CN113848417A - Rail transit power supply equipment fault prediction method and device - Google Patents
Rail transit power supply equipment fault prediction method and device Download PDFInfo
- Publication number
- CN113848417A CN113848417A CN202110911227.1A CN202110911227A CN113848417A CN 113848417 A CN113848417 A CN 113848417A CN 202110911227 A CN202110911227 A CN 202110911227A CN 113848417 A CN113848417 A CN 113848417A
- Authority
- CN
- China
- Prior art keywords
- power supply
- data
- supply equipment
- equipment
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 57
- 230000006870 function Effects 0.000 claims description 22
- 238000013145 classification model Methods 0.000 claims description 20
- 238000012549 training Methods 0.000 claims description 18
- 238000000354 decomposition reaction Methods 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 9
- 210000002569 neuron Anatomy 0.000 claims description 8
- 238000001514 detection method Methods 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 5
- 238000003062 neural network model Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 239000006185 dispersion Substances 0.000 claims description 2
- 238000012706 support-vector machine Methods 0.000 abstract description 40
- 230000006403 short-term memory Effects 0.000 abstract description 13
- 230000007787 long-term memory Effects 0.000 abstract description 9
- 230000007774 longterm Effects 0.000 abstract description 8
- 238000013528 artificial neural network Methods 0.000 description 15
- 238000012423 maintenance Methods 0.000 description 10
- 238000003745 diagnosis Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 9
- 230000000306 recurrent effect Effects 0.000 description 7
- 230000006872 improvement Effects 0.000 description 4
- 238000007781 pre-processing Methods 0.000 description 4
- 210000004027 cell Anatomy 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000010438 heat treatment Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 230000004913 activation Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 2
- 230000006378 damage Effects 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000001934 delay Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000008676 import Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 230000015654 memory Effects 0.000 description 2
- 230000003252 repetitive effect Effects 0.000 description 2
- 230000002194 synthesizing effect Effects 0.000 description 2
- 230000009897 systematic effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 230000003449 preventive effect Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/08—Locating faults in cables, transmission lines, or networks
- G01R31/088—Aspects of digital computing
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Supply And Distribution Of Alternating Current (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
本发明公开了一种轨道交通供电设备故障预测方法及装置，率先采用LSTM+SVM模型对供电系统故障进行预测。首先采用LSTM算法和系统设备历史数据生成供电系统故障预测模型，再将系统设备的当前数据进行去噪和标准化处理，然后将处理后的数据输入预测模型得到目标设备的预测数据，最后将预测模型输出的数据用SVM模型进行分类，输出设备状态结果。本发明结合了长短期记忆模型能够进行长时间预测和支持向量机具有良好的非线性分类的优点，率先采用LSTM+SVM模型对供电系统故障进行预测，为轨道交通供电系统建立了一种精准预测故障的方法，可有效改善轨道交通供电系统的稳定性和安全性。
The invention discloses a fault prediction method and device for rail transit power supply equipment, and takes the lead in using the LSTM+SVM model to predict the power supply system fault. First, the LSTM algorithm and the historical data of the system equipment are used to generate the power supply system fault prediction model, then the current data of the system equipment is denoised and standardized, and then the processed data is input into the prediction model to obtain the prediction data of the target equipment, and finally the prediction model is used. The output data is classified by the SVM model, and the equipment status results are output. The invention combines the advantages of long-term and short-term memory model for long-term prediction and support vector machine with good nonlinear classification. It takes the lead in using LSTM+SVM model to predict power supply system failures, and establishes an accurate prediction for rail transit power supply system. The fault method can effectively improve the stability and safety of the rail transit power supply system.
Description
技术领域technical field
本发明涉及轨道交通电力系统故障预测领域，具体涉及一种基于LSTM神经网 络模型的对轨道交通供设备进行故障预测方法。The invention relates to the field of rail transit power system fault prediction, in particular to a fault prediction method for rail transit supply equipment based on an LSTM neural network model.
背景技术Background technique
随着我国高铁和城市轨道交通的飞速发展，列车的规模和列车运输频次不断的提高，供电系统的故障对社会经济和人身安全的危害越来越严重，轨道交通供电系 统正面临着严峻的挑战。精准、快速的对轨道交通供电系统故障进行预测，是避免 供电系统发生故障，进而避免带来一系列严重后果的有效方法。研究可靠性高、预 测速度快的轨道交通供电设备故障预测方法对保障整个轨道交通供电系统的安全 性和经济性有着十分重要的意义。With the rapid development of my country's high-speed rail and urban rail transit, the scale of trains and the frequency of train transportation continue to increase, and the failure of the power supply system has become more and more serious harm to social economy and personal safety. The rail transit power supply system is facing severe challenges. . Accurate and fast prediction of rail transit power supply system failures is an effective way to avoid power supply system failures and a series of serious consequences. It is of great significance to study the fault prediction method of rail transit power supply equipment with high reliability and fast prediction speed to ensure the safety and economy of the entire rail transit power supply system.
前人在针对轨道交通供电系统设备故障诊断方法做了很多研究，但是并没有将设备未来数据预测后再对设备状态进行精准分类。现有技术是结合人工智能领域中 的人工神经网络、贝叶斯网络、专家系统、数据挖掘等技术对供电系统设备进行故 障诊断。现有技术的方法是结合了人工智能技术通过归纳专家经验对已发生的故障 进行智能诊断，能够对故障数据进行高维度分类，具有较高的准确率。但是该方法 存在只能分析故障的类型以及产生的原因，并不能进行故障预测，属于先故障后诊 断的模式。The predecessors have done a lot of research on the fault diagnosis method of the rail transit power supply system equipment, but they have not predicted the future data of the equipment and then accurately classified the equipment status. The existing technology combines artificial neural network, Bayesian network, expert system, data mining and other technologies in the field of artificial intelligence to perform fault diagnosis on power supply system equipment. The method in the prior art combines artificial intelligence technology to perform intelligent diagnosis of faults that have occurred by summarizing expert experience, and can classify fault data in high dimensions with high accuracy. However, this method can only analyze the type and cause of the fault, but cannot predict the fault, which belongs to the mode of diagnosing the fault first.
发明内容SUMMARY OF THE INVENTION
针对现有技术中的缺陷，本发明所要解决的技术问题是：轨道交通供电系统无 法在供电设备发生故障之前进行精准预测来避免供电事故的发生。In view of the defects in the prior art, the technical problem to be solved by the present invention is that the rail transit power supply system cannot accurately predict before the failure of the power supply equipment to avoid the occurrence of power supply accidents.
为解决上述问题，本发明通过下述技术方案来实现：In order to solve the above-mentioned problems, the present invention realizes through the following technical solutions:
本发明提供轨道交通供电设备故障预测方法，包括以下步骤：The present invention provides a fault prediction method for rail transit power supply equipment, comprising the following steps:
S1，获取t时刻的目标供电设备的供电设备数据，其中所述供电设备数据包括 电压、电流、功率和设备运行温度；S1, obtain the power supply equipment data of the target power supply equipment at time t, wherein the power supply equipment data includes voltage, current, power and equipment operating temperature;
S2，将所述供电设备数据进行去噪处理得到无干扰元素的设备数据集，对所述 设备数据集进行标准化处理得到t时刻的输入参数；S2, carrying out denoising processing to the power supply equipment data to obtain the equipment data set without interference elements, and standardizing the equipment data set to obtain the input parameter at time t;
S3，对所述输入参数输入到预测时间步长为n的供电设备数据LSTM预测模型， 并输出t+n时刻的所述目标设备运行参数预测值；S3, inputting the input parameters to the power supply equipment data LSTM prediction model with a prediction time step of n, and outputting the predicted value of the target equipment operating parameters at time t+n;
S4将所述预测值作为SVM模型的输入值对其进行数据分类，分析t+n时刻设备 是否会发生故障，若否，则返回步骤S1进行循环检测；S4 uses described predicted value as the input value of SVM model to carry out data classification to it, analyzes whether equipment can fail at time t+n, if not, then returns to step S1 to carry out loop detection;
S5：如果是，则说明所述目标供电设备具有发生故障的可能性。S5: If yes, it indicates that the target power supply device has the possibility of failure.
进一步地，所述LSTM预测模型的构建步骤包括：Further, the construction steps of the LSTM prediction model include:
设置LSTM输入维度和预测时间步长；Set the LSTM input dimension and prediction time step;
设置模型优化器以及学习速率；Set the model optimizer and learning rate;
设置隐含层神经元节点个数；Set the number of hidden layer neurons;
将所述目标设备运行历史数据集导入到LSTM神经网络模型中进行训练；Importing the target device operation history data set into the LSTM neural network model for training;
在训练过程中对LSTM模型的超参数进行不断的优化调整，得到优化后的LSTM预测模型。During the training process, the hyperparameters of the LSTM model are continuously optimized and adjusted to obtain the optimized LSTM prediction model.
进一步地，所述SVM分类模型的构建步骤包括：Further, the steps of constructing the SVM classification model include:
选取所述目标设备历史数据中的正常运行时参数和故障时运行参数合并组成数据 集；Selecting the normal operation parameters in the historical data of the target equipment and the operation parameters during failure are combined to form a data set;
将所述数据集进行标准化处理，处理后的所述数据集作为SVM模型训练时的输入数 据；Standardized processing is carried out to the data set, and the processed data set is used as the input data during SVM model training;
对构建的SVM模型进行训练，训练过程包括：调整SVM的核函数，调整类别权重， 调整迭代次数参数，优化分类模型的性能，得到优化后的SVM分类模型。The constructed SVM model is trained, and the training process includes: adjusting the kernel function of the SVM, adjusting the category weight, adjusting the iteration number parameter, optimizing the performance of the classification model, and obtaining the optimized SVM classification model.
进一步地，所获取的所述目标设备数据的数值受到所述目标供电设备的类型和/或 所述目标设备在供电系统中所起的作用的影响。Further, the acquired value of the target device data is affected by the type of the target power supply device and/or the role played by the target device in the power supply system.
所述步骤S2中对数据进行去噪处理采用的方法为小波去噪法。The method used for denoising the data in the step S2 is the wavelet denoising method.
所述小波去噪法的步骤包括：The steps of the wavelet denoising method include:
构造函数空间，将信号分解到函数空间中进行计算，获取有用的数据；Construct the function space, decompose the signal into the function space for calculation, and obtain useful data;
重构返回原始信号，其中，数据分解和重构公式如下：The reconstruction returns the original signal, where the data decomposition and reconstruction formulas are as follows:
分解公式：A0[f(t)]＝f(t)Decomposition formula: A 0 [f(t)]=f(t)
重构公式：
式中：t为时间序列，f(t)为原始信号，j为分解的层数，H、G为时域中的小波分解滤波器，h、g为时域中小波重构滤波器，Aj为信号f(t)在第j层的低频部分的小波系数， Dj为信号f(t)在第j层的高频部分的小波系数。In the formula: t is the time series, f(t) is the original signal, j is the number of layers to be decomposed, H and G are the wavelet decomposition filters in the time domain, h and g are the wavelet reconstruction filters in the time domain, A j is the wavelet coefficient of the low frequency part of the signal f(t) in the jth layer, D j is the wavelet coefficient of the high frequency part of the signal f(t) in the jth layer.
进一步地，所述步骤S2中的进行标准化处理采用离差标准化的方式进行标准处理，采 用以下公式进行建立：Further, in the described step S2, the standardization process is carried out in the mode of standardization of dispersion, and the following formula is used to establish:
其中，
本发明的另一方面，提供一个轨道交通供电设备故障预测装置，所述轨道交通供电 设备故障预测装置包括：Another aspect of the present invention provides a rail transit power supply equipment failure prediction device, and the rail transit power supply equipment failure prediction device includes:
数据获取单元，用于获取t时刻的目标供电设备的供电设备数据，其中所述供电设备数据包括电压、电流、功率和设备运行温度；a data acquisition unit, configured to acquire power supply equipment data of the target power supply equipment at time t, wherein the power supply equipment data includes voltage, current, power, and equipment operating temperature;
去噪标准化单元，用于将所述供电设备数据进行去噪处理得到无干扰元素的设备数 据集，对所述设备数据集进行标准化处理得到t时刻的输入参数；A denoising standardization unit, used for denoising the power supply equipment data to obtain a device data set without interference elements, and standardizing the device data set to obtain an input parameter at time t;
预测单元，用于对所述输入参数输入到预测时间步长为n的供电设备数据LSTM预测模型，并输出t+n时刻的所述目标设备运行参数预测值；a prediction unit, configured to input the input parameters into the LSTM prediction model of the power supply equipment data with a prediction time step of n, and output the predicted value of the target equipment operating parameters at time t+n;
分析单元，用于将所述预测值作为SVM模型的输入值对其进行数据分类，分析t+n时刻设备是否会发生故障；an analysis unit, used for classifying the predicted value as an input value of the SVM model, and analyzing whether the equipment will fail at time t+n;
第一执行单元，用于在分析出t+n时刻设备不会发生故障时，执行返回所述数据处理单元的进行循环检测。The first execution unit is configured to execute loop detection that returns to the data processing unit when it is analyzed that the device will not fail at time t+n.
第二执行单元，用于在分析出t+n时刻设备会发生故障时，示出所述目标供电设备具有发生故障的可能性。The second execution unit is configured to indicate that the target power supply device has the possibility of failure when it is analyzed that the device will fail at time t+n.
本发明的技术效果：Technical effect of the present invention:
本发明运用一种LSTM预测模型与SVM分类模型相结合；所述LSTM预测模型输入维度包括：目标供电设备的电压、电流、功率和温度经过预处理后的数据；所述LSTM预 测模型输出维度为预测后的供电设备数据；所述SVM分类模型的输入维度为LSTM预测 模型输出的预测数据；所述SVM模型的输出维度为目标供电设备的状态，包括正常、维 护、故障三个状态。本发明所采用的供电设备故障预测方法通过对目标设备未来参数的 预测后进行状态分类，评估设备未来状态。数据预处理算法将当前时刻的设备的参数数 据进行预处理后得到标准化数据；预测模型通过将当前时刻的标准化数据预处理后作为 输入分析，输出预测的下一时刻的数据；进一步，分类模型将预测模型的输出作为输入， 经过对输入数据进行状态分类，评估出目标设备未来时刻的状态，通过输出评估结果的 正常、维护、故障三个状态来对设备进行下一步的动作；如果预测设备为正常状态，则 可以继续正常运行；若为维护状态，说明设备已经处在故障边缘状态，近期需要维护检 查；预测设备状态为故障时，则需要进行停机检修，避免发生电力事故，保障人、车和 供电系统的安全。本发明的目的是提供一种基于机器视觉的供热管道泄漏检测装置， 能够自动识别供热管道上的损害情况，及时做出防范措施并通知工作人员解决具有 相关缺陷的供热管道。The present invention combines an LSTM prediction model with an SVM classification model; the input dimension of the LSTM prediction model includes: the preprocessed data of the voltage, current, power and temperature of the target power supply equipment; the output dimension of the LSTM prediction model is: The predicted power supply equipment data; the input dimension of the SVM classification model is the predicted data output by the LSTM prediction model; the output dimension of the SVM model is the state of the target power supply equipment, including three states of normal, maintenance, and fault. The power supply equipment failure prediction method adopted in the present invention evaluates the future state of the equipment by classifying the state after predicting the future parameters of the target equipment. The data preprocessing algorithm preprocesses the parameter data of the device at the current moment to obtain standardized data; the prediction model outputs the predicted data at the next moment by preprocessing the standardized data at the current moment as input analysis; further, the classification model will The output of the prediction model is used as the input. After classifying the input data, the state of the target device in the future is estimated, and the next action of the device is carried out by outputting the three states of normal, maintenance and failure of the evaluation result; if the predicted device is If it is in the normal state, it can continue to operate normally; if it is in the maintenance state, it means that the equipment is in a state of failure and needs maintenance and inspection in the near future; when the equipment state is predicted to be faulty, it needs to be shut down for maintenance to avoid electrical accidents and protect people and vehicles. and safety of the power supply system. The purpose of the present invention is to provide a leak detection device for heating pipes based on machine vision, which can automatically identify the damage on the heating pipes, take preventive measures in time, and notify the staff to solve the heating pipes with related defects.
附图说明Description of drawings
通过阅读参照以下附图对非限制性实施例所作的详细描述，本发明的其它特征、目的和优点将会变得更明显：Other features, objects and advantages of the present invention will become more apparent by reading the detailed description of non-limiting embodiments with reference to the following drawings:
图1为本发明提供轨道交通供电设备故障预测方法的流程图；1 is a flowchart of a method for predicting faults of rail transit power supply equipment provided by the present invention;
图2为本发明一个实施例中结合LSTM模型和SVM模型对供电系统故障预测的系 统流程图；Fig. 2 is a system flow chart of power supply system fault prediction in conjunction with LSTM model and SVM model in one embodiment of the present invention;
图3为本发明一个实施例中神经网络输入层、隐藏层、输出层的结构示意图；3 is a schematic structural diagram of an input layer, a hidden layer, and an output layer of a neural network in an embodiment of the present invention;
图4为本发明一个实施例中循环神经网络结构示意图；4 is a schematic structural diagram of a recurrent neural network in an embodiment of the present invention;
图5为本发明一个实施例中长短期记忆模型结构示意图；5 is a schematic structural diagram of a long-term and short-term memory model in an embodiment of the present invention;
图6为本发明一个实施例中长短期记忆模型内部数据处理方式模型图；6 is a model diagram of an internal data processing mode of a long-term and short-term memory model in an embodiment of the present invention;
图7为本发明一个实施例中采用模型故障实际预测结果图。FIG. 7 is a diagram showing an actual prediction result of a fault using a model in an embodiment of the present invention.
具体实施方式Detailed ways
下面结合具体实施例对本发明进行详细说明。以下实施例将有助于本领域的技术人 员进一步理解本发明，但不以任何形式限制本发明。应当指出的是，对本领域的普通技术人员来说，在不脱离本发明构思的前提下，还可以做出若干变形和改进。这些都属于 本发明的保护范围。The present invention will be described in detail below with reference to specific embodiments. The following examples will help those skilled in the art to further understand the present invention, but do not limit the present invention in any form. 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. These all belong to the protection scope of the present invention.
如图1所示为本发明提供轨道交通供电设备故障预测方法的流程图，本发明提 供轨道交通供电设备故障预测方法包括以下步骤：As shown in Figure 1, the present invention provides a flow chart of a rail transit power supply equipment failure prediction method, and the present invention provides a rail transit power supply equipment failure prediction method comprising the following steps:
S1，获取t时刻的目标供电设备的供电设备数据，其中所述供电设备数据包括 电压、电流、功率和设备运行温度；S1, obtain the power supply equipment data of the target power supply equipment at time t, wherein the power supply equipment data includes voltage, current, power and equipment operating temperature;
S2，将所述供电设备数据进行去噪处理得到无干扰元素的设备数据集，对所述 设备数据集进行标准化处理得到t时刻的输入参数；S2, carrying out denoising processing to the power supply equipment data to obtain the equipment data set without interference elements, and standardizing the equipment data set to obtain the input parameter at time t;
S3，对所述输入参数输入到预测时间步长为n的供电设备数据LSTM预测模型， 并输出t+n时刻的所述目标设备运行参数预测值；S3, inputting the input parameters to the power supply equipment data LSTM prediction model with a prediction time step of n, and outputting the predicted value of the target equipment operating parameters at time t+n;
S4，将所述预测值作为SVM模型的输入值对其进行数据分类，分析t+n时刻设 备是否会发生故障，若否，则返回步骤S1进行循环检测。S4, take the predicted value as the input value of the SVM model to perform data classification on it, analyze whether the equipment will fail at time t+n, if not, then return to step S1 for loop detection.
S5：如果是，则说明所述目标供电设备具有发生故障的可能性。S5: If yes, it indicates that the target power supply device has the possibility of failure.
图2为本发明结合LSTM模型和SVM模型对供电系统故障预测的系统流程图；图 3为神经人工神经网络的结构图，包含输入层、隐藏层以及输出层；图4为RNN模型 示意图，该模型也可以通过对前一状态的数据进行未来状态的数据预测，但是只能通过 细胞状态记忆信息，所以只能处理短期依赖问题；图5为长短期记忆模型结构示意图， 图6为长短期记忆模型内部数据处理方式模型图，LSTM通过输入门、遗忘门、输出门 引入sigmoid函数并结合tanh函数，添加求和操作，减少梯度消失和梯度爆炸的可能性， 既能够处理短期依赖问题，又能够处理长期依赖问题；图7为该模型故障实际预测结果， 通过历史数据对该系统进行训练后，可以对轨道交通供电设备的状态进行准确的预测。Fig. 2 is the system flow chart that the present invention combines LSTM model and SVM model to power supply system fault prediction; Fig. 3 is the structure diagram of neural artificial neural network, including input layer, hidden layer and output layer; Fig. 4 is RNN model schematic diagram, this The model can also predict the data of the future state through the data of the previous state, but it can only memorize the information through the cell state, so it can only deal with the short-term dependency problem; Figure 5 is a schematic diagram of the structure of the long-term and short-term memory model, and Figure 6 is the long-term and short-term memory model. The model diagram of the internal data processing method of the model, LSTM introduces the sigmoid function through the input gate, forget gate, and output gate and combines the tanh function to add a sum operation to reduce the possibility of gradient disappearance and gradient explosion. It can not only deal with short-term dependency problems, but also can To deal with the long-term dependence problem; Figure 7 shows the actual prediction result of the model failure. After training the system through historical data, the state of the rail transit power supply equipment can be accurately predicted.
在本发明的一个实施例中，进一步地，对采集的设备数据进行去噪处理是采用的小 波去噪法，该方法分为数据分解步骤与数据重构步骤，其公式为：In one embodiment of the present invention, further, the wavelet denoising method is adopted to denoise the collected equipment data, and the method is divided into a data decomposition step and a data reconstruction step, and its formula is:
分解公式：A0[f(t)]＝f(t)Decomposition formula: A 0 [f(t)]=f(t)
重构公式：
式中：t为时间序列，f(t)为原始信号，j为分解的层数，H、G为时域中的小波分解滤波 器，h、g为时域中小波重构滤波器，Aj为信号f(t)在第j层的低频部分的小波系数，Dj为信号f(t)在第j层的高频部分的小波系数。In the formula: t is the time series, f(t) is the original signal, j is the number of layers to be decomposed, H and G are the wavelet decomposition filters in the time domain, h and g are the wavelet reconstruction filters in the time domain, A j is the wavelet coefficient of the low frequency part of the signal f(t) in the jth layer, and D j is the wavelet coefficient of the high frequency part of the signal f(t) in the jth layer.
进一步地，所述步骤S3中的对去噪后的设备数据进行规范化处理是采用以下公式进行 建立：Further, in the described step S3, the equipment data after denoising is normalized and is established by using the following formula:
其中，
所述的轨道交通供电设备故障预测方法，运用一种LSTM预测模型与SVM分类模 型相结合；所述LSTM预测模型输入维度包括：目标供电设备的电压、电流、功率和温 度经过预处理后的数据；所述LSTM预测模型输出维度为预测后的供电设备数据；所述 SVM分类模型的输入维度为LSTM预测模型输出的预测数据；所述SVM模型的输出维 度为目标供电设备的状态，包括正常、维护、故障三个状态。The described rail transit power supply equipment fault prediction method uses a combination of an LSTM prediction model and an SVM classification model; the input dimensions of the LSTM prediction model include: preprocessed data of voltage, current, power and temperature of the target power supply equipment The output dimension of the LSTM prediction model is the predicted power supply equipment data; the input dimension of the SVM classification model is the prediction data output by the LSTM prediction model; the output dimension of the SVM model is the state of the target power supply equipment, including normal, Maintenance and failure three states.
本发明所采用的供电设备故障预测方法通过对目标设备未来参数的预测后进行状 态分类，评估设备未来状态。数据预处理算法将当前时刻的设备的参数数据进行预处理后得到标准化数据；预测模型通过将当前时刻的标准化数据预处理后作为输入分析，输 出预测的下一时刻的数据；进一步，分类模型将预测模型的输出作为输入，经过对输入 数据进行状态分类，评估出目标设备未来时刻的状态，通过输出评估结果的正常、维护、 故障三个状态来对设备进行下一步的动作；如果预测设备为正常状态，则可以继续正常 运行；若为维护状态，说明设备已经处在故障边缘状态，近期需要维护检查；预测设备 状态为故障时，则需要进行停机检修，避免发生电力事故，保障人、车和供电系统的安 全。The power supply equipment failure prediction method adopted in the present invention evaluates the future state of the equipment by classifying the state after predicting the future parameters of the target equipment. The data preprocessing algorithm preprocesses the parameter data of the device at the current moment to obtain standardized data; the prediction model outputs the predicted data at the next moment by preprocessing the standardized data at the current moment as input analysis; further, the classification model will The output of the prediction model is used as input. After classifying the input data, the state of the target device in the future is estimated, and the next action of the device is carried out by outputting the three states of normal, maintenance and failure of the evaluation result; if the predicted device is If it is in the normal state, it can continue to operate normally; if it is in the maintenance state, it means that the equipment is in a state of failure and needs maintenance and inspection in the near future; when the equipment state is predicted to be faulty, it needs to be shut down for maintenance to avoid electrical accidents and protect people and vehicles. and safety of the power supply system.
另一方面，本发明还提供一种轨道交通供电设备预测装置，所述轨道交通供电设备 故障预测装置包括：On the other hand, the present invention also provides a rail transit power supply equipment prediction device, and the rail transit power supply equipment failure prediction device includes:
数据获取单元，用于获取t时刻的目标供电设备的供电设备数据，其中所述供电设备数据包括电压、电流、功率和设备运行温度；a data acquisition unit, configured to acquire power supply equipment data of the target power supply equipment at time t, wherein the power supply equipment data includes voltage, current, power, and equipment operating temperature;
去噪标准化单元，用于将所述供电设备数据进行去噪处理得到无干扰元素的设备数 据集，对所述设备数据集进行标准化处理得到t时刻的输入参数；A denoising standardization unit, used for denoising the power supply equipment data to obtain a device data set without interference elements, and standardizing the device data set to obtain an input parameter at time t;
预测单元，用于将所述输入参数输入到预测时间步长为n的供电设备数据LSTM预测模型，并输出t+n时刻的所述目标设备运行参数预测值；a prediction unit, configured to input the input parameters into the power supply equipment data LSTM prediction model with a prediction time step of n, and output the predicted value of the target equipment operating parameters at time t+n;
分析单元，用于将所述预测值作为SVM模型的输入值对其进行数据分类，分析t+n时刻设备是否会发生故障；an analysis unit, used for classifying the predicted value as an input value of the SVM model, and analyzing whether the equipment will fail at time t+n;
第一执行单元，用于在分析出t+n时刻设备不会发生故障时，执行返回所述数据处理单元的进行循环检测。The first execution unit is configured to execute loop detection that returns to the data processing unit when it is analyzed that the device will not fail at time t+n.
第二执行单元，用于在分析出t+n时刻设备会发生故障时，示出所述目标供电设备具有发生故障的可能性。The second execution unit is configured to indicate that the target power supply device has the possibility of failure when it is analyzed that the device will fail at time t+n.
另一方面，本发明还提供一种轨道交通供电设备预测系统，主要包括供电设备参数 单元、供电设备数据去噪单元、供电设备数据规范化单元、供电设备数据预测单元、供电设备数据运行状态分类单元；On the other hand, the present invention also provides a rail transit power supply equipment prediction system, which mainly includes a power supply equipment parameter unit, a power supply equipment data denoising unit, a power supply equipment data normalization unit, a power supply equipment data prediction unit, and a power supply equipment data operation state classification unit ;
所述供电设备参数单元，用于获取供电设备结构参数信息；The power supply equipment parameter unit is used to obtain the structural parameter information of the power supply equipment;
所述供电设备数据去噪单元，通过小波去噪法将设备数据中的干扰因素去掉，保证 所取数据与目标设备的高度相关性；The power supply equipment data denoising unit removes the interference factor in the equipment data by wavelet denoising method to ensure the high correlation between the data taken and the target equipment;
所述供电设备数据规范化单元，用于对去噪后的数据进行规范化处理，避免因为数 据量纲不同影响数据分析的结果；The power supply equipment data normalization unit is used to normalize the denoised data to avoid affecting the results of data analysis due to different data dimensions;
所述供电设备数据预测单元，用于综合所述权利要求3中目标设备所处环境，通过将设备运行实时数据作为输入对设备未来运行参数进行预测，得到目标设备预测数据。The power supply equipment data prediction unit is used for synthesizing the environment in which the target equipment is located in the claim 3, and predicting the future operation parameters of the equipment by taking the equipment operation real-time data as an input to obtain the target equipment prediction data.
所述供电设备数据运行状态分类单元，用于将供电设备预测数据进行状态分类，分 析得到目标设备未来的运行状态。The power supply equipment data operation state classification unit is used for classifying the power supply equipment prediction data by state, and analyzes and obtains the future operation state of the target equipment.
本发明提供的轨道交通供电设备数据预测和状态分类系统，可以有效改善现有供电 系统故障诊断方法所存在的不足。现有的轨道交通供电系统故障诊断的方法主要是对故 障发生后产生的设备参数进行分析分类，然后将故障进行诊断。这种诊断的方式在故障发生后，此时已经产生故障影响，不但会对线路造成大规模的延误造成经济损失，也会 危害人车安全。The rail transit power supply equipment data prediction and state classification system provided by the present invention can effectively improve the shortcomings of the existing power supply system fault diagnosis methods. The existing method of fault diagnosis of rail transit power supply system is mainly to analyze and classify the equipment parameters generated after the fault occurs, and then diagnose the fault. After the fault occurs, this method of diagnosis has already produced the impact of the fault, which will not only cause large-scale delays to the line and cause economic losses, but also endanger the safety of people and vehicles.
本发明还提供一种数据获取预测目标设备数据的LSTM模型和对设备数据进行状态分类的SVM模型的方法，包括以下内容：The present invention also provides a method for obtaining an LSTM model for predicting target device data and an SVM model for state classification of device data, including the following content:
所述获取预测目标设备数据的LSTM模型的构建是由以下步骤来实现的：The construction of the LSTM model for obtaining and predicting the target device data is achieved by the following steps:
A1：设置LSTM输入维度和预测时间步长；A1: Set the LSTM input dimension and prediction time step;
A2：设置模型优化器以及学习速率；A2: Set the model optimizer and learning rate;
A3：设置隐含层神经元节点个数；A3: Set the number of hidden layer neurons;
A4：将目标设备运行历史数据集导入到LSTM神经网络模型中进行训练；A4: Import the target device operation history data set into the LSTM neural network model for training;
A5：在训练过程中对LSTM模型的超参数进行不断的优化调整，得到最优模型。A5: During the training process, the hyperparameters of the LSTM model are continuously optimized and adjusted to obtain the optimal model.
所述步骤A4中的神经网络LSTM模型采用tanh函数作为激活函数，tanh函数公式为：The neural network LSTM model in the step A4 adopts the tanh function as the activation function, and the tanh function formula is:
其中，x表示隐含层神经元输入特征向量的分量，即目标设备参数经规范化后的数据特征；ω表示输入分量的权重；θ表示神经元的阈值。Among them, x represents the component of the input feature vector of the hidden layer neuron, that is, the normalized data feature of the target device parameters; ω represents the weight of the input component; θ represents the threshold of the neuron.
所述对设备数据进行状态分类的SVM模型的构建是由以下步骤来实现的：The construction of the SVM model for state classification of equipment data is realized by the following steps:
B1：选取目标设备历史数据中的正常运行时参数和故障时运行参数合并组成数据集；B1: Select the normal operation parameters and failure operation parameters in the historical data of the target device and combine them to form a data set;
B2：将组成的数据集进行标准化处理，处理后的数据集作为SVM模型训练时的输入数据；B2: Standardize the composed data set, and use the processed data set as the input data for SVM model training;
B3：对构建的SVM模型进行训练，在此过程中不断调整SVM的核函数、类别权重、 迭代次数等参数，优化分类模型的性能，得到最优分类模型。B3: Train the constructed SVM model, and continuously adjust parameters such as the kernel function, category weight, and number of iterations of the SVM in the process, optimize the performance of the classification model, and obtain the optimal classification model.
所述步骤B2中的时间序列分类SVM模型采用sigmoid函数作为核函数：The time series classification SVM model in the described step B2 adopts the sigmoid function as the kernel function:
所述的获取预测目标设备数据的长短期记忆网络(LSTM)模型的方法，其特征在于， LSTM是一种循环神经网络的改进，增加了一条细胞状态的通路结构，比循环神经网络结构更为复杂，使得LSTM解决了循环神经网络不能记忆长期状态的缺点，能够有效的 通过对长期的历史状态分析来预测未来状态，使得预测结果更为准确，更加适合轨道交 通供电设备数据的预测。The method for obtaining a long short-term memory network (LSTM) model for predicting target device data is characterized in that, LSTM is an improvement of a recurrent neural network, adding a path structure of the cell state, which is more efficient than a recurrent neural network structure. The complexity makes LSTM solve the shortcoming that the recurrent neural network cannot memorize the long-term state, and can effectively predict the future state by analyzing the long-term historical state, making the prediction result more accurate and more suitable for the prediction of rail transit power supply equipment data.
长短期记忆网络(LSTM)是深度学习中能预测未知时长延时时间序列的深层网络模 型。故障预测系统需要各类传感器输入大量的参数信息，针对数据量大且数据重复率高的特点，长短期记忆的忘记阶段以及选择记忆阶段可以对故障预测网络中输入的大量参数进行筛选，提取有用的信息，忽略重复性高的信息，大大减少参数的运算量。提取时 间序列中有用信息后，该模型可以分析时间序列中参数的关联关系、预测时间序列数据 的趋势，同时也克服了其他神经网络无法长时间预测的缺陷。所述的获取设备数据进行 状态分类的支持向量机(SVM)模型的方法，其特征在于，SVM是一种高效准确的分 类模型，通过设置不同类型的核函数，可以对不同输入维度的数据进行有效的分类，并 且可以调节其惩罚参数来调节分类模型的泛化能力。本实施例融合了多种神经网络算法， 制定了精确的步骤，最终可以预测供电设备未来是否为故障状态。另外，本发明通过融 合多维度输入数据进行模型训练，得到供电设备的预测参数数据，运用支持向量机模型 将预测的未来时刻参数数据进行设备状态分类。率先结合了长短期记忆模型能够进行长 时间预测和支持向量机具有良好的非线性分类的优点，采用系统化的设计，为轨道交通 供电系统建立了一种精准预测故障的方法，有效改善轨道交通供电系统的稳定性和安全 性，为乘客、列车、供电系统的安全提供保障。Long short-term memory network (LSTM) is a deep network model in deep learning that can predict time series of unknown duration. The fault prediction system requires various types of sensors to input a large amount of parameter information. According to the characteristics of large amount of data and high data repetition rate, the forgetting stage of long-term and short-term memory and the selection memory stage can screen a large number of parameters input in the fault prediction network and extract useful information. information, ignoring highly repetitive information, greatly reducing the amount of parameter calculation. After extracting the useful information in the time series, the model can analyze the relationship between the parameters in the time series and predict the trend of the time series data, and also overcome the shortcomings that other neural networks cannot predict for a long time. The described method for obtaining a support vector machine (SVM) model for state classification by obtaining equipment data is characterized in that, SVM is an efficient and accurate classification model, and by setting different types of kernel functions, data of different input dimensions can be processed. Effective classification, and its penalty parameter can be adjusted to adjust the generalization ability of the classification model. This embodiment integrates various neural network algorithms, formulates precise steps, and finally can predict whether the power supply equipment will be in a fault state in the future. In addition, the present invention performs model training by fusing multi-dimensional input data, obtains the predicted parameter data of the power supply equipment, and uses the support vector machine model to classify the predicted future moment parameter data for equipment state classification. It takes the lead in combining the advantages of long-term and short-term memory model for long-term prediction and support vector machine with good nonlinear classification. It adopts a systematic design to establish a method for accurately predicting faults for rail transit power supply systems, effectively improving rail transit. The stability and safety of the power supply system provide a guarantee for the safety of passengers, trains and power supply systems.
本发明还提供一种轨道交通供电设备预测和状态分类系统，主要包括供电设备参数 单元、供电设备数据去噪单元、供电设备数据规范化单元、供电设备数据预测单元、供电设备数据运行状态分类单元；所述供电设备参数单元，用于获取供电设备结构参数信息；所述供电设备数据去噪单元，通过小波去噪法将设备数据中的干扰因素去掉，保证 所取数据与目标设备的高度相关性；供电设备数据规范化单元，用于对去噪后的数据进 行规范化处理，避免因为数据量纲不同影响数据分析的结果；The invention also provides a rail transit power supply equipment prediction and status classification system, which mainly includes a power supply equipment parameter unit, a power supply equipment data denoising unit, a power supply equipment data normalization unit, a power supply equipment data prediction unit, and a power supply equipment data operation state classification unit; The power supply equipment parameter unit is used to obtain the structural parameter information of the power supply equipment; the power supply equipment data denoising unit removes the interference factors in the equipment data through the wavelet denoising method to ensure the high correlation between the acquired data and the target equipment ; The data normalization unit of power supply equipment is used to normalize the denoised data to avoid affecting the results of data analysis due to different data dimensions;
所述供电设备数据预测单元，用于综合目标设备所处环境，通过将设备运行实时数 据作为输入对设备未来运行参数进行预测，得到目标设备预测数据。The power supply equipment data prediction unit is used for synthesizing the environment in which the target equipment is located, and predicting the future operation parameters of the equipment by taking the equipment operation real-time data as input to obtain the target equipment prediction data.
所述供电设备数据运行状态分类单元，用于将供电设备预测数据进行状态分类，分 析得到目标设备未来的运行状态。The power supply equipment data operation state classification unit is used for classifying the power supply equipment prediction data by state, and analyzes and obtains the future operation state of the target equipment.
本发明提供的轨道交通供电设备数据预测和状态分类系统，可以有效改善现有供电 系统故障诊断方法所存在的不足。现有的轨道交通供电系统故障诊断的方法主要是对故 障发生后产生的设备参数进行分析分类，然后将故障进行诊断。这种诊断的方式在故障发生后，此时已经产生故障影响，不但会对线路造成大规模的延误造成经济损失，也会 危害人车安全。The rail transit power supply equipment data prediction and state classification system provided by the present invention can effectively improve the shortcomings of the existing power supply system fault diagnosis methods. The existing method of fault diagnosis of rail transit power supply system is mainly to analyze and classify the equipment parameters generated after the fault occurs, and then diagnose the fault. After the fault occurs, this method of diagnosis has already produced the impact of the fault, which will not only cause large-scale delays to the line and cause economic losses, but also endanger the safety of people and vehicles.
本发明还提供一种数据获取预测目标设备数据的LSTM模型和对设备数据进行状态分类的SVM模型的方法，包括以下内容：The present invention also provides a method for obtaining an LSTM model for predicting target device data and an SVM model for state classification of device data, including the following content:
所述获取预测目标设备数据的LSTM模型的构建是由以下步骤来实现的：The construction of the LSTM model for obtaining and predicting the target device data is achieved by the following steps:
A1：设置LSTM输入维度和预测时间步长；A1: Set the LSTM input dimension and prediction time step;
A2：设置模型优化器以及学习速率；A2: Set the model optimizer and learning rate;
A3：设置隐含层神经元节点个数；A3: Set the number of hidden layer neurons;
A4：将目标设备运行历史数据集导入到LSTM神经网络模型中进行训练；A4: Import the target device operation history data set into the LSTM neural network model for training;
A5：在训练过程中对LSTM模型的超参数进行不断的优化调整，得到最优模型。A5: During the training process, the hyperparameters of the LSTM model are continuously optimized and adjusted to obtain the optimal model.
所述步骤A4中的神经网络LSTM模型采用tanh函数作为激活函数，tanh函数公式为：The neural network LSTM model in the step A4 adopts the tanh function as the activation function, and the tanh function formula is:
其中，x表示隐含层神经元输入特征向量的分量，即目标设备参数经规范化后的数据特征；ω表示输入分量的权重；θ表示神经元的阈值。Among them, x represents the component of the input feature vector of the hidden layer neuron, that is, the normalized data feature of the target device parameters; ω represents the weight of the input component; θ represents the threshold of the neuron.
所述对设备数据进行状态分类的SVM模型的构建是由以下步骤来实现的：The construction of the SVM model for state classification of equipment data is realized by the following steps:
B1：选取目标设备历史数据中的正常运行时参数和故障时运行参数合并组成数据集；B1: Select the normal operation parameters and failure operation parameters in the historical data of the target device and combine them to form a data set;
B2：将组成的数据集进行标准化处理，处理后的数据集作为SVM模型训练时的输入数据；B2: Standardize the composed data set, and use the processed data set as the input data for SVM model training;
B3：对构建的SVM模型进行训练，在此过程中不断调整SVM的核函数、类别权重、 迭代次数等参数，优化分类模型的性能，得到最优分类模型。B3: Train the constructed SVM model, and continuously adjust parameters such as the kernel function, category weight, and iteration times of the SVM in the process, optimize the performance of the classification model, and obtain the optimal classification model.
所述步骤B2中的时间序列分类SVM模型采用sigmoid函数作为核函数：The time series classification SVM model in the described step B2 adopts the sigmoid function as the kernel function:
所述的获取预测目标设备数据的长短期记忆网络(LSTM)模型的方法，其特征在于， LSTM是一种循环神经网络的改进，增加了一条细胞状态的通路结构，比循环神经网络结构更为复杂，使得LSTM解决了循环神经网络不能记忆长期状态的缺点，能够有效的 通过对长期的历史状态分析来预测未来状态，使得预测结果更为准确，更加适合轨道交 通供电设备数据的预测。The method for obtaining a long short-term memory network (LSTM) model for predicting target device data is characterized in that, LSTM is an improvement of a recurrent neural network, adding a path structure of the cell state, which is more efficient than a recurrent neural network structure. The complexity makes LSTM solve the shortcoming that the recurrent neural network cannot memorize the long-term state, and can effectively predict the future state by analyzing the long-term historical state, making the prediction result more accurate and more suitable for the prediction of rail transit power supply equipment data.
长短期记忆网络(LSTM)是深度学习中能预测未知时长延时时间序列的深层网络模 型。故障预测系统需要各类传感器输入大量的参数信息，针对数据量大且数据重复率高的特点，长短期记忆的忘记阶段以及选择记忆阶段可以对故障预测网络中输入的大量参数进行筛选，提取有用的信息，忽略重复性高的信息，大大减少参数的运算量。提取时 间序列中有用信息后，该模型可以分析时间序列中参数的关联关系、预测时间序列数据 的趋势，同时也克服了其他神经网络无法长时间预测的缺陷。Long short-term memory network (LSTM) is a deep network model in deep learning that can predict time series of unknown duration. The fault prediction system requires various types of sensors to input a large amount of parameter information. According to the characteristics of large amount of data and high data repetition rate, the forgetting stage of long-term and short-term memory and the selection memory stage can screen a large number of parameters input in the fault prediction network and extract useful information. information, ignoring highly repetitive information, greatly reducing the amount of parameter calculation. After extracting the useful information in the time series, the model can analyze the relationship between the parameters in the time series and predict the trend of the time series data, and also overcome the defects that other neural networks cannot predict for a long time.
所述的获取设备数据进行状态分类的支持向量机(SVM)模型的方法，其特征在于，SVM是一种高效准确的分类模型，通过设置不同类型的核函数，可以对不同输入维度 的数据进行有效的分类，并且可以调节其惩罚参数来调节分类模型的泛化能力。The described method for obtaining a support vector machine (SVM) model for state classification by obtaining equipment data is characterized in that, SVM is an efficient and accurate classification model, and by setting different types of kernel functions, data of different input dimensions can be processed. Effective classification, and its penalty parameter can be adjusted to adjust the generalization ability of the classification model.
本发明通过融合多维度输入数据进行模型训练，得到供电设备的预测参数数据，运 用支持向量机模型将预测的未来时刻参数数据进行设备状态分类。率先结合了长短期记 忆模型能够进行长时间预测和支持向量机具有良好的非线性分类的优点，采用系统化的 设计，为轨道交通供电系统建立了一种精准预测故障的方法，有效改善轨道交通供电系统的稳定性和安全性，为乘客、列车、供电系统的安全提供保障。The present invention performs model training by fusing multi-dimensional input data, obtains the predicted parameter data of the power supply equipment, and uses the support vector machine model to classify the predicted future moment parameter data for equipment state classification. It takes the lead in combining the advantages of long-term and short-term memory model for long-term prediction and support vector machine with good nonlinear classification. It adopts a systematic design to establish a method for accurately predicting faults for rail transit power supply systems, effectively improving rail transit. The stability and safety of the power supply system provide a guarantee for the safety of passengers, trains and power supply systems.
以上所述的仅为本发明具体实施方式，并不用于限制本发明，仅是对本发明的目的、 技术方案和有益效果进行了更进一步详细说明，The above are only specific embodiments of the present invention, and are not intended to limit the present invention, but merely describe the purpose, technical solutions and beneficial effects of the present invention in further detail.
凡在本发明的精神和原则之内，倘若所做任何的修改、等同替换、改进等，均应包含在本发明的保护范围之内。Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.
以上对本发明的具体实施例进行了描述。需要理解的是，本发明并不局限于上 述特定实施方式，本领域技术人员可以在权利要求的范围内做出各种变形或修改， 这并不影响本发明的实质内容。Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the above-mentioned specific embodiments, and those skilled in the art can make various variations or modifications within the scope of the claims, which do not affect the essential content of the present invention.
Claims (9)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110911227.1A CN113848417A (en) | 2021-08-09 | 2021-08-09 | Rail transit power supply equipment fault prediction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110911227.1A CN113848417A (en) | 2021-08-09 | 2021-08-09 | Rail transit power supply equipment fault prediction method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113848417A true CN113848417A (en) | 2021-12-28 |
Family
ID=78975696
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110911227.1A Pending CN113848417A (en) | 2021-08-09 | 2021-08-09 | Rail transit power supply equipment fault prediction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113848417A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115037630A (en) * | 2022-04-29 | 2022-09-09 | 电子科技大学长三角研究院(湖州) | Weighted network link prediction method based on structural disturbance model |
| CN115034285A (en) * | 2022-04-22 | 2022-09-09 | 广西电网有限责任公司 | SVM-WS based intelligent electric meter fault prediction method |
| CN115063048A (en) * | 2022-08-16 | 2022-09-16 | 天津津轨汇海科技发展有限公司 | Intelligent management system for rail transit power supply operation and maintenance |
| CN115149653A (en) * | 2022-09-05 | 2022-10-04 | 天津津轨汇海科技发展有限公司 | Safety management system for rail transit power supply equipment |
| CN116150601A (en) * | 2022-12-15 | 2023-05-23 | 无锡同方人工环境有限公司 | Fault diagnosis method and system for air conditioner compressor |
| CN116340875A (en) * | 2023-05-30 | 2023-06-27 | 单县鑫和纺织有限公司 | Roving frame operation fault prediction system based on data analysis |
| CN116433226A (en) * | 2023-06-14 | 2023-07-14 | 浙江亿视电子技术有限公司 | Alternating current power supply equipment management method and system based on electrical parameter AI analysis |
| CN116662862A (en) * | 2023-06-05 | 2023-08-29 | 西安工程大学 | Transformer Fault Early Warning Method Based on Prediction and Diagnosis of Dissolved Gas Content in Oil |
| CN117516927A (en) * | 2024-01-05 | 2024-02-06 | 四川省机械研究设计院(集团)有限公司 | Gearbox fault detection method, system, equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108090606A (en) * | 2017-12-12 | 2018-05-29 | 上海应用技术大学 | Equipment fault finds method and system |
| CN109858345A (en) * | 2018-12-25 | 2019-06-07 | 华中科技大学 | A kind of intelligent failure diagnosis method suitable for pipe expanding equipment |
| CN110161343A (en) * | 2019-06-12 | 2019-08-23 | 中南大学 | A kind of non-intrusion type real-time dynamic monitoring method of intelligence train exterior power receiving device |
| CN111239672A (en) * | 2020-03-19 | 2020-06-05 | 东南大学 | Gradual fault prediction method of optical fiber current transformer based on machine learning algorithm |
| CN111368890A (en) * | 2020-02-26 | 2020-07-03 | 珠海格力电器股份有限公司 | Fault detection method and device and information physical fusion system |
-
2021
- 2021-08-09 CN CN202110911227.1A patent/CN113848417A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108090606A (en) * | 2017-12-12 | 2018-05-29 | 上海应用技术大学 | Equipment fault finds method and system |
| CN109858345A (en) * | 2018-12-25 | 2019-06-07 | 华中科技大学 | A kind of intelligent failure diagnosis method suitable for pipe expanding equipment |
| CN110161343A (en) * | 2019-06-12 | 2019-08-23 | 中南大学 | A kind of non-intrusion type real-time dynamic monitoring method of intelligence train exterior power receiving device |
| CN111368890A (en) * | 2020-02-26 | 2020-07-03 | 珠海格力电器股份有限公司 | Fault detection method and device and information physical fusion system |
| CN111239672A (en) * | 2020-03-19 | 2020-06-05 | 东南大学 | Gradual fault prediction method of optical fiber current transformer based on machine learning algorithm |
Non-Patent Citations (1)
| Title |
|---|
| 杨秀璋，颜娜: "Python 网络数据爬取及分析从入门到精通 分析篇", 北京航空航天大学出版社, pages: 147 - 148 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115034285A (en) * | 2022-04-22 | 2022-09-09 | 广西电网有限责任公司 | SVM-WS based intelligent electric meter fault prediction method |
| CN115037630A (en) * | 2022-04-29 | 2022-09-09 | 电子科技大学长三角研究院(湖州) | Weighted network link prediction method based on structural disturbance model |
| CN115037630B (en) * | 2022-04-29 | 2023-10-20 | 电子科技大学长三角研究院(湖州) | Weighted network link prediction method based on structure disturbance model |
| CN115063048A (en) * | 2022-08-16 | 2022-09-16 | 天津津轨汇海科技发展有限公司 | Intelligent management system for rail transit power supply operation and maintenance |
| CN115149653A (en) * | 2022-09-05 | 2022-10-04 | 天津津轨汇海科技发展有限公司 | Safety management system for rail transit power supply equipment |
| CN116150601A (en) * | 2022-12-15 | 2023-05-23 | 无锡同方人工环境有限公司 | Fault diagnosis method and system for air conditioner compressor |
| CN116340875A (en) * | 2023-05-30 | 2023-06-27 | 单县鑫和纺织有限公司 | Roving frame operation fault prediction system based on data analysis |
| CN116340875B (en) * | 2023-05-30 | 2023-10-17 | 单县鑫和纺织有限公司 | Roving frame operation fault prediction system based on data analysis |
| CN116662862A (en) * | 2023-06-05 | 2023-08-29 | 西安工程大学 | Transformer Fault Early Warning Method Based on Prediction and Diagnosis of Dissolved Gas Content in Oil |
| CN116433226A (en) * | 2023-06-14 | 2023-07-14 | 浙江亿视电子技术有限公司 | Alternating current power supply equipment management method and system based on electrical parameter AI analysis |
| CN117516927A (en) * | 2024-01-05 | 2024-02-06 | 四川省机械研究设计院(集团)有限公司 | Gearbox fault detection method, system, equipment and storage medium |
| CN117516927B (en) * | 2024-01-05 | 2024-04-05 | 四川省机械研究设计院(集团)有限公司 | Gearbox fault detection method, system, device and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113848417A (en) | Rail transit power supply equipment fault prediction method and device | |
| Huang et al. | PM2. 5 concentration forecasting at surface monitoring sites using GRU neural network based on empirical mode decomposition | |
| Ajagekar et al. | Quantum computing based hybrid deep learning for fault diagnosis in electrical power systems | |
| Cao et al. | A parallel GRU with dual-stage attention mechanism model integrating uncertainty quantification for probabilistic RUL prediction of wind turbine bearings | |
| CN111813084B (en) | Mechanical equipment fault diagnosis method based on deep learning | |
| Tan et al. | A new sensor fault diagnosis method for gas leakage monitoring based on the naive Bayes and probabilistic neural network classifier | |
| Shang et al. | A hybrid method for traffic incident detection using random forest-recursive feature elimination and long short-term memory network with Bayesian optimization algorithm | |
| CN115688035A (en) | Time sequence power data anomaly detection method based on self-supervision learning | |
| CN112000084A (en) | An intelligent BIT design method for controller module based on 1D-CNN and GRU-SVM | |
| CN118312722B (en) | Power time sequence big data real-time cleaning-based energy consumption diagnosis method and system | |
| CN110020637A (en) | A kind of analog circuit intermittent fault diagnostic method based on more granularities cascade forest | |
| Li et al. | Intelligent fault diagnosis of aeroengine sensors using improved pattern gradient spectrum entropy | |
| CN119357562B (en) | Method and system for evaluating encapsulation tightness of integrated circuit under high-temperature environment | |
| Li et al. | A multi-factor combination prediction model of carbon emissions based on improved CEEMDAN | |
| Liu et al. | Air Pollutant Concentration Forecasting Using Long Short‐Term Memory Based on Wavelet Transform and Information Gain: A Case Study of Beijing | |
| Zhou et al. | Real-time anomaly detection in distribution grids using long short term memory network | |
| CN117609885A (en) | High-speed rail intrusion monitoring method based on distributed optical fiber sensing and fused neural network | |
| CN111881299B (en) | Outlier Event Detection and Recognition Method Based on Replication Neural Network | |
| Sahoo et al. | Online fault detection and classification of 3-phase long transmission line using machine learning model | |
| CN110046663A (en) | A kind of complex electromechanical systems fault critical state discrimination method | |
| Li et al. | Efficient battery fault monitoring in electric vehicles: Advancing from detection to quantification | |
| Qu et al. | Remaining useful life prediction of high-speed railroad contact network based on stacking integrated attention-lstm-cnn deep learning | |
| Chen et al. | A data-driven approach for assessing aero-engine health status | |
| Dang et al. | seq2graph: discovering dynamic dependencies from multivariate time series with multi-level attention | |
| Shi et al. | A new adaptive feature fusion and selection network for intelligent transportation systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |