A rail vehicle RAMS data analysis and prediction method and device, electronic equipment and storage medium

Abstract

A rail vehicle RAMS data analysis and prediction method and device, electronic equipment and storage medium, compared with the prior art, the present application can better learn and capture the complex nonlinear relationship between the rail vehicle operating state and the RAMS index by modeling a large amount of time series data using a deep learning model, can effectively improve the accuracy and automation degree of the rail vehicle RAMS index prediction, and provide strong support for vehicle operating condition monitoring and maintenance decision.

Description

Technical Field

[0001] This invention relates to the field of RAMS data analysis technology, and in particular to a method and apparatus for analyzing and predicting RAMS data of rail vehicles, electronic equipment, and storage medium. Background Technology

[0002] In recent years, with the continuous development of urban rail transit systems, the number of rail vehicles in operation has also been increasing. However, various malfunctions and accidents are inevitable during rail vehicle operation, which will have a significant impact on rail transit operations, affecting safety, reliability, and operational efficiency. In order to better monitor and predict the reliability of rail vehicles, many cities are conducting RAMS (Reliability, Availability, Maintainability, and Safety) data collection and analysis.

[0003] However, current RAMS data analysis for rail vehicles mainly relies on traditional statistical methods and empirical models. These methods can only analyze historical data and are difficult to predict potential future problems. On the other hand, due to the complexity of rail vehicle systems, considering the reliability of a single component or system cannot fully reflect the overall reliability of the vehicle. Summary of the Invention

[0004] Currently, the analysis of RAMS data for rail vehicles mainly relies on traditional statistical methods and empirical models. These methods can only analyze historical data and are difficult to predict potential future problems.

[0005] To address the aforementioned problems, this invention provides a method and apparatus for analyzing and predicting RAMS data of rail vehicles, an electronic device, and a computer-readable storage medium.

[0006] The specific technical solution adopted in this application is as follows:

[0007] A method for analyzing and predicting RAMS data of rail vehicles, comprising:

[0008] Step 1: Measure the time series data of the rail vehicle under different RAMS indicators; the time series data of the rail vehicle includes: real-time vehicle speed time series, power system traction time series, fuel tank temperature time series, acceleration time series, car temperature time series and car smoke concentration time series;

[0009] Step 2: Preprocess the time series of the rail vehicles to obtain the preprocessed time series of the rail vehicles;

[0010] Step 3: Calculate the Pearson coefficient for the time series and RAMS index of each rail vehicle;

[0011] Step 4: Remove the time series data of the corresponding rail vehicles that are below the preset threshold to obtain training samples;

[0012] Step 5: Input the training samples into the deep learning model for training to obtain the rail vehicle RAMS index estimation model;

[0013] Step 6: Input the time series data of the target vehicle into the rail vehicle RAMS index estimation model to obtain the total RAMS index of the target vehicle.

[0014] Preferably, step 2: preprocessing the time series of the rail vehicles to obtain a preprocessed time series of the rail vehicles specifically includes:

[0015] Step 2.1: Use basis functions to decompose the time series signal of the rail vehicle into sub-signals of different frequencies;

[0016] Step 2.2: Obtain the transform coefficients corresponding to different frequency sub-signals;

[0017] Step 2.3: Calculate the variance of the transformation coefficients at each decomposition scale;

[0018] Step 2.4: Construct a sequence preprocessing model based on the variance of the transformation coefficients;

[0019] Step 2.5: Use the sequence preprocessing model to preprocess the time series of the rail vehicle to obtain the preprocessed time series of the rail vehicle.

[0020] Preferably, step 2.4: constructing a sequence preprocessing model based on the variance of the transform coefficients, includes:

[0021] Step 2.4.1: Estimate the standard deviation of the anomalous noise based on the variance of the transformation coefficients;

[0022] Step 2.4.2: Construct a coefficient removal threshold based on the standard deviation of the noise; wherein, the coefficient removal threshold is:

[0023]

[0024] Where t represents the coefficient removal threshold, σ0 represents the variance of the transformation coefficients, and σ p The standard deviation of anomalous noise, median(d) p ) represents the median of the transformation coefficients at the p-th decomposition scale;

[0025] Step 2.4.3: Use the coefficients to remove the threshold and construct a sequence preprocessing model.

[0026] Preferably, in step 2.4.3, the sequence preprocessing model is as follows:

[0027]

[0028] Where α represents the transformation coefficient, w j,k Let m represent the transformation coefficients at the j-th decomposition scale, m represent the convergence coefficients, sgn represent the sign function, and wj represent the transformation coefficients at the j-th decomposition scale. ,k This represents the transformation coefficients used to remove outliers.

[0029] Preferably, step 2.5: using the sequence preprocessing model to preprocess the time series of the rail vehicle to obtain the preprocessed time series of the rail vehicle;

[0030] Step 2.5 includes:

[0031] Step 2.5.1: Set the initial transformation coefficients and convergence coefficients for the sequence preprocessing model to obtain the set sequence preprocessing model;

[0032] Step 2.5.2: Use the pre-processed sequence model to process the transform coefficients to obtain transform coefficients with outliers removed;

[0033] Step 2.5.3: Reconstruct the transform coefficients after removing outliers to obtain the reconstructed signal;

[0034] Step 2.5.4: Calculate the signal-to-noise ratio between the reconstructed signal and the original rail vehicle time series signal;

[0035] Step 2.5.5: When the signal-to-noise ratio is not within the preset range, reset the transformation coefficient and convergence coefficient until the signal-to-noise ratio is maintained within the preset range.

[0036] Preferably, step 3: calculating the Pearson coefficient of the time series and RAMS index for each rail vehicle; includes:

[0037] Step 3.1: Calculate the mean of the time series of rail vehicles;

[0038] Step 3.2: Calculate the mean of the RAMS index corresponding to the time series of rail vehicles;

[0039] Step 3.3: Construct a formula for calculating the Pearson coefficient based on the mean of the rail vehicle time series and the mean of the RAMS index; wherein, the formula for calculating the Pearson coefficient is:

[0040]

[0041] Where r represents the Pearson coefficient, X i Let X represent the value of the i-th data point in the time series of rail vehicles, and let Y represent the mean of the time series of rail vehicles. i Y represents the value of the i-th RAMS index.i This represents the mean of the RAMS index;

[0042] Step 3.4: Calculate the Pearson coefficient of each rail vehicle's time series and RAMS index using the Pearson coefficient calculation formula.

[0043] The present invention also provides a device for analyzing and predicting RAMS data of rail vehicles, comprising:

[0044] The time series data acquisition module is used to measure the time series data of rail vehicles under different RAMS indicators; the time series data of the rail vehicles includes: real-time vehicle speed time series, power system traction time series, fuel tank temperature time series, acceleration time series, car temperature time series and car smoke concentration time series;

[0045] The preprocessing module is used to preprocess the time series of the rail vehicles to obtain the preprocessed time series of the rail vehicles.

[0046] The Pearson coefficient calculation module is used to calculate the Pearson coefficient of each rail vehicle's time series and RAMS index.

[0047] The training sample construction module is used to remove the time series of corresponding rail vehicles that are below a preset threshold to obtain training samples.

[0048] The training module is used to input the training samples into the deep learning model for training to obtain the RAMS index estimation model for rail vehicles;

[0049] The RAMS index evaluation module is used to input the time series data of the target vehicle into the rail vehicle RAMS index estimation model to obtain the total RAMS index of the target vehicle.

[0050] The present invention also provides an electronic device, including a bus, a transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor. The transceiver, the memory, and the processor are connected via the bus. When the computer program is executed by the processor, it implements the steps in the above-described method for analyzing and predicting RAMS data of a rail vehicle.

[0051] The present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the steps in the above-described method for analyzing and predicting RAMS data of a rail vehicle.

[0052] This invention relates to a method and apparatus for analyzing and predicting RAMS data of rail vehicles, an electronic device, and a computer-readable storage medium. Compared with the prior art, this invention utilizes a deep learning model to model a large amount of time series data, which can better learn and capture the complex nonlinear relationship between the operating status of rail vehicles and RAMS indicators. This can effectively improve the accuracy and automation of predicting the RAMS indicators of rail vehicles, and provide strong support for vehicle operating status monitoring and maintenance decisions. Attached Figure Description

[0053] Figure 1 A flowchart of a method for analyzing and predicting RAMS data of a rail vehicle provided in an embodiment of the present invention;

[0054] Figure 2 A schematic diagram of a rail vehicle RAMS data analysis and prediction device provided in an embodiment of the present invention.

[0055] Please see Figure 1 A method for analyzing and predicting RAMS data of rail vehicles, comprising:

[0056] Step 1: Measure the time series data of the rail vehicle under different RAMS indicators; the time series data of the rail vehicle includes: real-time vehicle speed time series, power system traction time series, fuel tank temperature time series, acceleration time series, car temperature time series and car smoke concentration time series;

[0057] Step 2: Preprocess the time series of the rail vehicles to obtain the preprocessed time series of the rail vehicles;

[0058] Furthermore, step 2 includes:

[0059] Step 2.1: Use basis functions to decompose the time series signal of the rail vehicle into sub-signals of different frequencies;

[0060] In the embodiments of the present invention, Haar wavelet, Daubechies (dbN) wavelet, Mexican Hat (mexh) wavelet, Morlet wavelet, Meyer wavelet, etc. can be used.

[0061] Step 2.2: Obtain the transform coefficients corresponding to different frequency sub-signals;

[0062] Step 2.3: Calculate the variance of the transformation coefficients at each decomposition scale;

[0063] Step 2.4: Construct a sequence preprocessing model based on the variance of the transformation coefficients;

[0064] Step 2.4 includes:

[0065] Step 2.4.1: Estimate the standard deviation of the anomalous noise based on the variance of the transformation coefficients;

[0066] Step 2.4.2: Construct a coefficient removal threshold based on the standard deviation of the noise; wherein, the coefficient removal threshold is:

[0067]

[0068] Where t represents the coefficient removal threshold, σ0 represents the variance of the transformation coefficients, and σ p The standard deviation of anomalous noise, median(d) p ) represents the median of the transformation coefficients at the p-th decomposition scale;

[0069] Step 2.4.3: Construct a sequence preprocessing model using the coefficients to remove the threshold. In step 2.4.3, the sequence preprocessing model is:

[0070]

[0071] Where α represents the transformation coefficient, w j,k Let m represent the transformation coefficients at the j-th decomposition scale, m represent the convergence coefficients, sgn represent the sign function, and wj represent the transformation coefficients at the j-th decomposition scale. ,k This represents the transformation coefficients used to remove outliers.

[0072] This invention utilizes wavelet transform to convert the time-series signal of a railway vehicle into sub-signals of different frequencies. The aforementioned sequence preprocessing model is continuous at time t, with a function value of (1-α)t and first-order differentiability, its derivative at time t being 1-α. The convergence speed can be adjusted by the value of m; a larger m results in faster convergence. The degree of contraction of the wavelet coefficients within ±t can be adjusted by the value of α, allowing the aforementioned function to be transformed between conventional hard-thresholding and soft-thresholding functions. This better adapts to the characteristics of the time-series signal of the railway vehicle, thereby perfectly removing noise interference.

[0073] Step 2.5: Use the aforementioned sequence preprocessing model to preprocess the time series of the rail vehicles to obtain the preprocessed time series of the rail vehicles;

[0074] Step 2.5 includes:

[0075] Step 2.5.1: Set the initial transformation coefficients and convergence coefficients for the sequence preprocessing model to obtain the set sequence preprocessing model;

[0076] Step 2.5.2: Use the pre-processed sequence model to process the transform coefficients to obtain transform coefficients with outliers removed;

[0077] Step 2.5.3: Reconstruct the transform coefficients after removing outliers to obtain the reconstructed signal;

[0078] Step 2.5.4: Calculate the signal-to-noise ratio between the reconstructed signal and the original rail vehicle time series signal;

[0079] Step 2.5.5: When the signal-to-noise ratio is not within the preset range, reset the transformation coefficient and convergence coefficient until the signal-to-noise ratio is maintained within the preset range.

[0080] Step 3: Calculate the Pearson coefficient for the time series and RAMS index of each rail vehicle;

[0081] Furthermore, step 3 includes:

[0082] Step 3.1: Calculate the mean of the time series of rail vehicles;

[0083] Step 3.2: Calculate the mean of the RAMS index corresponding to the time series of rail vehicles;

[0084] Step 3.3: Construct a formula for calculating the Pearson coefficient based on the mean of the rail vehicle time series and the mean of the RAMS index; wherein, the formula for calculating the Pearson coefficient is:

[0085]

[0086] Where r represents the Pearson coefficient, X i This represents the value of the i-th data point in the time series of rail vehicles. Y represents the mean of the time series of rail vehicles. i This represents the value of the i-th RAMS indicator. This represents the mean of the RAMS index.

[0087] Step 3.4: Calculate the Pearson coefficient of each rail vehicle's time series and RAMS index using the Pearson coefficient calculation formula.

[0088] Step 4: Remove the time series data of the corresponding rail vehicles that are below the preset threshold to obtain training samples;

[0089] This invention removes vehicle time series with low correlation to RAMS index, allowing focus on vehicles with higher correlation to RAMS index. This facilitates the establishment of a more accurate prediction model and reduces the size of the training sample, thus improving the model training speed.

[0090] Step 5: Input the training samples into the deep learning model for training to obtain the rail vehicle RAMS index estimation model;

[0091] Step 6: Input the time series data of the target vehicle into the rail vehicle RAMS index estimation model to obtain the total RAMS index of the target vehicle.

[0092] This invention utilizes deep learning models to model large amounts of time-series data, enabling better learning and capture of the complex nonlinear relationship between the operating status of rail vehicles and RAMS indicators. This effectively improves the accuracy and automation of rail vehicle RAMS indicator prediction, providing strong support for vehicle operating status monitoring and maintenance decisions.

[0093] The present invention also provides a device for analyzing and predicting RAMS data of rail vehicles, comprising:

[0094] The time series data acquisition module is used to measure the time series data of rail vehicles under different RAMS indicators; the time series data of the rail vehicles includes: real-time vehicle speed time series, power system traction time series, fuel tank temperature time series, acceleration time series, car temperature time series and car smoke concentration time series;

[0095] The preprocessing module is used to preprocess the time series of the rail vehicles to obtain the preprocessed time series of the rail vehicles.

[0096] The Pearson coefficient calculation module is used to calculate the Pearson coefficient of each rail vehicle's time series and RAMS index.

[0097] The training sample construction module is used to remove the time series of corresponding rail vehicles that are below a preset threshold to obtain training samples.

[0098] The training module is used to input the training samples into the deep learning model for training to obtain the RAMS index estimation model for rail vehicles;

[0099] The RAMS index evaluation module is used to input the time series data of the target vehicle into the rail vehicle RAMS index estimation model to obtain the total RAMS index of the target vehicle.

[0100] The present invention also provides an electronic device, including a bus, a transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the transceiver, the memory, and the processor are connected via the bus, characterized in that the computer program, when executed by the processor, implements the steps in the above-described method for analyzing and predicting RAMS data of a rail vehicle.

[0101] Compared with the prior art, the beneficial effects of the electronic device provided by the present invention are the same as those of the rail vehicle RAMS data analysis and prediction method described in the above technical solution, and will not be repeated here.

[0102] The present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the steps in the above-described method for analyzing and predicting RAMS data of a rail vehicle.

[0103] Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by the present invention are the same as the beneficial effects of the rail vehicle RAMS data analysis and prediction method described in the above technical solution, and will not be repeated here.

[0104] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for analyzing and predicting RAMS data of rail vehicles, characterized in that, include: Step 1: Measure the time series data of rail vehicles under different RAMS indices; The time series data of the rail vehicle includes: real-time vehicle speed time series, power system traction time series, fuel tank temperature time series, acceleration time series, car temperature time series, and car smoke concentration time series; Step 2: Preprocess the time series of the rail vehicles to obtain the preprocessed time series of the rail vehicles; Step 3: Calculate the Pearson coefficient for the time series and RAMS index of each rail vehicle; Step 4: Remove the time series data of the corresponding rail vehicles that are below the preset threshold to obtain training samples; Step 5: Input the training samples into the deep learning model for training to obtain the rail vehicle RAMS index estimation model; Step 6: Input the time series data of the target vehicle into the rail vehicle RAMS index estimation model to obtain the total RAMS index of the target vehicle; Step 2 specifically includes: Step 2.1: Use basis functions to decompose the time series signal of the rail vehicle into sub-signals of different frequencies; Step 2.2: Obtain the transform coefficients corresponding to different frequency sub-signals; Step 2.3: Calculate the variance of the transformation coefficients at each decomposition scale; Step 2.4: Construct a sequence preprocessing model based on the variance of the transformation coefficients; Step 2.5: Use the aforementioned sequence preprocessing model to preprocess the time series of the rail vehicles to obtain the preprocessed time series of the rail vehicles; Step 2.5 specifically includes: Step 2.5.1: Set the initial transformation coefficients and convergence coefficients for the sequence preprocessing model to obtain the set sequence preprocessing model; Step 2.5.2: Use the pre-processed sequence model to process the transform coefficients to obtain transform coefficients with outliers removed; Step 2.5.3: Reconstruct the transform coefficients after removing outliers to obtain the reconstructed signal; Step 2.5.4: Calculate the signal-to-noise ratio between the reconstructed signal and the original rail vehicle time series signal; Step 2.5.5: When the signal-to-noise ratio is not within the preset range, reset the transformation coefficient and convergence coefficient until the signal-to-noise ratio is maintained within the preset range.

2. The method for analyzing and predicting RAMS data of rail vehicles according to claim 1, characterized in that, Step 2.4: Constructing a sequence preprocessing model based on the variance of the transform coefficients, including: Step 2.4.1: Estimate the standard deviation of the anomalous noise based on the variance of the transformation coefficients; Step 2.4.2: Construct a coefficient removal threshold based on the standard deviation of the noise; wherein, the coefficient removal threshold is: in, Indicates the threshold for removing coefficients. This represents the variance of the transformation coefficients. The standard deviation of abnormal noise, This represents the median of the transformation coefficients at the p-th decomposition scale; Step 2.4.3: Construct a sequence preprocessing model using the coefficients to remove the threshold; In step 2.4.3, the sequence preprocessing model is as follows: in, Represents the transformation coefficients. Denotes the transformation coefficients at the j-th decomposition scale. Represents the convergence coefficient. Represents a symbolic function. This represents the transformation coefficients used to remove outliers.

3. The method for analyzing and predicting RAMS data of rail vehicles according to claim 1, characterized in that, Step 3: Calculate the Pearson coefficient of the time series and RAMS index for each rail vehicle; include: Step 3.1: Calculate the mean of the time series of rail vehicles; Step 3.2: Calculate the mean of the RAMS index corresponding to the time series of rail vehicles; Step 3.3: Construct a formula for calculating the Pearson coefficient based on the mean of the rail vehicle time series and the mean of the RAMS index; wherein, the formula for calculating the Pearson coefficient is: in, Represents the Pearson coefficient. This represents the value of the i-th data point in the time series of rail vehicles. This represents the mean of the time series of rail vehicles. This represents the value of the i-th RAMS indicator. This represents the mean of the RAMS index; Step 3.4: Calculate the Pearson coefficient of each rail vehicle's time series and RAMS index using the Pearson coefficient calculation formula.

4. A rail vehicle RAMS data analysis and prediction device, employing the method described in any one of claims 1-3, characterized in that, include: The time series data acquisition module is used to measure the time series data of rail vehicles under different RAMS indices. The time series data of the rail vehicle includes: real-time vehicle speed time series, power system traction time series, fuel tank temperature time series, acceleration time series, car temperature time series, and car smoke concentration time series; The preprocessing module is used to preprocess the time series of the rail vehicles to obtain the preprocessed time series of the rail vehicles. The Pearson coefficient calculation module is used to calculate the Pearson coefficient of each rail vehicle's time series and RAMS index. The training sample construction module is used to remove the time series of corresponding rail vehicles that are below a preset threshold to obtain training samples. The training module is used to input the training samples into the deep learning model for training to obtain the RAMS index estimation model for rail vehicles; The RAMS index evaluation module is used to input the time series data of the target vehicle into the rail vehicle RAMS index estimation model to obtain the total RAMS index of the target vehicle.

5. An electronic device, characterized in that, The method includes a bus, a transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor. The transceiver, the memory, and the processor are connected via the bus. When the computer program is executed by the processor, it implements the steps of any one of claims 1-3.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method of any one of claims 1-3.

Images