MVB bus data real-time processing method based on cloud computing

Abstract

The invention relates to the field of data processing, in particular to an MVB bus data real-time processing method based on cloud computing, and the method comprises the steps: collecting the dimension data of a traction motor in real time through MVB bus interface equipment; dividing the dimension data of the traction motor into an acceleration stage, a constant speed stage and a deceleration stage, and calculating the importance of each dimension data in each stage; allocating a weight to each piece of dimension data based on the importance of each piece of dimension data in the current stage to obtain a corresponding dimension weight; performing weighted adjustment on the path length of the dimension data according to the dimension weight to obtain a weighted path length, and calculating an abnormal score of the dimension data based on the weighted path length; and performing anomaly detection according to the anomaly score. According to the method, the problem that the difference between the dimensions is easily ignored in simple average path length calculation is effectively solved, so that the abnormal condition in the data can be identified more accurately, and the accuracy of anomaly detection is improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and more specifically, to a real-time processing method of MVB bus data based on cloud computing. Background Art

[0002] As a communication protocol used in the transportation field, the Multifunction Vehicle Bus (MVB) can connect various devices in the train, such as the traction control unit, braking system, door control, air conditioning system, etc., to achieve real-time communication and data exchange between devices. As an important part of the train communication network, the MVB bus collects data containing various key information during the train operation process, which is crucial for the safe operation and fault diagnosis of the train.

[0003] The Isolation Forest Algorithm is an unsupervised anomaly detection algorithm based on a tree model. In the scenario of MVB bus data collection, the Isolation Forest Algorithm can analyze the collected data and quickly identify abnormal data that deviates greatly from the normal data pattern. These abnormal data are of great significance in reminding relevant personnel to conduct inspections and repairs in a timely manner, which helps to improve the reliability and safety of equipment.

[0004] However, when performing fault warning, data of different dimensions have different sensitivities to faults. For example, data from some sensors may be more likely to reflect specific fault types, while data from other sensors may be less sensitive. Therefore, if the anomaly score is simply calculated based on the average path length of each data point in all isolated trees, the differences between these dimensions are easily ignored, resulting in inaccurate calculation results, which in turn affects the accuracy of the warning results. Summary of the invention

[0005] In order to solve the above-mentioned technical problem of not taking into account the differences in sensitivity of data of different dimensions to faults, resulting in inaccurate processing results of the collected data, the present invention provides the following technical solution.

[0006] The real-time processing method of MVB bus data based on cloud computing includes: Use the MVB bus interface device to collect the dimensional data of the traction motor in real time; use the improved isolation forest algorithm to calculate the anomaly score of each dimensional data; perform anomaly detection based on the anomaly score; The improved process is: The dimensional data of the traction motor is divided into acceleration, constant speed and deceleration stages, and the importance of each dimensional data in each stage is calculated; a weight is assigned to each dimensional data based on its importance in the current stage to obtain the corresponding dimensional weight; the path length of the dimensional data is weighted and adjusted according to the dimensional weight to obtain the weighted path length, and the anomaly score of the dimensional data is calculated based on the weighted path length.

[0007] The present invention first divides the dimensional data of the traction motor into stages, avoiding unified processing of all data, which helps to identify the behavioral characteristics of the motor in different operating stages, and further more accurately understand the importance of each dimensional data in different stages. Then, the importance of each dimensional data in each stage is calculated through the behavioral characteristics of the motor in each stage. It can be clarified which dimensions have a significant impact on the operating state of the motor in a specific stage, so that important dimensions will be given higher weights when calculating the path length, thereby more accurately reflecting the actual abnormality of the data points.

[0008] Through the above operations, the algorithm can more accurately reflect the behavioral characteristics of the traction motor in different operating stages, more sensitively capture abnormal data points, and reduce false positives and negatives caused by ignoring differences between dimensions.

[0009] Preferably, the dimensional data includes at least one of electrical parameters, rotation speed and temperature.

[0010] Preferably, the process of dividing the dimensional data of the traction motor into acceleration, constant speed and deceleration stages includes: The speed data of the train is synchronously collected based on the collected dimensional data of the traction motor, and is sorted in chronological order to construct a speed sequence; Starting from the second item of the speed sequence, calculate the difference value with the previous item in sequence. If the difference value is positive, the train is in the acceleration stage at the moment corresponding to the speed data. If the difference value is zero, the train is in the constant speed stage at the moment corresponding to the speed data. If the difference value is negative, the train is in the deceleration stage at the moment corresponding to the speed data.

[0011] Through the differential value of speed data, it is possible to accurately determine whether the train is in the acceleration, constant speed or deceleration stage at a certain moment, which helps to understand the operating status of the train more accurately and provides a basis for subsequent performance analysis of the traction motor under different working conditions.

[0012] Preferably, the importance calculation process includes: Calculate the degree of change of all sub-data under any dimension data, and take the mean of the degree of change of all sub-data as the importance of the dimension data.

[0013] By calculating the degree of change of the sub-data and finding the average to indicate the importance of the dimensional data, the contribution of each dimension to the overall data change can be intuitively displayed. In the dimensional data of the traction motor, data of different dimensions such as current, voltage, and temperature reflect the motor operating status to different degrees. For those dimensions that change more drastically during operation and have a critical impact on the motor performance, such as the change of current during the acceleration phase, their importance will be highlighted, and then quickly identify which dimensions are the key factors affecting the operation of the traction motor and which are relatively minor factors.

[0014] Preferably, the process of obtaining the dimension weight includes: Select any dimension and take the ratio of the importance of this dimension to the sum of the importance of all dimensions as the dimension weight of the data of this dimension.

[0015] In the traditional isolation forest algorithm, all dimensions are given the same weight, which may cause dimensions with low importance to have unnecessary influence on the anomaly detection results, thereby increasing the possibility of misjudgment. By assigning weights to each dimension, the algorithm is able to reduce the impact of unimportant dimensions, thereby reducing misjudgments and improving the reliability of detection results.

[0016] Preferably, the weighted path length satisfies the relationship: ; In the formula, For the The weighted path length of the dimension data, For the The dimension data is in The path length in an isolated tree, For the The dimension weight of the dimension data corresponding to an isolated tree, is the total number of isolated trees.

[0017] Preferably, performing anomaly detection according to anomaly scores includes: When the abnormal score of the dimensional data of the traction motor exceeds the preset abnormal threshold, the early warning mechanism is triggered.

[0018] Preferably, the process of obtaining the degree of change includes: In any dimension, select a set number of sub-data as target data; Calculate the numerical difference and time difference between each sub-data and the target data; The product of the normalized numerical difference and the time difference is taken as the degree of change of the corresponding sub-data.

[0019] For each sub-data, the numerical difference between it and the target data is calculated, which reflects the degree of change in the numerical value of the sub-data; at the same time, the time difference between the sub-data and the target data is calculated, which takes into account the time factor of the data change, which helps to understand the timeliness of the data change.

[0020] Preferably, the importance calculation process includes: Calculate the degree of change of all sub-data under any dimension data, sort them from small to large, and select the degree of change corresponding to its quartile as the importance of the dimension data.

[0021] Preferably, the process of dividing the dimensional data of the traction motor into acceleration, constant speed and deceleration stages includes: The speed data of the train is synchronously collected based on the collected dimensional data of the traction motor, and is sorted in chronological order to construct a speed sequence; For each moment in the speed sequence, the speed change rate is calculated. According to the positive, negative and zero values ​​of the speed change rate, the train's operating state is preliminarily divided into acceleration stage, constant speed stage and deceleration stage. If the continuous moments of a certain stage are less than the preset threshold, the stage is merged into the previous stage.

[0022] Taking into account the possible short-term abnormal situations or inaccurate data collection in actual operation, by setting the continuous moment threshold, those stages that may be abnormal or inaccurate can be merged into the previous more reasonable stage, thereby reducing the impact of data collection errors or temporary abnormalities on the overall division results, reflecting a certain degree of fault tolerance.

[0023] The beneficial effects of the present invention are: The present invention adopts an improved isolation forest algorithm, which divides the operation phases of the traction motor (acceleration, constant speed, deceleration), calculates the importance and weight of each dimension data in each phase, and then adjusts the path length and calculates the anomaly score. This method effectively solves the problem that the simple average path length calculation easily ignores the difference between dimensions, and can more accurately identify anomalies in the data, thereby improving the accuracy of anomaly detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a method flow chart of steps S1 to S3 in the real-time processing method of MVB bus data based on cloud computing in 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: to collect the operating data of the traction motor of the train in real time through the MVB bus interface, to analyze and process these data using an improved anomaly detection algorithm to identify anomalies or potential faults, and to immediately issue a warning signal when an anomaly is detected.

[0027] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0028] Reference Figure 1 The real-time processing method of MVB bus data based on cloud computing includes steps S1 to S3, which are as follows: S1: Use the MVB bus interface device to collect the dimensional data of the traction motor in real time.

[0029] The traction motor is a key component of the train's power output. If the traction motor fails, such as being overloaded or overheated, it may cause the train's power to be interrupted or even cause serious accidents such as fire. Therefore, real-time monitoring of the traction motor is crucial.

[0030] In one embodiment, an interface device supporting MVB bus communication, such as an MVB board or adapter, is first selected, the MVB bus interface device is connected to the control system of the traction motor, and the communication link is ensured to be normal.

[0031] The electrical parameter sensor, speed sensor and temperature sensor are installed at the corresponding positions of the traction motor respectively, and the MVB bus interface device is started to collect and transmit the multi-dimensional data of the traction motor in real time with a fixed sampling period (for example, 10 milliseconds), including electrical parameters (voltage, current, power, etc.), speed and temperature.

[0032] Among them, the collected multi-dimensional data is processed, such as filtering, amplification, etc., to improve the accuracy and stability of the data.

[0033] S2: Use the improved isolation forest algorithm to calculate the anomaly score of each dimension of data.

[0034] In one embodiment, the MVB bus interface is used to collect the dimensional data of the traction motor and the speed data of the train at the same time according to the collection frequency set in S1.

[0035] It should be noted that in order to ensure data synchronization, the train speed data and dimension data are time-stamped during the collection process to ensure that each data point accurately corresponds to its collection time.

[0036] The acquired train speed data are sorted according to their marked timestamps to form a speed sequence. Starting from the second item in the speed sequence, the difference values ​​with the previous item are calculated in sequence.

[0037] Among them, if the differential term is positive, it means that the train speed at the current moment is increasing, that is, the train is in the acceleration stage. If the differential term is 0, it means that the train speed at the current moment remains unchanged, that is, the train is in the constant speed stage. If the differential term is negative, it means that the train speed at the current moment is decreasing, that is, the train is in the deceleration stage, and the train's operating state is divided into acceleration, constant speed and deceleration stages.

[0038] Furthermore, according to the dimensional data of the traction motor collected at the same time as the train speed sequence, these dimensional data are also divided into acceleration, constant speed and deceleration stages.

[0039] It should be noted that the data of the traction motor is classified according to different operating stages to avoid mixing the data of all stages together for calculation, which can reduce the calculation error caused by mixing different operating states.

[0040] In one embodiment, the variation degree of sub-data of each data in a specific dimension is calculated.

[0041] Exemplarily, taking temperature dimension data as an example, for each sub-data under the temperature dimension data, the absolute value of the difference between each sub-data and all other sub-data is calculated respectively, and these difference absolute values ​​are sorted in ascending order to obtain a difference sequence, and the sub-data corresponding to the first K smallest differences in the difference sequence are selected as the target data.

[0042] Furthermore, the sum of the numerical differences between each sub-data and all target data is calculated, and normalized to obtain the numerical difference of each sub-data; the square root of the mean of the square of the difference between the acquisition time of each sub-data and all target data is calculated, and normalized to obtain the time difference of each sub-data.

[0043] The product of the above numerical difference and time difference is taken as the change degree of the corresponding sub-data, and the change degree satisfies the relationship:

[0044] In the formula, The temperature dimension The degree of change of sub-data, The temperature dimension The value of each sub-data, For the The target data of the sub-data The value of each sub-data, is the total number of target data, The temperature dimension The collection time corresponding to each sub-data is For the The target data of the sub-data The collection time corresponding to each sub-data is Indicates normalization.

[0045] Among them, if The greater the difference between the values ​​of the sub-data and their target data, and the greater the difference between the sampling time of these target data and the The sampling time of each sub-data is also very different, so The higher the degree of change of each sub-data, the The more obvious the change characteristics of individual data are.

[0046] Furthermore, according to the above The calculation method of the change degree of each sub-data is similar to that of the temperature dimension to obtain the change degree of all sub-data.

[0047] Furthermore, the mean of the change degree of all sub-data under the temperature dimension is taken as the importance of the temperature dimension.

[0048] Furthermore, according to the calculation method of the importance of the temperature dimension mentioned above, the importance of all dimensions can be obtained in the same way.

[0049] At different stages of train operation, data of different dimensions have different sensitivities to faults, and the degree of change of data points can be used as an indicator to measure the importance of the dimension. The greater the degree of change, the more sensitive the data of that dimension is to faults, that is, the more important the dimension is in fault detection and diagnosis.

[0050] In one embodiment, a weight is assigned to each dimension data based on the importance of each dimension data at the current stage to obtain a corresponding dimension weight. For example, still taking the temperature dimension as an example, the dimension weight of the temperature dimension satisfies the relationship:

[0051] In the formula, is the dimension weight of the temperature dimension, The importance of the temperature dimension, is the sum of the importance of all dimensions.

[0052] Furthermore, for the dimensional data of each stage, the genetic algorithm constructs multiple isolated trees, each of which is composed of continuously segmented data of the same dimension. For each data point, the path length in all isolated trees is weighted and summed based on the corresponding dimensional weight, that is, the relationship is satisfied:

[0053] In the formula, For the The weighted path length of the dimension data, For the The dimension data is in The path length in an isolated tree, For the The dimension weight of the dimension data corresponding to an isolated tree, is the total number of isolated trees.

[0054] Then the weighted path length of all dimensional data is obtained. According to the path length of each dimensional data, the genetic algorithm will assign an abnormality score to each dimensional data. The higher the score, the more likely the dimensional data is an outlier.

[0055] It should be noted that obtaining anomaly scores is part of the isolation forest algorithm and will not be described in detail here.

[0056] In another embodiment, when dividing the train running status, the error tolerance space is considered.

[0057] Specifically, the acquired train speed data is sorted according to its marked timestamp to form a speed sequence, and the train speed is calculated. The speed change rate at the moment (usually the The ratio of the speed difference between the moment and the previous moment to the time difference), if the The speed change rate at the moment is positive, indicating that the train is accelerating. The speed change rate at the moment is zero, indicating that the train is in a constant speed stage. The speed change rate at the moment is negative, indicating that the train is decelerating.

[0058] According to The speed change rate at each moment can be used to divide the train's running state into acceleration phase, constant speed phase and deceleration phase. However, in order to avoid fine-tuning the train during operation and affecting the judgment of the train's running phase, by analyzing the duration of each phase, that is, the number of consecutive moments is less than the preset threshold (obtained through experiments and debugging), it is considered that this phase is caused by the fine-tuning of the train, and then this phase is merged into the previous phase.

[0059] In another embodiment, according to the above The calculation method of the degree of change of each sub-data is similar to that of obtaining the degree of change of all sub-data under the temperature dimension, and then sorting all the degrees of change from small to large, and selecting the degree of change corresponding to its quartile as the importance of the temperature dimension.

[0060] S3: Perform anomaly detection based on anomaly scores.

[0061] In one embodiment, the preset abnormality threshold is 0.85. When the abnormality score of the dimensional data of the traction motor exceeds the preset abnormality threshold, the system triggers an early warning mechanism to notify relevant staff to check possible problems and take necessary measures to prevent potential failures or accidents.

[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 real-time processing method for MVB bus data based on cloud computing, characterized in that: include: Use MVB bus interface equipment to collect traction motor dimensional data in real time; The improved isolation forest algorithm is used to calculate the anomaly score of each dimension of data; Anomaly detection based on anomaly scores; The improved process is: The dimensional data of the traction motor is divided into acceleration, constant speed and deceleration stages, and the importance of each dimensional data in each stage is calculated; a weight is assigned to each dimensional data based on its importance in the current stage to obtain the corresponding dimensional weight; the path length of the dimensional data is weighted and adjusted according to the dimensional weight to obtain the weighted path length, and the anomaly score of the dimensional data is calculated based on the weighted path length.

2. The method for real-time processing of MVB bus data based on cloud computing according to claim 1 is characterized in that: The dimensional data includes at least one of electrical parameters, rotation speed, and temperature.

3. The method for real-time processing of MVB bus data based on cloud computing according to claim 2 is characterized in that: The process of dividing the dimensional data of the traction motor into acceleration, constant speed and deceleration stages includes: The speed data of the train is synchronously collected based on the collected dimensional data of the traction motor, and is sorted in chronological order to construct a speed sequence; Starting from the second item of the speed sequence, calculate the difference value with the previous item in sequence. If the difference value is positive, the train is in the acceleration stage at the moment corresponding to the speed data. If the difference value is zero, the train is in the constant speed stage at the moment corresponding to the speed data. If the difference value is negative, the train is in the deceleration stage at the moment corresponding to the speed data.

4. The method for real-time processing of MVB bus data based on cloud computing according to claim 3 is characterized in that: The calculation process of the importance includes: Calculate the degree of change of all sub-data under any dimension data, and take the mean of the degree of change of all sub-data as the importance of the dimension data.

5. The method for real-time processing of MVB bus data based on cloud computing according to claim 4 is characterized in that: The process of obtaining the dimension weight includes: Select any dimension and take the ratio of the importance of this dimension to the sum of the importance of all dimensions as the dimension weight of the data of this dimension.

6. The method for real-time processing of MVB bus data based on cloud computing according to claim 5 is characterized in that: The weighted path length satisfies the relationship: ; In the formula, For the The weighted path length of the dimension data, For the The dimension data is in The path length in an isolated tree, For the The dimension weight of the dimension data corresponding to an isolated tree, is the total number of isolated trees.

7. The method for real-time processing of MVB bus data based on cloud computing according to claim 6 is characterized in that: The anomaly detection according to the anomaly score comprises: When the abnormal score of the dimensional data of the traction motor exceeds the preset abnormal threshold, the early warning mechanism is triggered.

8. The method for real-time processing of MVB bus data based on cloud computing according to claim 4 is characterized in that: The process of obtaining the degree of change includes: In any dimension, select a set number of sub-data as target data; Calculate the numerical difference and time difference between each sub-data and the target data; The product of the normalized numerical difference and the time difference is taken as the degree of change of the corresponding sub-data.

9. The method for real-time processing of MVB bus data based on cloud computing according to claim 1, characterized in that: The calculation process of the importance includes: Calculate the degree of change of all sub-data under any dimension data, sort them from small to large, and select the degree of change corresponding to its quartile as the importance of the dimension data.

10. The method for real-time processing of MVB bus data based on cloud computing according to claim 1, characterized in that: The process of dividing the dimensional data of the traction motor into acceleration, constant speed and deceleration stages includes: The speed data of the train is synchronously collected based on the collected dimensional data of the traction motor, and is sorted in chronological order to construct a speed sequence; For each moment in the speed sequence, the speed change rate is calculated. According to the positive, negative and zero values ​​of the speed change rate, the train's operating state is preliminarily divided into acceleration stage, constant speed stage and deceleration stage. If the continuous moments of a certain stage are less than the preset threshold, the stage is merged into the previous stage.

Images