Data analysis method for diesel engine oil consumption test

Abstract

The invention relates to the field of data processing, in particular to a data analysis method for a diesel engine oil consumption test, and the method comprises the steps: constructing a data set which comprises a plurality of samples, and each sample comprises multiple dimensions of oil consumption data; dividing the data set into sliding windows with various lengths, for one sliding window, calculating the fluctuation degree of data in the sliding window, for one sample, calculating the LOF score of the sample in each sliding window by using an improved LOF algorithm, and combining the fluctuation degree of each sliding window to obtain the weighted LOF score of the sample through weighted summation; and marking the samples of which the weighted LOF scores are greater than or equal to a preset threshold value as abnormal samples, and processing the abnormal samples. According to the method, the fluctuation degree of the data in the sliding window is calculated, and the weighted LOF score is combined to identify the abnormal sample, so that the abnormal point can be positioned more accurately.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and more specifically, to a data analysis method for a diesel engine fuel consumption test. Background Art

[0002] A vehicle diesel engine is a power unit based on an internal combustion engine, which is mainly used in the automotive field and generates power through the combustion of diesel. Among them, a vehicle diesel engine can only return to its initial state after going through four thermodynamic processes: intake, compression, expansion (work), and exhaust, so that the diesel engine can continuously generate mechanical work. The performance indicators of a vehicle diesel engine mainly include dynamic indicators (such as effective torque, effective power, speed, etc.) and fuel consumption rate, which directly affect the performance and efficiency of the diesel engine. By analyzing the fuel consumption test data, we can find out the shortcomings of vehicle diesel engines in terms of fuel consumption, and then guide the design optimization of the engine.

[0003] In the anomaly detection of fuel consumption data of automotive diesel engines, the LOF algorithm can identify abnormal fuel consumption data points that are significantly different from the majority of data points. These anomalies may be caused by engine failure, abnormal driving behavior or environmental factors.

[0004] However, there are significant differences in the fuel consumption performance of diesel engines under different working conditions. For example, under heavy load or full load conditions, the diesel engine needs to output maximum power and torque, and the fuel consumption is usually higher; under light load or idling conditions, the fuel consumption is relatively low; in addition, transient conditions with turbocharging, low temperature starting conditions, and plateau or high altitude conditions will also affect the fuel consumption of diesel engines. The density of fuel consumption data under these conditions (that is, the distribution and degree of aggregation of data points) is also different. Normal fuel consumption data under a certain condition may be misjudged as anomalies by the LOF algorithm because its density is significantly different from that of data under other conditions. Conversely, some data points that appear abnormal under other conditions are considered normal under specific conditions, resulting in misjudgment of the traditional LOF algorithm in anomaly detection. Summary of the invention

[0005] In order to solve the technical problem that the traditional LOF algorithm has limitations and causes misjudgment in anomaly detection due to differences in fuel consumption performance and data density of the above-mentioned diesel engine under different working conditions, the present invention provides the following technical solution.

[0006] A data analysis method for a diesel engine fuel consumption test, comprising: Collect the fuel consumption data of the diesel engine in multiple dimensions at a set time, integrate the multi-dimensional fuel consumption data corresponding to the same time into one sample, and sort the samples corresponding to each collected time into one data set in chronological order; The data set is divided into sliding windows of various lengths. For a sliding window, the fluctuation degree of the data in the sliding window is calculated. For a sample, the LOF score in each sliding window is calculated using an improved LOF algorithm, and the weighted LOF score of the sample is obtained by weighted summation in combination with the fluctuation degree of each sliding window; the K value in the improved LOF algorithm is positively correlated with the sparsity of the data set; The samples whose weighted LOF scores are greater than or equal to the preset threshold are marked as abnormal samples and processed.

[0007] The present invention first integrates and analyzes multi-dimensional fuel consumption data, divides the data set formed after integration into sliding windows of various lengths, and then captures the local characteristics of the data in different time periods, further calculates the degree of fluctuation of the data in the sliding window, and dynamically adjusts the K value of the LOF algorithm according to the change of data density, and then uses the improved LOF algorithm to calculate the LOF score of the sample in each sliding window, and combines the degree of fluctuation of each sliding window to obtain the weighted LOF score of the sample through weighted summation, which can more accurately identify abnormal samples. The traditional LOF algorithm may have the problem of inaccurate local density estimation or insufficient sensitivity to abnormalities of different scales, and the improved algorithm can better adapt to the characteristics of different data sets and improve the accuracy of abnormality detection by considering the positive correlation between the K value and the sparsity of the data set and the weighted summation method.

[0008] Preferably, the multi-dimensional fuel consumption data includes one or more of fuel consumption, speed, torque, and coolant temperature.

[0009] Preferably, the sparsity of the data set satisfies the relationship: ; In the formula, is the sparsity of the dataset, is the number of dimensional fuel consumption data contained in the dataset, is the number of samples contained in the data set, For the The first The fuel consumption data of the dimension is compared with the first The mean Euclidean distance between the fuel consumption data of each dimension, Indicates normalization processing.

[0010] By calculating the average of the average Euclidean distances of each dimension in the dataset, the sparsity of the dataset can be evaluated. If the sparsity of the dataset is large, it means that the samples in the dataset are sparsely distributed in each dimension; if the sparsity of the dataset is small, it means that the samples in the dataset are densely distributed in each dimension.

[0011] Preferably, the K value in the improved LOF algorithm satisfies the relationship: ; In the formula, represents the K value in the improved LOF algorithm, is the preset initial K value, is the sparsity of the dataset, is the logarithmic function with base e.

[0012] For different data sets, especially those with uneven density distribution, a fixed K value may not be suitable for all situations. By introducing the sparsity of the data set, the improved K value can be dynamically adjusted according to the distribution characteristics of the data. When the data set is more sparse, the sparsity value of the data set will be larger, so that the K value increases to better capture abnormal points in sparse areas.

[0013] Preferably, said dividing the data set into sliding windows of various lengths comprises: The data set is divided into sliding windows of different lengths, with the minimum length being half the number of samples in the data set and the maximum length being the number of samples in the data set.

[0014] By setting the minimum length to half the number of samples in the dataset, we can ensure that each window contains enough data points to reflect certain basic features of the data, avoiding feature loss due to a window that is too short. By setting the maximum length to the number of samples in the dataset, we can avoid computational inefficiency and waste of resources due to a window that is too long. At the same time, this also ensures that each data point is included in at least one window, thereby making full use of the dataset.

[0015] Preferably, the process of obtaining the fluctuation degree includes: In the current sliding window, for each dimension of fuel consumption data, calculate the mean value of the dimension of fuel consumption data in the current sliding window. For each sample in the current sliding window, calculate the absolute value of the difference between the dimension of fuel consumption data and its mean value. Sum the absolute values ​​of the differences of all samples in the current sliding window to obtain the sum of fluctuations of the dimension in the current sliding window. In the current sliding window, the slope of the fuel consumption data of this dimension is calculated, and the average slope of the fuel consumption data of this dimension over all samples is calculated; The product of the sum of the fluctuations of the fuel consumption data of each dimension and the mean of the slope is summed and normalized to obtain the degree of fluctuation of the data in the sliding window.

[0016] By calculating the absolute value of the difference between the fuel consumption data of each sample and the mean in the current sliding window, and summing them up to get the total fluctuation, this step measures the degree of discreteness of the fuel consumption data of this dimension in the current window. The larger the total fluctuation, the greater the deviation of the data point from the mean, that is, the greater the fluctuation of the data; calculating the slope of the fuel consumption data and finding the average value can reflect the changing trend of the data within a certain time range. The larger the slope mean, the more obvious the changing trend of the data.

[0017] The product of the sum of the fluctuations of the fuel consumption data of each dimension and the mean of the slope is summed and normalized to obtain the overall fluctuation degree of the data in the sliding window, which comprehensively considers the discrete degree and change trend of the data, so as to more comprehensively and accurately evaluate the fluctuation of the data.

[0018] Preferably, the weighted LOF score satisfies the relationship: ; In the formula, is the weighted LOF score of the sample, is the number of sliding windows that contain the sample, Indicates The degree of fluctuation of the data in a sliding window Indicates that the sample is in LOF score of a sliding window.

[0019] By taking a weighted average of the LOF scores in multiple sliding windows, a comprehensive, smooth and reliable anomaly score is provided, which can better capture and reflect the anomalies of samples in different time periods while reducing the noise impact caused by data fluctuations.

[0020] Preferably, processing the abnormal sample includes: According to the characteristics of abnormal samples, fault troubleshooting is carried out, and according to the troubleshooting results, the diesel engine is adjusted or maintained.

[0021] Preferably, the sparsity of the data set satisfies the relationship: ; In the formula, is the sparsity of the dataset, is the number of dimensional fuel consumption data contained in the dataset, For the The information entropy of fuel consumption data in three dimensions.

[0022] Preferably, after collecting the fuel consumption data of the diesel engine in multiple dimensions at a set time, all the fuel consumption data are standardized or normalized.

[0023] The beneficial effects of the present invention are: The present invention realizes a comprehensive, accurate and efficient analysis of diesel engine fuel consumption data through multi-dimensional data integration, dynamic anomaly detection, adaptive K value selection, multiple sliding window divisions, and a combination of fluctuation degree and weighted LOF score, thus providing strong support for diesel engine performance evaluation, fault diagnosis and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a method flow chart of steps S1 to S3 in a data analysis method for a diesel engine fuel consumption test according to an embodiment of the present invention. DETAILED DESCRIPTION

[0025] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments.

[0026] The application scenario of the present invention is: using the improved LOF algorithm to perform abnormality detection on the fuel consumption data of a diesel engine.

[0027] Reference Figure 1 A diesel engine fuel consumption test data analysis method includes steps S1 to S3, which are as follows: S1: Collect the fuel consumption data of the diesel engine in multiple dimensions at a set time, integrate the multi-dimensional fuel consumption data corresponding to the same time into one sample, and sort the samples corresponding to each collected time into one data set in chronological order.

[0028] In one embodiment, multiple key data of the diesel engine are collected simultaneously according to the set time, including but not limited to: fuel consumption, speed, torque, coolant temperature, exhaust temperature, intake pressure and exhaust flow. Specifically, a fuel consumption meter is used to measure the fuel consumption of the diesel engine, a speed sensor is used to monitor the speed of the diesel engine, a torque sensor is used to measure the torque of the diesel engine, a temperature sensor is used to record the temperature of the coolant, a thermocouple is used to monitor the exhaust temperature, an intake pressure sensor is used to measure the intake pressure, and a vortex flowmeter is used to measure the exhaust flow.

[0029] The above 7 characteristic data collected at the same time are integrated into one sample. Each sample contains a complete description of the working status of the diesel engine at that moment. The samples corresponding to each moment are sorted in sequence according to the time sequence of collection. Since the 7 characteristic data collected have different units and formats, in order to conduct unified analysis and processing, we need to standardize these data. Standardization processing usually includes data cleaning (removing outliers, filling missing values, etc.), data conversion (such as logarithmic transformation, Z-score standardization, etc.) and data normalization (scaling the data to a specific range, such as between 0 and 1). The standardized data is integrated into a structured diesel engine fuel consumption test data set.

[0030] In summary, this dataset not only contains the fuel consumption data of the diesel engine at each moment, but also contains information related to it in multiple dimensions such as speed, torque, temperature, pressure, flow, etc., and reflects the dynamic changes of the diesel engine during the entire test process.

[0031] S2: Divide the data set into sliding windows of various lengths. For a sliding window, calculate the fluctuation degree of the data in the sliding window. For a sample, use the improved LOF algorithm to calculate its LOF score in various sliding windows, and combine the fluctuation degree of each sliding window to obtain the weighted LOF score of the sample through weighted summation; the K value in the improved LOF algorithm is positively correlated with the sparsity of the data set.

[0032] In the diesel engine fuel consumption test, the density of fuel consumption data collected in different test stages (such as cold start, steady state, transient) is different. For example, the data density in the steady state stage is high, the data is relatively stable, and the change is small; the data density in the transient stage is low, the data change is large, and the fluctuation is obvious; the data in the cold start stage may be between steady state and transient, and may also have its own unique volatility.

[0033] The traditional LOF algorithm uses a fixed K value (i.e., the number of nearest neighbors) for anomaly detection. However, this method of fixing the K value will have the following problems in scenarios with different data densities: In areas with high data density (such as the steady-state stage), a larger K value may cause the LOF score of abnormal data or noise to be too low, thus being mistakenly identified as normal data; in areas with low data density (such as the transient stage), a smaller K value may cause the LOF score of normal data to be too high, thus being mistakenly identified as abnormal data.

[0034] Therefore, the K value is adaptively adjusted by calculating the sparsity of the data. A data set with a greater degree of sparsity (i.e., a data set with a smaller density) should use a larger K value to improve the accuracy of anomaly detection.

[0035] Specifically, first, for each dimensional fuel consumption data and each sample in the data set, the degree of difference between different samples of each dimensional fuel consumption data is analyzed, and then the average difference between different samples of the dimensional fuel consumption data is quantified, and the average sparsity of all features is averaged to obtain the sparsity of the entire data set.

[0036] In one embodiment, the sparsity of the data set satisfies the relationship:

[0037] In the formula, is the sparsity of the dataset, is the number of dimensional fuel consumption data contained in the dataset, is the number of samples contained in the data set, For the The first The fuel consumption data of the dimension is compared with the first The mean Euclidean distance between the fuel consumption data of each dimension, Indicates normalization processing.

[0038] in, Reflects the The difference between different samples in fuel consumption data of each dimension, Reflects the The average difference of fuel consumption data in different dimensions between different samples, The average sparsity of all features is averaged to obtain the sparsity of the entire data set.

[0039] The greater the sparsity, the smaller the density of the data set. This means that the eigenvalues ​​in the data set vary greatly between different samples, and the data points are more dispersed; conversely, the smaller the sparsity, the greater the density of the data set. This means that the eigenvalues ​​in the data set vary less between different samples, and the data points are more concentrated.

[0040] In another embodiment, the sparsity of the data set also satisfies the relationship:

[0041]

[0042] In the formula, is the sparsity of the dataset, is the number of dimensional fuel consumption data contained in the dataset, For the The information entropy of fuel consumption data in each dimension, is the number of samples contained in the data set, For the The first The fuel consumption data of the dimension is compared with the first The mean Euclidean distance between the fuel consumption data of each dimension, is the logarithmic function with base e.

[0043] Furthermore, the K value in the LOF algorithm is adaptively adjusted according to the sparsity of the data set, thereby obtaining an improved LOF algorithm. The K value in the improved LOF algorithm satisfies the following relationship:

[0044] In the formula, represents the K value in the improved LOF algorithm, is the preset initial K value, is the sparsity of the dataset, is the logarithmic function with base e.

[0045] Among them, by multiplying , so that the initial K value can be appropriately adjusted according to the sparsity of the data set. If the data set is very sparse (i.e., its sparsity is very large), then will also be larger, making the final On the contrary, when the data set is not so sparse, is close to 1, making the final The value is close to the initial K value.

[0046] In another embodiment, the K value in the improved LOF algorithm also satisfies the relationship:

[0047] In the formula, represents the K value in the improved LOF algorithm, is the preset initial K value, is the sparsity of the dataset, and are all constants obtained by fitting the data set.

[0048] After obtaining the improved LOF algorithm, the collected data set is divided into sliding windows of different lengths, the fluctuation degree of the data in each sliding window is calculated, and the sample anomaly score is obtained by weighting according to the fluctuation degree and the LOF score of the sample in the sliding window.

[0049] Specifically, the length of the sliding window is first set, with the minimum length being half of the number of samples in the data set and the maximum length being the number of samples in the data set.

[0050] Then, within the current sliding window, for each dimension of fuel consumption data, calculate the mean of the fuel consumption data of this dimension within the current sliding window; for each sample in the current sliding window, calculate the absolute value of the difference between the fuel consumption data of this dimension and its mean; sum the absolute values ​​of the differences of all samples in the current sliding window to obtain the sum of the fluctuations of this dimension within the current sliding window; within the current sliding window, calculate the slope of the fuel consumption data of this dimension (that is, the change in the fuel consumption data of this dimension between adjacent samples), and further calculate the average value of the slope of the fuel consumption data of this dimension over all samples.

[0051] The fluctuation degree of the data in the sliding window is further measured according to the slope of the fuel consumption data of each dimension and the absolute difference between the fuel consumption data of this dimension in the sliding window and the mean of the fuel consumption data of this dimension of all samples, that is, the relationship is satisfied:

[0052] In the formula, is the fluctuation degree of the data in the sliding window, is the number of dimensional fuel consumption data contained in the dataset, The first In the sample The value of fuel consumption data in each dimension, The first The mean of fuel consumption data in each dimension, The first The slope mean of the fuel consumption data in each dimension, is the number of samples contained in the sliding window, Indicates normalization processing.

[0053] The above fluctuation degree reflects the change of data in the sliding window. When the fluctuation degree is large, it means that the data has a large change in the sliding window. This may cause the abnormal sample to be far away from the dense area, making its local density much lower than that of its neighbors, resulting in a high LOF score, and normal samples may be misjudged as abnormal samples; when the fluctuation degree is small, it means that the data changes little in the sliding window, and the abnormal sample may still be in a relatively dense area, making its local density close to that of its neighbors, resulting in a low LOF score, and the abnormal sample may be misjudged as a normal sample.

[0054] The fluctuation degree of the data in each sliding window can be calculated similarly according to the calculation formula of the fluctuation degree of the data in the sliding window.

[0055] In one embodiment, for each sample's position in each sliding window, its LOF (local outlier factor) score is calculated, and then the LOF score of each sample is adjusted according to the degree of fluctuation of the data in each sliding window to obtain a weighted LOF score of the sample, that is, the relationship is satisfied:

[0056] In the formula, is the weighted LOF score of the sample, is the number of sliding windows that contain the sample, Indicates The degree of fluctuation of the data in a sliding window Indicates that the sample is in LOF score of a sliding window.

[0057] When the fluctuation is greater, The smaller the fluctuation, the smaller the weight of the LOF score of the sliding window. This means that the area with large fluctuations (which may be normal areas) has less impact on the final anomaly score, avoiding misjudging normal samples as abnormal. The larger the value, the greater the weight of the LOF score of the sliding window. This means that areas with small fluctuations (possibly abnormal areas) have a greater impact on the final anomaly score, avoiding missing abnormal samples.

[0058] The weighted LOF scores of all samples can be calculated according to the calculation formula of the weighted LOF scores of the above samples.

[0059] S3: Mark samples whose weighted LOF scores are greater than or equal to a preset threshold as abnormal samples, and process the abnormal samples.

[0060] In one embodiment, the threshold is set to 1.5, and samples with a weighted LOF score greater than or equal to the preset threshold are marked as abnormal samples. Furthermore, detailed information of the abnormal samples is recorded, including timestamps, abnormal features, etc., and troubleshooting is performed based on the features of the abnormal samples, for example, checking whether the diesel engine sensor is faulty or whether the fuel system is abnormal.

[0061] According to the inspection results, the diesel engine is adjusted or maintained accordingly, for example, replacing faulty parts, adjusting fuel injection parameters, etc.

[0062] It should be noted that, for those skilled in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these modifications and improvements all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A data analysis method for diesel engine fuel consumption test, characterized in that: include: Collect the fuel consumption data of the diesel engine in multiple dimensions at a set time, integrate the multi-dimensional fuel consumption data corresponding to the same time into one sample, and sort the samples corresponding to each collected time into one data set in chronological order; The data set is divided into sliding windows of various lengths. For a sliding window, the fluctuation degree of the data in the sliding window is calculated. For a sample, the LOF score in each sliding window is calculated using an improved LOF algorithm, and the weighted LOF score of the sample is obtained by weighted summation in combination with the fluctuation degree of each sliding window; the K value in the improved LOF algorithm is positively correlated with the sparsity of the data set; The samples whose weighted LOF scores are greater than or equal to the preset threshold are marked as abnormal samples and processed.

2. A diesel engine fuel consumption test data analysis method according to claim 1, characterized in that: The multi-dimensional fuel consumption data includes one or more of fuel consumption, rotation speed, torque, and coolant temperature.

3. A diesel engine fuel consumption test data analysis method according to claim 2, characterized in that: The sparsity of the data set satisfies the relationship: ; In the formula, is the sparsity of the dataset, is the number of dimensional fuel consumption data contained in the dataset, is the number of samples contained in the data set, For the The first The fuel consumption data of the dimension is compared with the first The mean Euclidean distance between the fuel consumption data of each dimension, Indicates normalization processing.

4. A diesel engine fuel consumption test data analysis method according to claim 3, characterized in that: The K value in the improved LOF algorithm satisfies the relationship: ; In the formula, represents the K value in the improved LOF algorithm, is the preset initial K value, is the sparsity of the dataset, is the logarithmic function with base e.

5. A diesel engine fuel consumption test data analysis method according to claim 4, characterized in that: The step of dividing the data set into sliding windows of various lengths comprises: The data set is divided into sliding windows of different lengths, with the minimum length being half the number of samples in the data set and the maximum length being the number of samples in the data set.

6. A diesel engine fuel consumption test data analysis method according to claim 5, characterized in that: The process of obtaining the fluctuation degree includes: In the current sliding window, for each dimension of fuel consumption data, calculate the mean value of the dimension of fuel consumption data in the current sliding window. For each sample in the current sliding window, calculate the absolute value of the difference between the dimension of fuel consumption data and its mean value. Sum the absolute values ​​of the differences of all samples in the current sliding window to obtain the sum of fluctuations of the dimension in the current sliding window. In the current sliding window, the slope of the fuel consumption data of this dimension is calculated, and the average slope of the fuel consumption data of this dimension over all samples is calculated; The product of the sum of the fluctuations of the fuel consumption data of each dimension and the mean of the slope is summed and normalized to obtain the degree of fluctuation of the data in the sliding window.

7. A diesel engine fuel consumption test data analysis method according to claim 6, characterized in that: The weighted LOF score satisfies the relationship: ; In the formula, is the weighted LOF score of the sample, is the number of sliding windows that contain the sample, Indicates The degree of fluctuation of the data in a sliding window Indicates that the sample is in LOF score of a sliding window.

8. A diesel engine fuel consumption test data analysis method according to claim 7, characterized in that: Processing of abnormal samples includes: According to the characteristics of abnormal samples, fault troubleshooting is carried out, and according to the troubleshooting results, the diesel engine is adjusted or maintained.

9. A diesel engine fuel consumption test data analysis method according to claim 2, characterized in that: The sparsity of the data set satisfies the relationship: ; In the formula, is the sparsity of the dataset, is the number of dimensional fuel consumption data contained in the dataset, For the The information entropy of fuel consumption data in three dimensions.

10. A diesel engine fuel consumption test data analysis method according to claim 1, characterized in that: After collecting the fuel consumption data of the diesel engine in multiple dimensions at a set time, all the fuel consumption data are standardized or normalized.

Images