A user vehicle maintenance status prediction management method based on big data analysis
Through big data analysis and neural network model, the vehicle loss risk is judged in real time and the maintenance time is adjusted, which solves the problems of untimely maintenance and high maintenance costs in the existing technology, and improves the quality of vehicle maintenance and driving safety.
Patent Information
- Application Number
- CN202411169214.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-23
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-08-23
AI Technical Summary
The existing vehicle maintenance methods mainly rely on regular maintenance and post-malfunction repairs, resulting in problems such as untimely maintenance and high maintenance costs.
Adopt the user vehicle maintenance status prediction management method based on big data analysis, and collect vehicle usage information and driving habit information in real time, calculate vehicle usage status indicators and driving habit indicators, and determine whether there is a risk of loss exceeding the standard. If it exists, shorten the maintenance time, otherwise Maintain the original time. This method uses vehicle-mounted sensors, GPS positioning system and neural network model to analyze vehicle status and driving habit data in real time and adjust maintenance cycles.
It improves the accuracy and safety of vehicle maintenance, reduces maintenance costs, avoids the problems of untimely maintenance and high maintenance costs, and improves the efficiency of vehicle usage and driving safety.
Smart Images

Figure CN118674436B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of vehicle management, and in particular to a method for predicting and managing the maintenance status of a user's vehicle based on big data analysis. Background Art
[0002] With the continuous innovation and popularization of information technology, especially the widespread application of technologies such as the Internet of Things (IoT), cloud computing, and big data, strong technical support has been provided for vehicle maintenance management. These technologies make it possible to collect, transmit, store, and process vehicle operation data in real time, providing a rich data source for the prediction of vehicle maintenance status.
[0003] Modern vehicles are equipped with a large number of sensors and intelligent devices, such as OBD interfaces, GPS positioning systems, on-board cameras, etc. These devices can collect real-time data such as vehicle speed, fuel consumption, location, engine status, mileage, etc. These data not only reflect the operating status of the vehicle, but also contain rich vehicle performance information and potential fault information.
[0004] Good vehicle maintenance can significantly improve vehicle reliability and safety, reduce maintenance costs, and improve vehicle efficiency; however, existing vehicle maintenance methods are still mainly based on regular maintenance and post-fault repairs. Compared with predictive maintenance, this method obviously has the problems of untimely maintenance and high maintenance costs. Summary of the invention
[0005] The purpose of the present invention is to provide a user vehicle maintenance status prediction management method based on big data analysis, so as to solve the problem that the existing vehicle maintenance method is still mainly based on regular maintenance and post-fault repair. Compared with predictive maintenance, this method obviously has the technical problems of untimely maintenance and high maintenance costs.
[0006] The purpose of the present invention can be achieved through the following technical solutions:
[0007] A method for predicting and managing the maintenance status of user vehicles based on big data analysis comprises the following steps:
[0008] Step 1: Collect the usage information of each vehicle and the changes of the usage information of the vehicle over time in multiple consecutive monitoring cycles in real time;
[0009] Step 2: Calculate the vehicle usage status index based on the usage information of each vehicle and the change of the vehicle usage information over time in multiple consecutive monitoring cycles;
[0010] Step 3: Obtain driving habit indicators based on the information processing and analysis of the driving habits of the vehicle owner during multiple monitoring periods;
[0011] Step 4: Determine whether the wear and tear of each vehicle is at risk of exceeding the standard based on the vehicle usage status index and driving habit index. If there is a risk of exceeding the standard, shorten the next vehicle maintenance time; otherwise, maintain the next vehicle maintenance time unchanged.
[0012] As a further technical solution, the specific steps of step 2 are as follows:
[0013] Collect every usage information of the vehicle based on vehicle-mounted sensors and GPS positioning system;
[0014] Substitute each usage information into the formula: Calculate the real-time usage status coefficient of the current vehicle ;
[0015] Then through the formula: Calculate the current vehicle's stage usage status coefficient ;
[0016] Then through the formula: Calculate the current vehicle usage status index ;
[0017] in, is the vehicle real-time usage status coefficient, Using state coefficients for the vehicle phase, For the The parameter value of the item usage information, For the preset The warning value of the item usage information, , is the preset weight coefficient, ~ is the starting point and end point of the monitoring period, is the number of items of usage information, For the The curve of item usage information changing over time, For the A reference curve showing the change of item usage information over time. , is the preset scale factor, is the number of monitoring cycles, For the The reference coefficient for a monitoring period.
[0018] As a further technical solution, the specific steps of step three are as follows:
[0019] Input each driving habit information of the car owner during multiple monitoring cycles into the pre-trained driving habit model and output the driving habit index ;
[0020] The driving habit model includes a fully connected neural network and a recurrent neural network.
[0021] As a further technical solution, a method for judging whether the loss of each vehicle is at risk of exceeding the standard according to the vehicle usage status index and driving habit index is as follows:
[0022] The calculated vehicle usage status index and driving habits indicators The corresponding set thresholds , Make comparisons;
[0023] like and , it is judged that the current vehicle loss is at risk of exceeding the standard;
[0024] like or , then the risk of the vehicle's loss exceeding the standard is to be determined;
[0025] like and , it is judged that there is no risk of exceeding the standard for the current vehicle's loss.
[0026] As a further technical solution, or When the vehicle usage status index is calculated and driving habits indicators Substitute the following formula: Calculate the current vehicle's over-limit risk prediction index ;
[0027] in, , is the judgment function, when hour, ,when hour, ;when hour, ,when hour, 0;
[0028] Exceeding the standard risk prediction index Compared with the preset risk prediction threshold Make a comparison and determine whether the risk of excessive wear and tear on the current vehicle is high or low based on the comparison results.
[0029] As a further technical solution, the process of judging whether the probability of the current vehicle's loss exceeding the standard risk is large or small according to the comparison result is as follows:
[0030] when When , it is judged that the risk probability of the current vehicle's loss exceeding the standard is high;
[0031] when When , it is judged that the risk probability of the current vehicle's loss exceeding the standard is small.
[0032] As a further technical solution, the process of shortening the next vehicle maintenance time is as follows:
[0033] Get the next maintenance time based on the vehicle system , get the current time when the risk of exceeding the standard is judged, get the current time when the risk of exceeding the standard is judged ;
[0034] By formula: Calculate the maintenance interval of the current vehicle ;
[0035] The maintenance interval Substituting into the formula: Calculate the adjusted maintenance interval of the current vehicle ;in, is the judgment function, is the dynamic coefficient, is the preset dynamic threshold. < hour, ,when ≥ hour, ;
[0036] Then through Calculate the next maintenance time of the current vehicle .
[0037] As a further technical solution, the process of obtaining the dynamic coefficient is:
[0038] By formula: Calculate the dynamic coefficient of the current vehicle ;
[0039] in, is the mean value of vehicle usage status index, is the mean of driving habit index, , is the preset weight factor, is the total number of sampling points, For the sampling points.
[0040] Beneficial effects of the present invention:
[0041] The present invention obtains a vehicle usage status index and a driving habit index respectively according to the vehicle usage information and the driving habit information of the owner, wherein the driving habit index is obtained by first inputting each driving habit information of the owner in multiple monitoring cycles into a pre-trained neural network model, and then outputting the driving habit index. Since the important factors affecting the vehicle maintenance cycle are the vehicle usage information and the driving habit information of the owner, a maintenance prediction model is established after the vehicle itself and the driving habit of the owner are associated through the two, so as to realize accurate judgment on whether the vehicle maintenance time should be adjusted, thereby achieving the purpose of improving the vehicle maintenance quality and improving the vehicle driving safety, and preventing the problems of untimely maintenance and high maintenance costs caused by regular maintenance and post-fault repair. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] The present invention will be further described below in conjunction with the accompanying drawings.
[0043] Figure 1 It is a diagram of the method steps of the present invention. DETAILED DESCRIPTION
[0044] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0045] See also Figure 1 As shown, the present invention is a method for predicting and managing the maintenance status of user vehicles based on big data analysis, comprising the following steps:
[0046] Step 1: Collect the usage information of each vehicle and the changes of the usage information of the vehicle over time in multiple consecutive monitoring cycles in real time;
[0047] Step 2: Calculate the vehicle usage status index based on the usage information of each vehicle and the change of the vehicle usage information over time in multiple consecutive monitoring cycles;
[0048] Step 3: Obtain driving habit indicators based on the information processing and analysis of the driving habits of the vehicle owner during multiple monitoring periods;
[0049] Specifically, each driving habit information of the car owner during multiple monitoring cycles is input into the pre-trained driving habit model, and the driving habit index is output. ;
[0050] The driving habit model includes a fully connected neural network and a recurrent neural network. The process of constructing a driving habit model composed of a fully connected neural network FCN and a recurrent neural network RNN is as follows:
[0051] Data preprocessing: preprocessing the owner’s driving habits information, including normalization, standardization, padding or truncating time series data, and converting non-sequence data into a format suitable for FCN;
[0052] Construct the RNN layer, select the RNN variant LSTM, determine the number of RNN layers and the number of units in each layer to process the time series data in the owner's driving habit information;
[0053] Connect the FCN layer, use the output of the RNN layer as the input of the FCN, and build a fully connected layer for feature extraction and classification;
[0054] Model compilation, select loss function, optimizer and evaluation metric to compile driving habit model;
[0055] Model training: Use the training data to train the model and validate it on the validation set to adjust the hyperparameters;
[0056] Model evaluation and tuning: Evaluate the performance of the model on the test set and tune it as needed.
[0057] In the field of deep learning, in order to process complex data and tasks, it is often necessary to combine different types of neural networks to make full use of their respective advantages. The composite model of RNN and FCN is one of the mature technologies. Therefore, the construction process of the above model is an existing technology and will not be elaborated here.
[0058] The driving habit information of the vehicle owner includes but is not limited to the number of accelerations, brakings, turns, emergency stops, sharp turns, emergency accelerations, driving time and driving distance per day; the above driving habit information can be obtained through vehicle-mounted sensors or video monitoring data, which is existing technology and will not be elaborated here;
[0059] Step 4: Determine whether the wear and tear of each vehicle is at risk of exceeding the standard based on the vehicle usage status index and driving habit index. If there is a risk of exceeding the standard, shorten the next vehicle maintenance time; otherwise, maintain the next vehicle maintenance time unchanged.
[0060] The specific steps of step 2 are as follows:
[0061] Based on the vehicle-mounted sensors and GPS positioning system, each item of vehicle usage information is collected; the usage information includes mileage, usage frequency, usage time period, average speed, maximum speed, acceleration, engine temperature scale, engine oil pressure, engine speed, number of brakes, braking distance, braking speed, exhaust temperature scale, exhaust pressure, exhaust flow, temperature, humidity and air quality. Each of the above usage information is monitored and obtained through existing sensors or detection instruments, and will not be elaborated here;
[0062] Substitute each usage information into the formula: Calculate the real-time usage status coefficient of the current vehicle ;
[0063] Then through the formula: Calculate the current vehicle's stage usage status coefficient ;
[0064] Then through the formula: Calculate the current vehicle usage status index ;
[0065] in, is the vehicle real-time usage status coefficient, Using state coefficients for the vehicle phase, For the The parameter value of the item usage information, For the preset The warning value of the item usage information, , The preset weight coefficient is formulated based on historical data and experimental data. ~ is the starting point and end point of the monitoring period, is the number of items of usage information, For the The curve of item usage information changing over time, For the A reference curve of item usage information changing over time, obtained based on big data analysis. , It is a preset proportionality factor, which is selected and determined based on historical data and empirical data. is the number of monitoring cycles, For the The reference coefficients for each monitoring period are formulated based on historical experimental data.
[0066] In this embodiment, in order to solve the problem that the existing vehicle maintenance method is still mainly based on regular maintenance and post-fault repair, which obviously has the problems of untimely maintenance and high maintenance cost compared with predictive maintenance, the vehicle usage status index and driving habit index are obtained according to the vehicle usage information and the owner's driving habit information, wherein the driving habit index is obtained by first inputting each driving habit information of the owner in multiple monitoring cycles into a pre-trained neural network model, and then outputting the driving habit index. It is found that the important factors affecting the vehicle maintenance cycle are vehicle usage information and the owner's driving habits. Therefore, a maintenance prediction model is established by associating the vehicle itself with the owner's driving habits to accurately determine whether the vehicle maintenance time should be adjusted, thereby achieving the purpose of improving vehicle maintenance quality and improving vehicle driving safety.
[0067] At the same time, a method for obtaining vehicle usage status indicators is also provided. Specifically, various vehicle usage information is first substituted into , calculate the real-time usage status coefficient of the current vehicle It is obvious that if the ratio between the parameter value of each usage information and the warning value is larger, it means that each usage information is closer to the warning value, so the usage information is more dangerous. Then, the ratio between the parameter value of each usage information and the warning value is accumulated by summing up, so as to obtain the overall usage of the current vehicle, the real-time usage status coefficient of the current vehicle It can quickly and in real time reflect the status of the vehicle. Once the usage information exceeds the warning value, it means that the next maintenance interval of the current vehicle is shorter. Otherwise, the vehicle condition is likely to deteriorate, posing a driving safety hazard.
[0068] Then through the formula Calculate the current vehicle's stage usage status coefficient The stage usage status coefficient reflects the changes in various usage information of the vehicle in multiple historical monitoring cycles; it can be clearly seen that The larger the value, the more The greater the deviation between the curve of item usage information changing over time and the reference curve, the greater the cumulative change, indicating that the The more unstable the usage information is, that is, there is a sudden increase or decrease in the vehicle usage status, which indicates that it is more likely to have safety problems. Therefore, the deviation of each usage information is summed up by accumulation, and then multiple historical monitoring cycles are accumulated to obtain a more accurate usage status of the vehicle connection stage; finally, the formula is used Calculate the current vehicle usage status index Obviously, through the above technical solution, the vehicle real-time usage status coefficient and the vehicle stage usage status coefficient are multiplied as the numerator and weighted as the denominator, so that the change range of the numerator is greater than the change range of the denominator. Even if one of the coefficients changes slightly, it can be amplified and reflected, thereby achieving the purpose of accurately obtaining the current vehicle usage status index. Obviously, the larger the current vehicle usage status index is, the worse the current vehicle usage status is, and vice versa.
[0069] The method for judging whether the loss of each vehicle is at risk of exceeding the standard based on the vehicle usage status index and driving habit index is as follows:
[0070] The calculated vehicle usage status index and driving habits indicators The corresponding set thresholds , Make comparisons;
[0071] like and , it is judged that the current vehicle loss is at risk of exceeding the standard;
[0072] like or , then the risk of the vehicle's loss exceeding the standard is to be determined;
[0073] like and , it is judged that there is no risk of exceeding the standard for the current vehicle's loss.
[0074] In this embodiment, a specific method for judging whether the loss of each vehicle has a risk of exceeding the standard is provided, that is, by calculating the vehicle usage status index and driving habits indicators The corresponding set thresholds , By comparing, it is obvious that and , it indicates that the vehicle is in poor condition and the owner has bad driving habits, which together lead to a high risk of excessive vehicle wear and tear. Therefore, it is judged that the current vehicle wear and tear has a risk of exceeding the standard; when or When one of the conditions is met, it indicates that the vehicle is in good condition or the owner has good driving habits, and the risk of excessive wear and tear of the current vehicle is to be determined; when and When , it indicates that the vehicle is in good condition and the owner's driving habits are good, which leads to a low risk of excessive wear and tear of the vehicle. It is judged that there is no risk of excessive wear and tear of the current vehicle. Through the above judgment and analysis process, it can provide more accurate judgment support for the subsequent adjustment of whether to shorten the maintenance time of the vehicle, while taking into account timely maintenance to maintain driving safety, reducing the cost increase and time waste caused by ineffective shortening of maintenance time.
[0075] when or When the vehicle usage status index is calculated and driving habits indicators Substitute the following formula: Calculate the current vehicle's over-limit risk prediction index ;
[0076] in, , is the judgment function, when hour, ,when hour, ;when hour, ,when hour, 0;
[0077] Exceeding the standard risk prediction index Compared with the preset risk prediction threshold Make a comparison and determine whether the risk of excessive wear and tear on the current vehicle is high or low based on the comparison results.
[0078] The process of judging whether the risk probability of excessive loss of the current vehicle is high or low based on the comparison results is as follows:
[0079] when When , it is judged that the risk probability of the current vehicle's loss exceeding the standard is high;
[0080] when When , it is judged that the risk probability of the current vehicle's loss exceeding the standard is small.
[0081] In this embodiment, a method for predicting and analyzing a vehicle in a predetermined state is provided. Specifically, the vehicle usage state index obtained by calculation is first and driving habits indicators Substitute the following formula:
[0082] The over-standard risk prediction index of the current vehicle can be calculated Obviously, when the vehicle usage status index and driving habit index exceed the corresponding set threshold, it means that the current vehicle usage status is worse and the owner's driving habits are worse, so the vehicle is more likely to have a higher risk of excessive wear and tear in the subsequent driving process. On the contrary, the trend towards normal wear and tear is stronger, and the excess risk prediction index is Compared with the preset risk prediction threshold For comparison, When , it is judged that the current vehicle has a high probability of exceeding the loss standard risk. When , it is judged that the risk probability of excessive wear and tear of the current vehicle is small; in order to realize the probability of excessive wear and tear risk of the vehicle in the pending state in the subsequent driving process, the purpose of early prevention and timely maintenance is achieved.
[0083] The process of shortening the next vehicle maintenance time is as follows:
[0084] Get the next maintenance time based on the vehicle system , get the current moment when the risk of exceeding the standard is determined ;
[0085] By formula: Calculate the maintenance interval of the current vehicle ;
[0086] The maintenance interval Substituting into the formula: Calculate the adjusted maintenance interval of the current vehicle ;in, is the judgment function, is the dynamic coefficient, is the preset dynamic threshold. < hour, ,when ≥ hour, ;
[0087] Then through Calculate the next maintenance time of the current vehicle .
[0088] The process of obtaining the dynamic coefficient is:
[0089] By formula: Calculate the dynamic coefficient of the current vehicle ;
[0090] in, is the mean value of vehicle usage status index, is the mean of driving habit index, , is a preset weight factor, determined based on historical data analysis. is the total number of sampling points, For the sampling points.
[0091] In this embodiment, vehicles are divided into dangerous vehicles with a risk of excessive wear and tear, pending vehicles with a high probability of excessive wear and tear, pending vehicles with a low probability of excessive wear and tear, and safe vehicles with no risk of excessive wear and tear; among them, dangerous vehicles and pending vehicles with a high probability of excessive wear and tear are judged to have a risk of excessive wear and tear and are marked in red, pending vehicles with a low probability of excessive wear and tear are marked in yellow, and safe vehicles with no risk of excessive wear and tear are marked in green;
[0092] Therefore, all vehicles marked in red are judged to have the risk of exceeding the standard, and then the formula is used: Calculate the maintenance interval of the current vehicle ; Then the maintenance interval Substituting into the formula: and The adjusted maintenance interval of the current vehicle is calculated ;in, is the judgment function, is the dynamic coefficient, is the preset dynamic threshold. < hour, ,when ≥ hour, ;
[0093] Last passed Calculate the next maintenance time of the current vehicle Through the above technical solution, a specific method for adjusting the maintenance time point of vehicles with excessive risk is realized, which is close to the actual situation of the vehicle. Obtain the fluctuations in the vehicle's usage status and the owner's driving habits. Obviously, the greater the fluctuation, the more unstable the vehicle's status is, and therefore the higher the probability of driving risks. Therefore, timely maintenance is needed to achieve a stable vehicle status, thereby preventing driving safety risks.
[0094] It should be noted that the calculation formulas and various parameters involved in the calculation in the present invention have been dimensionally processed in advance, and the process of dimensionless processing is well known in the industry and will not be described here.
[0095] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.
Claims
1. A method for predicting and managing the maintenance status of user vehicles based on big data analysis, characterized in that: The steps include: Step 1: Collect the usage information of each vehicle and the changes of the usage information of the vehicle over time in multiple consecutive monitoring cycles in real time; Step 2: Calculate the vehicle usage status index based on the usage information of each vehicle and the change of the vehicle usage information over time in multiple consecutive monitoring cycles; Step 3: Obtain driving habit indicators based on the information processing and analysis of the driving habits of the vehicle owner during multiple monitoring periods; Step 4: Determine whether the wear and tear of each vehicle is at risk of exceeding the standard based on the vehicle usage status index and the driving habit index. If there is a risk of exceeding the standard, shorten the next vehicle maintenance time; otherwise, keep the next vehicle maintenance time unchanged; The specific steps of step 2 are as follows: Collect every usage information of the vehicle based on vehicle-mounted sensors and GPS positioning system; Substitute each usage information into the formula: Calculate the current vehicle's real-time usage status coefficient ; Then through the formula: Calculate the current vehicle's stage usage status coefficient ; Then through the formula: Calculate the current vehicle usage status index ; in, is the vehicle real-time usage status coefficient, Using state coefficients for the vehicle phase, For the The parameter value of the item usage information, For the preset The warning value of the item usage information, , is the preset weight coefficient, ~ is the starting point and end point of the monitoring period, is the number of items of usage information, For the The curve of item usage information changing over time, For the A reference curve showing the change of item usage information over time. , is the preset scale factor, is the number of monitoring cycles, For the Reference coefficients for each monitoring period; The specific steps of step three are as follows: Input each driving habit information of the car owner during multiple monitoring cycles into the pre-trained driving habit model and output the driving habit index ; The driving habit model includes a fully connected neural network and a recurrent neural network; The method for judging whether the loss of each vehicle is at risk of exceeding the standard based on the vehicle usage status index and driving habit index is as follows: The calculated vehicle usage status index and driving habits indicators The corresponding set thresholds , Make comparisons; like and , it is judged that the current vehicle loss is at risk of exceeding the standard; like or , then the risk of the vehicle's loss exceeding the standard is to be determined; like and , it is judged that the current vehicle loss does not have the risk of exceeding the standard; when or When the vehicle usage status index is calculated and driving habits indicators Substitute the following formula: Calculate the current vehicle's over-limit risk prediction index ; in, , is the judgment function, when hour, ,when hour, ; when hour, ,when hour, 0; Exceeding the standard risk prediction index Compared with the preset risk prediction threshold Make a comparison and determine whether the risk of excessive wear and tear on the current vehicle is high or low based on the comparison results.
2. The user vehicle maintenance status prediction management method based on big data analysis according to claim 1 is characterized in that: The process of judging whether the risk probability of excessive loss of the current vehicle is high or low based on the comparison results is as follows: when When , it is judged that the risk probability of the current vehicle's loss exceeding the standard is high; when When , it is judged that the risk probability of the current vehicle's loss exceeding the standard is small.
3. The user vehicle maintenance status prediction management method based on big data analysis according to claim 1 or 2 is characterized in that: The process of shortening the next vehicle maintenance time is as follows: Get the next maintenance time based on the vehicle system , get the current moment when the risk of exceeding the standard is determined ; By formula: Calculate the maintenance interval of the current vehicle ; The maintenance interval Substituting into the formula: Calculate the adjusted maintenance interval of the current vehicle ; in, is the judgment function, is the dynamic coefficient, is the preset dynamic threshold. < hour, ,when ≥ hour, ; Then through Calculate the next maintenance time of the current vehicle .
4. The user vehicle maintenance status prediction management method based on big data analysis according to claim 3 is characterized in that: The process of obtaining the dynamic coefficient is: By formula: Calculate the dynamic coefficient of the current vehicle ; in, is the mean value of vehicle usage status index, is the mean of driving habit index, , is the preset weight factor, is the total number of sampling points, For the sampling points.
