CN111178553A - Industrial equipment health trend analysis method and system based on ARIMA and LSTM algorithms - Google Patents
Industrial equipment health trend analysis method and system based on ARIMA and LSTM algorithms Download PDFInfo
- Publication number
- CN111178553A CN111178553A CN201911292039.4A CN201911292039A CN111178553A CN 111178553 A CN111178553 A CN 111178553A CN 201911292039 A CN201911292039 A CN 201911292039A CN 111178553 A CN111178553 A CN 111178553A
- Authority
- CN
- China
- Prior art keywords
- data
- equipment
- industrial equipment
- trend
- lstm
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
Abstract
The invention relates to an industrial equipment health trend analysis method and system based on ARIMA and LSTM algorithms. The method comprises the following steps: 1) acquiring running state information data of industrial equipment through a sensor; 2) preprocessing the acquired data, removing noise data and ensuring the usability and accuracy of the data; 3) performing feature selection on the preprocessed data; 4) predicting the variation trend of a single feature in the features obtained in the step 3) by utilizing an ARIMA model; 5) inputting the prediction result of the step 4) into the LSTM model as training data, and training the LSTM model; 6) and inputting historical data of the industrial equipment into the trained LSTM model to obtain a prediction result of the health trend of the industrial equipment. The method is used as a support technology for the operation and maintenance of key components of an intelligent factory, can realize the analysis and prediction of the state trend of the equipment, realizes the intellectualization of the operation and maintenance of the equipment, and ensures the high-efficiency operation of the equipment.
Description
Technical Field
The invention belongs to the technical field of information technology and intelligent manufacturing, and particularly relates to an industrial equipment health trend analysis method and system based on ARIMA and LSTM algorithms.
Background
With the fusion of informatization technology and automation technology, intelligent factories formed by assets such as industrial robots and large numerical control machines put higher demands on the availability of key equipment. Once an unexpected shutdown occurs, it will have a great impact on production efficiency. Therefore, the traditional after-the-fact maintenance and periodic maintenance cannot meet the requirements of real-time, intelligent and networked equipment maintenance management. The preventive maintenance technologies with the core of feature identification, service life prediction, fault analysis and maintenance planning become the inevitable trend of technical development, the technologies introduce the degradation performance factors of equipment, integrate the fault diagnosis information and the prediction information of the equipment, research the dynamic prediction maintenance planning modeling method of online monitoring by taking the total maintenance cost and the total maintenance time as the target, continuously update the optimal preventive maintenance strategy based on the online state monitoring data of an actual operation system, and optimize the maintenance cost.
At present, most industrial equipment is not provided with built-in modules for fault diagnosis, fault prediction and the like, so that external sensors for acceleration, displacement and the like are needed for data acquisition, and the running state of the equipment is judged by extracting the characteristics of the equipment during running through machine learning, deep learning and training of a fault diagnosis model. At present, the fault diagnosis of equipment adopts a threshold processing mode which is more conventional, utilizes a correlation analysis method to analyze the correlation between a plurality of characteristics and the equipment state, judges key factors influencing the equipment state and further sets a threshold for main characteristic factors. In the running process of the equipment, the state of the equipment is judged by monitoring data of some collection points and comparing the data with a preset result. In the aspects of trend analysis and fault prediction of equipment, algorithms such as a support vector machine, linear regression and ANN are well applied to the field, but the prediction effect is not obvious due to the fact that time dimension factors are not introduced, and the algorithms are difficult to use in production.
Disclosure of Invention
Aiming at the future development trend of intelligent, service and social manufacturing industries, the invention solves the outstanding challenges of two aspects in the intelligent manufacturing environment:
(1) industrial equipment does not have a perfect data acquisition module at present, and a complete response processing mechanism does not exist for processing data and real-time operation data of the equipment in the production and manufacturing process. According to the invention, the running data of the equipment is acquired through the three-dimensional acceleration sensor, the displacement sensor, the temperature sensor and the like, and the characteristic variable with larger contribution weight to the health state of the equipment is obtained based on a plurality of big data processing modes such as characteristic extraction correlation analysis and the like, so that the purity and the reliability of the subsequent input model data are ensured.
(2) The equipment state data acquisition and processing are completed in a time domain, and the accuracy rate is low and the performance effect is poor for result prediction by using regression algorithms such as deep learning and machine learning and trained algorithm models.
According to the method, through an ARIMA (automatic Integrated Moving Average Autoregressive model), under the condition that no external factors are introduced, the trend prediction of a single characteristic is completed, the accuracy of data can be ensured in a small range, the error is kept in an acceptable range, and then the trend of the equipment state along with the time in historical data is learned based on an LSTM (long short-Term Memory) Memory neural network, so that the trend prediction of the equipment state is completed.
The technical scheme adopted by the invention is as follows:
an industrial equipment health trend analysis method based on ARIMA and LSTM algorithms comprises the following steps:
1) acquiring running state information data of industrial equipment through a sensor;
2) preprocessing the acquired data, removing noise data and ensuring the usability and accuracy of the data;
3) performing feature selection on the preprocessed data;
4) predicting the variation trend of a single feature in the features obtained in the step 3) by utilizing an ARIMA model;
5) inputting the prediction result of the step 4) into the LSTM model as training data, and training the LSTM model;
6) and inputting historical data of the industrial equipment into the trained LSTM model to obtain a prediction result of the health trend of the industrial equipment.
Further, in the step 1), the three-dimensional acceleration sensor, the temperature sensor, the angular velocity sensor and the displacement sensor are used for acquiring the running state information data of the industrial equipment, including X, Y, Z acceleration, motor temperature, axle center track and torque signals in three directions, the frequency of data acquisition of the sensors is unified, and the synchronism of data acquisition of the equipment is ensured.
Further, the preprocessing of step 2) includes data smoothing, normalization, and missing value processing.
Further, step 3) selects features according to the contribution degree of the features, wherein the contribution degree of the features is the linear correlation strength between continuous variables measured by a Pearson correlation coefficient r, and the method comprises the following steps: 0 ═ r ═ 0.3, low correlation; 0.3 ═ r ═ 0.8, moderately relevant; 0.8 ═ r | < ═ 1, highly correlated.
Further, step 4) comprises:
a) data obtained by data preprocessing and feature selection are used as data sources, and stability inspection is carried out on data distribution;
b) if the data is unstable, processing the data in a differential operation mode until the data distribution shows a stable trend;
c) filtering noise data by a white noise detection method to obtain a data prediction value of a single variable on a time sequence; and if the white noise test is not passed, fitting the ARIMA model, and carrying out white noise detection again until the white noise test is passed.
Further, the LSTM model comprises an information filtering layer, an information storage layer and an information output layer.
Further, the prediction result of the health trend of the industrial equipment comprises the health state of the industrial equipment, which is divided into three types: a) serious faults including failure of equipment to work normally and shutdown; b) sending an alarm when the medium fault comprises the abnormal condition of the partial indexes of the equipment; c) and in a normal state, the equipment operates normally.
An ARIMA and LSTM algorithm based industrial equipment health trend analysis system, comprising:
the data acquisition module is responsible for acquiring the running state information data of the industrial equipment through the sensor;
the data preprocessing module is responsible for preprocessing the acquired data, removing noise data and ensuring the usability and accuracy of the data;
the characteristic selection module is responsible for carrying out characteristic selection on the preprocessed data;
the single variable prediction module is responsible for predicting the variation trend of a single feature in the features obtained by the feature selection module by utilizing the ARIMA model;
and the trend prediction module is responsible for inputting the prediction result of the single variable prediction module into the LSTM model as training data, training the LSTM model, and inputting the historical data of the industrial equipment into the LSTM model after training to obtain the prediction result of the health trend of the industrial equipment.
The key point of the invention is that the ARIMA model is applied to single variable data prediction and is simultaneously used as training data to be input into the LSTM model for trend analysis.
The invention has the following beneficial effects:
according to the invention, through an ARIMA algorithm, the data change trend at the next moment is deduced only through endogenous variables without the help of other exogenous variables, and then according to the predicted value, the prediction of a fault model is completed by using an LSTM algorithm in combination with historical data. The intelligent operation and maintenance support device is used as a support technology for the operation and maintenance of key components of intelligent plant assets, can realize the analysis and prediction of the state trend of equipment, realize the intellectualization of the operation and maintenance of the equipment, ensure the high-efficiency operation of the equipment, and reduce the maintenance cost and abnormal downtime caused by faults, thereby embodying the intellectualized operation and maintenance of the equipment.
Drawings
FIG. 1 is a flow chart of the overall steps of the method of the present invention.
FIG. 2 is a flow chart of predicting a trend of a single variable data.
FIG. 3 is a schematic structural diagram of the LSTM model.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, the present invention shall be described in further detail with reference to the following detailed description and accompanying drawings.
The main contents and objects of the present invention include: 1) estimating future values based on the single variable; 2) training a model based on time sequence data to predict the health degree of the equipment; 3) early warning of equipment faults is realized; 4) and the equipment fault downtime is reduced.
The complete technical scheme provided by the invention is shown in figure 1. The method comprises the steps of collecting data through a sensor, screening key features as data sources according to a feature selection method, training based on an ARIMA + LSTM algorithm to obtain a prediction model, and predicting the trend of the equipment state based on the prediction model.
1. Data acquisition
The running state information data of the equipment is acquired by using a three-dimensional acceleration sensor, a temperature sensor, an angular velocity sensor, a displacement sensor and the like, the running state information data mainly comprises X, Y, Z acceleration in three directions, motor temperature, an axis track, a torque signal and the like, the frequency of data acquisition of the sensors is unified, and the synchronism of data acquisition of the equipment is ensured.
2. Data pre-processing
According to the collected data, noise data are removed through various preprocessing methods such as data smoothing, standardization and missing value processing, and the usability and accuracy of the data are guaranteed.
3. Feature selection
The pre-processed data is subjected to correlation analysis, and a smaller number of uncorrelated variables are formed from the large data set. The maximum amount of variance is explained using the minimum number of principal components, avoiding multiple collinearity. And selecting the features according to the contribution degree of the data features to obtain key features, and abandoning non-key features.
The linear correlation relationship mainly adopts a Pearson correlation coefficient r to measure the linear correlation strength between continuous variables as the contribution degree of the characteristics. The method comprises the following specific steps:
0 ═ r ═ 0.3, low correlation;
0.3 ═ r ═ 0.8, moderately relevant;
0.8 ═ r ═ 1, highly correlated;
and further researching the correlation between the highly correlated variable and a Label (target value) variable aiming at the highly correlated variable, calculating a correlation coefficient and determining a characteristic variable.
4. Single variable prediction
And (3) predicting the variation trend of the single variable data (the single feature obtained after feature selection in the step (3)) by using an ARIMA model based on historical data. Even with vibration data, the data is guaranteed to be in the time dimension.
The specific process of using the ARIMA model to predict the variation trend of the single variable data is shown in fig. 2, and comprises the following steps:
1) data obtained by data preprocessing and feature selection are used as data sources, and stability inspection is carried out on data distribution;
2) if the data is unstable, the data needs to be processed in a differential operation mode until the data distribution shows a stable trend;
3) then, filtering noise data by a white noise detection method to obtain a data prediction value of a single variable on a time sequence; if the white noise test fails, the ARIMA model is fitted, the model parameters are reselected, and then white noise detection is performed again until the white noise test passes.
5. Trend prediction
And (4) inputting the selected characteristic data (namely the single variable prediction result obtained in the step (4)) into the LSTM model, and training model parameters to obtain the LSTM model after training. And inputting historical data of the equipment to be predicted into the LSTM model after training is completed, and obtaining the health evaluation score of the equipment at the current time so as to predict the service life.
The LSTM model, as shown in fig. 3, is mainly divided into three parts: the information filtering layer, the information storage layer and the information output layer select useful information of the data through the model. In fig. 3, "a" denotes a neural network node, and σ denotes a sigmoid function.
An information filtering layer: decided by Sigmoid layer, it looks at ht-1(previous output) and xt(currently entered) and is cell state Ct-1Each digit in (last state) outputs a number between 0 and 1. 1 represents a complete reservation and 0 represents a complete deletion.
An information storage layer: it is first decided by the Sigmoid layer called the "input gate layer" which values are to be updated. The next tanh layer creates a candidate vector Ct which will be added to the state of the cell.
An information output layer: a Sigmoid layer is run that determines which parts of the cell state to export. The cell state is then passed through tanh (normalizing the value to between-1 and 1) and multiplied by the Sigmoid gate output, to which valid data information screened by the model is output.
Training process:
1. the output value of each neuron is calculated forward.
2. The error term value for each neuron is calculated in reverse. The back propagation of the LSTM error term also includes two directions: one is the backward propagation along the time, namely, the error term of each moment is calculated from the current t moment; one is to propagate the error term up one layer.
3. The gradient of each weight is calculated according to the corresponding error term.
And (3) actual prediction process:
the model construction part mainly comprises the construction of an input layer, an LSTM layer, an output layer, loss, an optimizer and the like:
1. input layer
In the data pre-processing stage, a mini-batch segmentation function is defined, with the size of the input layer depending on the size at which the batch is set.
Layer of LSTM
The LSTM layer is a key part of the whole neural network, and after the LSTMCell is constructed by using the basic module BasiclSSTMCell, in order to prevent overfitting, dropout regulation is added to a hidden layer of the LSTM layer.
3. The output layer uses softmax, which is fully connected to the LSTM.
4. To one-hot encode targets (target values), we use softmax _ cross _ entropy _ with _ locations cross entropy to calculate loss.
5. And (4) model training, namely setting parameters according to experience and performing iterative training.
The parameters are as follows:
batch size number of sequences in a Single batch
num steps number of characters in a single sequence
lstm _ size, number of hidden layer nodes
num _ layers: LSTM layer number
learning rate
keep _ prob-proportion of nodes remaining in dropout layer during training
6. The specific application is as follows:
the health condition of the equipment is predicted through historical data. The health status is mainly divided into three types: major failure (equipment can not work normally, shut down); moderate fault (abnormal condition exists in the index of the equipment part, and alarm is given); and (3) normal state: the equipment operates normally.
The general idea is as follows: and collecting operation data of the equipment, predicting a single variable by using the ARIMA model, inputting the single variable serving as training data into the LSTM time sequence model, and completing the prediction of a result.
For example, industrial robot bearings are parts that are prone to wear, and failure of a robot bearing can result in loss of downtime, and therefore failure detection of equipment is of paramount importance. By adopting the method, the vibration data of the equipment is acquired through the sensor, and because the vibration data of the equipment is time sequence data, the vibration data of the equipment is slightly influenced by other factors, under the condition of not changing a time domain, the trend of the data is mined through an ARIMA model, the vibration value in a fixed time period in the future is predicted, and the vibration data is used as a data source to diagnose the fault of the bearing and predict the health state by combining other characteristics.
Based on the same inventive concept, another embodiment of the present invention provides an ARIMA and LSTM algorithm-based industrial equipment health trend analysis system, which includes:
the data acquisition module is responsible for acquiring the running state information data of the industrial equipment through the sensor;
the data preprocessing module is responsible for preprocessing the acquired data, removing noise data and ensuring the usability and accuracy of the data;
the characteristic selection module is responsible for carrying out characteristic selection on the preprocessed data;
the single variable prediction module is responsible for predicting the variation trend of a single feature in the features obtained by the feature selection module by utilizing the ARIMA model;
and the trend prediction module is responsible for inputting the prediction result of the single variable prediction module into the LSTM model as training data, training the LSTM model, and inputting the historical data of the industrial equipment into the LSTM model after training to obtain the prediction result of the health trend of the industrial equipment.
The specific operation of each module is described in the foregoing description of the method of the present invention.
The above embodiments are only intended to illustrate the technical solution of the present invention and not to limit the same, and a person skilled in the art can modify the technical solution of the present invention or substitute the same without departing from the principle and scope of the present invention, and the scope of the present invention should be determined by the claims.
Claims (10)
1. An industrial equipment health trend analysis method based on ARIMA and LSTM algorithms is characterized by comprising the following steps:
1) acquiring running state information data of industrial equipment through a sensor;
2) preprocessing the acquired data, removing noise data and ensuring the usability and accuracy of the data;
3) performing feature selection on the preprocessed data;
4) predicting the variation trend of a single feature in the features obtained in the step 3) by utilizing an ARIMA model;
5) inputting the prediction result of the step 4) into the LSTM model as training data, and training the LSTM model;
6) and inputting historical data of the industrial equipment into the trained LSTM model to obtain a prediction result of the health trend of the industrial equipment.
2. The method as claimed in claim 1, wherein step 1) uses three-dimensional acceleration sensors, temperature sensors, angular velocity sensors and displacement sensors to obtain information data of the operating state of the industrial equipment, including X, Y, Z acceleration, motor temperature, axial locus and torque signals in three directions, and unifies the frequency of data collection by the sensors to ensure the synchronism of data collection of the equipment.
3. The method of claim 1, wherein the preprocessing of step 2) comprises data smoothing, normalization, missing value processing.
4. The method of claim 1, wherein step 3) performs feature selection according to the contribution degree of the features, wherein the contribution degree of the features is the linear correlation strength between continuous variables measured by a Pearson correlation coefficient r, and the method comprises the following steps: 0 ═ r ═ 0.3, low correlation; 0.3 ═ r ═ 0.8, moderately relevant; 0.8 ═ r | < ═ 1, highly correlated.
5. The method of claim 1, wherein step 4) comprises:
a) data obtained by data preprocessing and feature selection are used as data sources, and stability inspection is carried out on data distribution;
b) if the data is unstable, processing the data in a differential operation mode until the data distribution shows a stable trend;
c) filtering noise data by a white noise detection method to obtain a data prediction value of a single variable on a time sequence;
and if the white noise test is not passed, fitting the ARIMA model, and carrying out white noise detection again until the white noise test is passed.
6. The method of claim 1, wherein the LSTM model comprises an information filtering layer, an information storage layer, and an information output layer.
7. The method of claim 1, wherein the predicted outcome of the health trend of the industrial equipment comprises a health status of the industrial equipment, and is classified into three types: a) serious faults including failure of equipment to work normally and shutdown; b) sending an alarm when the medium fault comprises the abnormal condition of the partial indexes of the equipment; c) and in a normal state, the equipment operates normally.
8. An industrial equipment health trend analysis system based on ARIMA and LSTM algorithms, comprising:
the data acquisition module is responsible for acquiring the running state information data of the industrial equipment through the sensor;
the data preprocessing module is responsible for preprocessing the acquired data, removing noise data and ensuring the usability and accuracy of the data;
the characteristic selection module is responsible for carrying out characteristic selection on the preprocessed data;
the single variable prediction module is responsible for predicting the variation trend of a single feature in the features obtained by the feature selection module by utilizing the ARIMA model;
and the trend prediction module is responsible for inputting the prediction result of the single variable prediction module into the LSTM model as training data, training the LSTM model, and inputting the historical data of the industrial equipment into the LSTM model after training to obtain the prediction result of the health trend of the industrial equipment.
9. The system of claim 8, wherein the data acquisition module acquires running state information data of the industrial equipment by using a three-dimensional acceleration sensor, a temperature sensor, an angular velocity sensor and a displacement sensor, wherein the running state information data comprises X, Y, Z acceleration signals in three directions, motor temperature signals, shaft center track signals and torque signals, and the frequency of data acquisition by the sensors is unified, so that the synchronism of equipment data acquisition is ensured.
10. The system of claim 8, wherein the single variable prediction module:
a) data obtained by data preprocessing and feature selection are used as data sources, and stability inspection is carried out on data distribution;
b) if the data is unstable, processing the data in a differential operation mode until the data distribution shows a stable trend;
c) filtering noise data by a white noise detection method to obtain a data prediction value of a single variable on a time sequence;
and if the white noise test is not passed, fitting the ARIMA model, and carrying out white noise detection again until the white noise test is passed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911292039.4A CN111178553A (en) | 2019-12-16 | 2019-12-16 | Industrial equipment health trend analysis method and system based on ARIMA and LSTM algorithms |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911292039.4A CN111178553A (en) | 2019-12-16 | 2019-12-16 | Industrial equipment health trend analysis method and system based on ARIMA and LSTM algorithms |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111178553A true CN111178553A (en) | 2020-05-19 |
Family
ID=70652032
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911292039.4A Pending CN111178553A (en) | 2019-12-16 | 2019-12-16 | Industrial equipment health trend analysis method and system based on ARIMA and LSTM algorithms |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111178553A (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111638672A (en) * | 2020-06-05 | 2020-09-08 | 昆山艾控智能科技有限公司 | Automatic control system of industrial machine table |
| CN111881298A (en) * | 2020-08-04 | 2020-11-03 | 上海交通大学 | Semi-structured text processing and analyzing method |
| CN112270449A (en) * | 2020-11-04 | 2021-01-26 | 国网(苏州)城市能源研究院有限责任公司 | Industrial system time delay determination and controlled quantity prediction method based on time correlation |
| CN112329341A (en) * | 2020-11-02 | 2021-02-05 | 浙江智昌机器人科技有限公司 | Fault diagnosis system and method based on AR and random forest model |
| CN112417701A (en) * | 2020-12-01 | 2021-02-26 | 江苏南高智能装备创新中心有限公司 | Numerical control machine tool residual life prediction method and device based on autoregressive summation moving average model and network side server |
| CN112418277A (en) * | 2020-11-03 | 2021-02-26 | 西安电子科技大学 | Method, system, medium, and apparatus for predicting remaining life of rotating machine component |
| CN113435725A (en) * | 2021-06-21 | 2021-09-24 | 国网宁夏电力有限公司信息通信公司 | Power grid host dynamic threshold setting method based on FARIMA-LSTM prediction |
| CN113867262A (en) * | 2020-06-30 | 2021-12-31 | 华晨宝马汽车有限公司 | Apparatus, method and medium for monitoring operation state of spindle device |
| CN114065651A (en) * | 2021-11-30 | 2022-02-18 | 重庆忽米网络科技有限公司 | Fault time prediction method for rotary equipment |
| CN115096627A (en) * | 2022-06-16 | 2022-09-23 | 中南大学 | Method and system for fault diagnosis and operation and maintenance in manufacturing process of hydraulic forming intelligent equipment |
| CN115307896A (en) * | 2022-09-20 | 2022-11-08 | 重庆忽米网络科技有限公司 | Equipment health state detection method based on machine learning |
| CN115893167A (en) * | 2022-11-16 | 2023-04-04 | 青岛城市轨道交通科技有限公司 | Escalator vibration trend prediction method based on big data |
| CN117688464A (en) * | 2024-02-04 | 2024-03-12 | 国网上海市电力公司 | Hidden danger analysis method and system based on multi-source sensor data |
| CN117688464B (en) * | 2024-02-04 | 2024-04-19 | 国网上海市电力公司 | Hidden danger analysis method and system based on multi-source sensor data |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108181556A (en) * | 2017-12-18 | 2018-06-19 | 国网浙江省电力有限公司检修分公司 | Porcelain insulator zero value detection method based on chapeau de fer temperature difference time series analysis |
| CN109472110A (en) * | 2018-11-29 | 2019-03-15 | 南京航空航天大学 | A kind of aero-engine remaining life prediction technique based on LSTM network and ARIMA model |
-
2019
- 2019-12-16 CN CN201911292039.4A patent/CN111178553A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108181556A (en) * | 2017-12-18 | 2018-06-19 | 国网浙江省电力有限公司检修分公司 | Porcelain insulator zero value detection method based on chapeau de fer temperature difference time series analysis |
| CN109472110A (en) * | 2018-11-29 | 2019-03-15 | 南京航空航天大学 | A kind of aero-engine remaining life prediction technique based on LSTM network and ARIMA model |
Non-Patent Citations (1)
| Title |
|---|
| 钟令枢: "基于ARIMA-LSTM的架空线状态数据挖掘", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 * |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111638672A (en) * | 2020-06-05 | 2020-09-08 | 昆山艾控智能科技有限公司 | Automatic control system of industrial machine table |
| CN111638672B (en) * | 2020-06-05 | 2024-01-16 | 昆山艾控智能科技有限公司 | Automatic control system of industrial machine |
| CN113867262A (en) * | 2020-06-30 | 2021-12-31 | 华晨宝马汽车有限公司 | Apparatus, method and medium for monitoring operation state of spindle device |
| CN111881298A (en) * | 2020-08-04 | 2020-11-03 | 上海交通大学 | Semi-structured text processing and analyzing method |
| CN112329341A (en) * | 2020-11-02 | 2021-02-05 | 浙江智昌机器人科技有限公司 | Fault diagnosis system and method based on AR and random forest model |
| CN112418277A (en) * | 2020-11-03 | 2021-02-26 | 西安电子科技大学 | Method, system, medium, and apparatus for predicting remaining life of rotating machine component |
| CN112418277B (en) * | 2020-11-03 | 2024-04-09 | 西安电子科技大学 | Method, system, medium and equipment for predicting residual life of rotating machine parts |
| CN112270449B (en) * | 2020-11-04 | 2023-12-26 | 国网(苏州)城市能源研究院有限责任公司 | Industrial system time delay determination and controlled quantity prediction method based on time correlation |
| CN112270449A (en) * | 2020-11-04 | 2021-01-26 | 国网(苏州)城市能源研究院有限责任公司 | Industrial system time delay determination and controlled quantity prediction method based on time correlation |
| CN112417701A (en) * | 2020-12-01 | 2021-02-26 | 江苏南高智能装备创新中心有限公司 | Numerical control machine tool residual life prediction method and device based on autoregressive summation moving average model and network side server |
| CN112417701B (en) * | 2020-12-01 | 2024-01-23 | 江苏南高智能装备创新中心有限公司 | Method and device for predicting residual life of numerical control machine tool and network side server |
| CN113435725A (en) * | 2021-06-21 | 2021-09-24 | 国网宁夏电力有限公司信息通信公司 | Power grid host dynamic threshold setting method based on FARIMA-LSTM prediction |
| CN114065651A (en) * | 2021-11-30 | 2022-02-18 | 重庆忽米网络科技有限公司 | Fault time prediction method for rotary equipment |
| CN115096627B (en) * | 2022-06-16 | 2023-04-07 | 中南大学 | Method and system for fault diagnosis and operation and maintenance in manufacturing process of hydraulic forming intelligent equipment |
| CN115096627A (en) * | 2022-06-16 | 2022-09-23 | 中南大学 | Method and system for fault diagnosis and operation and maintenance in manufacturing process of hydraulic forming intelligent equipment |
| CN115307896A (en) * | 2022-09-20 | 2022-11-08 | 重庆忽米网络科技有限公司 | Equipment health state detection method based on machine learning |
| CN115893167A (en) * | 2022-11-16 | 2023-04-04 | 青岛城市轨道交通科技有限公司 | Escalator vibration trend prediction method based on big data |
| CN117688464A (en) * | 2024-02-04 | 2024-03-12 | 国网上海市电力公司 | Hidden danger analysis method and system based on multi-source sensor data |
| CN117688464B (en) * | 2024-02-04 | 2024-04-19 | 国网上海市电力公司 | Hidden danger analysis method and system based on multi-source sensor data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111178553A (en) | Industrial equipment health trend analysis method and system based on ARIMA and LSTM algorithms | |
| Cai et al. | Artificial intelligence enhanced two-stage hybrid fault prognosis methodology of PMSM | |
| US20180003588A1 (en) | Machine learning device which learns estimated lifetime of bearing, lifetime estimation device, and machine learning method | |
| CN110763997A (en) | Early fault early warning method for synchronous motor stator | |
| CN110757510A (en) | Method and system for predicting remaining life of robot | |
| CN108961460B (en) | Fault prediction method and device based on sparse ESGP (Enterprise service gateway) and multi-objective optimization | |
| Zurita et al. | Distributed neuro-fuzzy feature forecasting approach for condition monitoring | |
| CN116305614A (en) | Rolling bearing residual service life prediction method based on residual convolution neural network and LSTM | |
| CN115081585A (en) | Man-machine-object cooperation abnormal state detection method for reinforcing heterogeneous graph neural network | |
| KR102474332B1 (en) | Prediction method of machine health stability in smart factory system, and recording medium thereof | |
| Liu et al. | A hybrid LSSVR-HMM based prognostics approach | |
| CN116020879B (en) | Technological parameter-oriented strip steel hot continuous rolling space-time multi-scale process monitoring method and device | |
| CN117273440A (en) | Engineering construction Internet of things monitoring and managing system and method based on deep learning | |
| CN117332347A (en) | Intelligent operation and maintenance method and system for power grid by integrating machine patrol and AI | |
| Salah et al. | Echo state network and particle swarm optimization for prognostics of a complex system | |
| CN114486262B (en) | CNN-AT-LSTM-based bearing residual service life prediction method | |
| Ak et al. | Ensemble neural network model for predicting the energy consumption of a milling machine | |
| Fernandez-Anakabe et al. | An Attribute Oriented Induction based Methodology for Data Driven Predictive Maintenance | |
| Martins et al. | Prediction maintenance based on vibration analysis and deep learning–A case study of a drying press supported on a hidden Markov model | |
| CN114299338A (en) | Fault prediction and health management system for system management and related method | |
| Wang et al. | Predictive maintenance and sensitivity analysis for equipment with multiple quality states | |
| Banerjee et al. | Application of IoT in industry 4.0 for predictive analytics | |
| El Jamal et al. | Similarity-based brownian motion approach for remaining useful life prediction | |
| Zhang et al. | A Novel Quality-Related Distributed Fault Diagnosis Framework for Large-Scale Sequential Manufacturing Processes | |
| CN115047313B (en) | ZPW-2000R orbit circuit fault diagnosis method and device based on OC-SVM and DNN |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200519 |