Operation and maintenance time sequence data filtering method and device, equipment and medium

Abstract

The invention relates to an operation and maintenance time sequence data filtering method and device, equipment and a medium. The method comprises the following steps: constructing a recurrent neural network; obtaining original time sequence data from the operation and maintenance system; initializing one piece of initialization time sequence data; and taking the original time sequence data and the initialized time sequence data as the input of a recurrent neural network, outputting filtered time sequence data, taking the filtered time sequence data output in the current iteration as the initialized time sequence data of the next iteration, and outputting a final data filtering result until the training precision reaches a preset accuracy rate. Compared with the prior art, the filtering effect and accuracy of the time sequence data can be improved, and the characteristic value in the original time sequence data can be reserved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method, device, equipment and medium for filtering operation and maintenance time series data based on a recurrent neural network. Background Art

[0002] In modern multi-professional operation and maintenance systems, the system generates a large amount of time series data, which contains key information such as the system's operating status, alarm information, performance indicators, etc. However, due to various factors, these data often contain a large amount of noise and outliers, which can interfere with the monitoring and analysis of the operation and maintenance system, and even lead to false alarms and misjudgments.

[0003] In the multi-professional operation and maintenance system of rail transit, a huge amount of time series data is generated. These data come from a wide range of sources, including sensor monitoring data, equipment operation status data, operation record data, etc. However, the original time series data often contains a lot of noise, outliers, and data that is irrelevant to analysis and decision-making. These interference factors not only increase the burden of data storage and transmission, but also affect the accuracy and efficiency of subsequent data-based operation and maintenance analysis, fault diagnosis and prediction.

[0004] Traditional time series data filtering methods, such as threshold-based filtering, median filtering, mean filtering, low-pass filtering, high-pass filtering, etc., can remove noise to a certain extent, but these methods often cannot fully utilize the contextual information in the time series data, nor can they fully retain the characteristic values ​​in the time series data. For complex and changeable operation and maintenance system time series data, their filtering effect is not ideal.

[0005] Some technologies have also proposed the use of machine learning algorithms for data filtering. For example, CN114004148A discloses a method for filtering failure data of mining catalytic sensors based on neural networks, which belongs to the field of sensor data filtering and includes the following steps: S1: obtaining the response curve of mining catalytic sensors; S2: taking the response curve as sample material for interval sampling, obtaining multiple points from it as a set of sample data, and sampling multiple sets of sample data; S3: inputting the sample data into the neural network, constructing and training a failure data recognition model; S4: simulating the failure data recognition model; S5: applying the trained failure data recognition model to the sensor system to filter failure data. The neural network is applied to the sensor system to determine the failure data and filter it, which greatly improves the detection accuracy of the sensor. However, the essence of this method is to use the neural network to identify the failure data and then remove the data to complete the filtering, rather than directly using the neural network to complete the data filtering; and this method does not consider the timing problem of the data. Summary of the invention

[0006] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a method, device, equipment and medium for filtering operation and maintenance time series data based on a recurrent neural network, apply the recurrent neural network to the time series data filtering of the rail transit operation and maintenance system, make full use of its time series data processing capabilities, improve the filtering effect and accuracy, and more accurately retain the characteristic values ​​of the original data.

[0007] The purpose of the present invention can be achieved by the following technical solutions:

[0008] According to a first aspect of the present invention, a method for filtering operation and maintenance time series data based on a recurrent neural network is provided, the method comprising the following steps:

[0009] Build a recurrent neural network;

[0010] Obtain original time series data from the operation and maintenance system;

[0011] Initialize an initialization timing data;

[0012] The original time series data and the initialized time series data are used as the input of the recurrent neural network, and the filtered time series data is output. The filtered time series data output in the current iteration is used as the initialized time series data for the next iteration until the training accuracy reaches the preset accuracy rate, and the final data filtering result is output.

[0013] As a preferred technical solution, the original time series data includes system operation status, alarm information and performance indicator time series data obtained from a multi-professional operation and maintenance system.

[0014] As a preferred technical solution, the original time series data and the initialized time series data are both one-dimensional time series data.

[0015] As a preferred technical solution, the recurrent neural network adopts a three-layer RNN network, including an input layer, a hidden layer and an output layer, wherein the number of nodes in the input layer is 2, the number of nodes in the first hidden layer is 100, the number of nodes in the second layer is 50, the number of nodes in the third layer is 25, the number of nodes in the fourth layer is 10, and the number of nodes in the output layer is 1.

[0016] As a preferred technical solution, the input layer stacks the input original time series data and the initialized time series data in the column direction to obtain input features.

[0017] As a preferred technical solution, the activation function of the hidden layer of the recurrent neural network adopts the ReLU function.

[0018] As a preferred technical solution, the optimization algorithm of the recurrent neural network adopts the Adam algorithm.

[0019] As a preferred technical solution, the loss function of the recurrent neural network adopts the DTW algorithm.

[0020] According to a second aspect of the present invention, a device for filtering operation and maintenance time series data based on a recurrent neural network is provided, for implementing the method described, the device comprising:

[0021] Network building module: building a recurrent neural network;

[0022] Input data acquisition module: obtain original time series data from the operation and maintenance system; initialize an initial time series data;

[0023] Data filtering module: The original time series data and the initialized time series data are used as the input of the recurrent neural network, and the filtered time series data are output. The filtered time series data output in the current iteration is used as the initialized time series data for the next iteration until the training accuracy reaches the preset accuracy rate, and the final data filtering result is output.

[0024] As a preferred technical solution, the device also includes a visualization module for visually displaying the data filtering results.

[0025] According to a third aspect of the present invention, there is provided an electronic device, comprising a memory and a processor, wherein a computer program is stored in the memory, and the method described above is implemented when the processor executes the program.

[0026] According to a fourth aspect of the present invention, there is provided a computer-readable storage medium having a computer program stored thereon, wherein the program implements the method described when executed by a processor.

[0027] Compared with the prior art, the present invention utilizes the sensitivity of recurrent neural network time series data to process time series data in multi-professional operation and maintenance systems, so that it can not only improve the filtering effect and accuracy of time series data, but also retain the characteristic values ​​in the original time series data, thereby improving the accuracy and efficiency of data-based operation and maintenance analysis, fault diagnosis and prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 is a flow chart of the method of the present invention;

[0029] Figure 2 Schematic diagram of the RNN network used in the present invention. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0031] Unless otherwise defined, the technical terms or scientific terms involved in this application should be understood by people with ordinary skills in the technical field to which this application belongs. The words "one", "a", "a", "the" and the like involved in this application do not indicate a quantitative limitation, and may represent the singular or plural. The terms "include", "comprise", "have" and any of their variations involved in this application are intended to cover non-exclusive inclusions; for example, a process, method, system, product or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units that are not listed, or may also include other steps or units inherent to these processes, methods, products or devices. The words "connect", "connected", "coupled" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "multiple" involved in this application refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there may be three relationships, for example, "A and / or B" can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the objects before and after are in an "or" relationship. The terms "first", "second", "third", etc. involved in this application are only used to distinguish similar objects and do not represent a specific ordering of the objects.

[0032] This embodiment first provides an operation and maintenance time series data filtering method based on a recurrent neural network. Figure 1 As shown, the method comprises the following steps:

[0033] S1, build a recurrent neural network (RNN).

[0034] A recurrent neural network is a neural network model specifically used to process time series data. It can capture the time series information and contextual relationships in the data, and by using a specific loss function, it can fully retain the feature values ​​of the original data. Therefore, it has achieved remarkable results in speech recognition, machine translation, time series prediction and other fields. This embodiment will use the RNN network to implement operation and maintenance data filtering.

[0035] like Figure 2As shown, the recurrent neural network adopts a three-layer RNN network, including an input layer, a hidden layer and an output layer, wherein the number of nodes in the input layer is 2, the number of nodes in the first hidden layer is 100, the number of nodes in the second hidden layer is 50, the number of nodes in the third layer is 25, the number of nodes in the fourth layer is 10, and the number of nodes in the output layer is 1.

[0036] S2, obtains the original time series data from the operation and maintenance system.

[0037] The original time series data includes sensor monitoring data, system operation status, alarm information, operation record data, performance indicators and other time series data obtained from multi-professional operation and maintenance systems.

[0038] S3, initialize an initialization time series data.

[0039] Both the original time series data and the initialized time series data are one-dimensional time series data, and their dimensions are consistent.

[0040] S4, using the original time series data and the initialized time series data as the input of the recurrent neural network, outputting the filtered time series data, and using the filtered time series data output in the current iteration as the initialized time series data for the next iteration, until the training accuracy reaches the preset accuracy, and outputting the final data filtering result.

[0041] The input layer of the RNN network stacks the input original time series data and the initialized time series data in the column direction to obtain the input features, which are then sent to the hidden layer. The activation function of the hidden layer uses the ReLU function. In this embodiment, the optimization algorithm uses the Adam algorithm. In order to allow the filtered time series data to retain the characteristic values ​​of the original time series data as much as possible, this embodiment uses the DTW (Dynamic Time Warping) algorithm as the loss function of the recurrent neural network. DTW is an algorithm based on a dynamic programming strategy that nonlinearly adjusts the time domain alignment of two time series in order to correctly calculate the similarity between the two. It is widely used in applications such as time series classification and clustering.

[0042] The filtered time series data output by the RNN network is displayed to the user, and can be used as basic data for some operation and maintenance analysis, fault diagnosis and prediction, which can improve the accuracy of subsequent analysis.

[0043] This embodiment also provides an operation and maintenance time series data filtering device based on a recurrent neural network, which is used to implement the method described, and the device includes:

[0044] Network building module: building a recurrent neural network;

[0045] Input data acquisition module: obtain original time series data from the operation and maintenance system; initialize an initial time series data;

[0046] Data filtering module: takes the original time series data and the initialization time series data as the input of the recurrent neural network, outputs the filtered time series data, and uses the filtered time series data output in the current iteration as the initialization time series data for the next iteration until the training accuracy reaches the preset accuracy rate, and outputs the final data filtering result;

[0047] The visualization module is used to visualize the data filtering results.

[0048] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the described module can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0049] The electronic device of the present invention includes a central processing unit (CPU), which can perform various appropriate actions and processes according to computer program instructions stored in a read-only memory (ROM) or loaded from a storage unit into a random access memory (RAM). In the RAM, various programs and data required for device operation can also be stored. The CPU, ROM and RAM are connected to each other via a bus. An input / output (I / O) interface is also connected to the bus.

[0050] Multiple components in the device are connected to the I / O interface, including: input units, such as keyboards, mice, etc.; output units, such as various types of displays, speakers, etc.; storage units, such as disks, optical disks, etc.; and communication units, such as network cards, modems, wireless communication transceivers, etc. The communication unit allows the device to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunication networks.

[0051] The processing unit performs the various methods and processes described above, such as methods S1 to S4. For example, in some embodiments, methods S1 to S4 may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit. In some embodiments, part or all of the computer program may be loaded and / or installed on the device via a ROM and / or a communication unit. When the computer program is loaded into the RAM and executed by the CPU, one or more steps of methods S1 to S4 described above may be performed. Alternatively, in other embodiments, the CPU may be configured to execute methods S1 to S4 in any other appropriate manner (e.g., by means of firmware).

[0052] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0053] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0054] In the context of the present invention, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0055] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for filtering operation and maintenance time series data based on recurrent neural network, characterized in that: The method comprises the following steps: Build a recurrent neural network; Obtain original time series data from the operation and maintenance system; Initialize an initialization timing data; The original time series data and the initialized time series data are used as the input of the recurrent neural network, and the filtered time series data is output. The filtered time series data output in the current iteration is used as the initialized time series data for the next iteration until the training accuracy reaches the preset accuracy rate, and the final data filtering result is output.

2. According to the method for filtering operation and maintenance time series data based on recurrent neural network in claim 1, it is characterized in that: The original time series data includes system operation status, alarm information and performance indicator time series data obtained from multiple professional operation and maintenance systems.

3. According to the method for filtering operation and maintenance time series data based on recurrent neural network in claim 1, it is characterized in that: The original time series data and the initialized time series data are both one-dimensional time series data.

4. According to the method for filtering operation and maintenance time series data based on recurrent neural network in claim 1, it is characterized in that: The recurrent neural network adopts a three-layer RNN network, including an input layer, a hidden layer and an output layer, wherein the number of nodes in the input layer is 2, the number of nodes in the first hidden layer is 100, the number of nodes in the second hidden layer is 50, the number of nodes in the third hidden layer is 25, the number of nodes in the fourth hidden layer is 10, and the number of nodes in the output layer is 1.

5. According to the method for filtering operation and maintenance time series data based on recurrent neural network in claim 4, it is characterized in that: The input layer stacks the input original time series data and the initialized time series data in the column direction to obtain input features.

6. The method for filtering operation and maintenance time series data based on a recurrent neural network according to claim 1, characterized in that: The activation function of the hidden layer of the recurrent neural network adopts the ReLU function.

7. The method for filtering operation and maintenance time series data based on a recurrent neural network according to claim 1, characterized in that: The optimization algorithm of the recurrent neural network adopts the Adam algorithm.

8. The method for filtering operation and maintenance time series data based on a recurrent neural network according to claim 1, characterized in that: The loss function of the recurrent neural network adopts the DTW algorithm.

9. An operation and maintenance time series data filtering device based on recurrent neural network, characterized in that: For implementing the method according to any one of claims 1 to 8, the device comprises: Network building module: building a recurrent neural network; Input data acquisition module: obtain original time series data from the operation and maintenance system; initialize an initial time series data; Data filtering module: The original time series data and the initialized time series data are used as the input of the recurrent neural network, and the filtered time series data are output. The filtered time series data output in the current iteration is used as the initialized time series data for the next iteration until the training accuracy reaches the preset accuracy rate, and the final data filtering result is output.

10. The operation and maintenance time series data filtering device based on recurrent neural network according to claim 9, characterized in that: The device also includes a visualization module for visually displaying the data filtering results.

11. An electronic device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the program, the method according to any one of claims 1 to 8 is implemented.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.

Images