Vehicle air conditioner fault prediction method and system based on local outlier factor algorithm

Abstract

The application belongs to the field of new energy vehicle intelligent networking, and particularly relates to a vehicle-mounted air conditioner fault prediction method and system based on a local outlier factor algorithm, which comprises the following steps: identifying outliers in key indicator data of a full-quantity vehicle in the same city as a to-be-tested vehicle by using a LOF algorithm to obtain a fault vehicle list; identifying key indicator data of each vehicle in the fault vehicle list in a current day and a recent time period of the to-be-tested vehicle by using the LOF algorithm to obtain a first prediction result; identifying key indicator data of each vehicle in the fault vehicle list in a current day and a historical same-period time period of the to-be-tested vehicle by using the LOF algorithm to obtain a second prediction result; and determining a final prediction result according to the first prediction result and the second prediction result. Thus, the application solves the problems of low accuracy, high cost and poor applicability of the existing air conditioner fault prediction method.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of intelligent networking technology of new energy vehicles, and particularly relates to a vehicle-mounted air conditioner fault prediction method and system based on a local outlier factor algorithm. BACKGROUND

[0002] In summer, a car is exposed to the sun for a long time outdoors at 40℃, the temperature in the car is above 70 degrees, and the highest temperature can reach 90 degrees. The cooling effect of the car air conditioner is crucial. For a passenger car, the normal operation of the air conditioner can ensure the normal operation of the customer, and the source of the cooling system of the battery motor is also the refrigeration function of the air conditioner.

[0003] In summary, the high frequency of use of the air conditioner causes the failure rate to rise accordingly, and at the same time, due to the fact that the air conditioner contains numerous parts and is distributed throughout the space on the roof of the vehicle, the difficulty of maintenance and repair is also greatly increased. If the real-time data reported through the Internet of Vehicles can be used to identify air conditioner failures in advance, it will have important guiding significance for the normal operation of customers and the improvement of product design.

[0004] At present, the processing method for vehicle-mounted air conditioner faults mainly remains in the form of customer repair, and the driver or passenger has discovered that the air conditioner cannot be turned on or has an abnormal noise, does not cool, and cannot achieve the cooling effect during driving or maintenance, which has affected the normal operation of the vehicle, and the manufacturer's after-sales service is notified to repair. At this time, the fault is already quite serious, the repair cost is high, and there is a risk of passenger complaints.

[0005] Some air conditioner manufacturers install various temperature, humidity, air pressure, and vibration sensors to monitor the working state of the air conditioner. However, due to the variable working conditions during driving, road bumps, passenger numbers, and signal strength, and many external disturbances, the collected data fluctuates greatly, the prediction accuracy is low, and not only cannot be applied to the existing vehicles, but also increases the cost of the vehicle.

[0006] In addition, the application of some algorithms to air conditioner fault prediction is based on air conditioners installed on buildings, and the surrounding environment is stable, which cannot be popularized to vehicle-mounted air conditioners whose environment is constantly changing, and the applicability is poor. For example, the method for predicting air conditioner faults using a distance algorithm disclosed in the Chinese patent application publication No. CN111412579A, but this method is based on air conditioners installed on buildings, and the surrounding environment is almost constant, which cannot be applied to vehicle-mounted air conditioners, and requires model training before application. Different types of air conditioner working parameters differ greatly, and different algorithms are needed to support different types of air conditioner units. The operation and maintenance cost is high, and the accuracy is not high, resulting in invalid after-sales operation and maintenance cost.

[0007] In summary, the existing air conditioner fault prediction method has the problems of low accuracy, high cost, and poor applicability. SUMMARY

[0008] The application aims to provide a vehicle-mounted air conditioner fault prediction method and system based on a local outlier factor algorithm, so as to solve the problems of low accuracy, high cost and poor applicability of the air conditioner fault prediction method in the prior art.

[0009] To solve the above technical problems, the technical solutions provided by the application and the beneficial effects corresponding to the technical solutions are as follows:

[0010] The application provides a vehicle-mounted air conditioner fault prediction method based on a local outlier factor algorithm, which comprises the following steps:

[0011] 1) Obtain indexes related to air conditioner faults to obtain a key index set;

[0012] 2) Select a set number of vehicles in a geographical area to which the to-be-tested vehicle belongs as sample vehicles, and obtain key index data of the sample vehicles; use a LOF outlier factor algorithm to identify outliers in the key index data of the sample vehicles to obtain a fault vehicle list, wherein the fault vehicle list includes fault vehicles and fault information;

[0013] 3) Use the LOF outlier factor algorithm to identify outliers in the first data and the second data to obtain a first prediction result; use the LOF outlier factor algorithm to identify outliers in the third data and the fourth data to obtain a second prediction result; the first data includes key index data of the to-be-tested vehicle on the current day and in a recent time period; the second data includes key index data of each vehicle in the fault vehicle list on the current day and in a corresponding recent time period; the third data includes key index data of the to-be-tested vehicle on the current day and in a historical same period; and the fourth data includes key index data of each vehicle in the fault vehicle list on the current day and in a corresponding historical same period;

[0014] 4) Determine a final air conditioner fault result according to the first prediction result and the second prediction result.

[0015] The beneficial effects of the above technical solutions are: 1) the present application selects a large number of sample vehicles from the geographical area to which the vehicle to be tested belongs, identifies outliers in the key indicator data of the sample vehicles according to the LOF anomaly factor algorithm, obtains a small number of faulty vehicles and stores them in the fault vehicle list, and the present application selects vehicles in the geographical area to which the vehicle to be tested belongs as sample data, reduces the influence of environmental changes, and purposefully reduces the sample size according to the key indicators and the LOF anomaly factor algorithm, thereby reducing the calculation amount and improving the prediction accuracy; 2) the LOF anomaly factor algorithm identifies outliers in the data of the vehicles in the fault vehicle list, the data of the vehicle to be tested on the same day, and the data in the recent time period, obtains the first prediction result, the data range identified is accurate and wide, and the accuracy of the prediction result is improved; 3) the LOF anomaly factor algorithm is used to identify outliers in the data of the vehicles in the fault vehicle list, the data of the vehicle to be tested on the same day, and the data in the historical same period, to obtain the second prediction result, and the second prediction result is used to verify the prediction result again, thereby further improving the accuracy.

[0016] In summary, in the first aspect, the present application is suitable for fault prediction of all vehicle-mounted air conditioners and fault prediction of air conditioners installed in buildings, has a wide application range and strong applicability, solves the industry status of passive discovery of vehicle part faults, obtains air conditioner fault information in advance, avoids further serious faults, reduces maintenance costs, and improves vehicle operation income. In the second aspect, the present application is based on a verification method of three different sample sets (data of vehicles in the geographical area to which the vehicle to be tested belongs, data in the recent time period, and data in the historical same period), has high prediction result accuracy, and does not need to periodically train models and optimize model thresholds as in most algorithms in the prior art, thereby solving the high maintenance cost of different prediction models corresponding to different models due to different control strategies of different air conditioner models.

[0017] Further, in order to improve the accuracy, in step 4), if the first prediction result and the second prediction result are both outliers, the compressor of the vehicle to be tested has a fault, otherwise the compressor of the vehicle to be tested is normal.

[0018] Further, in order to improve the accuracy, in step 1), the following method is used to obtain the key indicator set:

[0019] Sample data of air conditioner repair orders are obtained, the sample data of the air conditioner repair orders are subjected to fault correlation analysis based on the basic indicator set, indicators with strong correlation are selected as key indicators, and a key indicator set is obtained; the key indicator set includes the proportion of the compressor working end not reaching the set temperature and the average end temperature difference.

[0020] Further, the basic index set is derived from the basic data sent by the vehicle end; the basic data includes: the temperature in the vehicle, the air conditioner setting temperature, the air conditioner working state, the air conditioner compressor working state, the whole vehicle driving state, the door opening and closing state, the compressor working end not reaching the set temperature ratio, and the end temperature difference interval average.

[0021] Further, the basic index set includes: the compressor working end not reaching the set temperature ratio and the end temperature difference average; the basic index set further includes at least one of the compressor working time length, the temperature in the vehicle start value and end value, the air conditioner setting temperature start value and end value, the start temperature difference, the end temperature difference, the temperature in the vehicle change value, the end temperature difference, the working time length, and the compressor working end not reaching the set temperature ratio.

[0022] The calculation formula of the compressor working end not reaching the set temperature ratio is: Σ (the end temperature difference>0 and the working time is less than n minutes) / Σ (the compressor working times), n∈(5,15);

[0023] The end temperature difference average is obtained in the following way:

[0024] When the temperature difference is <-2, the first temperature difference is -2; when the temperature difference is >=-2 and the temperature difference is <=5, the second temperature difference is the actual value; when the temperature difference is >5, the third temperature difference is 5; the end temperature difference average is the average of the first temperature difference, the second temperature difference, and the third temperature difference;

[0025] The compressor working time length is obtained according to the air conditioner working state; the temperature in the vehicle start value and end value is obtained according to the temperature in the vehicle and the corresponding time point; the air conditioner setting temperature start value and end value is obtained according to the air conditioner setting temperature; the start temperature difference is the temperature in the vehicle at the air conditioner working start time minus the air conditioner setting temperature; the end temperature difference is the temperature in the vehicle at the end time minus the air conditioner setting temperature; the working time length is obtained according to the air conditioner working state and the corresponding time point.

[0026] Further, in order to improve the accuracy, the basic data is the data after removing the abnormal invalid data, and the abnormal invalid data includes empty data and extreme data.

[0027] Further, in order to improve the accuracy, in step 2), the recent time period is from the current day to the previous k days of the current day, k∈[25,35]; the historical same period time period is the same time period of the previous year.

[0028] Further, in order to improve the accuracy, in step 2), the geographical area to which the vehicle to be tested belongs is the same city as the vehicle to be tested, and the set number is all vehicles in the same city.

[0029] In order to solve the above problems, the application further provides a vehicle-mounted air conditioner fault prediction system based on a local outlier factor algorithm, which comprises a processor used for executing computer instructions to realize a vehicle-mounted air conditioner fault prediction method based on a local outlier factor algorithm of the application, and achieves the same beneficial effects. BRIEF DESCRIPTION OF DRAWINGS

[0030] Figure 1 is a flowchart of a vehicle-mounted air conditioner fault prediction method based on a local outlier factor algorithm of the application;

[0031] Figure 2 is a prediction result distribution map based on the prediction results of all vehicles in the same city by using the LOF algorithm in the method embodiment of the application;

[0032] Figure 3 is a prediction result distribution map based on the air conditioner data of the vehicle to be tested on the current day and in the past 30 days by using the LOF algorithm in the method embodiment of the application;

[0033] Figure 4 is a prediction result distribution map based on the air conditioner working data of the vehicle to be tested on the current day and in the same period of last year by using the LOF algorithm in the method embodiment of the application;

[0034] Figure 5 is a method schematic diagram for confirming the vehicle to be tested as a fault vehicle in the method embodiment of the application. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical scheme and advantages of the application more clear and obvious, the application is further described in detail below in combination with the drawings and embodiments.

[0036] Method embodiment:

[0037] An embodiment of a vehicle-mounted air conditioner fault prediction method based on a local outlier factor algorithm of the application is based on the LOF outlier factor algorithm of big data, and only uses the existing reported data of the vehicle to realize the prediction of air conditioner faults. The application realizes fault prediction based on the evaluation of the cooling effect of the air conditioner.

[0038] Among them, LOF: Local outlier factor, that is, local outlier factor. LOF mainly judges whether a point p is an abnormal point by comparing the density of each point p and its neighborhood point. If the density of point p is lower, it is more likely to be identified as an abnormal point, and the density is obtained by calculating the distance between points. The farther the distance, the lower the density, and the closer the distance, the higher the density.

[0039] The application will be described below in combination with Figure 1 and specific steps:

[0040] Step 1: Construct an index set.

[0041] According to the basic data reported by the vehicle end, the business derivative index is calculated, and the basic index set is constructed according to the business derivative index.

[0042] 1) The existing vehicle end is used to collect basic data, and the basic data is uploaded to the cloud end through the 4G / 5G network for data analysis and storage. The basic data includes the temperature in the vehicle, the air conditioner set temperature, the air conditioner working state, the air conditioner compressor working state, the vehicle driving state and the door opening and closing state. The business derivative index includes the compressor working time, the initial and final values of the temperature in the vehicle, the initial and final values of the air conditioner set temperature, the initial temperature difference, the final temperature difference, the temperature change value in the vehicle, the working time, the proportion of the compressor working end not reaching the set temperature, and the average of the final temperature difference.

[0043] The compressor working time is calculated according to the air conditioner working state. The initial and final values of the temperature in the vehicle are obtained by filtering according to the temperature in the vehicle and the corresponding time. The final temperature difference is the temperature in the vehicle at the end time minus the set temperature. The initial and final values of the air conditioner set temperature: the first set temperature and the last set temperature set during the air conditioner working time. The initial temperature difference is the temperature in the vehicle at the air conditioner working start time minus the set temperature. The temperature change value in the vehicle is the temperature in the vehicle at the air conditioner working start time minus the temperature in the vehicle at the air conditioner working end time.

[0044] The compressor working time is obtained according to the air conditioner working state.

[0045] The average of the final temperature difference is Σ (the final temperature difference > 0 and the working time is less than n minutes) / Σ (the number of times the compressor works), n ∈ (5, 15).

[0046] 2) In the cloud, the collected basic data also needs to be processed to exclude abnormal and invalid data to obtain the finally constructed basic index set. The abnormal and invalid data includes empty data and extreme data.

[0047] Step 2: Based on the real market air conditioner repair single sample data, the correlation between each index in the basic index set and the air conditioner fault is analyzed, and the key index set strongly related to the fault is selected. In this embodiment, Pearson method is used to analyze the correlation between the basic index data and the air conditioner fault.

[0048] The selected key index set includes the proportion of the compressor working end not reaching the set temperature and the average of the final temperature difference interval, which evaluates the cooling effect of the air conditioner through the above key index. The calculation formula of the proportion of the compressor working end not reaching the set temperature is:

[0049] Σ (end_t_s > 0 and working time < 10) / Σ (compressor working times). In other embodiments, the working time can be less than a set value, which can be in the range of (5, 15), and preferably, the working time is less than 10 minutes.

[0050] The calculation method of the average of the end temperature difference (end_t_s) is as follows:

[0051] When the temperature difference is <-2, the temperature difference is -2;

[0052] When the temperature difference is >=-2 and <=5, the temperature difference is the actual value;

[0053] When the temperature difference is >5, the temperature difference is 5. The average of the end temperature difference interval is the average of the above temperature differences.

[0054] Step 3: Based on the LOF anomaly factor algorithm, identify the outlying points of the full-vehicle in the city.

[0055] As shown in FIGS. Figure 2 , Figure 2 , Figure 3 and Figure 4 , the horizontal axis represents the end temperature difference, and the vertical axis represents the proportion of the compressor working end not reaching the set temperature. Most of the air conditioners work normally, and the compressor stops working when the temperature inside the vehicle is lower than the set temperature. The proportion of the compressor working end not reaching the set temperature is between 0 and 0.5, the end temperature difference is less than 0, and the density is concentrated in the left area. A small number of outliers on the right have an end temperature difference between 1 and 5, and the proportion of the compressor working end not reaching the set temperature is between 0.6 and 1. (Note: "+" in the figure represents a specific date of a fault vehicle, and this vehicle is used as a verification tag).

[0056] Obtain the key indicator data of the full-vehicle in the city, identify the outlying points in the key indicator data of the full-vehicle in the city using the LOF anomaly factor algorithm, and output the calculation results:

[0057] 1) Outlier factor tag [-1 -1 -1 -1 -1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 -11]; Corresponding output fault vehicle list list_1 (V1, V2, V3, V4 …… Vn). The fault vehicle list includes fault vehicles, fault states, and fault dates.

[0058] 2) Outlier score, reflecting the abnormality of the sample, the larger the absolute value, the more abnormal:

[0059] [-13.72300608 -17.56860982 -11.83203106 -17.61255772 -13.71363421

[0060] -0.99817859 -1.00569839 -0.96221402 -1.34529239 -1.02148983

[0061] -1.12360409 -0.96431497 -1.11256157 -1.01836114 -0.97290984

[0062] -0.99499985 -1.10426975 -1.32305091 -1.02964179 -1.1547492

[0063] -0.97075492 -0.98762276 -3.56539482 -1.41396739 -1.04845145

[0064] -1.36399049 -0.99099528 -1.03203924 -0.98961872 -1.00552565

[0065] In this step, based on the key indicator data, the LOF anomaly factor algorithm is used to select the vehicles that have faults on a certain day from a large number of data of vehicles in the same city, so as to reduce the number of vehicles, reduce the subsequent calculation amount, and the fault vehicle list contains vehicles that have faults on a certain day. The data of the vehicle is representative and can be used to verify the fault state of the to-be-tested vehicle to improve the accuracy.

[0066] Fourth step: based on the fault vehicle list list_1 (V1, V2, V3, V4……Vn) output in the third step, the air conditioner working data of the vehicle Vi on the fault day and the recent 30 days (which can also be the recent k days, k∈[25, 35]) are verified in turn. The LOF algorithm is used to judge the outlying situation of the data on the fault day, and the prediction result is as shown in the following table. Figure 3 Then, the LOF algorithm is used to judge the outlying situation of the air conditioner working data of the to-be-tested vehicle on the operation day and the recent 30 days.

[0067] Specifically, the following steps are included:

[0068] 1) The key indicator data of the to-be-tested vehicle on the operation day and the key indicator data in the recent 30 days are identified as outliers by using the LOF algorithm.

[0069] 2) The key indicator data of each vehicle in the fault vehicle list on the fault day and the recent 30 days of the fault day are identified as outliers again by using the LOF algorithm. The identification results of each vehicle in the fault vehicle list contain outliers, which are used to verify whether the identification result of the to-be-tested vehicle on the day is near the outliers.

[0070] Compared with the identification result of the vehicle to be tested in the last 30 days and the identification result of each vehicle in the fault vehicle list, it is judged whether the identification result of the vehicle to be tested on the running day is in an outlier.

[0071] The same city vehicle is selected for comparison, because the same city environment and temperature are the same, the influence of environmental changes is reduced, and the data of the vehicle to be tested in the last 30 days is selected to exclude the influence of different vehicle models.

[0072] Step 5: Based on the fault vehicle list list_1 (V1, V2, V3, V4, …, Vn) output in step 3, the air conditioner working data of the vehicle Vi on the fault day and in the same period of last year are verified in turn, the LOF algorithm is used to judge the outlier of the data on the fault day, and then the LOF algorithm is used to judge the outlier of the air conditioner working data of the vehicle to be tested on the running day and in the same period of last year, and the prediction result is as shown in Figure 4

[0073] 1) The key indicator data of the vehicle to be tested on the running day and the key indicator data in the same period of last year are identified again using the LOF algorithm to identify outliers;

[0074] 2) The key indicator data of each vehicle in the fault vehicle list on the fault day and in the same period of last year are identified again using the LOF algorithm to identify outliers, and the outliers in the identification result are used to verify whether the identification result of the vehicle to be tested on the day is near the outliers.

[0075] Compared with the identification result of the vehicle to be tested in the last 30 days and the identification result of each vehicle in the fault vehicle list, it is judged whether the identification result of the vehicle to be tested on the running day is in an outlier.

[0076] The external factors in the last 30 days and the same period of last year are basically the same, which excludes the interference of external factors, and the accuracy is improved by comparing the vehicle to be tested with the same city vehicle and comparing the vehicle to be tested with the data in the same period of last year.

[0077] Step 6: If the fourth step and the fifth step are verified as outliers, it is confirmed that the vehicle is a fault vehicle, and after-sales order maintenance is performed.

[0078] The first outlier is obtained based on the LOF algorithm for identifying the key indicator data of all vehicles in the same city, the second outlier is obtained based on the LOF algorithm for identifying the key indicator data of the vehicle to be tested in the same period of last year, and the third outlier is obtained based on the LOF algorithm for identifying the key indicator data of the vehicle to be tested in the last 30 days. If the prediction result of the vehicle to be tested on the day is in the intersection region of the first outlier, the second outlier and the third outlier, it is considered that the vehicle is faulty. The intersection region is as shown in the middle diagonal cross section. Figure 5

[0079] ​​Step 7: The air conditioner fault of the full amount of new energy vehicles is predicted every day, and the fault information is pushed to the after-sales for maintenance and repair. The prediction model is optimized regularly based on the repair results.

[0080] The vehicle-mounted air conditioner fault prediction method based on the local anomaly factor algorithm is suitable for fault prediction of all vehicle-mounted air conditioners, solves the industry status of passive discovery of vehicle part faults, obtains air conditioner fault information in advance, avoids further serious faults, reduces maintenance cost, and improves vehicle operation income. It is also suitable for air conditioner fault prediction installed in buildings.

[0081] The present application uses an unsupervised three-different-sample-set verification method, uses three different sample sets based on unsupervised algorithms, and uses the method of taking the intersection of the results to solve the disadvantages of most algorithms that need to train the model regularly and optimize the model threshold. At the same time, it solves the high maintenance cost of different models corresponding to different prediction models caused by different control strategies of different air conditioner models.

[0082] The present application can output compressor working score for measuring air conditioner refrigeration effect, which is helpful for evaluating air conditioner performance and assisting after-sales intelligent extended warranty service.

[0083] System embodiment:

[0084] The vehicle-mounted air conditioner fault prediction system embodiment based on the local anomaly factor algorithm of the present application comprises a memory, a processor and an internal bus, and the processor and the memory communicate and exchange data through the internal bus. The memory includes at least one software function module stored in the memory, and the processor executes various function applications and data processing by running the software program and module stored in the memory, and realizes the vehicle-mounted air conditioner fault prediction method based on the local anomaly factor algorithm introduced in the method embodiment of the present application. The processor can be a microprocessor MCU, a programmable logic device FPGA and other processing devices. The memory can be various memories that store information by using electric energy, such as RAM, ROM, etc.

Claims

1. A vehicle air conditioner fault prediction method based on a local outlier factor algorithm, characterized by: The method comprises the following steps: 1) obtaining indicators with strong correlation with air conditioner faults to obtain a key indicator set; 2) selecting a set number of vehicles in a geographical area to which the vehicle to be tested belongs as sample vehicles, and obtaining key indicator data of the sample vehicles; using the LOF anomaly factor algorithm to identify outliers in the key indicator data of the sample vehicles to obtain a fault vehicle list, the fault vehicle list including fault vehicles and fault information; 3) using the LOF anomaly factor algorithm to identify outliers in the first data and the second data to obtain a first prediction result; using the LOF anomaly factor algorithm to identify outliers in the third data and the fourth data to obtain a second prediction result; the first data including key indicator data of the vehicle to be tested on the current day and in a recent time period; the second data including key indicator data of each vehicle in the fault vehicle list on the current day and in the corresponding recent time period; the third data including key indicator data of the vehicle to be tested on the current day and in a historical same period; and the fourth data including key indicator data of each vehicle in the fault vehicle list on the current day and in the corresponding historical same period; 4) determining a final air conditioner fault result based on the first prediction result and the second prediction result; if both the first prediction result and the second prediction result are outliers, the compressor of the vehicle to be tested has a fault, otherwise the compressor of the vehicle to be tested is normal. 2.The vehicle air conditioner fault prediction method based on the local outlier factor algorithm of claim 1, wherein: The method further comprises pushing the fault information obtained in step 4) to after-sales maintenance. 3.The method of claim 1, wherein: In step 1), the key indicator set is obtained as follows: Sample data of air conditioner repair orders is obtained, and the sample data of the air conditioner repair orders is subjected to fault correlation analysis based on a basic indicator set to screen indicators with strong correlation as key indicators to obtain the key indicator set; The key indicator set includes a compressor working end temperature not reaching a set temperature ratio and an end temperature difference average value.

4. The method according to claim 3, wherein the method is characterized by: The basic indicator set is derived from basic data sent by a vehicle end; the basic data includes: an in-vehicle temperature, an air conditioner set temperature, an air conditioner working state, an air conditioner compressor working state, an entire vehicle driving state, a door opening and closing state, a compressor working end temperature not reaching a set temperature ratio, and an end temperature difference interval average value.

5. The method according to claim 4, wherein the method is characterized by: The basic indicator set includes: the compressor working end temperature not reaching a set temperature ratio and the end temperature difference average value; the basic indicator set further includes at least one of a compressor working time length, an in-vehicle temperature start value and an end value, an air conditioner set temperature start value and an end value, a start temperature difference, an end temperature difference, an in-vehicle temperature change value, an end temperature difference, a working time length, and a compressor working end temperature not reaching a set temperature ratio; A calculation formula of the compressor working end temperature not reaching a set temperature ratio is: Σ (end temperature difference > 0 and working time is less than n minutes) / Σ (compressor working times), n ∈ (5, 15); The end temperature difference average value is obtained as follows: When the temperature difference is <-2, a first temperature difference is -2; when the temperature difference is >=-2 and <=5, a second temperature difference is an actual value; when the temperature difference is >5, a third temperature difference is 5; and the end temperature difference average value is an average value of the first temperature difference, the second temperature difference, and the third temperature difference. The compressor working time length is obtained according to the air conditioner working state; the vehicle interior temperature starting value and the ending value are obtained according to the vehicle interior temperature and the corresponding time point; the air conditioner setting temperature starting value and the ending value are obtained according to the air conditioner setting temperature; the starting temperature difference is the vehicle interior temperature corresponding to the air conditioner starting time minus the air conditioner setting temperature; the ending temperature difference is the vehicle interior temperature corresponding to the ending time minus the air conditioner setting temperature; and the working time length is obtained according to the air conditioner working state and the corresponding time point.

6. The method of claim 4, wherein the method is based on a local outlier factor algorithm. The basic data is data after removing invalid data, and the invalid data includes empty data and extreme data.

7. The method according to any one of claims 1 to 6, wherein the method is characterized by: In step 2, the recent time period is from the current day to the previous k days, and k is an integer in the range of 25 to 35; and the historical same period time period is the same time period of the previous year.

8. The method according to any one of claims 1 to 6, wherein the method is based on a local outlier factor algorithm. In step 2, the geographical area to which the vehicle to be tested belongs is the same city as the vehicle to be tested, and the set number is all vehicles in the same city. 9.A vehicle air conditioner fault prediction system based on a local outlier factor algorithm, characterized by: The system comprises a processor configured to execute computer instructions to implement the vehicle-mounted air conditioner fault prediction method based on the local outlier factor algorithm according to any one of claims 1 to 8.

Images