New energy automobile fuse life expectancy prediction system based on current sensing

Abstract

The invention discloses a new energy automobile fuse life expectancy prediction system based on current sensing, and belongs to the technical field of power transmission and distribution monitoring. In order to solve the problems that life early warning of a dynamic adjustment mechanism is lacked and variable factors in a complex working environment cannot be handled, the system can recognize different load modes by accurately collecting current, temperature and vibration data related to a fuse and combining multi-dimensional factors such as temperature and vibration, and the service life of the fuse is accurately adjusted. The performance of the fuse under various working conditions is comprehensively evaluated, thermal load and loss conditions of the fuse can be estimated by combining a loss model and temperature influence calculation, the accuracy of life prediction is improved, an early warning threshold value is adjusted in real time through a dynamic threshold value adjusting mechanism, and the service life of the fuse can be predicted by dynamically adjusting residual life threshold values at different stages. According to the utility model, the health state of the fuse can be accurately grasped, the maintenance and replacement time can be reasonably adjusted, unnecessary shutdown or delay can be avoided, the fault risk can be effectively reduced, and the safety and reliability of a new energy automobile system can be enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of power transmission and distribution monitoring, and in particular to a new energy vehicle fuse life expectancy prediction system based on current sensing. Background Art

[0002] With the rapid development of new energy vehicles, the performance requirements of electrical equipment such as battery management systems and motor drive systems are getting higher and higher. Among them, fuses, as electrical protection devices, play a vital role. The main function of the fuse is to prevent circuit overload or short circuit from causing fire or equipment damage. However, premature or delayed failure of the fuse will seriously affect the safety and reliability of the vehicle, especially in working environments with large current load fluctuations and drastic temperature changes. Therefore, how to accurately predict the remaining life of the fuse has become one of the key technologies to improve the safety of new energy vehicles and reduce maintenance costs.

[0003] However, there are still the following problems in actual operation:

[0004] In the prior art, fuse life prediction usually relies on simple assumptions or static models, and fails to fully consider the impact of multiple complex factors such as current waveform, load changes, temperature fluctuations, etc. At the same time, traditional fuse life prediction systems usually set fixed alarm thresholds and fail to adjust alarm strategies in time according to actual usage conditions (such as load changes, temperature fluctuations, etc.), ignoring the complex changes in the working environment and the impact of environmental factors such as temperature changes and vibrations on fuse loss. Summary of the invention

[0005] The purpose of the present invention is to provide a new energy vehicle fuse life expectancy prediction system based on current sensing to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: a new energy vehicle fuse life expectancy prediction system based on current sensing, comprising:

[0007] Current state acquisition module, used for:

[0008] Real-time collection of current signal data and sensor data related to fuses in new energy vehicles, wherein the sensor data includes temperature data and vibration data, and the current state acquisition module collects current signals through at least one sensor, wherein the sensor is a Hall effect sensor, a shunt resistor or a current transformer;

[0009] Load pattern recognition module for:

[0010] Extract the current waveform based on the current signal data, extract the current spectrum characteristics from the current waveform, analyze the collected current spectrum characteristics, identify the load pattern in the current, and extract the feature data based on the load pattern;

[0011] A temperature influence calculation module, configured to:

[0012] Establish a loss model of the fuse according to the current signal data and temperature data, calculate the heat load and loss situation of the fuse based on the loss model, and the temperature influence calculation module estimates the fuse loss based on the heat loss model and the temperature-loss relationship;

[0013] A remaining life prediction module, configured to:

[0014] Predict the remaining life of the fuse based on the current waveform, load pattern, and loss data;

[0015] An alarm and early warning module, configured to:

[0016] Trigger an alarm when the remaining life of the fuse is lower than a set threshold. The alarm and early warning module includes a multi-level alarm mechanism, automatically adjusts the early warning level according to the fuse life prediction result and triggers an early warning. The alarm methods include text messages, emails, and App push notifications;

[0017] A data recording module, configured to:

[0018] Record all the collected data, the working state of the fuse, the remaining life prediction result, and the alarm information, store the data through a time-series database, and provide a graphical interface for data display.

[0019] Furthermore, the load pattern recognition module includes:

[0020] A current waveform extraction unit, configured to:

[0021] Sample the real-time collected current signal through a digital converter, perform spectrum analysis on the current signal through fast Fourier transform and extract the current waveform, and extract the current spectrum characteristics from the current waveform. The current spectrum characteristics include the maximum current value, frequency distribution, waveform period, and change frequency;

[0022] A load pattern classification unit, configured to:

[0023] Analyze the current spectrum characteristics, perform unsupervised classification on the extracted current spectrum characteristics through clustering and identify different current patterns. The current patterns include overload, normal load, and transient load. Based on the classification result, generate corresponding load pattern labels.

[0024] Furthermore, the load pattern recognition module further includes:

[0025] A feature data extraction unit, configured to:

[0026] Extract current spectrum feature data from the classified current waveforms, analyze the identified load patterns, and extract the current peak value, current fluctuation frequency, and duration as feature data;

[0027] Perform multi-dimensional correlation analysis on the current characteristics in combination with vehicle usage conditions to generate high-dimensional feature data for fuse life prediction, where the vehicle usage conditions include vehicle speed and driving environment, and the high-dimensional feature data includes average vehicle speed, vehicle speed fluctuation, ambient temperature, humidity, road type, harmonic distortion, fundamental wave current, harmonic currents of each order, starting current, steady-state current, load switching current, current change rate, current fluctuation energy, root mean square value of current, current standard deviation, number of acceleration times, and number of braking times.

[0028] Further, the temperature influence calculation module includes:

[0029] A loss model establishment unit, configured to:

[0030] Establish a heat loss model based on the physical characteristics of the fuse and the current-temperature relationship, where the physical characteristics include material and resistance, input the real-time collected current signal and temperature data into the model, and calculate the heat load of the fuse under different working conditions;

[0031] Output the loss data of the fuse and generate a loss curve;

[0032] A loss relationship calculation unit, configured to:

[0033] Dynamically adjust the loss value according to the relationship between the collected temperature data and the fuse loss model, and based on the temperature-loss relationship, calculate the correction value of the temperature change to the fuse loss in real time through curve fitting.

[0034] Further, the temperature influence calculation module further includes:

[0035] A loss data correction unit, configured to:

[0036] Perform feedback adjustment based on historical data and actual loss conditions, regularly correct the loss calculation model, and correct the parameters in the loss model.

[0037] Further, the life prediction module includes:

[0038] A prediction model training unit, configured to:

[0039] Train the fuse life prediction model using historical data and generate a prediction model, where the historical data includes current waveforms, load patterns, and loss data;

[0040] A real-time life prediction unit, configured to:

[0041] Input the real-time collected current signal, load pattern, and temperature data into the trained life prediction model, calculate the remaining life of the fuse through the model, and generate a life prediction report;

[0042] A prediction result output unit, configured to:

[0043] Output the prediction result of the remaining life of the fuse to the user side through a graphical interface and an API interface, and generate a historical life prediction curve.

[0044] Further, the real-time life prediction unit includes:

[0045] A data acquisition sub-unit, configured to acquire the real-time collected current signal, load pattern, and temperature data, and predict the remaining life of the fuse according to the real-time collected current signal, load pattern, and temperature data, specifically including:

[0046] A first calculation sub-unit, configured to calculate the remaining life of the fuse according to the following formula;

[0047] S2 = S1 - β * [(Q1 - Q2) * φ] * ∫(I 实际 / I 额定 ) n * dt;

[0048] Wherein, S2 represents the remaining life of the fuse; S1 represents the theoretical life of the fuse; β represents the influence coefficient of the load pattern, and the value range is (1.01, 1.02); Q1 represents the actual temperature of the fuse; Q2 represents the reference temperature of the fuse; φ represents the temperature influence coefficient, and the value range is (0.98, 1.02); I 实际 represents the actual current of the fuse; I 额定 represents the rated current of the fuse; n represents the current influence coefficient, and the value range is (0, 1); dt represents the time integration;

[0049] A second calculation sub-unit, configured to obtain the theoretical remaining life, and calculate the comprehensive remaining life of the fuse according to the remaining life and the theoretical remaining life of the fuse;

[0050]

[0051] Wherein, μ represents the comprehensive remaining life of the fuse;

[0052] A qualification determination sub-unit, configured to:

[0053] Judge whether the comprehensive remaining life of the fuse is consistent with the remaining life of the fuse;

[0054] If the remaining life of the fuse is consistent with the comprehensive remaining life of the fuse, the life prediction is determined to be qualified;

[0055] Otherwise, the life prediction is determined to be unqualified. At the same time, an alarm signal for alarming the life prediction model is generated.

[0056] Furthermore, the prediction model training unit is further configured to:

[0057] Collect historical data, and perform cleaning, denoising, and normalization processing on the historical data;

[0058] Extract the necessary features for fuse life prediction from the historical data. The necessary features include the peak value of the current waveform, the frequency of current fluctuation, the change frequency and amplitude of the load pattern, and the temperature change rate. Input the extracted necessary features into the fuse life prediction model and train according to the historical data labels, where the historical data labels are the known fuse lives;

[0059] Verify the accuracy of the model by comparing with the actual fuse life data, and save the trained fuse life prediction model.

[0060] Furthermore, the real-time life prediction unit is further configured to:

[0061] Collect current waveform data in real time, and combine with the load pattern recognition module to obtain real-time load pattern data. Extract prediction features from the real-time data. The prediction features include the instantaneous value of the current, the frequency features of the waveform, the change of the load pattern, and the temperature rise rate;

[0062] Input the extracted prediction features into the trained fuse life prediction model, and the model calculates the remaining life of the fuse according to the input data;

[0063] Generate a remaining life prediction report for the fuse according to the calculation result of the model. The remaining life prediction report includes the current state of the fuse, the predicted value of the remaining life, and the prediction confidence interval.

[0064] Furthermore, the alarm and early warning module includes:

[0065] The trigger condition determination unit is configured to:

[0066] Set the remaining life threshold of the fuse. If the life prediction value is lower than the remaining life threshold, trigger the alarm condition, and dynamically adjust the alarm trigger condition based on the prediction result of the remaining life;

[0067] The multi-level alarm mechanism unit is configured to:

[0068] Set different alarm levels according to the remaining life. The alarm levels include severe failure, early warning, and normal. Automatically select and trigger the corresponding alarm level according to the predicted life value;

[0069] An alarm notification unit for:

[0070] When the alarm condition is triggered, push the triggered alarm information to the user and the maintenance personnel via SMS, email, or App, and provide the alarm history record and detailed information in the alarm message.

[0071] Furthermore, dynamically adjusting the alarm trigger condition also includes:

[0072] Set an initial threshold. Set a basic remaining life threshold according to the fuse specifications, historical data, and environmental factors. Among them, the basic remaining life threshold is divided into three stage intervals. When the remaining life is less than 20%, it is regarded as a severe failure. When the remaining life is between 20% and 50%, it is an early warning. When the remaining life is more than 50%, it is in a normal state;

[0073] Set a threshold adjustment period, and perform periodic threshold adjustment based on the adjustment period. During each adjustment period, collect real-time data;

[0074] Based on the change trend of the remaining life of the fuse, when the remaining life continuously decreases or changes abnormally, dynamically adjust the remaining life threshold;

[0075] Among them, when the system detects that the remaining life of the fuse continues to decrease and the decrease amplitude exceeds 5% per cycle, then dynamically reduce the remaining life threshold and reduce the remaining life threshold of each stage interval by 5%;

[0076] When the system detects that the change in the remaining life of the fuse shows a sharp decrease or irregular change, then dynamically reduce the remaining life threshold and reduce the remaining life threshold of each stage interval by 10%;

[0077] Dynamically weighted based on environmental factors. When the load mode or temperature changes, dynamically adjust the remaining life threshold;

[0078] Among them, when the system detects frequent load fluctuations or a large increase in the load current value, then dynamically reduce the remaining life threshold and reduce the remaining life threshold of each stage interval by 10%;

[0079] If the operating temperature of the fuse exceeds the preset safety range, then dynamically reduce the remaining life threshold and reduce the remaining life threshold of each stage interval by 10%;

[0080] When the load increases and the temperature exceeds the safe range, the remaining life threshold is dynamically reduced, and the reduction amplitude of the threshold is weighted, reducing the remaining life threshold of each stage interval by 10 - 20%.

[0081] Furthermore, the data recording module includes:

[0082] A recording dimension determination unit, configured to obtain the recording dimension for recording the data to be recorded, construct a dimension label according to the recording dimension, and construct a corresponding data recording interval according to the dimension label, where the dimension label and the data recording interval are in one-to-one correspondence;

[0083] A data mapping unit, configured to determine a data recording mapping relationship according to the dimension label and the data to be recorded, and map the data to be recorded in the corresponding data recording interval according to the mapping relationship according to the dimension label;

[0084] A sorting unit, configured to sort the target recorded data in each data recording interval according to the time sequence according to the mapping result, and obtain a time sequence data sequence for each data recording interval;

[0085] An index generation unit, configured to obtain the time points of the time sequence data sequence, generate a column index according to the dimension label, and generate a row index according to the time points, where the time nodes and the row index are in one-to-one correspondence;

[0086] A data storage unit, configured to:

[0087] Integrate all data recording intervals based on the column index and the row index to generate a data storage table;

[0088] Store the data storage table in the time series database;

[0089] A data display unit, configured to:

[0090] When the data display requirement is input based on the user terminal, extract the time keyword and the data dimension keyword of the data display requirement;

[0091] Match the time keyword with the column index to determine the target column index consistent with the time keyword;

[0092] Match the data dimension keyword with the row index to determine the target row index consistent with the data dimension keyword;

[0093] Match the target row index with a target display template in a preset display template library;

[0094] Retrieve the data to be displayed consistent with the data display requirement from the time series database according to the target row index and the target column index;

[0095] The data to be displayed is mapped to a target display template, a target graphical interface is obtained, and the target graphical interface is displayed.

[0096] Compared with the prior art, the present invention has the following beneficial effects:

[0097] 1. The present invention uses sensors to accurately and in real time collect current signals, temperature data and vibration data related to the fuse, providing a reliable input basis for subsequent fuse life prediction. By integrating data such as temperature and vibration, it can comprehensively consider the performance of the fuse under different working conditions and provide more accurate life assessment information. By comprehensively considering multiple factors such as current, load, temperature, etc., the system can accurately predict the life of the fuse.

[0098] 2. The present invention accurately identifies different load modes according to the characteristics of the current spectrum, which helps to better understand the working status of the fuse under different load conditions. By identifying the current mode and extracting relevant characteristic data, it provides rich characteristic data for the subsequent life prediction module, significantly improving the accuracy of life prediction. Through the establishment of the loss model and the calculation of the temperature influence, the thermal load and loss of the fuse under different temperature and current conditions can be accurately estimated, thereby providing early warning of the failure risk of the fuse. By dynamically adjusting the loss value and considering the influence of temperature, this temperature-loss relationship makes the life prediction of the fuse more accurate, especially in an environment with drastic temperature changes. The life prediction model can accurately predict the remaining life of the fuse and adapt to changes in different working environments.

[0099] 3. The present invention sets a dynamic threshold adjustment mechanism. The system can adjust the remaining life threshold through a dynamic weighting mechanism according to factors such as the remaining life change of the fuse, load fluctuations and temperature changes. When the load current fluctuates greatly or the temperature exceeds the safe range, the system can adjust the remaining life threshold through a dynamic weighting mechanism, thereby identifying potential faults at an early stage and reducing the risk of faults. By dynamically adjusting the thresholds of each stage interval, the system can accurately grasp the health status of the fuse, reasonably adjust the maintenance or replacement time, avoid unnecessary downtime or delays, and dynamically adjust the remaining life threshold according to different situations. By considering the remaining life change trend of the fuse, load mode, temperature change and other factors, the system can intelligently adjust the alarm threshold, issue early warnings, and ensure that the fuse can issue an alarm in time under any abnormal situation to ensure the safety of the vehicle.

[0100] 4. By acquiring the real-time collected current signal, load mode and temperature data, and effectively predicting the remaining life of the fuse based on the real-time collected current signal, load mode and temperature data, by calculating the comprehensive remaining life and making a life prediction qualification judgment, the judgment of the life prediction model is effectively guaranteed, and then when it is unqualified, an alarm signal is generated to alarm the life prediction model, which is conducive to timely grasping the status of the life prediction model.

[0101] 5. By obtaining the recording dimension for recording the data to be recorded, the construction of the data recording interval can be effectively realized. Furthermore, through the data recording mapping relationship, the partition recording of the data to be recorded can be effectively realized. The target recording data in the data recording interval is sorted according to the time sequence, and then the time series data sequence can be effectively obtained, providing effective assistance for subsequent data storage, extraction, and display. By generating the row index and column index, the data can be stored in the time series database, effectively ensuring the accuracy of data extraction and improving the efficiency of data extraction. By matching the target display template consistent with the target row index, the accurate and efficient display of data under each recording dimension can be effectively ensured, making the data display more convenient, efficient, and intelligent, and improving the user experience. Brief Description of the Drawings

[0102] Figure 1 It is a schematic diagram of the fuse expected life prediction system module of the present invention. Detailed Embodiments

[0103] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0104] In order to solve the technical problems in the prior art that the life prediction of fuses usually relies on simple assumptions or static models, and fails to fully consider the influence of multiple complex factors such as current waveforms, load changes, and temperature fluctuations. At the same time, traditional fuse life prediction systems usually set fixed alarm thresholds and fail to adjust the alarm strategy in a timely manner according to actual usage conditions (such as load changes, temperature fluctuations, etc.), ignoring the complex changes in the working environment and the influence of environmental factors such as temperature changes and vibrations on fuse losses. Please refer to Figure 1 , the present invention provides the following technical solutions:

[0105] A new energy vehicle fuse expected life prediction system based on current sensing, including:

[0106] A current state acquisition module, used for:

[0107] Real-time collecting current signal data and sensing data related to the fuse in the new energy vehicle, where the sensing data includes temperature data and vibration data. The current state acquisition module realizes the acquisition of current signals through at least one sensor, and the sensor is a Hall effect sensor, a shunt resistor, or a current transformer;

[0108] A load mode recognition module, configured to:

[0109] Extract a current waveform based on the current signal data, extract current spectrum features based on the current waveform, analyze the collected current spectrum features, identify the load mode in the current, and extract feature data based on the load mode;

[0110] A temperature influence calculation module, configured to:

[0111] Establish a loss model of the fuse according to the current signal data and temperature data, calculate the thermal load and loss conditions of the fuse based on the loss model, and the temperature influence calculation module estimates the fuse loss based on the thermal loss model and the temperature-loss relationship;

[0112] A remaining life prediction module, configured to:

[0113] Predict the remaining life of the fuse based on the current waveform, load mode, and loss data;

[0114] An alarm warning module, configured to:

[0115] Trigger an alarm when the remaining life of the fuse is lower than a set threshold. The alarm warning module includes a multi-level alarm mechanism, automatically adjusts the warning level according to the fuse life prediction result and triggers a warning. The alarm methods include text messages, emails, and App push notifications;

[0116] A data recording module, configured to:

[0117] Record all collected data, the working state of the fuse, the remaining life prediction result, and alarm information, store the data through a time series database, and provide a graphical interface for data display.

[0118] In the above embodiment, through sensors such as Hall effect sensors, shunt resistors, or current transformers, current signals, temperature data, and vibration data related to the fuse can be accurately and real-time collected, providing a reliable input basis for subsequent fuse life prediction. By integrating data such as temperature and vibration, the performance of the fuse under different working conditions can be comprehensively considered, providing more accurate life assessment information.

[0119] In the above embodiments, by comprehensively considering multiple factors such as current, load, and temperature, the system can accurately predict the lifespan of the fuse and dynamically adjust the alarms and thresholds, improving the safety and reliability of the entire new energy vehicle system. The system can process a large amount of data in real time and make adaptive adjustments based on data analysis, such as load changes and temperature fluctuations, providing users with intelligent maintenance suggestions and alarm information, reducing manual intervention and errors. By predicting the fuse lifespan in advance and issuing alarms in a timely manner, the system can effectively reduce the vehicle downtime and maintenance costs caused by fuse failures, while increasing the overall service life of the vehicle.

[0120] The load mode recognition module includes:

[0121] The current waveform extraction unit is used for:

[0122] Sampling the real-time collected current signal through a digital converter, performing spectral analysis on the current signal through fast Fourier transform and extracting the current waveform, and extracting the current spectrum features based on the current waveform. The current spectrum features include the maximum current value, frequency distribution, waveform period, and change frequency;

[0123] The load mode classification unit is used for:

[0124] Analyzing the current spectrum features, performing unsupervised classification on the extracted current spectrum features through clustering and identifying different current modes. The current modes include overload, normal load, and transient load. Based on the classification results, generating corresponding load mode labels;

[0125] The feature data extraction unit is used for:

[0126] Extracting the current spectrum feature data from the classified current waveforms, analyzing the identified load modes, and extracting the current peak value, current fluctuation frequency, and duration as feature data;

[0127] Performing multi-dimensional correlation analysis on the current characteristics in combination with the vehicle usage conditions to generate high-dimensional feature data for fuse lifespan prediction. Among them, the vehicle usage conditions include vehicle speed and driving environment, and the high-dimensional feature data includes average vehicle speed, vehicle speed fluctuation, environmental temperature, humidity, road type, harmonic distortion, fundamental wave current, each harmonic current, starting current, steady-state current, load switching current, current change rate, current fluctuation energy, root mean square value of current, current standard deviation, number of accelerations, and number of brakings.

[0128] In the above embodiment, the current waveform extraction unit and the load mode classification unit are used to accurately identify different load modes according to the current spectrum characteristics, which helps to better understand the working status of the fuse under different load conditions. By identifying the current mode and extracting related characteristic data, rich characteristic data is provided for the subsequent life prediction module, which significantly improves the accuracy of life prediction.

[0129] Temperature impact calculation module, including:

[0130] Loss model building unit for:

[0131] Based on the physical characteristics of the fuse and the current-temperature relationship, a heat loss model is established. The physical characteristics include material and resistance. The real-time collected current signal and temperature data are input into the model to calculate the heat load of the fuse under different working conditions.

[0132] Output the loss data of the fuse and generate the loss curve;

[0133] Loss relationship calculation unit for:

[0134] According to the relationship between the collected temperature data and the fuse loss model, the loss value is adjusted dynamically. Based on the temperature-loss relationship, the correction value of the fuse loss due to temperature change is calculated in real time through curve fitting;

[0135] Loss data correction unit for:

[0136] Feedback adjustments are made based on historical data and actual loss conditions, and the loss calculation model is regularly calibrated to correct the parameters in the loss model.

[0137] In the above embodiment, by establishing a loss model and calculating the influence of temperature, the thermal load and loss of the fuse under different temperature and current conditions can be accurately estimated, thereby providing early warning of the risk of fuse failure. By dynamically adjusting the loss value and taking into account the influence of temperature, this temperature-loss relationship makes the life prediction of the fuse more accurate, especially in an environment with drastic temperature changes. The existence of the loss data correction unit can continuously optimize the loss calculation model based on feedback adjustments of historical data and actual loss conditions, thereby further improving the accuracy of life prediction.

[0138] Life prediction module, including:

[0139] Prediction model training unit, used for:

[0140] Using historical data to train a fuse life prediction model and generate a prediction model, the historical data including current waveform, load mode and loss data;

[0141] Collect historical data and perform cleaning, denoising, and normalization on the historical data;

[0142] Extract the necessary features for fuse life prediction from the historical data. The necessary features include the peak value of the current waveform, the frequency of current fluctuations, the change frequency and its amplitude of the load pattern, and the temperature change rate. Input the extracted necessary features into the fuse life prediction model and train it according to the historical data labels, where the historical data labels are the known fuse lives;

[0143] Verify the accuracy of the model by comparing it with the actual fuse life data and save the trained fuse life prediction model;

[0144] A real-time life prediction unit for:

[0145] Input the real-time collected current signal, load pattern, and temperature data into the trained life prediction model, calculate the remaining life of the fuse through the model, and generate a life prediction report;

[0146] Collect real-time current waveform data, and combine with the load pattern recognition module to obtain real-time load pattern data. Extract prediction features from the real-time data. The prediction features include the instantaneous value of the current, the frequency features of the waveform, the change of the load pattern, and the temperature rise rate;

[0147] Input the extracted prediction features into the trained fuse life prediction model, and the model calculates the remaining life of the fuse according to the input data;

[0148] Generate a remaining life prediction report for the fuse according to the calculation result of the model. The remaining life prediction report includes the current state of the fuse, the predicted value of the remaining life, and the prediction confidence interval;

[0149] A prediction result output unit for:

[0150] Output the prediction result of the remaining life of the fuse to the user side through a graphical interface and an API interface, and generate a historical life prediction curve.

[0151] In the above embodiment, through the cleaning, denoising, and normalization of historical data, the trained life prediction model can accurately predict the remaining life of the fuse and adapt to the changes in different working environments. The real-time life prediction unit can dynamically predict the remaining life of the fuse by analyzing the real-time collected data and comparing it with the trained model, and generate a life prediction report according to the real-time data, providing instant feedback. The generated life prediction report not only includes the remaining life value but also provides the prediction confidence interval, enabling users to more clearly understand the health status of the fuse and make corresponding maintenance or replacement decisions.

[0152] Alarm warning module, including:

[0153] Trigger condition determination unit, for:

[0154] Set the remaining life threshold of the fuse. If the life prediction value is lower than the remaining life threshold, trigger the alarm condition, and dynamically adjust the alarm trigger condition based on the prediction result of the remaining life;

[0155] Multi-level alarm mechanism unit, for:

[0156] Set different alarm levels according to the different remaining lives. The alarm levels include serious failure, early warning, and normal. Automatically select and trigger the corresponding alarm level according to the predicted life value;

[0157] Alarm notification unit, for:

[0158] When the alarm condition is triggered, push the triggered alarm information to the user and the maintenance personnel by text message, email or App, and provide the alarm history record and detailed information in the alarm information.

[0159] Among them, dynamically adjusting the alarm trigger condition also includes:

[0160] Set the initial threshold. Set a basic remaining life threshold according to the specifications, historical data and environmental factors of the fuse. Among them, the basic remaining life threshold is divided into three stage intervals. When the remaining life is less than 20%, it is regarded as a serious failure. When the remaining life is between 20% and 50%, it is an early warning. When the remaining life is more than 50%, it is in a normal state;

[0161] Set the threshold adjustment period, and perform periodic threshold adjustment based on the adjustment period. In each adjustment period, collect real-time data;

[0162] Based on the change trend of the remaining life of the fuse, when the remaining life continuously decreases or changes abnormally, dynamically adjust the remaining life threshold;

[0163] Among them, when the system detects that the remaining life of the fuse continues to decline, and the decline amplitude exceeds 5% per cycle, the remaining life threshold is dynamically reduced, and the remaining life thresholds of each stage interval are reduced by 5%;

[0164] When the system detects that the change of the remaining life of the fuse shows a sharp decline or irregular change, the remaining life threshold is dynamically reduced, and the remaining life thresholds of each stage interval are reduced by 10%;

[0165] Dynamically weighted based on environmental factors, and dynamically adjust the remaining life threshold when the load mode or temperature changes;

[0166] Among them, when the system detects frequent load fluctuations or a significant increase in the load current value, the remaining life threshold is dynamically reduced, and the remaining life threshold for each stage interval is reduced by 10%;

[0167] If the operating temperature of the fuse exceeds the preset safe range, the remaining life threshold is dynamically reduced, and the remaining life threshold for each stage interval is reduced by 10%;

[0168] When the load increases and the temperature exceeds the safe range, the remaining life threshold is dynamically reduced, and at the same time, the reduction amplitude of the threshold is weighted, and the remaining life threshold for each stage interval is reduced by 10 - 20%

[0169] In the above embodiments, based on factors such as the change trend of the remaining life, the load pattern, and the temperature change, the alarm trigger condition can be dynamically adjusted, which can avoid premature or late alarms, improve the sensitivity and accuracy of the alarms. The system automatically adjusts the alarm level according to different stages of the remaining life, so as to achieve multi-level alarm management, which is convenient for users to take different countermeasures according to different alarm levels. Alarm notifications are sent to users and maintenance personnel through methods such as text messages, emails, and App push, and at the same time, historical records and detailed information are provided to help users promptly grasp the status of the fuse and conduct effective maintenance and prevention.

[0170] Here, the present invention is further described through the following examples of threshold adjustment rules:

[0171] 1. When the load increases: If the load increases and the current predicted remaining life value of the fuse is 40%, the system may lower the threshold to 30%, that is, an alarm is triggered when the remaining life is lower than 30%.

[0172] 2. When the temperature rises: Assume that the current remaining life of the fuse is 60%, but due to the temperature exceeding 80°C, the system will reduce the threshold from 50% to 40% according to the temperature weighting mechanism to give an early warning of the impact of the temperature rise.

[0173] 3. When the downward trend of the remaining life is significant: If the remaining life of the fuse continuously drops by more than 5% (for example, from 60% to 50%), the system will dynamically reduce the threshold to 45% to trigger an early warning in a timely manner.

[0174] Final threshold adjustment logic: When the remaining life of the fuse continues to drop or the environmental factors deteriorate (such as load increase, temperature rise, severe load fluctuations, etc.), the system will lower the threshold to give an early alarm to prevent the fuse from failing when the remaining life is too low; if the predicted remaining life value of the fuse recovers and tends to be stable, and the load and environmental factors are relatively normal, the system can appropriately increase the threshold to reduce the alarm frequency and avoid excessive alarms.

[0175] By setting up a dynamic threshold adjustment mechanism, the system can, based on factors such as the remaining life change of the fuse, load fluctuation, and temperature change, adjust the remaining life threshold through a dynamic weighting mechanism in cases where the load current fluctuates significantly or the temperature exceeds the safe range. Thus, potential faults can be identified at an early stage, reducing the risk of faults. By dynamically adjusting the thresholds in each stage interval, the system can accurately grasp the health status of the fuse, reasonably adjust the maintenance or replacement time, avoid unnecessary downtime or delays, and dynamically adjust the remaining life threshold according to different situations. By considering factors such as the remaining life change trend of the fuse, load pattern, and temperature change, the system can intelligently adjust the alarm threshold, give early warnings, and ensure that the fuse can issue alarms in a timely manner under any abnormal circumstances, guaranteeing the safety of the vehicle.

[0176] In one embodiment, a prediction system for the expected life of a new energy vehicle fuse based on current sensing is provided. The real-time life prediction unit includes:

[0177] A data acquisition sub-unit, configured to acquire the current signal, load pattern, and temperature data collected in real time, and predict the remaining life of the fuse based on the current signal, load pattern, and temperature data collected in real time. Specifically, it includes:

[0178] A first calculation sub-unit, configured to calculate the remaining life of the fuse according to the following formula;

[0179] S2 = S1 - β * [(Q1 - Q2) * φ] * ∫(I 实际 / I 额定 ) n * dt;

[0180] Wherein, S2 represents the remaining life of the fuse; S1 represents the theoretical life of the fuse; β represents the influence coefficient of the load pattern, and its value range is (1.01, 1.02); Q1 represents the actual temperature of the fuse; Q2 represents the reference temperature of the fuse; φ represents the temperature influence coefficient, and its value range is (0.98, 1.02); I 实际 represents the actual current of the fuse; I 额定 represents the rated current of the fuse; n represents the current influence coefficient, and its value range is (0, 1); dt represents the time integral;

[0181] A second calculation sub-unit, configured to obtain the theoretical remaining life and calculate the comprehensive remaining life of the fuse according to the remaining life and the theoretical remaining life of the fuse;

[0182]

[0183] Wherein, μ represents the comprehensive remaining life of the fuse;

[0184] A qualified judgment sub-unit, configured to:

[0185] Determine whether the comprehensive remaining life of the fuse is consistent with the remaining life of the fuse;

[0186] If the remaining life of the fuse is consistent with the comprehensive remaining life of the fuse, it is determined that the life prediction is qualified;

[0187] Otherwise, it is determined that the life prediction is unqualified. At the same time, an alarm signal for alarming the life prediction model is generated.

[0188] In this embodiment, the alarm signal can be a combination of sound, light and vibration.

[0189] The working principle and beneficial effects of the above technical solution are as follows: By acquiring the current signal, load pattern and temperature data collected in real time, and effectively predicting the remaining life of the fuse according to the current signal, load pattern and temperature data collected in real time, calculating the comprehensive remaining life, and determining the qualification of the life prediction effectively guarantee the determination of the life prediction model. Furthermore, when it is unqualified, an alarm signal for alarming the life prediction model is generated, which is beneficial to timely grasp the state of the life prediction model.

[0190] In one embodiment, a system for predicting the expected life of a new energy vehicle fuse based on current sensing is provided. The data recording module includes:

[0191] A record dimension determination unit, configured to obtain the record dimension for recording the data to be recorded, construct a dimension label according to the record dimension, and construct a corresponding data record interval according to the dimension label, where the dimension label and the data record interval are in one-to-one correspondence;

[0192] A data mapping unit, configured to determine a data record mapping relationship according to the dimension label and the data to be recorded, and map the data to be recorded in the corresponding data record interval according to the mapping relationship according to the dimension label;

[0193] A sorting unit, configured to sort the target record data in each data record interval according to the time sequence according to the mapping result to obtain a time sequence data sequence for each data record interval;

[0194] An index generation unit, configured to obtain the time points of the time sequence data sequence, generate a column index according to the dimension label, and generate a row index according to the time points, where the time nodes and the row indexes are in one-to-one correspondence;

[0195] A data storage unit, configured to:

[0196] Integrate all data record intervals based on the column index and the row index to generate a data storage table;

[0197] Store the data storage table in the time series database;

[0198] A data display unit, configured to:

[0199] When a data display requirement is input based on a user terminal, extract the time keyword and data dimension keyword of the data display requirement;

[0200] Match the time keyword with a column index to determine a target column index that is consistent with the time keyword;

[0201] Match the data dimension keyword with a row index to determine a target row index that is consistent with the data dimension keyword;

[0202] Match a target display template in a preset display template library according to the target row index;

[0203] Retrieve data to be displayed that is consistent with the data display requirement from a time series database according to the target row index and the target column index;

[0204] Map the data to be displayed to the target display template to obtain a target graphical user interface, and display the target graphical user interface.

[0205] In this embodiment, the record dimensions include: collected data, fuse working status, life prediction result, and alarm information.

[0206] In this embodiment, the dimension label can be used as a characterization identifier for distinguishing each record dimension, so as to implement the construction of a data record interval, that is, the data record interval corresponds to the record dimension one by one.

[0207] In this embodiment, the record mapping relationship is to classify the data to be recorded according to the dimension label, that is, a data record connection between a dimension label and a type (such as an alarm information type).

[0208] In this embodiment, the time series data sequence can be the result of sorting the target record data in chronological order in each data record interval.

[0209] In this embodiment, the row index is used as a marker for pointing to time.

[0210] In this embodiment, the column index is used as a marker for pointing to the dimension label.

[0211] In this embodiment, the preset display template library can be set in advance and used to store data display templates corresponding to different dimension labels.

[0212] The working principle and beneficial effects of the above technical solution are as follows: By obtaining the recording dimension for the data to be recorded, the construction of the data recording interval can be effectively achieved. Furthermore, through the data recording mapping relationship, the partition recording of the data to be recorded can be effectively realized. The target recorded data in the data recording interval is sorted according to the time sequence, and then the time series data sequence can be effectively obtained, providing effective assistance for subsequent data storage, extraction, and display. By generating the row index and column index, the data can be stored in the time series database, effectively ensuring the accuracy of data extraction and improving the efficiency of data extraction. By matching the target display template consistent with the target row index, the accurate and efficient display of data under each recording dimension is effectively guaranteed, making the data display more convenient, efficient, and intelligent, and improving the user experience.

[0213] The above is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, making equivalent substitutions or changes should be covered within the protection scope of the present invention.

Claims

1. A predicted life expectancy system for a fuse of a new energy vehicle based on current sensing, characterized in that, Including: A current state acquisition module, configured to: Real-time collect current signal data and sensing data related to the fuse in a new energy vehicle, where the sensing data includes temperature data and vibration data. The current state acquisition module realizes the acquisition of current signals through at least one sensor, and the sensor is a Hall effect sensor, a shunt resistor or a current transformer; A load mode recognition module, configured to: Extract the current waveform based on the current signal data, extract the current spectrum characteristics based on the current waveform, analyze the collected current spectrum characteristics, identify the load mode in the current, and extract feature data based on the load mode; A temperature impact calculation module, configured to: Establish a loss model of the fuse according to the current signal data and temperature data, calculate the thermal load and loss conditions of the fuse based on the loss model. The temperature impact calculation module estimates the fuse loss based on the thermal loss model and the temperature-loss relationship; A remaining life prediction module, configured to: Predict the remaining life of the fuse based on the current waveform, load mode, and loss data; An alarm and early warning module, configured to: Trigger an alarm when the remaining life of the fuse is lower than a set threshold. The alarm and early warning module includes a multi-level alarm mechanism, automatically adjusts the early warning level and triggers an early warning according to the fuse life prediction result, and the alarm methods include text messages, emails, and App push notifications; A data recording module, configured to: Record all collected data, the working state of the fuse, the life prediction result, and alarm information, store the data through a time series database, and provide a graphical interface for data display.

2. The current-sensing-based new energy vehicle fuse expected life prediction system according to claim 1, wherein The load mode recognition module includes: A current waveform extraction unit, configured to: Sample the real-time collected current signal through a digital converter, perform spectrum analysis on the current signal through fast Fourier transform and extract the current waveform, and extract the current spectrum characteristics based on the current waveform. The current spectrum characteristics include the maximum current value, frequency distribution, waveform period, and change frequency; A load mode classification unit, configured to: Analyze the current spectrum characteristics, perform unsupervised classification on the extracted current spectrum characteristics through clustering and identify different current modes. The current modes include overload, normal load, and transient load. Based on the classification result, generate corresponding load mode labels; A feature data extraction unit, configured to: Extract current spectrum feature data from the classified current waveforms, analyze the identified load modes, and extract the current peak value, current fluctuation frequency, and duration as feature data; Perform multi-dimensional correlation analysis on the current characteristics in combination with vehicle usage conditions to generate high-dimensional feature data for fuse life prediction. Among them, the vehicle usage conditions include vehicle speed and driving environment, and the high-dimensional feature data includes average vehicle speed, vehicle speed fluctuation, environmental temperature, humidity, road type, harmonic distortion, fundamental wave current, each harmonic current, starting current, steady-state current, load switching current, current change rate, current fluctuation energy, root mean square value of current, current standard deviation, number of acceleration times, and number of braking times.

3. The predicted service life system of the fuse for new energy vehicles based on current sensing according to claim 1, characterized in that, The temperature impact calculation module includes: A loss model establishment unit, configured to: Based on the physical characteristics of the fuse and the current-temperature relationship, a heat loss model is established. The physical characteristics include the material and resistance. The real-time collected current signal and temperature data are input into the model to calculate the heat load of the fuse under different working conditions; Output the loss data of the fuse and generate a loss curve; The loss relationship calculation unit is used for: Dynamically adjust the loss value according to the relationship between the collected temperature data and the fuse loss model. Based on the temperature-loss relationship, calculate the correction value of the temperature change on the fuse loss in real time through curve fitting; The loss data correction unit is used for: Based on the historical data and the actual loss situation, perform feedback adjustment, regularly correct the loss calculation model, and correct the parameters in the loss model.

4. The new energy vehicle fuse expected life prediction system based on current sensing according to claim 1, characterized in that, The life prediction module includes: The prediction model training unit is used for: Train the fuse life prediction model using historical data and generate a prediction model. The historical data includes current waveforms, load patterns, and loss data; The real-time life prediction unit is used for: Input the real-time collected current signal, load pattern, and temperature data into the trained life prediction model, calculate the remaining life of the fuse through the model, and generate a life prediction report; The prediction result output unit is used for: Output the prediction result of the remaining life of the fuse to the user side through a graphical interface and an API interface, and generate a historical life prediction curve.

5. The predicted service life system of the fuse for new energy vehicles based on current sensing according to claim 4, wherein The real-time life prediction unit includes: The data acquisition sub-unit is used to acquire the real-time collected current signal, load pattern, and temperature data, and predict the remaining life of the fuse according to the real-time collected current signal, load pattern, and temperature data. Specifically, it includes: The first calculation sub-unit is used to calculate the remaining life of the fuse according to the following formula; S2 = S1 - β * [(Q1 - Q2) * φ] * ∫(I 实际 / I 额定 ) n * dt; Among them, S2 represents the remaining life of the fuse; S1 represents the theoretical life of the fuse; β represents the influence coefficient of the load mode, and its value range is (1.01, 1.02); Q1 represents the actual temperature of the fuse; Q2 represents the reference temperature of the fuse; φ represents the temperature influence coefficient, and its value range is (0.98, 1.02); I 实际 represents the actual current of the fuse; I 额定 represents the rated current of the fuse; n represents the current influence coefficient, and its value range is (0, 1); dt represents the integration with respect to time; The second calculation sub-unit is used to obtain the theoretical remaining life and calculate the comprehensive remaining life of the fuse according to the remaining life and the theoretical remaining life of the fuse; Where, μ represents the comprehensive remaining life of the fuse; The qualified determination sub-unit is used for: Judge whether the comprehensive remaining life of the fuse is consistent with the remaining life of the fuse; If the remaining life of the fuse is consistent with the comprehensive remaining life of the fuse, it is determined that the life prediction is qualified; Otherwise, it is determined that the life prediction is unqualified. At the same time, an alarm signal for alarming the life prediction model is generated.

6. The current-sensing-based new energy vehicle fuse expected life prediction system according to claim 4, wherein The prediction model training unit is also used for: Collect historical data, and perform cleaning, denoising, and normalization processing on the historical data; Extract the necessary features for fuse life prediction from the historical data. The necessary features include the peak value of the current waveform, the frequency of current fluctuation, the change frequency and amplitude of the load pattern, and the temperature change rate. Input the extracted necessary features into the fuse life prediction model and train according to the historical data label. The historical data label is the known fuse life; Verify the accuracy of the model by comparing with the actual fuse life data, and save the trained fuse life prediction model.

7. The current-sensing-based new energy vehicle fuse expected life prediction system according to claim 4, wherein The real-time life prediction unit is also used for: Collect real-time current waveform data, and combine it with the load mode recognition module to obtain real-time load mode data. Extract prediction features from the real-time data, where the prediction features include the instantaneous value of the current, the frequency characteristics of the waveform, the change of the load mode, and the temperature rise rate. Input the extracted prediction features into the trained fuse life prediction model, and the model calculates the remaining life of the fuse based on the input data. Generate a remaining life prediction report for the fuse according to the calculation result of the model. The remaining life prediction report includes the current state of the fuse, the predicted value of the remaining life, and the prediction confidence interval.

8. The current-sensing-based new energy vehicle fuse expected life prediction system according to claim 1, wherein The alarm and early warning module includes: The trigger condition determination unit is used for: Set the remaining life threshold of the fuse. If the life prediction value is lower than the remaining life threshold, trigger the alarm condition, and dynamically adjust the alarm trigger condition based on the prediction result of the remaining life. The multi-level alarm mechanism unit is used for: Set different alarm levels according to the different remaining lives. The alarm levels include severe fault, early warning, and normal. Automatically select and trigger the corresponding alarm level according to the predicted life value. The alarm notification unit is used for: When the alarm condition is triggered, push the triggered alarm information to the user and the maintenance personnel by SMS, email or App push, and provide the alarm history record and detailed information in the alarm information.

9. The current-sensing-based new energy vehicle fuse expected life prediction system according to claim 8, wherein Dynamically adjusting the alarm trigger condition also includes: Set the initial threshold. Set a basic remaining life threshold according to the fuse specifications, historical data and environmental factors. Among them, the basic remaining life threshold is divided into three stage intervals. When the remaining life is less than 20%, it is regarded as a severe fault. When the remaining life is between 20% and 50%, it is an early warning. When the remaining life is more than 50%, it is in a normal state. Set the threshold adjustment period, and perform periodic threshold adjustment based on the adjustment period. In each adjustment period, collect real-time data. Based on the change trend of the remaining life of the fuse, when the remaining life continuously decreases or changes abnormally, dynamically adjust the remaining life threshold. Among them, when it is detected that the remaining life of the fuse continues to decrease, and the decrease amplitude exceeds 5% per cycle, then dynamically reduce the remaining life threshold, and reduce the remaining life threshold of each stage interval by 5%. When it is detected that the change of the remaining life of the fuse shows a sharp decrease or irregular change, then dynamically reduce the remaining life threshold, and reduce the remaining life threshold of each stage interval by 10%. Dynamically weighted based on environmental factors. When the load mode or temperature changes, dynamically adjust the remaining life threshold. Among them, when it is detected that the load fluctuates frequently or the load current value increases significantly, then dynamically reduce the remaining life threshold, and reduce the remaining life threshold of each stage interval by 10%. If the working temperature of the fuse exceeds the preset safe range, then dynamically reduce the remaining life threshold, and reduce the remaining life threshold of each stage interval by 10%. When the load increases and the temperature exceeds the safe range, then dynamically reduce the remaining life threshold, and at the same time weight the threshold reduction amplitude, and reduce the remaining life threshold of each stage interval by 10%-20%.

10. The expected life prediction system of a fuse for a new energy vehicle based on current sensing according to claim 1, characterized in that, The data recording module includes: A record dimension determination unit, configured to obtain a record dimension for recording data to be recorded, construct a dimension label according to the record dimension, and construct a corresponding data record interval according to the dimension label, where the dimension label and the data record interval are in one-to-one correspondence; A data mapping unit, configured to determine a data record mapping relationship according to the dimension label and the data to be recorded, and map the data to be recorded in the corresponding data record interval according to the mapping relationship according to the dimension label; A sorting unit, configured to sort the target record data in each data record interval according to the time sequence according to the mapping result, and obtain a time sequence data sequence for each data record interval; An index generation unit, configured to obtain the time points of the time sequence data sequence, generate a column index according to the dimension label, and generate a row index according to the time points, where the time nodes and the row index are in one-to-one correspondence; A data storage unit, configured to: Integrate all data record intervals based on the column index and the row index to generate a data storage table; Store the data storage table in a time series database; A data display unit, configured to: When a data display requirement is input based on a user terminal, extract the time keyword and the data dimension keyword of the data display requirement; Match the time keyword with the column index to determine a target column index that is consistent with the time keyword; Match the data dimension keyword with the row index to determine a target row index that is consistent with the data dimension keyword; Match a target display template in a preset display template library according to the target row index; Retrieve data to be displayed that is consistent with the data display requirement from the time series database according to the target row index and the target column index; Map the data to be displayed to the target display template to obtain a target graphical interface, and display the target graphical interface.

Images