieve and match monitoring data, issue early warning signals in time and take corresponding response measures.

[0021] The alarm database generation module 10 analyzes and mines the monitoring data of the electric tricycle, and finally generates a structured alarm database to provide basic data support for subsequent early warning signal determination and early warning response.

[0022] The data preprocessing module 20 is used to retrieve the sensor monitoring data, transmit it back to the data processing model, perform data preprocessing and integrate to determine the effective monitoring data. The sensor monitoring data is timestamped. The data processing model includes three layers of fully connected data correction layer, heterogeneous conversion layer and de-redundancy layer.

[0023] Specifically, the data preprocessing module 20 obtains the real-time sensor monitoring data of the electric tricycle through various sensors, such as speed sensors, battery status sensors, temperature sensors, etc. These sensor monitoring data are marked with timestamps to track the time sequence and relevance of the data. The acquired sensor monitoring data is transmitted to the data processing model for data preprocessing, useful monitoring data is screened out, invalid or duplicate data is removed, and the validity and integrity of the data are ensured. The data processing model is a model for processing and converting sensor monitoring data, which is embedded in the data preprocessing module 20 and includes a three-layer fully connected structure, namely, a data correction layer, a heterogeneous conversion layer, and a de-redundancy layer. The data correction layer is used to correct the data through a fully connected neural network to improve the accuracy of the data, for example, by identifying and correcting the abnormal points in the sensor data through an outlier detection algorithm, and using a smoothing algorithm to reduce the noise in the data. The heterogeneous conversion layer is responsible for standardizing and converting different types of sensor data so that various data formats and units are consistent, which is convenient for subsequent processing; at the same time, time synchronization processing can also be performed to ensure that all data points are on the same time basis. For example, voltage data may be recorded once per second, while speed data may be recorded once per minute. This layer synchronizes them to the same time scale. De-redundancy is used to identify and remove redundant information in the data through a fully connected neural network to reduce the amount of data and improve processing efficiency. For example, if multiple sensors record similar values ​​at the same time point, the system can identify and remove duplicate records and retain only one representative data.

[0024] The data preprocessing module 20 processes and integrates the real-time sensor monitoring data of the electric tricycle through the data processing model to determine the effective monitoring data. These data will be used in the subsequent analysis and decision-making process, ensuring the high quality and reliability of the data, thereby improving the monitoring and decision-making capabilities of the entire system.

[0025] The early warning signal determination module 30 is used to train an adaptive decision module based on the alarm database, make abnormal control decisions and control trend predictions on the effective monitoring data, and determine the early warning signal. The early warning signal has an impact compensation based on the operating environment.

[0026] Specifically, the adaptive decision module is a model that can dynamically adjust decision parameters according to real-time monitoring data, and can adapt to different operating environments and conditions. It is embedded in the early warning signal determination module 30. Abnormal control decision refers to a series of measures taken by the system when abnormal data is detected to prevent or mitigate possible failures or accidents, such as deceleration and stopping. The adaptive decision module establishes models for abnormal detection and trend prediction by learning threshold information in the alarm database. These models can be dynamically adjusted to continuously optimize and update their own decision criteria based on new input data.

[0027] The warning signal determination module 30 extracts the threshold information of various monitoring parameters from the alarm database, including independent thresholds and collaborative thresholds. Using these threshold information, the adaptive decision module is trained using a machine learning algorithm so that it can accurately identify abnormal situations and predict future trends. The training process can use a variety of machine learning algorithms, such as decision trees, random forests or neural networks, to improve the accuracy and robustness of the model. At the same time, the adaptive decision module uses time series analysis methods, such as ARIMA models, moving averages, etc., to perform trend analysis on historical monitoring data and identify possible trend changes; based on the analysis results, predict abnormal situations that may occur in the future. For example, by analyzing the battery voltage data of the past month, the system identifies the trend of the voltage gradually decreasing, and predicts that the voltage may drop below the critical value within the next week, and the system will issue a warning signal in advance. The effective monitoring data obtained by the data preprocessing module 20 is input into the adaptive decision module for real-time analysis to determine whether there is an abnormal situation and make corresponding control decisions. For example, if the speed data exceeds the preset threshold in the alarm database, the system will immediately identify the data as abnormal, and issue an alarm or execute corresponding control decisions, such as automatically reducing the vehicle speed. At the same time, future control trends are predicted and possible trend changes are identified.

[0028] Based on the abnormal control decision and control trend prediction results, the warning signal determination module 30 generates corresponding warning signals, which include specific abnormality types, warning levels and recommended countermeasures. At the same time, considering that different environmental conditions, such as temperature, humidity, road conditions, etc., may affect the sensor data and thus affect the accuracy of abnormality detection, the warning signal determination module 30 will also compensate the warning signal according to the environmental data. For example, in a high temperature environment, the battery temperature threshold may need to be adjusted appropriately to avoid false alarms.

[0029] The warning signal determination module 30 can monitor the running status of the electric tricycle in real time and make adaptive decisions based on the monitoring data and the information in the alarm database. The determination of the warning signal takes into account the influence of the running environment, thereby improving the practicality and reliability of the warning and ensuring the safe operation of the electric tricycle.

[0030] The early warning response module 40 is used to perform signal analysis and response processing on the early warning signal to determine the early warning response list.

[0031] Specifically, the warning response column refers to a record containing response measures for the warning signal, including the type of response, the time and method of execution, etc. The warning response module 40 receives the warning signal determined by the warning signal determination module 30, analyzes the warning signal, and identifies the type, source and severity of the warning signal. For example, if the warning signal is caused by an abnormal battery voltage, the module will check the voltage value, change trend and possible causes, such as over-discharge or ambient temperature influence. According to the results of the signal analysis, the warning response module 40 module formulates corresponding response measures. The response measures can be automated measures such as vehicle deceleration or manual measures such as reminding the operator to check the vehicle. The specific measures depend on the type and severity of the signal. For example, for a slight voltage abnormality, the system may only issue a warning; while for a serious abnormality, emergency measures may be triggered, such as immediate parking and inspection. The signal analysis and response processing results are recorded in the warning response column and sent to the operation control warning management module 50. Exemplarily, the early warning response module 40 receives an early warning signal of temperature exceeding the standard, analyzes the early warning signal, determines the severity of the temperature exceeding the standard, checks the rate of temperature rise and the current ambient temperature, and generates response processing measures according to the analysis results as follows: Decide to decelerate immediately to reduce the motor load, and notify the operator to check. If the temperature is very high, the operator may be required to stop the vehicle immediately for inspection. The response processing measures are recorded as a single early warning response column.

[0032] The early warning response module 40 analyzes each early warning signal, responds to and processes the early warning signal in a timely manner, and generates a corresponding early warning response list, thereby effectively managing the safety and operating status of the electric tricycle and improving the reliability and safety of the system.

[0033] The operation control warning management module 50 is used to execute the warning response sequence and perform operation control warning management on the electric tricycle.

[0034] Specifically, the operation control warning management module 50 is connected to the control system of the electric tricycle and can directly execute or issue instructions. For example, it communicates with the ECU (electronic control unit) of the vehicle through the CAN bus protocol to control the speed and braking system of the vehicle. The operation control warning management module 50 can also interact with the user through buttons on the steering wheel and menus on the display screen. The operation control warning management module 50 first receives the effective monitoring data and warning signals provided by the data preprocessing module 20 and the warning signal determination module 30, and then issues instructions to the vehicle control system through the interface with the electric tricycle control system according to the specific instructions in the warning response column, and monitors the execution effect in real time, dynamically adjusts the control strategy according to the actual operation situation, and ensures the safe and stable operation of the vehicle. For example, when it is detected that the battery voltage is lower than the safety threshold, the operation control warning management module 50 will start the deceleration program, turn off non-essential loads, switch to the backup power supply, and display a battery voltage abnormality warning on the vehicle display screen, and notify the driver through the buzzer and mobile phone APP; if it is detected that the motor temperature exceeds the set threshold, the module will automatically reduce the motor load, start the cooling system, and prompt the display screen that the temperature is too high.

[0035] Through the above steps, the operation control warning management module 50 ensures that safety management and fault handling can be carried out in a timely and effective manner under abnormal circumstances, thereby improving the safety and reliability of the electric tricycle.

[0036] Further, such as Figure 2 As shown, the alarm database generation module 10 in the embodiment of the present application is also used to perform the following steps:

[0037] P11: Determine the monitoring feature points based on the mechanical control dimension and electrical control dimension of the electric tricycle.

[0038] P12: Retrieve and call historical control records, perform multi-level clustering on the monitoring feature points for single feature points and multiple collaborative feature points, and determine multiple clustering clusters, wherein the multiple clustering clusters have overlapping records, and the multiple collaborative feature points include at least two monitoring feature points.

[0039] P13: traverse the multiple clusters, mine critical feature values ​​that meet the preset frequency, and integrate to determine the alarm database.

[0040] Preferably, the alarm database generation module 10 of the embodiment of the present application is also used to perform the following steps:

[0041] P31: Based on the static features and the dynamic features, determine the trend threshold and the vector threshold that meet the preset frequency.

[0042] P32: traverse the single feature point, map the trend threshold and the vector threshold and associate them, and integrate to generate the independent threshold sequence.

[0043] P33: traverse the multiple collaborative feature points, map the trend threshold and the vector threshold, associate the feature points, and integrate to generate the collaborative threshold sequence.

[0044] Specifically, the mechanical control dimension refers to the mechanical control characteristics of the electric tricycle, such as steering, braking, tire pressure, etc. The electrical control dimension refers to the electrical control characteristics of the electric tricycle, such as battery status, motor speed, current sensor, etc.

[0045] Combined with the mechanical control dimension and the electrical control dimension, the specific feature points that need to be monitored are determined. Then, the control data records of the electric tricycle during past operation, that is, the historical control records, are retrieved and called. Through different clustering algorithms (such as K-means, DBSCAN, etc.), single feature points and multiple collaborative feature points are multi-level clustered to find natural groupings in the data and determine multiple clusters, each of which represents a specific operating state or abnormal situation. The information of these clusters will be used for subsequent abnormal threshold mining and alarm database generation to provide basic data support for real-time monitoring and early warning of electric tricycles. Specifically, the data of single feature points can be clustered at a primary level to identify common patterns and abnormal patterns under different operating conditions; then the data of multiple collaborative feature points can be further clustered to identify the association patterns and abnormal patterns of these monitoring point combinations under different states. A single feature point refers to a single monitoring feature point, while multiple collaborative feature points refer to a combination of at least two monitoring feature points, and these monitoring points have a certain correlation. The records of clusters overlap, that is, some monitoring feature points may belong to multiple clusters at the same time. For example, a certain speed and battery voltage combination may appear in both normal operation mode and abnormal mode, so their records will exist in multiple clusters at the same time. Such overlapping records can help more accurately identify and distinguish different operating states and abnormal situations.

[0046] Traverse these clusters and mine critical eigenvalues ​​that meet the preset frequency. The preset frequency refers to the frequency of occurrence of the pre-set eigenvalues. Only the eigenvalues ​​that meet this frequency will be identified as critical eigenvalues. Critical eigenvalues ​​refer to eigenvalues ​​with specific threshold conditions mined in the clusters. These critical eigenvalues ​​are classified and integrated to determine the final alarm database. In the classification and integration process, eigenvalues ​​can be classified according to their types and their associated monitoring points. Independent eigenvalues ​​will be classified into independent threshold sequences, while collaborative eigenvalues ​​will be classified into collaborative threshold sequences.

[0047] Mining critical feature values ​​that meet the preset frequency First, based on static features and dynamic features, determine the trend threshold and vector threshold that meet the preset frequency. Static features refer to features that do not change over time, such as fixed parameters of the vehicle, including vehicle weight, maximum load, body size, etc. Dynamic features refer to features that change over time, such as real-time speed, temperature, battery voltage and current, etc. These dynamic features reflect the changing state information of the vehicle during operation. For example, the speed of the vehicle will change with driving conditions and environmental conditions, and the temperature may change due to different external weather or internal working conditions of the vehicle. The trend threshold refers to the critical value of the feature value under a certain trend. For example, if the vehicle speed rises rapidly in a short period of time and exceeds a certain acceleration critical value, it may indicate a potential dangerous situation. The vector threshold refers to the threshold in a multidimensional space, involving the comprehensive relationship between multiple features, for example, the combination of multiple dynamic features such as speed, temperature and battery voltage changes in different dimensions. By analyzing these multidimensional features, a comprehensive threshold space can be determined.

[0048] First, collect a large amount of historical data, and use data mining technology to analyze the change patterns of static and dynamic features. For static features, the normal range of their fixed parameters can be determined through statistical analysis; for dynamic features, time series analysis and trend detection algorithms, such as ARIMA model or exponential smoothing, can be used to identify the critical points of feature values ​​under trend changes. Compare these analysis results with the preset frequency to determine the trend change threshold that meets the preset frequency. For example, if a feature value in the historical data has a high frequency of trend changes exceeding a specific threshold, this feature value is marked as a trend change threshold. Similarly, through multidimensional data analysis, determine the vector threshold of multiple feature combinations under different conditions. For example, use principal component analysis or cluster analysis to identify abnormal patterns of multi-feature combinations and determine the corresponding vector threshold. When the monitoring data exceeds the preset vector threshold in this multidimensional space, the system will identify it as an abnormal situation.

[0049] Traverse the single feature point, map the trend threshold and the vector threshold, associate them, and integrate to generate an independent threshold sequence. By analyzing the data of a single feature point, its independent abnormal threshold is found. For example, for a speed sensor, its historical data can be analyzed to determine the maximum and minimum values ​​of the speed under normal operating conditions, as well as the critical value of the speed under trend changes, that is, the trend threshold and the vector threshold under comprehensive consideration of speed and other parameters such as time or ambient temperature. These thresholds are sorted into a sequence to form an independent threshold sequence, with each single feature point corresponding to a sequence.

[0050] Traverse multiple collaborative feature points, map the trend threshold and the vector threshold, associate the feature points, and integrate to generate a collaborative threshold sequence. Specifically, by analyzing the data of multiple collaborative feature points, find the correlation between them and the common abnormal threshold, and integrate these association values ​​into a collaborative threshold sequence. Different combinations of collaborative feature points correspond to a sequence. For example, by analyzing the data of the speed sensor and the battery voltage sensor, the mode of simultaneous change of speed and voltage under certain operating conditions is identified. If the combination of speed and voltage exceeds the preset vector threshold, it can be judged as an abnormal situation.

[0051] Finally, the multiple independent threshold sequences and multiple collaborative threshold sequences obtained in the above steps are integrated into the alarm database to provide accurate threshold standards for real-time monitoring and early warning of electric tricycles. When the real-time monitoring data of the vehicle exceeds these thresholds, the system will issue an alarm in time to ensure the safe operation of the vehicle.

[0052] Further, such as Figure 3 As shown, the data preprocessing module 20 in the embodiment of the present application is also used to perform the following steps:

[0053] P21: Based on the timestamp identifier, the sensor monitoring data is temporally and spatially integrated.

[0054] P22: For the integrated sensor monitoring data, identify the abnormal data part and correct it to determine the corrected data, wherein the correction method includes mean interpolation and trend interpolation, and the interpolation method is determined by using a preset data segment.

[0055] P23: Determine the main data structure, perform heterogeneous conversion and de-redundancy on the correction data, and determine the effective monitoring data.

[0056] Furthermore, the main data structure described in the embodiment of the present application is any data structure in the sensor monitoring data, and de-redundancy includes fusion of homologous data and elimination of invalid data.

[0057] Specifically, the collected sensor monitoring data is sorted according to the timestamp to ensure that all data points are arranged in chronological order, and the data of different sensors at the same time point are integrated together. For example, assuming that at a certain time point, the speed recorded by the speed sensor is 20 km / h, the temperature recorded by the temperature sensor is 25°C, and the voltage recorded by the voltage sensor is 48V. Through alignment, these data points are integrated into a comprehensive data point containing speed, temperature and voltage information. And the data in the spatial dimension are integrated together. For example, the speed, temperature and voltage data of the vehicle come from different sensor locations. Through the integration in the spatial dimension, these data can be associated to form a data set that comprehensively reflects the operation status of the vehicle. Through spatiotemporal integration, the consistency of sensor monitoring data in time and space can be ensured to form a comprehensive time series data set. This comprehensive data set not only contains the monitoring data of different sensors at various time points, but also can accurately reflect the changing trends of various parameters of the vehicle during the entire operation process.

[0058] Next, the integrated sensor monitoring data is used to identify abnormal data. Abnormal data may be caused by sensor failure, environmental interference, etc. Statistical analysis and machine learning algorithms can be used to identify abnormal data. For example, the standard deviation method is used to detect abnormal data points. If the value of a data point exceeds the normal range by multiple standard deviations, it can be regarded as an outlier. In addition, abnormal data identification can also be combined with time series analysis to detect abnormal points that do not conform to the overall trend of the data. Correct the identified abnormal data. During the correction process, common methods include mean interpolation and trend interpolation. Mean interpolation replaces the abnormal value with the average value of adjacent data points, while trend interpolation interpolates according to the overall trend of the data. Regression analysis or moving average can be used to predict the reasonable range of the abnormal value. For example, linear regression method is used to predict and correct the abnormal value according to the data trend over a period of time. The interpolation method is selected according to the preset data segment and data characteristics. For short-term anomalies, mean interpolation method can be used because this method is simple and effective; for long-term data missing or complex trend anomalies, trend interpolation method can be used to ensure data continuity and trend consistency. For example, if the speed sensor data is missing for a certain period of time, the average speed before and after the period can be used for interpolation correction; if the temperature data shows obvious abnormal changes over a period of time, the trend interpolation method can be used to correct it by analyzing the change trend of the data before and after. Through these correction methods, the accuracy and consistency of the sensor data can be ensured, and abnormal values ​​caused by sensor failure or environmental interference can be removed or corrected.

[0059] Determine the master data structure, which is any data structure in the sensor monitoring data, such as a time series data structure. Perform heterogeneous conversion and de-redundancy on the correction data, convert correction data of different types and formats into a unified format, and perform fusion based on homologous data and elimination of invalid data, so as to unify various types of correction data into a data set with consistent structure and unified format. For example, if different sensors record similar values ​​at the same time point, these data can be fused into a representative data, while eliminating those duplicate or invalid data, and finally determine the valid monitoring data.

[0060] Through these steps, the data preprocessing module 20 can effectively process and integrate the sensor monitoring data to ensure the accuracy, consistency and effectiveness of the data. High-quality preprocessed data not only improves the monitoring accuracy of the system, but also improves the efficiency of abnormal detection and response.

[0061] Furthermore, the warning signal determination module 30 of the embodiment of the present application is also used to perform the following steps:

[0062] P31: Acquire real-time traffic data, where the real-time traffic data includes road condition dimension and environment dimension.

[0063] P32: Based on the real-time operation data, determine the decision-making influencing characteristics, and the decision-making influencing characteristics are determined based on the characteristic correlation relationship between the operation characteristics and the monitoring control points.

[0064] P33: Based on the decision-making impact characteristics, assist in decision analysis of the effective monitoring data.

[0065] Specifically, the early warning signal determination module 30 obtains real-time operation data before making abnormal control decisions based on effective monitoring data. Real-time operation data refers to the road condition and environmental data collected in real time during the operation of the electric tricycle, including road condition dimensions and environmental dimensions. The road condition dimension includes the flatness, slope, number of bends, etc. of the road, and the environmental dimension includes temperature, humidity, rainfall, etc. These real-time operation data can provide comprehensive information about the operating environment of the electric tricycle.

[0066] Based on the real-time operation data obtained, the decision-making influencing features are determined. Decision-making influencing features refer to key features that can affect decisions, and are determined based on the correlation between operation characteristics and the characteristics of monitoring control points. For example, a road section with a large slope may have a greater impact on the battery voltage and motor temperature; in a high temperature environment, the working state of the battery and motor will also be affected. These correlations can be established and verified through historical data analysis and machine learning algorithms to extract key decision-making influencing features.

[0067] Based on the decision-making influencing features, assist in the decision-making analysis of the effective monitoring data. Before making an abnormal control decision, the system will analyze the effective monitoring data in combination with these decision-making influencing features. For example, if it is detected that the vehicle is traveling on a road section with a large slope, the system will adjust the battery voltage threshold based on this road condition feature to more accurately determine whether the battery has an abnormality. Similarly, if the ambient temperature is high, the system will adjust the motor temperature threshold accordingly to avoid misjudging abnormal conditions due to environmental factors.

[0068] For example, assuming that an electric tricycle is running in a high-temperature environment, the real-time operating data obtained by the system include the current temperature of 35 degrees Celsius (environmental dimension) and the road slope of 15 degrees (road condition dimension). Based on these real-time operating data, the system determines that high temperature and steep slope are decision-influencing features. Combined with these features, the system analyzes the monitoring data and finds that the battery voltage is slightly lower than usual, but considering the normal changes in battery performance in high-temperature environments, the system determines that this is a normal phenomenon rather than an abnormality, thereby avoiding false alarms. At the same time, the system predicts trends and prompts operators to avoid long-term high-load driving in high-temperature environments to protect the life of the battery and motor.

[0069] Through the above steps, the warning signal determination module 30 can dynamically adjust and compensate abnormal control decisions and warning signals according to real-time operation data and decision-making impact characteristics, ensuring the accuracy and reliability of abnormal control decisions, thereby improving the accuracy and effectiveness of warning signals.

[0070] Further, such as Figure 4 As shown, the early warning response module 40 in the embodiment of the present application is also used to perform the following steps:

[0071] P41: Configure the warning mode based on the exception type, and configure the warning response level based on the exception level and urgency.

[0072] P42: Based on the warning mode and the warning response level, build a warning response library.

[0073] P43: Based on the early warning response library, the early warning signal is matched and signal analysis is processed to generate the early warning response single column.

[0074] Specifically, the early warning response module 40 performs signal analysis on the early warning signal to identify and classify the abnormality type. Abnormality types include abnormal battery voltage, excessive temperature, excessive speed, etc. A corresponding early warning mode is configured for each abnormality type, which defines how the system should respond to a specific type of abnormality. For example, for abnormal battery voltage, the early warning mode may include monitoring the voltage change trend and detecting the battery health status. The early warning response module 40 evaluates the abnormality level and urgency of the early warning signal. The abnormality level indicates the severity of the abnormal situation, which is usually divided into mild, medium and severe. For example, a slight battery voltage fluctuation will not immediately affect the vehicle performance, while a serious battery voltage is extremely low and may cause the vehicle to fail to operate normally. The urgency indicates how quickly the abnormal situation needs to be responded to, which is usually divided into immediate response, fast response and delayed response. For example, a sudden and sharp drop in battery voltage requires an immediate response, while a slight speed fluctuation can be handled within the next maintenance cycle.

[0075] According to different early warning modes and early warning response levels, the early warning response module 40 constructs an early warning response library, and defines corresponding response measures for each early warning mode and early warning response level. For example, for abnormal battery voltage with a serious abnormality level and an immediate response level, the response measures may include immediate deceleration, notifying the operator to check the battery, switching to a backup power supply, etc. Then, these response measures are classified and organized according to the early warning mode and early warning response level, and a structured early warning response library is constructed using a database management system to ensure that the system can quickly take appropriate actions in different situations. At the same time, standardized response templates are established for different early warning modes and response levels to facilitate the rapid generation of early warning response columns in actual operations. The templates may include response types, such as deceleration, inspection, and notification; execution time, such as immediate, fast, and delayed; and execution methods, such as automatic and manual.

[0076] When a warning signal is detected, the warning response module 40 matches and analyzes the signal based on the warning response library. First, the corresponding warning mode is matched according to the abnormality type; then, the corresponding warning response measures are matched according to the abnormality level and urgency; finally, the warning response column is generated by integrating this information. The warning response column records the response measures for each warning signal in detail, including the response type, execution time and execution method. Exemplarily, when the warning response module 40 detects that the battery voltage of the electric tricycle is lower than the preset threshold, and the abnormality level is severe and the urgency is immediate response, the battery voltage data is first collected using a sensor, and then the data change trend is analyzed through a Python script, and the abnormal situation is identified using a preset decision tree model. Then, the corresponding response measures are extracted from the warning response library, and a warning response column is generated, including immediate deceleration and notification of the operator to check the battery. The system automatically controls the vehicle to decelerate and sends a notification through the on-board display screen.

[0077] The early warning response module 40 generates a corresponding early warning response column for each early warning signal by constructing an early warning response library, ensuring that each early warning signal can be responded to in a timely and appropriate manner, thereby effectively managing and controlling the safe operation of the electric tricycle.

[0078] Furthermore, the system described in the embodiment of the present application also includes a charge and discharge control warning management module, and the charge and discharge control warning management module is used to perform the following steps:

[0079] P61: Monitor the charging and discharging of the battery pack of the electric tricycle to determine a balancing coefficient, wherein the balancing coefficient is determined based on a balancing state of the battery cells and a strategic balancing coordination state.

[0080] P62: Determine whether the equalization coefficient meets the preset equalization degree.

[0081] P63: If not satisfied, perform charge and discharge control early warning management of the battery pack.

[0082] Specifically, in a feasible implementation, the embodiment of the present application also includes a charge and discharge control warning management module. The charge and discharge control warning management module monitors the charge and discharge status of the battery pack in real time through sensors and monitoring devices installed in the battery pack. These sensors may include voltage sensors, current sensors, temperature sensors, etc. Collect the charge and discharge data of each battery cell, including key parameters such as voltage, current, and temperature. According to the real-time monitoring data, evaluate the actual balance state of each battery cell. The balance state refers to the consistency of the voltage, capacity, health status and other indicators of the battery cell with other battery cells in the battery pack. According to the preset balance strategy, determine the ideal balance state that each battery cell should achieve, including indicators such as voltage difference, capacity difference, and health status difference between battery cells. Compare the actual balance state of each battery cell with the strategy balance coordination state, and calculate the balance coefficient of each battery cell. The balance coefficient is a quantitative indicator, usually expressed as a percentage or fraction, which reflects the difference between the actual balance state and the strategy balance coordination state. In order to facilitate comparison and evaluation, the balance coefficient is standardized to ensure that it is between 0 and 1.

[0083] The preset balance degree is a pre-set standard used to determine whether the balance state of the battery pack has reached the ideal level. Compare the balance coefficient with the preset balance degree to determine whether the balance state of the battery pack meets the preset standard, that is, check whether the state of each battery cell in the battery pack is within the allowable deviation range. If the balance coefficient does not meet the preset balance degree, the charge and discharge control early warning management of the battery pack is performed. At this time, the charge and discharge control early warning management module will take a series of measures to adjust the charge and discharge state of the battery cell to achieve overall balance, such as supplementing the charge or discharge of individual battery cells through active or passive balancing technology to ensure that the performance and life of the entire battery pack are optimized.

[0084] Through the above steps, the charge and discharge control warning management module can monitor and adjust the charge and discharge status of the battery pack of the electric tricycle in real time, promptly identify and handle imbalances, improve the reliability and service life of the battery pack, and thus ensure the safe operation of the electric tricycle.

[0085] In summary, the monitoring and early warning system for an electric tricycle controller provided in the embodiment of the present application has the following technical effects:

[0086] The alarm database generation module 10 deeply analyzes the historical monitoring data of the electric tricycle, performs multi-level clustering for single feature points and multiple collaborative feature points, mines critical feature values ​​that meet the preset frequency, obtains multiple independent threshold sequences and multiple collaborative threshold sequences and integrates the alarm database, and provides accurate threshold standards for real-time monitoring and early warning of the electric tricycle. The data preprocessing module 20 processes and integrates the real-time sensor monitoring data of the electric tricycle through the data processing model, determines effective monitoring data, ensures the accuracy, consistency and effectiveness of the data, and provides high-quality preprocessing data for subsequent analysis and decision-making processes, thereby improving the monitoring accuracy of the system and the efficiency of abnormal detection and response. The early warning signal determination module 30 monitors the operating status of the electric tricycle in real time, trains the adaptive decision module, makes adaptive decisions based on the monitoring data and the information in the alarm database, and dynamically adjusts and compensates according to the real-time operation data and decision-making impact characteristics, and generates an early warning signal with impact compensation for the operating environment, thereby ensuring the accuracy and reliability of abnormal control decisions and improving the accuracy and effectiveness of the early warning signal. The early warning response module 40 constructs an early warning response library, analyzes each early warning signal, responds to and processes the early warning signal in a timely manner, and generates a corresponding early warning response list for each early warning signal, ensuring that each early warning signal can receive a timely and appropriate response, thereby effectively managing and controlling the safe operation of the electric tricycle and improving the reliability and safety of the system. The operation control early warning management module 50 executes the early warning response list to ensure that safety management and fault handling can be carried out in a timely and effective manner under abnormal circumstances, thereby improving the safety and reliability of the electric tricycle. In addition, the charge and discharge control early warning management module monitors and adjusts the charge and discharge status of the battery pack of the electric tricycle in real time, identifies and handles imbalances in a timely manner, improves the reliability and service life of the battery pack, and thus ensures the safe operation of the electric tricycle.

[0087] Through the collaborative work of the above-mentioned modules, in-depth analysis and mining of monitoring data, the system described in the embodiment of the present application can intelligently identify and predict potential abnormal situations, adapt to different operating environment conditions, provide more accurate early warning signals and control decisions, and respond quickly, and take accurate control measures to deal with abnormal situations, thereby ensuring the safe and stable operation of the electric tricycle and extending its service life.

[0088] 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.

Claims

1. A monitoring and early warning system for an electric tricycle controller, characterized in that: The system comprises: An alarm database generation module, the alarm database generation module is used to determine the monitoring control points of the electric tricycle, perform abnormal parameter control threshold mining, and generate an alarm database, the alarm database includes an independent threshold sequence and a collaborative threshold sequence; A data preprocessing module, which is used to retrieve sensor monitoring data, transmit it back to the data processing model, perform data preprocessing and integrate to determine effective monitoring data. The sensor monitoring data is timestamped. The data processing model includes three layers of fully connected data correction layer, heterogeneous conversion layer and de-redundancy layer; An early warning signal determination module, the early warning signal determination module is used to train an adaptive decision module based on the alarm database, perform abnormal control decision and control trend prediction on the effective monitoring data, and determine an early warning signal, wherein the early warning signal has an impact compensation based on the operating environment; An early warning response module, the early warning response module is used to perform signal analysis and response processing on the early warning signal, and determine an early warning response column; An operation control warning management module, the operation control warning management module is used to execute the warning response list and perform operation control warning management on the electric tricycle; The alarm database generation module is also used to perform the following steps: Based on the mechanical control dimension and electrical control dimension of the electric tricycle, the monitoring characteristic points are determined; Retrieve and call historical control records, perform multi-level clustering on the monitoring feature points for single feature points and multiple collaborative feature points, and determine multiple clustering clusters, wherein the multiple clustering clusters have overlapping records, and the multiple collaborative feature points include at least two monitoring feature points; Traversing the plurality of clusters, mining critical feature values ​​that meet a preset frequency, and integrating and determining the alarm database; The alarm database generation module is also used to perform the following steps: Based on the static features and the dynamic features, determine the trend threshold and the vector threshold that meet the preset frequency; Traversing the single feature point, mapping the trend threshold and the vector threshold and associating them, integrating and generating the independent threshold sequence; The multiple collaborative feature points are traversed, the trend threshold and the vector threshold are mapped and the feature points are associated, and the collaborative threshold sequence is generated by integration.

2. The monitoring and early warning system for an electric tricycle controller as claimed in claim 1, characterized in that: The data preprocessing module is also used to perform the following steps: Based on the timestamp identification, the sensor monitoring data is temporally and spatially integrated; For the integrated sensor monitoring data, identify abnormal data parts and perform corrections to determine correction data, wherein the correction method includes mean interpolation and trend interpolation, and the correction method is determined by interpolation of preset data segments; The main data structure is determined, heterogeneous conversion and redundancy elimination are performed on the correction data, and the effective monitoring data is determined.

3. The monitoring and early warning system for an electric tricycle controller as claimed in claim 2, characterized in that: The main data structure is any data structure in the sensor monitoring data, and de-redundancy includes fusion based on homologous data and elimination of invalid data.

4. The monitoring and early warning system for an electric tricycle controller as claimed in claim 1, characterized in that: The warning signal determination module is also used to perform the following steps: Acquire real-time traffic data, wherein the real-time traffic data includes road condition dimension and environment dimension; Determine, based on the real-time operation data, decision-influencing features, wherein the decision-influencing features are determined based on a characteristic correlation relationship between operation features and monitoring control points; Based on the decision-influencing features, assist in decision analysis of the effective monitoring data.

5. The monitoring and early warning system for an electric tricycle controller as claimed in claim 1, characterized in that: The early warning response module is also used to perform the following steps: Configure the warning mode based on the abnormality type, and configure the warning response level based on the abnormality level and urgency; Based on the warning mode and the warning response level, building a warning response library; Based on the early warning response library, the early warning signal is matched and signal analyzed and processed to generate the early warning response single column.

6. The monitoring and early warning system for an electric tricycle controller as claimed in claim 1, characterized in that: The system also includes a charge and discharge control early warning management module, which is used to perform the following steps: Monitoring the charging and discharging of the battery pack of the electric tricycle to determine a balancing coefficient, wherein the balancing coefficient is determined based on a balancing state of the battery cells and a balancing coordination state of the strategy; Determining whether the equalization coefficient satisfies a preset equalization degree; If not satisfied, the battery pack’s charge and discharge control early warning management will be performed.

Images