Sensor fault analysis method and device, sensor fault analysis system and automobile

Abstract

The invention discloses a sensor fault analysis method and device, a sensor fault analysis system and an automobile. The method comprises the steps that first temperature data corresponding to the same engine are obtained, the first temperature data comprise first temperature data corresponding to a first temperature sensor and first temperature data corresponding to N-1 second temperature sensors, and N is larger than or equal to 2; determining predicted temperature data corresponding to the first temperature sensor based on the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to the N-1 second temperature sensors; based on the first temperature data corresponding to the first temperature sensor and the predicted temperature data, determining a residual mean value and a sample standard deviation corresponding to the first temperature sensor; and determining a target fault degree corresponding to the first temperature sensor based on the residual mean value and the sample standard deviation corresponding to the first temperature sensor. According to the method, accurate and effective fault analysis can be carried out on the temperature sensor, the safety risk is reduced, and the user viscosity is effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of fault analysis, and in particular to a sensor fault analysis method, equipment, a sensor fault analysis system and a car. Background Art

[0002] Performing fault analysis on the temperature sensor corresponding to the engine in the car in order to improve the temperature sensor is of great significance to improving the performance of the whole vehicle. Too high or too low engine water temperature can easily cause damage to the engine. Reasonable control of the engine water temperature can effectively reduce damage to the engine, improve the efficiency of the engine, and effectively exert the performance of the engine. At present, it is usually necessary to install a temperature sensor on the engine and control the engine water temperature according to the temperature data collected by the temperature sensor. However, with the aging of the temperature sensor and the related wiring harness, the temperature sensor will produce corresponding faults, which will lead to inaccurate engine water temperature control and affect the engine performance. There is no effective method for fault analysis of temperature sensors in the prior art. Therefore, how to accurately and effectively predict and analyze the faults of temperature sensors is a technical problem that needs to be solved urgently. Summary of the invention

[0003] Embodiments of the present invention provide a sensor fault analysis method, a device, a sensor fault analysis system and a vehicle to solve the technical problem of how to perform fault prediction analysis on a temperature sensor more accurately and effectively.

[0004] A sensor fault analysis method, comprising: Acquire first temperature data corresponding to the same engine, the first temperature data including first temperature data corresponding to the first temperature sensor and first temperature data corresponding to N-1 second temperature sensors, where N≥2; Determine predicted temperature data corresponding to the first temperature sensor based on first temperature data corresponding to the first temperature sensor and first temperature data corresponding to N-1 second temperature sensors; Determine a residual mean and a sample standard deviation corresponding to the first temperature sensor based on first temperature data and predicted temperature data corresponding to the first temperature sensor; A target fault degree corresponding to the first temperature sensor is determined based on a residual mean and a sample standard deviation corresponding to the first temperature sensor.

[0005] Preferably, the determining the predicted temperature data corresponding to the first temperature sensor based on the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to N-1 second temperature sensors includes: Determine the temperature prediction model corresponding to the first temperature sensor based on the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to N - 1 second temperature sensors; Determine the predicted temperature data corresponding to the first temperature sensor based on the temperature prediction model corresponding to the first temperature sensor and the first temperature data corresponding to N - 1 second temperature sensors.

[0006] Preferably, the temperature prediction model is a temperature fitting model between the first temperature sensor and N - 1 second temperature sensors; The determining the temperature prediction model corresponding to the first temperature sensor based on the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to N - 1 second temperature sensors includes: Based on a preset screening condition, screen the multiple first temperature data corresponding to the first temperature sensor and the multiple first temperature data corresponding to N - 1 second temperature sensors to obtain multiple second temperature data corresponding to the first temperature sensor and multiple second temperature data corresponding to N - 1 second temperature sensors; Based on the multiple second temperature data corresponding to the first temperature sensor and the multiple second temperature data corresponding to N - 1 second temperature sensors, fit an initial fitting model to obtain the temperature fitting model corresponding to the first temperature sensor.

[0007] Preferably, the preset screening condition is that the local anomaly factor corresponding to the first temperature data is within the anomaly value range; The anomaly value range is determined based on the local anomaly factor corresponding to the valid temperature data; The valid temperature data is the first temperature data with the current shutdown duration greater than a preset duration threshold, the current sampling moment within a preset time range, and the temperature value of the valid temperature data within a preset temperature range.

[0008] Preferably, the determining the target fault degree corresponding to the first temperature sensor based on the residual mean and sample standard deviation corresponding to the first temperature sensor includes: Determine the current fault type corresponding to the first temperature sensor based on the residual mean and sample standard deviation corresponding to the first temperature sensor; Determine the target fault degree corresponding to the first temperature sensor based on the current fault type corresponding to the first temperature sensor.

[0009] Preferably, the determining the current fault type corresponding to the first temperature sensor based on the residual mean and sample standard deviation corresponding to the first temperature sensor includes: If the absolute value of the residual mean is not greater than the lower limit of the residual, and the sample standard deviation is not greater than the lower limit of the standard deviation, it is determined that the current fault type of the first temperature sensor is no fault; If the absolute value of the residual mean is not greater than the lower limit of the residual, the sample standard deviation is greater than the lower limit of the standard deviation and not greater than the upper limit of the standard deviation, it is determined that the current fault type of the first temperature sensor is precision fault; If the absolute value of the residual mean is greater than the lower limit of the residual and not greater than the upper limit of the residual, and the sample standard deviation is not greater than the lower limit of the standard deviation, it is determined that the current fault type of the first temperature sensor is drift fault; If the absolute value of the residual mean is greater than the lower limit of the residual and not greater than the upper limit of the residual, and the sample standard deviation is greater than the lower limit of the standard deviation and not greater than the upper limit of the standard deviation, it is determined that the current fault type of the first temperature sensor is drift precision fault; If the absolute value of the residual mean is greater than the upper limit of the residual, and the sample standard deviation is greater than the upper limit of the standard deviation, it is determined that the current fault type of the first temperature sensor is complete fault.

[0010] Preferably, determining the target fault degree corresponding to the first temperature sensor based on the current fault type corresponding to the first temperature sensor includes: If the current fault type corresponding to the first temperature sensor is no fault, it is determined that the target fault degree corresponding to the first temperature sensor is 0; If the current fault type corresponding to the first temperature sensor is precision fault, the difference between the sample standard deviation and the lower limit of the standard deviation is determined as the first error, and the quotient of the first error and the normalization coefficient is determined as the target fault degree corresponding to the first temperature sensor; If the current fault type corresponding to the first temperature sensor is drift fault, the difference between the absolute value of the residual mean and the lower limit of the residual is determined as the second error, and the quotient of the second error and the normalization coefficient is determined as the target fault degree corresponding to the first temperature sensor; If the current fault type corresponding to the first temperature sensor is drift precision fault, the arithmetic square root of the sum of the squares of the first error and the second error is determined as the third error, and the quotient of the third error and the normalization coefficient is determined as the target fault degree corresponding to the first temperature sensor; If the current fault type corresponding to the first temperature sensor is complete fault, it is determined that the target fault degree corresponding to the first temperature sensor is 1; Wherein, the normalization coefficient is the arithmetic square root of the sum of the squares of the first difference and the second difference, the first difference is the difference between the upper limit of the residual and the lower limit of the residual, and the second difference is the difference between the upper limit of the standard deviation and the lower limit of the standard deviation.

[0011] Preferably, the N temperature sensors are respectively arranged on the engine water outlet, the engine cylinder head, the engine contact position with the environment, the engine oil pan, and the engine intake manifold; The first temperature data includes the water temperature at the water outlet, the water temperature of the engine cylinder head, the ambient temperature, the engine oil temperature, and the gas temperature of the intake manifold.

[0012] Preferably, after determining the target fault degree corresponding to the first temperature sensor, the sensor fault analysis method further includes: Based on M target analysis factors, analyze the target fault degree corresponding to the first temperature sensor to determine the sensor fault results corresponding to the M target analysis factors, where M≥1; Send the sensor fault results corresponding to the M target analysis factors to the target terminal.

[0013] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the above-mentioned sensor fault analysis method is implemented.

[0014] A sensor fault analysis system includes a first temperature sensor, a second temperature sensor, and the above-mentioned computer device. The computer device is respectively connected to the first temperature sensor and the second temperature sensor and is used to determine the target fault degree corresponding to the first temperature sensor.

[0015] A vehicle includes the above-mentioned sensor fault analysis system.

[0016] The above-mentioned sensor fault analysis method, device, sensor fault analysis system and vehicle determine any one of the N temperature sensors corresponding to the engine as the first temperature sensor, and determine the remaining N - 1 second temperature sensors having data correlation with the first temperature sensor as the second temperature sensors. According to the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to the N - 1 second temperature sensors, the predicted temperature data corresponding to the first temperature sensor is accurately determined. Based on the first temperature data and the predicted temperature data corresponding to the first temperature sensor, the residual mean value and the sample standard deviation corresponding to the first temperature sensor are determined, so as to facilitate the quantitative fault analysis of the first temperature sensor according to the residual mean value and the sample standard deviation, and accurately determine the target fault degree of the first temperature sensor. This method not only does not require relatively complex processing of temperature data, is relatively efficient and convenient, but also can perform quantitative analysis on the target fault degree of the first temperature sensor, is relatively accurate and effective, can achieve the purpose of accurately and effectively performing fault analysis on the temperature sensor, so as to facilitate the timely maintenance and performance improvement of the temperature sensor, thereby reducing the user's troubleshooting and maintenance waiting time, reducing the safety risks caused by temperature sensor failures, effectively enhancing user viscosity, and having a relatively broad application prospect. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required to be used in the description of the embodiments of the present invention. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0018] Figure 1 is a flowchart of a sensor fault analysis method in an embodiment of the present invention; Figure 2 is another flowchart of a sensor fault analysis method in an embodiment of the present invention; Figure 3 is another flowchart of a sensor fault analysis method in an embodiment of the present invention; Figure 4 is another flowchart of a sensor fault analysis method in an embodiment of the present invention; Figure 5 is another flowchart of a sensor fault analysis method in an embodiment of the present invention; Figure 6 is another flowchart of a sensor fault analysis method in an embodiment of the present invention; Figure 7 is another flowchart of a sensor fault analysis method in an embodiment of the present invention; Figure 8Schematic diagram of accuracy fault and drift fault corresponding to the first temperature sensor; Figure 9 Fault type diagram of the first temperature sensor determined by the upper residual limit, lower residual limit, upper standard deviation limit, and lower standard deviation limit; Figure 10 It is a schematic diagram of a computer device in an embodiment of the present invention. Detailed implementation manners

[0019] 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 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.

[0020] The sensor fault analysis method provided by the embodiments of the present invention is used to achieve the purpose of accurately and effectively predicting and analyzing the faults of temperature sensors. This sensor fault analysis method can be applied in a computer device, which can be an in-vehicle controller set on a vehicle, used to analyze the first temperature data collected by N temperature sensors corresponding to the same engine to determine the fault situation of any one of the N temperature sensors, so as to remind the target terminal to repair the temperature sensor with a fault in a timely manner according to the target fault degree of any one of the temperature sensors obtained from the fault analysis, and improve user viscosity. This computer device can also be an electronic device deployed outside the vehicle, such as the cloud, used to analyze the faults generally existing in the vehicles of multiple provinces or regions served by the target terminal, so as to enable the target terminal to improve and upgrade the temperature sensors in a targeted manner, improve the performance of the temperature sensors corresponding to the engines in the vehicles, and improve user trust.

[0021] In one embodiment, as Figure 1 shown, a sensor fault analysis method is provided. Taking the computer device in Figure 10 as an example for illustration, the method includes the following steps: S101: Obtain the first temperature data corresponding to the same engine, where the first temperature data includes the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to N - 1 second temperature sensors, where N ≥ 2; S102: Based on the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to N - 1 second temperature sensors, determine the predicted temperature data corresponding to the first temperature sensor; S103: Based on the first temperature data corresponding to the first temperature sensor and the predicted temperature data, determine the residual mean and sample standard deviation corresponding to the first temperature sensor; S104: Determine the target fault degree corresponding to the first temperature sensor based on the residual mean value and the sample standard deviation corresponding to the first temperature sensor.

[0022] Among them, the first temperature data refers to the temperature data collected by the temperature sensor for fault analysis. The first temperature sensor refers to the temperature sensor corresponding to the engine that needs to be analyzed for faults. The second temperature sensor is other temperature sensors related to the engine except the first temperature sensor, and can be specifically understood as the temperature sensor used for fault analysis of the first temperature sensor.

[0023] As an example, in step S101, the computer device obtains multiple groups of first temperature data collected by N temperature sensors corresponding to the same engine within a preset time period. Specifically, it obtains the first temperature data collected by 1 first temperature sensor at multiple sampling times and the first temperature data collected by N - 1 second temperature sensors at multiple sampling times within the preset time period, so as to perform fault analysis on the first temperature sensor based on the N second temperature data. It can be understood that temperature sensors are respectively set at N different positions corresponding to the engine to collect the first temperature data at the corresponding positions. Since there is a correlation between the first temperature data collected by different temperature sensors corresponding to the same engine, the first temperature data collected by N - 1 temperature sensors can be used to perform fault analysis on 1 of the temperature sensors. Therefore, when performing fault analysis on any temperature sensor corresponding to the engine, this temperature sensor is determined as the first temperature sensor, and the remaining N - 1 temperature sensors are determined as the second temperature sensors, which are used to perform fault analysis on the first temperature sensor according to the N - 1 second temperature sensors. For example, if it is necessary to perform fault analysis on the temperature sensor that collects the water temperature at the outlet of the engine, then the temperature sensor that collects the water temperature at the outlet of the engine is determined as the first temperature sensor, and the other N - 1 temperature sensors corresponding to the engine are determined as the second temperature sensors. In this example, the preset time period can be the time period from the first power-on moment of the engine to the current power-on moment, or the time period between any two preset power-on moments. The sampling time refers to the moment when the temperature sensor collects the temperature data. For example, the sampling time can be the moment when the engine is powered on each time, that is, the temperature sensor can collect the temperature data at the moment when the engine is powered on each time.

[0024] Among them, the predicted temperature data refers to the data generated by predicting the first temperature data corresponding to the first temperature sensor at the sampling moment. For example, if the power-on moment of the engine is determined as the sampling moment, for each power-on moment of the engine, both the first temperature sensor and the second temperature sensor can collect the first temperature data at the corresponding position. To perform a fault analysis on the first temperature sensor, for each power-on moment, by processing the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to N - 1 second temperature sensors, the temperature data collected by the first temperature sensor at each power-on moment can be predicted, and the predicted temperature data corresponding to the first temperature data of the first temperature sensor at each power-on moment can be obtained, so as to perform a fault analysis on the first temperature sensor based on the difference between the first temperature data and the predicted temperature data. In this example, to make a more accurate prediction, it is necessary to preliminarily screen the first temperature data to obtain more accurate first temperature data. For example, when the temperature value corresponding to the first temperature data obtained by the computer device is less than the preset temperature value (for example, -100 °C), to reduce the fault analysis error, the first temperature data corresponding to this sampling moment is discarded, that is, the first temperature data corresponding to this sampling moment is not recorded and analyzed.

[0025] As an example, in step S102, the computer device predicts the temperature data of the first temperature sensor at each sampling moment of the engine based on the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to N - 1 second temperature sensors, and obtains the predicted temperature data corresponding to the first temperature sensor at each sampling moment of the engine. That is, when performing a fault analysis on the first temperature sensor, each first temperature data of the first temperature sensor corresponds to a predicted temperature data. For example, the computer device can input the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to N - 1 second temperature sensors into a pre-trained temperature prediction model and output the predicted temperature data corresponding to the first temperature sensor. Among them, the temperature prediction model includes but is not limited to a neural network prediction model.

[0026] In this example, based on the first temperature data collected at the sampling moment, the predicted temperature data at each sampling moment corresponding to the first temperature sensor is predicted, so as to perform a fault analysis on the first temperature sensor based on the predicted temperature data corresponding to the first temperature sensor.

[0027] Among them, the residual mean refers to the value obtained by averaging the temperature residuals between each first temperature data actually collected by the first temperature sensor within a preset time period and the predicted temperature data corresponding to the same first temperature data actually collected by the first temperature sensor. The sample standard deviation refers to the standard deviation of all the first temperature data corresponding to the first temperature sensor.

[0028] As an example, in step S103, the computer device performs a difference process on each first temperature data actually collected by the first temperature sensor and the predicted temperature data corresponding to the same first temperature data actually collected by the first temperature sensor within a preset time period, to obtain a plurality of temperature residuals, performs a mean process on all the temperature residuals to obtain a residual mean value, and performs a standard deviation process on each first temperature data corresponding to the first temperature sensor and the residual mean value to obtain a sample standard deviation. It can be understood that the residual mean value and the sample standard deviation are used to reflect the difference between each first temperature data actually collected by the first temperature sensor and the predicted temperature data corresponding to each first temperature data actually collected by the first temperature sensor within a preset time period. If the difference between the first temperature data actually collected by the first temperature sensor and the predicted temperature data corresponding to the first temperature data is large, it indicates that the first temperature sensor may have a certain degree of failure. The failure degree of the first temperature sensor can be quantitatively analyzed according to the residual mean value and the sample standard deviation corresponding to the first temperature data and the predicted temperature data corresponding to the first temperature data. Therefore, determining the residual mean value and the sample standard deviation corresponding to the first temperature sensor can accurately perform a failure analysis on the first temperature sensor.

[0029] Wherein, the target failure degree refers to the failure degree of the first temperature sensor.

[0030] As an example, in step S104, the computer device processes the residual mean value and the sample standard deviation corresponding to the first temperature sensor to obtain the failure degree of the first temperature sensor, and determines this failure degree as the target failure degree corresponding to the first temperature sensor. For example, a two-dimensional mapping relationship table between the intervals corresponding to the residual mean value and the sample standard deviation and the target failure degree is pre-stored in the system database. The computer device queries the preset two-dimensional mapping relationship table according to the numerical values of the residual mean value and the sample standard deviation, and determines the target failure degree corresponding to the residual mean value and the sample standard deviation, which is the target failure degree of the first temperature sensor, which is relatively efficient and convenient. Another example is that the computer device uses a preset failure degree determination strategy to process the residual mean value and the sample standard deviation corresponding to the first temperature sensor to determine the target failure degree of the first temperature sensor, which is relatively efficient and convenient.

[0031] In this embodiment, any one of the N temperature sensors corresponding to the engine is determined as the first temperature sensor, and the remaining N - 1 second temperature sensors having data correlation with the first temperature sensor are determined as the second temperature sensors. According to the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to the N - 1 second temperature sensors, the predicted temperature data corresponding to the first temperature sensor is accurately determined. Based on the first temperature data and the predicted temperature data corresponding to the first temperature sensor, the residual mean and the sample standard deviation corresponding to the first temperature sensor are determined, so as to perform quantitative fault analysis on the first temperature sensor according to the residual mean and the sample standard deviation, and accurately determine the target fault degree of the first temperature sensor. This method not only does not require relatively complex processing of temperature data, is relatively efficient and convenient, but also can perform quantitative analysis on the target fault degree of the first temperature sensor, is relatively accurate and effective, can achieve the purpose of accurately and effectively predicting and analyzing the faults of temperature sensors, so as to facilitate timely maintenance and performance improvement of temperature sensors, thereby reducing the troubleshooting and maintenance waiting time of users, reducing the safety risks caused by temperature sensor failures, effectively improving user viscosity, and having a relatively broad application prospect.

[0032] In one embodiment, as Figure 2 shown, step S102, that is, based on the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to the N - 1 second temperature sensors, determining the predicted temperature data corresponding to the first temperature sensor, includes: S201: Based on the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to the N - 1 second temperature sensors, determine the temperature prediction model corresponding to the first temperature sensor; S202: Based on the temperature prediction model corresponding to the first temperature sensor and the first temperature data corresponding to the N - 1 second temperature sensors, determine the predicted temperature data corresponding to the first temperature sensor.

[0033] Among them, the temperature prediction model refers to a model used to predict the temperature of the first temperature sensor.

[0034] As an example, in step S201, the computer device processes the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to N - 1 second temperature sensors to obtain the temperature prediction model corresponding to the first temperature sensor. For example, the computer device uses the first temperature data corresponding to the first temperature sensor as the labeled data, inputs the first temperature data corresponding to N - 1 second temperature sensors within a preset time period into the model to be trained for model training, and after the model converges, obtains the temperature prediction model corresponding to the first temperature sensor. In this example, based on the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to N - 1 second temperature sensors, the temperature prediction model corresponding to the first temperature sensor is determined more accurately, so as to further determine the predicted temperature data corresponding to each first temperature data of the first temperature sensor according to the temperature prediction model.

[0035] As an example, in step S202, the computer device inputs the first temperature data corresponding to each sampling moment of N - 1 second temperature sensors within a preset time period into the temperature prediction model to obtain the predicted temperature data corresponding to the first temperature sensor at each sampling moment. For example, if within the preset time period, at the first sampling moment, the first temperature data corresponding to N - 1 second temperature sensors is , at the th sampling moment, the first temperature data corresponding to N - 1 second temperature sensors is . The computer device inputs the at the first sampling moment into the temperature prediction model, outputs the predicted temperature data corresponding to the first temperature data of the first temperature sensor at the first sampling moment, and inputs the at the th sampling moment into the temperature prediction model, and outputs the predicted temperature data corresponding to the first temperature data of the first temperature sensor at the th sampling moment.

[0036] In this embodiment, any one of the N temperature sensors corresponding to the engine is determined as the first temperature sensor, and the remaining N - 1 second temperature sensors having data correlation with the first temperature sensor are determined as the second temperature sensors. The temperature prediction model corresponding to the first temperature sensor can be trained according to the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to N - 1 second temperature sensors. The temperature prediction model can process the first temperature data actually collected by the remaining N - 1 second temperature sensors having data correlation with the first temperature sensor according to the data correlation between the first temperature sensor and the second temperature sensors, and predict the predicted temperature data corresponding to the first temperature sensor more accurately.

[0037] In one embodiment, the temperature prediction model is a temperature fitting model between the first temperature sensor and N - 1 second temperature sensors.

[0038] Among them, the temperature fitting model refers to a prediction model that takes the temperature data collected by N - 1 second temperature sensors as input and fits and outputs the temperature data of the first temperature sensor.

[0039] As Figure 3 shown, step S201, that is, based on the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to N - 1 second temperature sensors, determining the temperature prediction model corresponding to the first temperature sensor, includes: S301: Based on a preset screening condition, screen the multiple first temperature data corresponding to the first temperature sensor and the multiple first temperature data corresponding to N - 1 second temperature sensors to obtain multiple second temperature data corresponding to the first temperature sensor and multiple second temperature data corresponding to N - 1 second temperature sensors; S302: Based on the multiple second temperature data corresponding to the first temperature sensor and the multiple second temperature data corresponding to N - 1 second temperature sensors, fit the initial fitting model to obtain the temperature fitting model corresponding to the first temperature sensor.

[0040] Among them, the preset screening condition refers to a preset limiting condition for screening the first temperature data. The second temperature data refers to the temperature data used for fitting the initial fitting model. Among them, the initial fitting model refers to a model that has not been fitted.

[0041] As an example, in step S301, the computer device screens the first temperature data that meets the preset screening condition from the multiple first temperature data collected by the first temperature sensor at each sampling moment and the multiple first temperature data actually collected by N - 1 second temperature sensors at each sampling moment within a preset time period, and determines the first temperature data that meets the preset screening condition as the second temperature data corresponding to the first temperature sensor and the second temperature data corresponding to N - 1 second temperature sensors. Furthermore, multiple second temperature data corresponding to the first temperature sensor and multiple second temperature data corresponding to N - 1 second temperature sensors within the preset time period are obtained. It can be understood that the temperature prediction model is used to predict the temperature data of the first temperature sensor at the same sampling moment according to the temperature data of N - 1 second temperature sensors at a sampling moment, and obtain the predicted temperature data at the same sampling moment. This predicted temperature data is used to reflect the temperature data of the fault - free first temperature sensor at the sampling moment. To ensure the accuracy of the temperature prediction model, it is necessary to screen the first temperature data and eliminate the abnormal first temperature data to obtain normal second temperature data.

[0042] As an example, in step S302, the computer device establishes an initial fitting model. The input of the initial fitting model is the temperature data actually collected by N - 1 second temperature sensors at a sampling moment, and the output is the predicted data corresponding to the first temperature sensor at the same sampling moment. In this example, the computer device uses the multiple second temperature data collected by N - 1 second temperature sensors at the sampling moment and the multiple second temperature data actually collected by the first temperature sensor at the sampling moment to fit the initial fitting model, determines the coefficients and constant terms of the initial fitting model, and obtains a temperature prediction model that can predict the predicted temperature data of the first temperature sensor at the same sampling moment according to the temperature data actually collected by N - 1 second temperature sensors at a sampling moment.

[0043] In this example, the computer device fits the initial fitting model based on the multiple second temperature data corresponding to the first temperature sensor and the multiple second temperature data corresponding to N - 1 second temperature sensors, and obtains a temperature fitting model corresponding to the first temperature sensor, which specifically includes the following steps: (1) Establish an initial fitting model for predicting the temperature data of the first temperature sensor based on the temperature data actually collected by N - 1 second temperature sensors: , where , , , , , are coefficients, is the constant term.

[0044] (2) Use the multiple second temperature data corresponding to the first temperature sensor and the multiple second temperature data corresponding to N - 1 second temperature sensors to fit the initial fitting model , determine the coefficients in the initial fitting model, that is, determine , , , , , and etc. Among them, is the second temperature data of N - 1 second temperature sensors at the i-th sampling moment, is the second temperature data of the first temperature sensor at the i-th sampling moment. Specifically, the computer device determines , where = 1, 2, 3, 4, , , is the number of second temperature data. = , that is, the temperature prediction model obtained is Among them, , where = 1, 2, 3, 4, , , is the number of the second temperature data, = .

[0045] For example, if N = 5, the initial fitting model is: . The computer device fits the initial fitting model and obtains , , , and . The computer device determines , where = 1, 2, 3, and 4, is the number of the second temperature data, = . When N = 5, the temperature prediction model corresponding to the first temperature sensor is , where , , , , = , = 1, 2, 3, and 4.

[0046] In this embodiment, the first temperature data within the preset time period is screened to obtain normal second temperature data, and an initial fitting model is fitted based on the normal second temperature data to obtain a temperature fitting model that can accurately predict the temperature.

[0047] In one embodiment, the preset screening condition is that the local anomaly factor corresponding to the first temperature data is within the anomaly value range; the anomaly value range is determined based on the local anomaly factor corresponding to the effective temperature data; the effective temperature data is the first temperature data whose current downtime is greater than the preset duration threshold and whose current sampling moment is within the preset time range, and the temperature value of the effective temperature data is within the preset temperature range.

[0048] Among them, the current shutdown duration refers to the duration between the moment when the engine was last powered off and the next power-on moment adjacent to the last power-off moment. The current sampling moment refers to the sampling moment of the temperature data. The preset duration threshold refers to the preset threshold for judging the magnitude of the current shutdown duration. The preset time range refers to the preset time range for screening the current sampling moment. The preset temperature range refers to the preset temperature range for screening the temperature values of the valid temperature data. It can be understood that if the shutdown duration of the engine is short, it may lead to inaccurate acquisition of the first temperature data. By setting the preset duration threshold, the first temperature data is screened to obtain relatively accurate valid temperature data. If the current sampling moment exceeds the preset time range, it will result in exceeding the time range of the first temperature data. If the temperature value of the valid temperature data exceeds the preset temperature range, it will lead to abnormal data of the screened valid temperature data. Therefore, the preset time range and the preset temperature range are set to facilitate screening to obtain the first temperature data with normal valid temperature data.

[0049] In this embodiment, the valid temperature data refers to the first temperature data that satisfies that the current shutdown duration is not less than the preset duration threshold, the current sampling moment is within the preset time range, and its temperature value is within the preset temperature range. For example, when the preset duration threshold is 12 hours, the preset time range is within one month from the current sampling moment, and the preset temperature range is (-80, 150), if the current shutdown duration corresponding to the first temperature data is not less than 12 hours, the collection time point is within one month from the current sampling moment, and the temperature value is within the range of (-80, 150), then the first temperature data is determined as the valid temperature data.

[0050] In this embodiment, the first temperature data is , where is the first temperature data collected by N - 1 second temperature sensors at the th sampling moment, is the first temperature data collected by the first temperature sensor at the th sampling moment. Let the temperature data be any valid temperature data in the first temperature data . The local outlier factor of the temperature data is the ratio of the average reachability density of all temperature data within the K-nearest distance of the temperature data to the local reachability density of the temperature data .

[0051] Among them, the K-nearest distance of the temperature data refers to the kth temperature data closest to the temperature data . It can be understood that the temperature data The temperature data within the K-nearest distance from the temperature data and the temperature data The distance between them is less than or equal to the K-nearest distance of the temperature data p. Therefore, within the K-nearest distance, there are k data that are the closest to the temperature data in terms of distance.

[0052] Temperature data The local reachability density of the temperature data refers to the reciprocal of the average reachability distance of the temperature data within the K-nearest distance from the temperature data The average reachability distance refers to the average of the reachability distances between the temperature data and the temperature data within the K-nearest distance from the temperature data If a temperature data within the K-nearest distance corresponding to the temperature data is the reachability distance between the temperature data and the temperature data refers to the larger value between the distance between the temperature data and the temperature data and the K-nearest distance of the temperature data

[0053] Taking any valid temperature data (temperature data ) in the first temperature data as an example, the method for determining the outlier range is as follows: The temperature data within the K-nearest distance corresponding to the temperature data is denoted as , that is, ∈ . The K-nearest distance of the temperature data is denoted as , the K-nearest distance of the temperature data is denoted as , the distance between the temperature data and the temperature data is denoted as , the reachability distance between the temperature data and the temperature data is denoted as , then . The average reachability distance of the temperature data and the temperature data within the K-nearest distance is , then the local reachability density of the temperature data is the reciprocal of this average reachability distance, that is, the local reachability density of the temperature data is , temperature data The average reachable density of all temperature data within the K-nearest distance is , then the temperature data The local outlier factor of ( ) = . The computer device determines the local outlier factors of all valid temperature data according to the calculation method of the local outlier factor of the above temperature data ( ), and determines the numerical range corresponding to the local outlier factor of the valid temperature data as the outlier range.

[0054] In this embodiment, the computer device determines the local outlier factors of all first temperature data according to the calculation method of the local outlier factor of the above temperature data ( ). If the local outlier factor of the first temperature data is within the outlier range, it is determined that the first temperature data meets the preset screening conditions, and all the first temperature data that meet the preset screening conditions are determined as the second temperature data for model training to obtain a more accurate temperature prediction model. Understandably, if the local outlier factor is not within the outlier range, it indicates that the first temperature data is abnormal and needs to be discarded in order to obtain a more accurate temperature prediction model based on the normal second temperature data.

[0055] In one embodiment, as Figure 4 shown, step S104, that is, based on the residual mean and sample standard deviation corresponding to the first temperature sensor, determining the target fault degree corresponding to the first temperature sensor, includes: S401: Based on the residual mean and sample standard deviation corresponding to the first temperature sensor, determine the current fault type corresponding to the first temperature sensor; S402: Based on the current fault type corresponding to the first temperature sensor, determine the target fault degree corresponding to the first temperature sensor.

[0056] Among them, the current fault type refers to the fault type of the first temperature sensor. The fault types include, but are not limited to, no fault, drift fault, accuracy fault, drift accuracy fault, and complete fault. Among them, no fault means that there is no fault in the first temperature sensor. An accuracy fault means that as time goes by, the mean value of the first temperature data has not changed relative to the mean value of the predicted temperature data, while the standard deviation of the first temperature data has become larger relative to the standard deviation of the predicted temperature data, resulting in a decrease in the accuracy of the first temperature sensor. A drift fault means that the first temperature data actually collected by the first temperature sensor deviates from the predicted temperature data as time goes by. A drift accuracy fault means a fault situation where the first temperature sensor has both a drift fault and an accuracy fault. A complete fault means that the first temperature sensor has an irreparable fault.

[0057] Among them, as Figure 8 shown, it is a schematic diagram of the accuracy fault and drift fault corresponding to the first temperature sensor. From Figure 8 it can be seen that the manifestation of the drift fault is that as time goes by, the first temperature data collected by the first temperature sensor deviates from the predicted temperature data. The manifestation of the accuracy fault is that the data change trend of the first temperature data is overall consistent with the data change trend of the predicted temperature data, which will cause the mean value of the first temperature data not to change relative to the mean value of the predicted temperature data. As time goes by, at the same sampling moment, the difference between the first temperature data and the predicted temperature data becomes larger and larger, which will cause the standard deviation of the first temperature data to become larger relative to the standard deviation of the predicted temperature data.

[0058] As an example, in step S401, the computer device compares the residual mean and the sample standard deviation with the residual ranges and standard deviation ranges corresponding to different fault types respectively, determines the fault type corresponding to the residual mean and the sample standard deviation, and determines this fault type as the current fault type corresponding to the first temperature sensor. For example, the residual mean and the sample standard deviation are compared with the residual ranges and standard deviation ranges corresponding to no fault, drift fault, accuracy fault, drift accuracy fault, and complete fault respectively, and the fault type corresponding to the residual range where the residual mean is located and the standard deviation range where the sample standard deviation is located is determined as the current fault type.

[0059] As an example, in step S402, the computer device processes the mean residual and / or the mean standard deviation according to the current fault type corresponding to the first temperature sensor to determine the target fault degree corresponding to the first temperature sensor. Understandably, the methods for determining the target fault degree are different under different fault types, that is, the processing methods for the mean residual and / or the mean standard deviation are different. For example, since the precision fault is only caused by the abnormality of the standard deviation, if the fault type of the first temperature sensor is a precision fault, then it is necessary to focus on processing the sample standard deviation to determine the target fault degree of the first temperature sensor when the sample standard deviation is abnormal. Since the difference between the mean residual and the normal mean is large over time, which will cause a drift fault, if the fault type of the first temperature sensor is a drift fault, then it is necessary to focus on processing the mean residual to determine the target fault degree of the first temperature sensor when the mean residual is abnormal.

[0060] In this embodiment, based on the mean residual and the sample standard deviation corresponding to the first temperature sensor, the current fault type corresponding to the first temperature sensor is determined, and based on the current fault type corresponding to the first temperature sensor, the target fault degree corresponding to the first temperature sensor is determined, which can perform targeted fault analysis on the first temperature sensor according to different current fault types and can accurately determine the target fault degree corresponding to the first temperature sensor.

[0061] In one embodiment, as Figure 5 shown, step S401, that is, based on the mean residual and the sample standard deviation corresponding to the first temperature sensor, determining the current fault type corresponding to the first temperature sensor, includes: S501: If the absolute value of the mean residual is not greater than the residual lower limit and the sample standard deviation is not greater than the standard deviation lower limit, it is determined that the current fault type of the first temperature sensor is no fault; S502: If the absolute value of the mean residual is not greater than the residual lower limit, the sample standard deviation is greater than the standard deviation lower limit and not greater than the standard deviation upper limit, it is determined that the current fault type of the first temperature sensor is a precision fault; S503: If the absolute value of the mean residual is greater than the residual lower limit and not greater than the residual upper limit, and the sample standard deviation is not greater than the standard deviation lower limit, it is determined that the current fault type of the first temperature sensor is a drift fault; S504: If the absolute value of the mean residual is greater than the residual lower limit and not greater than the residual upper limit, and the sample standard deviation is greater than the standard deviation lower limit and not greater than the standard deviation upper limit, it is determined that the current fault type of the first temperature sensor is a drift precision fault; S505: If the absolute value of the mean residual is greater than the residual upper limit and the sample standard deviation is greater than the standard deviation upper limit, it is determined that the current fault type of the first temperature sensor is a complete fault.

[0062] Among them, the lower limit of the residual refers to the limit value corresponding to the absolute value of the mean residual measured in advance when the first temperature sensor is fault-free. The lower limit of the standard deviation refers to the limit value of the sample standard deviation measured in advance when the first temperature sensor is fault-free. Understandably, if the absolute value of the mean residual of the first temperature sensor exceeds the lower limit of the residual, and / or the sample standard deviation of the first temperature sensor exceeds the lower limit of the standard deviation, then the first temperature sensor has a fault.

[0063] As Figure 9 shown, it is a fault type diagram of the first temperature sensor determined by the upper limit of the residual, the lower limit of the residual, the upper limit of the standard deviation, and the lower limit of the standard deviation. From Figure 9 it can be seen that according to the magnitude relationship between the mean residual and the upper and lower limits of the residual, and between the sample standard deviation and the upper and lower limits of the standard deviation, different regions are divided, and different regions correspond to different fault types. The fault types corresponding to regions ① and ④ are drift accuracy faults, the fault types corresponding to regions ② and ③ are accuracy faults, the fault types corresponding to regions ⑤ and ⑧ are drift faults, regions ⑥ and ⑦ are fault-free, and regions ⑨ and ⑩ are complete faults.

[0064] Among them, the upper limit of the standard deviation refers to the critical value corresponding to the absolute value of the mean residual when the first temperature sensor has an irreparable fault measured in advance. The lower limit of the standard deviation refers to the critical value of the sample standard deviation when the first temperature sensor has an irreparable fault measured in advance. Understandably, if the absolute value of the mean residual of the first temperature sensor exceeds the upper limit of the residual, and the sample standard deviation of the first temperature sensor exceeds the upper limit of the standard deviation, then the first temperature sensor has an irreparable fault.

[0065] As an example, in step S501, when the computer device determines that the absolute value of the mean residual is greater than the lower limit of the residual , and the sample standard deviation is not greater than the lower limit of the standard deviation , it is determined that the current fault type of the first temperature sensor is fault-free. Understandably, as Figure 9 shown, if ≤ , and ≤ , corresponding to regions ⑥ and ⑦, it indicates that neither the mean residual nor the sample standard deviation of the first temperature sensor is abnormal, and the first temperature sensor is fault-free.

[0066] As an example, in step S502, when the computer device determines that the absolute value of the mean residual is not greater than the lower limit of the residual , and the sample standard deviation is greater than the lower limit of the standard deviation And not greater than the upper limit of the standard deviation , it is determined that the current fault type of the first temperature sensor is an accuracy fault. Figure 9 As shown, if ≤ ,and < ≤ , corresponding to areas ② and ③, it can be determined that the current fault type of the first temperature sensor is an accuracy fault.

[0067] As an example, in step S503, the computer device determines the residual mean The absolute value of Greater than the lower limit of residual And not greater than the upper limit of residual , sample standard deviation Not greater than the lower limit of standard deviation , it is determined that the current fault type of the first temperature sensor is a drift fault. Figure 9 It can be seen that if < ≤ ,and , corresponding to areas ⑤ and ⑧, are drift faults.

[0068] As an example, in step S504, the computer device determines the residual mean The absolute value of Greater than the lower limit of residual And not greater than the upper limit of residual , sample standard deviation Greater than the lower limit of standard deviation And not greater than the upper limit of the standard deviation , it is determined that the current fault type of the first temperature sensor is a drift accuracy fault. It can be understood that if Figure 9 As shown, if ≤ ,and < ≤ , corresponding to areas ① and ④, the corresponding current fault type is drift accuracy fault, that is, in this condition, the first temperature sensor has both drift fault and accuracy fault.

[0069] As an example, in step S505, the computer device determines the residual mean The absolute value of Greater than the upper limit of residual , and the sample standard deviation Greater than the upper standard deviation , it is determined that the current fault type corresponding to the first temperature sensor is a complete fault.Figure 9 It can be seen that if > , and > , corresponding to regions ⑨ and ⑩, that is, the residual mean deviates from the upper limit of the residual, and the sample standard deviation deviates from the upper limit of the standard deviation. The failure of the first temperature sensor is relatively serious and is an irreparable failure, that is, the current failure type is a complete failure.

[0070] In this embodiment, according to the magnitude relationship between the residual mean and the upper and lower limits of the residual, and the magnitude relationship between the sample standard deviation and the upper and lower limits of the standard deviation, the current failure type of the first temperature sensor can be determined more accurately. This method does not require complex processing of the residual mean and the sample standard deviation, and is relatively efficient and convenient.

[0071] In one embodiment, as Figure 6 shown, step S402, that is, based on the current failure type corresponding to the first temperature sensor, determining the target failure degree corresponding to the first temperature sensor, includes: S601: If the current failure type corresponding to the first temperature sensor is no failure, determine that the target failure degree corresponding to the first temperature sensor is 0.

[0072] S602: If the current failure type corresponding to the first temperature sensor is an accuracy failure, determine the difference between the sample standard deviation and the lower limit of the standard deviation as the first error, and determine the quotient of the first error and the normalization coefficient as the target failure degree corresponding to the first temperature sensor; S603: If the current failure type corresponding to the first temperature sensor is a drift failure, determine the difference between the absolute value of the residual mean and the lower limit of the residual as the second error, and determine the quotient of the second error and the normalization coefficient as the target failure degree corresponding to the first temperature sensor; S604: If the current failure type corresponding to the first temperature sensor is a drift accuracy failure, determine the arithmetic square root of the sum of the squares of the first error and the second error as the third error, and determine the quotient of the third error and the normalization coefficient as the target failure degree corresponding to the first temperature sensor; S605: If the current failure type corresponding to the first temperature sensor is a complete failure, determine that the target failure degree corresponding to the first temperature sensor is 1; Wherein, the normalization coefficient is the arithmetic square root of the sum of the squares of the first difference and the second difference. The first difference is the difference between the upper limit and the lower limit of the residual, and the second difference is the difference between the upper limit and the lower limit of the standard deviation.

[0073] In this example, the normalization coefficient is used to convert the target failure degree into a value between [0, 1], which is used to more intuitively reflect the target failure degree. The first difference is , the second difference is , and the normalization coefficient is . Among them, is the lower limit of the residual, is the upper limit of the residual, is the upper limit of the standard deviation, is the lower limit of the standard deviation.

[0074] As an example, in step S601, when the computer device determines that the current fault type corresponding to the first temperature sensor is no fault, it determines that the target fault degree corresponding to the first temperature sensor is 0. Understandably, if the first temperature sensor has no fault, the target fault degree is 0.

[0075] Among them, the first error is the sample standard deviation and the lower limit of the standard deviation . That is, the first error is .

[0076] As an example, in step S602, when the computer device determines that the current fault type corresponding to the first temperature sensor is an accuracy fault, it performs a difference operation on the sample standard deviation and the lower limit of the standard deviation to obtain the first error . The quotient of the first error and the normalization coefficient is determined as the target fault degree corresponding to the first temperature sensor. Understandably, if ≤ , and < ≤ , it indicates that the mean value of the residual is normal, while the sample standard deviation deviates from the normal standard deviation. Obtaining the difference between the sample standard deviation and the normal standard deviation can more accurately reflect the target fault degree of the first temperature sensor when the fault type is an accuracy fault.

[0077] The second error is the absolute value of the mean value of the residual and the lower limit of the residual . That is, the second error is - .

[0078] As an example, in step S603, when the computer device determines that the current fault type corresponding to the first temperature sensor is a drift fault, it determines the difference between the absolute value of the mean value of the residual and the lower limit of the residual as the second error - Divide the second error - by the normalization coefficient to obtain the quotient , which is determined as the target fault degree corresponding to the first temperature sensor. Understandably, if < ≤ , and , it indicates that the sample standard deviation is normal while the residual mean deviates from the normal value . Obtain the absolute value of the difference between and the lower limit of the residual as the second error, so as to more accurately determine the target fault degree of the first temperature sensor when the fault type is a drift fault based on the second error.

[0079] The third error is the arithmetic square root of the sum of the squares of the first error and the second error, that is, the third error is .

[0080] As an example, in step S604, when the computer device determines that the current fault type corresponding to the first temperature sensor is a drift precision fault, it determines the arithmetic square root of the sum of the squares of the first error and the second error - as the third error . Divide the third error by the normalization coefficient to obtain the quotient , which is determined as the target fault degree corresponding to the first temperature sensor. Understandably, if > , and > , it indicates that both the residual mean and the sample standard deviation deviate from the normal value. Therefore, considering the first error and the second error comprehensively, the arithmetic square root of the sum of the squares of the first error and the second error is used as the third error, so as to more accurately determine the target fault degree of the first temperature sensor when the fault type is a drift precision fault, that is, when there are both precision faults and drift faults.

[0081] As an example, in step S605, when the computer device determines that the current fault type corresponding to the first temperature sensor is a complete fault, it determines that the fault of the first temperature sensor is irreparable and relatively serious, and determines the target fault degree corresponding to the first temperature sensor as 1, so that the target fault degree corresponding to the complete fault is more intuitive.

[0082] In this embodiment, the fault types corresponding to each target fault degree are analyzed, and according to the fault types, the residual mean , the sample standard deviation , residual upper limit , residual lower limit , standard deviation lower limit and standard deviation upper limit By performing reasonable analysis on them, the target fault degree corresponding to each fault type can be determined more accurately and comprehensively.

[0083] In one embodiment, N temperature sensors are respectively arranged on the engine water outlet, the engine cylinder head, the engine contact position with the environment, the engine oil pan and the engine intake manifold; The first temperature data includes the water temperature at the water outlet, the water temperature of the engine cylinder head, the ambient temperature, the engine oil temperature and the gas temperature in the intake manifold.

[0084] Among them, the temperature sensor arranged on the engine water outlet is used to collect the water temperature at the water outlet of the engine, the temperature sensor arranged on the engine cylinder head is used to collect the water temperature of the engine cylinder head, the temperature sensor arranged at the engine contact position with the environment is used to collect the ambient temperature, the temperature sensor arranged on the engine oil pan is used to collect the engine oil temperature, and the temperature sensor arranged on the engine intake manifold is used to collect the gas temperature in the intake manifold.

[0085] Understandably, any one of the N temperature sensors can be determined as the first temperature sensor, and the remaining N - 1 temperature sensors are determined as the second temperature sensors, which are used to perform fault analysis on the first temperature sensor to determine the current fault type and the target fault degree of the first temperature sensor.

[0086] As an example, the temperature sensor arranged on the engine water outlet is determined as the first temperature sensor, and the second temperature sensors include the temperature sensor arranged on the engine cylinder head, the temperature sensor arranged at the engine contact position with the environment, the temperature sensor arranged on the engine oil pan, and the temperature sensor arranged on the engine intake manifold. Understandably, reasonably controlling the water temperature at the water outlet of the engine can effectively improve the performance of the engine. Therefore, real-time monitoring and analysis of the fault type and its target fault degree of the temperature sensor arranged at the engine water outlet are of relatively important significance for the control of the engine.

[0087] In this example, the computer device determines the temperature sensor set at the engine water outlet as the first temperature sensor, and determines the temperature sensors set at the engine cylinder head, the position where the engine contacts the environment, the engine oil pan, and the engine intake manifold as the second temperature sensors. By performing the above steps S101 to S104, steps S201 to S202, steps S301 to S302, and steps S401 to S405, it is possible to accurately and effectively determine the fault type and the target fault degree of the first temperature sensor set at the engine water outlet, achieving the purpose of accurately and effectively analyzing the fault of the temperature sensor at the engine water outlet.

[0088] In this embodiment, according to the data correlation between the N temperature sensors corresponding to the engine, any one of the N temperature sensors set at the engine water outlet, the engine cylinder head, the position where the engine contacts the environment, the engine oil pan, and the engine intake manifold can be determined as the first temperature sensor, achieving the purpose of analyzing the fault of the first temperature sensor with data correlation based on the N - 1 second temperature sensors.

[0089] In another embodiment, as Figure 7 shown, after step S104, that is, after determining the target fault degree corresponding to the first temperature sensor, the sensor fault analysis method further includes: S701: Analyze the target fault degree corresponding to the first temperature sensor based on M target analysis factors to determine the sensor fault results corresponding to the M target analysis factors, where M ≥ 1; S702: Send the sensor fault results corresponding to the M target analysis factors to the target terminal.

[0090] Among them, the target analysis factor refers to the factor used to analyze the target fault degree. For example, the vehicle mileage interval of the vehicle on which the engine is installed, the province to which the vehicle on which the engine is installed belongs, and the time interval of the sampling moment corresponding to the target fault degree. The vehicle mileage interval refers to the interval in which the current mileage of the vehicle on which the engine is installed is located. For example, if the driving mileage of the vehicle on which the engine corresponding to the first temperature sensor to be fault - analyzed is L, and L ∈ , then the vehicle mileage interval is . The time interval refers to the time period to which the sampling moment corresponding to the target fault degree belongs. For example, time intervals such as the last month, the last three months, and / or the last six months. The province to which it belongs is obtained by positioning the longitude and latitude of the vehicle on which the engine corresponding to the first temperature sensor is installed.

[0091] Among them, the sensor fault result refers to the analysis result obtained after analyzing the target fault degree according to the target analysis factors.

[0092] Among them, the target terminal is a terminal used for fault analysis, repair, and performance improvement of the temperature sensor.

[0093] In this embodiment, the computer device is set in the cloud and is used for analyzing the faults of the first temperature sensors corresponding to the engines installed in multiple vehicles. Understandably, the cloud can deploy the computer device to analyze the faults generally existing in the first temperature sensors corresponding to the engines in the vehicles in multiple provinces or regions served by the target terminal, so that the target terminal can specifically improve and upgrade the first temperature sensors to improve the performance of the first temperature sensors.

[0094] As an example, in step S701, the computer device obtains the target fault degree corresponding to M target analysis factors, and analyzes the target fault degree according to the dimension to which each target analysis factor belongs, and obtains the sensor fault result corresponding to each target analysis factor. For example, the computer device obtains the target fault degree of the first temperature sensor corresponding to the vehicle mileage interval of the vehicles in which multiple engines are installed, the province to which the vehicle in which the engine is installed belongs, and the time interval of the sampling moment corresponding to the target fault degree, so as to conduct an overall statistical analysis of the target fault degree of the first temperature sensors corresponding to the engines in vehicles with different vehicle mileage intervals, time intervals, and provinces to which they belong, and provide the corresponding statistical results for the target terminal, so that the target terminal can more reasonably improve the first temperature sensor and enhance the user's trust.

[0095] For example, the computer device performs a mean processing on the multiple target fault degrees corresponding to each vehicle mileage interval to obtain the average fault degree of each vehicle mileage interval, and generates a first average fault degree line chart according to each vehicle mileage interval and the average fault degree of each vehicle mileage interval, and determines the first average fault degree line chart as the sensor fault result. Among them, the first average fault degree line chart refers to the line chart obtained by performing a mean processing on the target fault degrees of different vehicle mileage intervals.

[0096] For example, the computer device performs a mean processing on the multiple target fault degrees corresponding to each time interval to obtain the average fault degree of each time interval, and generates a second average fault degree line chart according to each time interval and the average fault degree of each time interval, and determines the second average fault degree line chart as the sensor fault result. Among them, the second average fault degree line chart refers to the line chart obtained by performing a mean processing on the target fault degrees of different time intervals.

[0097] For example, a computer device classifies multiple target fault degrees into intervals, obtaining multiple fault degree intervals. Among them, a fault degree interval refers to an interval composed of target fault degrees of a certain magnitude.

[0098] The computer device can classify the target fault degrees within the range of [0, 1] according to the interval length c (where 0 < c < 1), obtaining a certain number of fault degree intervals. The computer device can also classify the target fault degrees according to the current fault type of the first temperature sensor, obtaining 5 fault degree intervals, namely, the fault degree interval corresponding to no fault type, the fault degree interval corresponding to a drift fault, the fault degree interval corresponding to a precision fault, the fault degree interval corresponding to a drift precision fault, and the fault degree interval corresponding to a complete fault.

[0099] The computer device counts the number of vehicles corresponding to all target fault degrees in each fault degree interval, generates a pie chart of the vehicle quantity ratio corresponding to each fault degree interval, and counts the provinces to which the vehicles equipped with the engines corresponding to the first temperature sensor belong in each fault degree interval, generating a bar chart of the vehicle quantities of the top H provinces with more vehicle quantities in each fault degree interval, where H ≥ 1. The pie chart of the vehicle quantity ratio corresponding to each fault degree interval and the bar chart of the vehicle quantities of the top H provinces with more vehicle quantities are determined as the sensor fault results.

[0100] In this example, the sensor fault results include but are not limited to the first average fault degree line chart, the second average fault degree line chart, the pie chart of the vehicle quantity ratio corresponding to each fault degree interval, and the bar chart of the vehicle quantities of the top H provinces with more vehicle quantities.

[0101] As an example, in step S702, the computer device sends the sensor fault results corresponding to M target analysis factors to the target terminal, so that the target terminal can specifically improve and upgrade the first temperature sensor based on the sensor fault results, improving the performance of the first temperature sensor.

[0102] In this example, based on M target analysis factors, the target fault degrees corresponding to the first temperature sensor are analyzed to determine the sensor fault results corresponding to M target analysis factors, and the sensor fault results corresponding to M target analysis factors are sent to the target terminal, so that the target terminal can more reasonably improve the first temperature sensor based on the sensor fault results, improving the performance of the first temperature sensor and enhancing the user's trust.

[0103] It should be understood that the sequence numbers of the steps in the above embodiments do not imply the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0104] In one embodiment, as Figure 10 shown, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the sensor fault analysis method in the above embodiments, such as Figure 1 the S101 - S104 shown, or Figures 2 to 7 as shown in [], for the sake of avoiding repetition, it will not be elaborated here.

[0105] In one embodiment, a sensor fault analysis system is provided, including a first temperature sensor, a second temperature sensor, and the above computer device. The computer device is respectively connected to the first temperature sensor and the second temperature sensor, and is used to determine the target fault degree corresponding to the first temperature sensor.

[0106] In one embodiment, an automobile is provided, including the above sensor fault analysis system.

[0107] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above division of each functional unit and module is used as an example. In practical applications, the above functions can be allocated to different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above.

[0108] The above - described embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A sensor failure analysis method, characterized in that: include: Acquire first temperature data corresponding to the same engine, the first temperature data including first temperature data corresponding to the first temperature sensor and first temperature data corresponding to N-1 second temperature sensors, where N≥2; Determine predicted temperature data corresponding to the first temperature sensor based on first temperature data corresponding to the first temperature sensor and first temperature data corresponding to N-1 second temperature sensors; Determine a residual mean and a sample standard deviation corresponding to the first temperature sensor based on first temperature data and predicted temperature data corresponding to the first temperature sensor; A target fault degree corresponding to the first temperature sensor is determined based on a residual mean and a sample standard deviation corresponding to the first temperature sensor.

2. The sensor failure analysis method according to claim 1, characterized in that: The determining, based on the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to N-1 second temperature sensors, predicted temperature data corresponding to the first temperature sensor includes: Determine a temperature prediction model corresponding to the first temperature sensor based on first temperature data corresponding to the first temperature sensor and first temperature data corresponding to N-1 second temperature sensors; Based on the temperature prediction model corresponding to the first temperature sensor and the first temperature data corresponding to N-1 second temperature sensors, predicted temperature data corresponding to the first temperature sensor is determined.

3. The sensor failure analysis method according to claim 2, characterized in that: The temperature prediction model is a temperature fitting model between the first temperature sensor and N-1 second temperature sensors; The determining, based on the first temperature data corresponding to the first temperature sensor and the first temperature data corresponding to N-1 second temperature sensors, a temperature prediction model corresponding to the first temperature sensor includes: Based on the preset screening condition, a plurality of first temperature data corresponding to the first temperature sensor and a plurality of first temperature data corresponding to the N-1 second temperature sensors are screened to obtain a plurality of second temperature data corresponding to the first temperature sensor and a plurality of second temperature data corresponding to the N-1 second temperature sensors; Based on the multiple second temperature data corresponding to the first temperature sensor and the multiple second temperature data corresponding to the N-1 second temperature sensors, the initial fitting model is fitted to obtain the temperature fitting model corresponding to the first temperature sensor.

4. The sensor failure analysis method according to claim 3, characterized in that: The preset screening condition is that the local abnormal factor corresponding to the first temperature data is within the abnormal value range; The abnormal value range is determined based on the local abnormal factor corresponding to the effective temperature data; The valid temperature data is the first temperature data when the current shutdown time is greater than a preset time threshold, the current sampling time is within a preset time range, and the temperature value of the valid temperature data is within a preset temperature range.

5. The sensor failure analysis method according to claim 1, characterized in that: The determining, based on the residual mean and sample standard deviation corresponding to the first temperature sensor, a target fault degree corresponding to the first temperature sensor, includes: Determine a current fault type corresponding to the first temperature sensor based on a residual mean and a sample standard deviation corresponding to the first temperature sensor; Based on a current fault type corresponding to the first temperature sensor, a target fault degree corresponding to the first temperature sensor is determined.

6. The sensor failure analysis method according to claim 5, characterized in that: The determining, based on the residual mean and the sample standard deviation corresponding to the first temperature sensor, the current fault type corresponding to the first temperature sensor includes: If the absolute value of the residual mean is not greater than the residual lower limit, and the sample standard deviation is not greater than the standard deviation lower limit, determining that the current fault type of the first temperature sensor is no fault; If the absolute value of the residual mean is not greater than the residual lower limit, and the sample standard deviation is greater than the standard deviation lower limit and not greater than the standard deviation upper limit, then it is determined that the current fault type of the first temperature sensor is an accuracy fault; If the absolute value of the residual mean is greater than the residual lower limit and not greater than the residual upper limit, and the sample standard deviation is not greater than the standard deviation lower limit, it is determined that the current fault type of the first temperature sensor is a drift fault; If the absolute value of the residual mean is greater than the residual lower limit but not greater than the residual upper limit, and the sample standard deviation is greater than the standard deviation lower limit but not greater than the standard deviation upper limit, then it is determined that the current fault type of the first temperature sensor is a drift accuracy fault; If the absolute value of the residual mean is greater than the residual upper limit, and the sample standard deviation is greater than the standard deviation upper limit, it is determined that the current fault type of the first temperature sensor is a complete fault.

7. The sensor failure analysis method according to claim 5, characterized in that: The determining, based on the current fault type corresponding to the first temperature sensor, a target fault degree corresponding to the first temperature sensor, includes: If the current fault type corresponding to the first temperature sensor is no fault, determining the target fault degree corresponding to the first temperature sensor to be 0; If the current fault type corresponding to the first temperature sensor is an accuracy fault, the difference between the sample standard deviation and the lower limit of the standard deviation is determined as the first error, and the quotient of the first error and the normalization coefficient is determined as the target fault degree corresponding to the first temperature sensor; If the current fault type corresponding to the first temperature sensor is a drift fault, the difference between the absolute value of the residual mean and the residual lower limit is determined as the second error, and the quotient of the second error and the normalization coefficient is determined as the target fault degree corresponding to the first temperature sensor; If the current fault type corresponding to the first temperature sensor is a drift accuracy fault, determining the arithmetic square root of the sum of the squares of the first error and the second error as a third error, and determining a quotient of the third error and a normalization coefficient as a target fault degree corresponding to the first temperature sensor; If the current fault type corresponding to the first temperature sensor is a complete fault, determining that the target fault degree corresponding to the first temperature sensor is 1; The normalization coefficient is the arithmetic square root of the sum of squares of the first difference and the second difference, the first difference is the difference between the upper limit of the residual and the lower limit of the residual, and the second difference is the difference between the upper limit of the standard deviation and the lower limit of the standard deviation.

8. The sensor failure analysis method according to claim 1, characterized in that: N temperature sensors are respectively arranged on the engine water outlet, the engine cylinder head, the position where the engine contacts the environment, the engine oil pan and the engine intake manifold; The first temperature data includes water outlet water temperature, engine cylinder head water temperature, ambient temperature, engine oil temperature and intake manifold gas temperature.

9. The sensor failure analysis method according to claim 1, characterized in that: After determining the target fault degree corresponding to the first temperature sensor, the sensor fault analysis method further includes: Based on M target analysis factors, analyzing the target fault degree corresponding to the first temperature sensor, and determining the sensor fault results corresponding to the M target analysis factors, where M≥1; The sensor failure results corresponding to the M target analysis factors are sent to the target terminal.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the sensor fault analysis method according to any one of claims 1 to 9 is implemented.

11. A sensor fault analysis system, comprising a first temperature sensor, a second temperature sensor and the computer device according to claim 10, characterized in that: The computer device is connected to the first temperature sensor and the second temperature sensor respectively, and is used to determine the target fault degree corresponding to the first temperature sensor.

12. A car, characterized in that: Includes the sensor fault analysis system as described in claim 11.

Images