CN111210092A - Stacking machine predictive maintenance method and system based on deep learning - Google Patents
Stacking machine predictive maintenance method and system based on deep learning Download PDFInfo
- Publication number
- CN111210092A CN111210092A CN202010143230.9A CN202010143230A CN111210092A CN 111210092 A CN111210092 A CN 111210092A CN 202010143230 A CN202010143230 A CN 202010143230A CN 111210092 A CN111210092 A CN 111210092A
- Authority
- CN
- China
- Prior art keywords
- stacker
- fault
- basic model
- data
- failure
- 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
- G06Q10/04—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
- 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
- G06N3/045—Combinations of networks
-
- 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
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Health & Medical Sciences (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Debugging And Monitoring (AREA)
Abstract
The invention discloses a stacker predictive maintenance method and a stacker predictive maintenance system based on deep learning, wherein the method comprises the following steps: collecting historical fault data of the stacker, wherein the historical fault data comprises a fault state parameter data set in a period of time before a fault occurs; establishing a basic model for analyzing the fault rate of the stacker according to historical fault data; collecting the state parameters of the stacker during working in real time, inputting the state parameters into a basic model, analyzing the state parameters of the stacker collected in real time based on the basic model, and outputting the failure rate value of the stacker; and when the failure rate value of the stacker exceeds a set threshold value, a failure alarm is sent out. The method is based on a basic model of the failure rate of the stacker, analyzes and processes the state parameters of the stacker during working and outputs the failure rate value of the stacker; whether the stacker breaks down or not is judged through the fault rate value so as to be maintained when the stacker breaks down in the first time, and therefore economic loss and potential safety hazard are reduced.
Description
Technical Field
The invention relates to the technical field of computers, in particular to a stacker predictive maintenance method and a stacker predictive maintenance system based on deep learning.
Background
The stacker is a commodity circulation storage equipment commonly used, and the wide application is got in the goods of various elevated vertical storehouses and is got and put, and it is comparatively central key equipment in the intelligent storage system, so also very important to the maintenance of stacker, need ensure that the stacker can maintain the very first time when breaking down, reduce the stacker sudden failure and the economic loss on the production line that leads to, even the potential safety hazard.
Most of the stacker equipment spare parts are expensive, the delivery cycle is long, and a lot of difficulties are brought to stacker maintenance. However, in the existing situation, since the stacker works frequently and is widely applied to an unmanned reservoir area, it is difficult for maintenance personnel to find a fault at the first time, and thus a great economic loss and even a potential safety hazard may be caused.
Disclosure of Invention
The invention aims to provide a stacker predictive maintenance method based on deep learning and a system thereof, aiming at solving the problems of economic loss and potential safety hazard caused by the fact that maintenance personnel are difficult to find and maintain faults at the first time when the existing stacker breaks down in work.
In a first aspect, an embodiment of the present invention provides a deep learning-based stacker predictive maintenance method, which includes:
collecting historical fault data of the stacker, wherein the historical fault data comprises a fault state parameter data set in a period of time before a fault occurs;
establishing a basic model for analyzing the fault rate of the stacker according to the historical fault data;
collecting the state parameters of the stacker during working in real time, inputting the state parameters into the basic model, analyzing the state parameters of the stacker collected in real time based on the basic model, and outputting the failure rate value of the stacker;
and when the failure rate value of the stacker exceeds a set threshold value, a failure alarm is sent out.
In a second aspect, an embodiment of the present invention further provides a deep learning-based stacker predictive maintenance system, which includes:
the system comprises an edge data processing terminal arranged on the stacker and a cloud server in communication connection with the edge data processing terminal;
the edge data processing terminal is used for collecting historical fault data of the stacker, wherein the historical fault data comprise a fault state parameter data set in a period of time before a fault occurs and are sent to the cloud server;
the cloud server is used for establishing a basic model for analyzing the failure rate of the stacker according to the historical failure data and sending the basic model to the edge data processing terminal;
the edge data processing terminal is also used for acquiring the state parameters of the stacker during working in real time, inputting the state parameters into the basic model, analyzing the state parameters of the stacker acquired in real time based on the basic model, and outputting the failure rate value of the stacker; and when the failure rate value of the stacker exceeds a set threshold value, a failure alarm is sent out.
The embodiment of the invention provides a stacker predictive maintenance method and a system thereof based on deep learning, wherein the stacker predictive maintenance method comprises the following steps: collecting historical fault data of the stacker, wherein the historical fault data comprises a fault state parameter data set in a period of time before a fault occurs; establishing a basic model for analyzing the fault rate of the stacker according to the historical fault data; collecting the state parameters of the stacker during working in real time, inputting the state parameters into the basic model, analyzing the state parameters of the stacker collected in real time based on the basic model, and outputting the failure rate value of the stacker; and when the failure rate value of the stacker exceeds a set threshold value, a failure alarm is sent out. The method is based on a basic model of the failure rate of the stacker, analyzes and processes the state parameters of the stacker during working and outputs the failure rate value of the stacker; whether the stacker breaks down or not is judged through the fault rate value so as to be maintained when the stacker breaks down in the first time, and therefore economic loss and potential safety hazard are reduced.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a schematic flowchart of a deep learning-based stacker predictive maintenance method according to an embodiment of the present invention;
fig. 2 is a sub-flow diagram of a deep learning-based stacker predictive maintenance method according to an embodiment of the present invention;
fig. 3 is another sub-flow diagram of the deep learning-based stacker predictive maintenance method according to the embodiment of the present invention;
fig. 4 is another sub-flow diagram of the deep learning-based stacker predictive maintenance method according to the embodiment of the present invention;
fig. 5 is another schematic flow chart of a deep learning-based stacker predictive maintenance method according to an embodiment of the present invention;
fig. 6 is a schematic block diagram of a deep learning-based stacker predictive maintenance system according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It will be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It is also to be understood that the terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in the specification of the present invention and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
It should be further understood that the term "and/or" as used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.
Referring to fig. 1, fig. 1 is a flowchart of a deep learning-based stacker predictive maintenance method according to an embodiment of the present invention.
As shown in fig. 1, the method includes steps S101 to S104.
S101, collecting historical fault data of the stacker, wherein the historical fault data comprise fault state parameter data groups in a period of time before a fault occurs.
In this embodiment, historical failure data of the stacker is collected, that is, a failure state parameter data set before the stacker fails is collected, and by analyzing the state parameters in this period, the change rule of the state parameters of the stacker before the stacker fails can be analyzed, so as to help predict in advance whether the stacker will fail, and to facilitate timely maintenance. The fault status parameter data set includes a current parameter, a voltage parameter, a temperature parameter, a vibration parameter, and the like.
In this embodiment, the historical fault data may be collected on the premise that it is clear that the stacker has failed. For example, when the voltage of the stacker increases, the current becomes zero, and the vibration becomes zero, or when the voltage becomes zero, the current increases, and the vibration becomes zero, the stacker itself or a worker may determine that the stacker has a fault. Before the stacker breaks down, the state parameters of the stacker can change regularly, and the signal acquisition module on the stacker continuously acquires the state parameters of the stacker, so that the embodiment can collect the state parameters of the stacker before the stacker breaks down, which are acquired by the signal acquisition module, form a fault state parameter data set, and provide data support for subsequent modeling.
And S102, establishing a basic model for analyzing the fault rate of the stacker according to the historical fault data.
In this embodiment, the basic model is established by deep learning of data by a neural network. The motivation of deep learning lies in establishing and simulating a neural network for analyzing and learning the human brain, and the neural network simulates the mechanism of the human brain to explain data; and taking the historical fault data as a standard for judging the fault of the stacker, and performing deep learning on the historical fault data so as to establish a basic model for analyzing the fault rate of the stacker.
In one embodiment, as shown in fig. 2, step S102 includes:
s201, preprocessing the historical fault data to obtain analysis data;
s202, importing the analysis data into a three-layer neural network based on open source framework Tensflow, and generating a basic model for analyzing the failure rate of the stacker.
In this embodiment, the collected historical fault data may have the problems of attribute mismatching, data incompleteness, transmission errors and the like, unprocessed original data is directly imported into the three-layer neural network, and accuracy and efficiency of subsequent fault rate analysis of the basic model may be affected, so the collected historical fault data needs to be preprocessed, and the preprocessing includes data cleaning, data integration, data conversion, data reduction and the like on the historical fault data, so that the historical fault data is converted into analysis data matched with the three-layer neural network analysis, and then the analysis data is imported into the three-layer neural network based on the open source framework Tensorflow, and the basic model for analyzing the fault rate of the stacker is generated through deep learning training.
In the open source framework Tensorflow-based three-layer neural network, the open source framework Tensorflow is a programming system and is used for building a three-layer neural network, and the three-layer neural network comprises:
a layer of neural network for receiving the analysis data; whenever a problem of machine learning is to be solved, the data set needs to be considered first, because the data set is the basis of all learning.
A two-layer neural network for training data; namely, the two-layer neural network is used for deep learning training analysis data.
A three-layer neural network for establishing a model; and after data are deeply learned and analyzed, generating a basic model for analyzing the fault rate of the stacker by using the three-layer neural network.
In one embodiment, as shown in fig. 3, step S202 includes:
s301, when the data volume of the analysis data reaches a limit value, importing the analysis data into a three-layer neural network based on open source framework Tensorflow;
s302, setting parameters of the three-layer neural network;
and S303, generating basic models of various fault problems, and numbering the basic models.
In this embodiment, the limit value is set to collect the analysis data of the stacker more comprehensively, and the modeling is based on a large amount of stacker fault information, so that the data acquisition needs to be kept for a certain time, the specific time is influenced by the base number of the stacker, and if the base number of the stacker is larger, the speed of various fault problems is higher, and the time spent is shorter; otherwise the longer it takes. When the data volume of the required analysis data reaches a limit value, the data volume can be guided into the three-layer neural network based on open source framework Tensflow, the three-layer neural network is subjected to parameter setting, the three-layer neural network outputs basic models of various fault problems, the basic models of each fault problem are numbered, corresponding prompt codes are set, and maintenance personnel can know the type of the fault problem in time conveniently.
S103, collecting the state parameters of the stacker during working in real time, inputting the state parameters into the basic model, analyzing the state parameters of the stacker collected in real time based on the basic model, and outputting the failure rate value of the stacker.
The action of acquiring the parameter state in real time is continued as long as the stacker is in a working state, and the embodiment inputs the state parameters acquired in real time into the basic model, and analyzes and outputs the fault rate value in real time based on the basic model. The fault rate value is a similarity degree ratio, i.e. a similarity degree value, between the input state parameter and a state parameter in the base model (which may be a standard state parameter obtained by the base model based on a large number of similar fault problems).
In one embodiment, as shown in fig. 4, step S103 includes:
s401, collecting the state parameters of the stacker during working in real time, inputting the state parameters into each basic model, and outputting a plurality of similarity values;
s402, sequencing the similarity values to obtain the maximum similarity value, and taking the maximum similarity value as a stacker fault rate value.
In this embodiment, each of the basic models corresponds to one or the same type of fault problem, the state parameters acquired in real time are input into each basic model, and the basic models can analyze and output corresponding similarity values in real time, so that a plurality of similarity values can be obtained through the plurality of basic models, and then the largest similarity value is selected as the fault rate value of the stacker. Of course, before the state parameters are input to each basic model, the state parameters may be preprocessed to obtain data to be predicted, and then the data to be predicted is input to each basic model for subsequent processing.
And S104, when the failure rate value of the stacker exceeds a set threshold value, sending out a failure alarm.
And sending out a fault alarm if the fault rate value exceeds a set threshold value, and not triggering the fault alarm if the fault rate value does not exceed the set threshold value, wherein the threshold value can be set according to the working condition of the stacker, namely the probability of the stacker failing after the threshold value is exceeded is higher, and the probability of the stacker failing when the fault rate value does not exceed the threshold value is lower.
In a specific application scenario, a plurality of threshold values can be set to judge the size of the fault rate value, the fault rate value is compared with each threshold value, and the fault probability of the stacker is determined according to the comparison result.
The thresholds may include a high threshold, a medium threshold, and a low threshold. And if the fault rate value exceeds the high threshold value, a red alarm is sent out, the indicator lamp of the stacker is red, if the fault rate value exceeds the medium threshold value and is lower than the high threshold value, a yellow alarm is sent out, the indicator lamp of the stacker is yellow, if the fault rate value exceeds the low threshold value and is lower than the medium threshold value, a blue alarm is sent out, and the indicator lamp of the stacker is blue.
Specifically, several percentage values may be preset, and then the percentage values respectively correspond to a threshold value, for example, if one percentage value is 80%, then it corresponds to a high threshold value, if one percentage value is 50%, then it corresponds to a medium threshold value of the high threshold value, and if one percentage value is 20%, then it corresponds to a low threshold value. If the output failure rate value is 60%, it exceeds the medium threshold and is below the high threshold, a yellow alarm is issued and the stacker indicator light is yellow. Therefore, the fault rate value can be intuitively fed back to maintenance personnel, and the maintenance personnel nearby can find the corresponding stacker according to the color of the indicator light and perform maintenance.
As shown in fig. 5, the method further includes steps S501 to S502.
S501, when the fault rate value of the stacker does not exceed a set threshold value and the stacker judges that a fault occurs, collecting fault data of the current fault of the stacker, wherein the fault data of the current fault comprises a fault state parameter data set in a period of time before the fault occurs;
s502, inputting the fault data of the fault into the basic model, training and updating the basic model to obtain an updated basic model.
When the failure rate value does not exceed the set threshold value, the method of this embodiment determines that the stacker is not failed, but in an actual situation, the stacker may still be failed (if the stacker is determined to be failed by a worker or determined to be failed by the stacker itself), at this time, failure data that is failed this time may be collected, preprocessing is formed to obtain analysis data, and the analysis data is input to the basic model to train and update the basic model, where, of course, an existing basic model may be updated, or a new basic model may be trained and updated. The reason is that the original basic model may have low prediction accuracy or a new fault problem, so that the fault value of the stacker does not exceed the set threshold but fails. Therefore, the original basic model is continuously updated by adding a continuous learning and updating process, so that the system can continuously and deeply learn, the data of the system is perfected, and the judgment capability of the system is improved.
Referring to fig. 6, fig. 6 is a schematic block diagram of a deep learning based stacker predictive maintenance system according to an embodiment of the present invention.
The embodiment of the invention provides a stacker predictive maintenance system based on deep learning, which comprises: an edge data processing terminal 611 arranged on the stacker 61 and a cloud server 62 in communication connection with the edge data processing terminal 611; the edge data processing terminal 611 is configured to collect historical fault data of the stacker 61, where the historical fault data includes a fault state parameter data set in a period of time before a fault occurs, and send the historical fault data to the cloud server 62; the cloud server 62 is configured to establish a basic model for analyzing the failure rate of the stacker according to the historical failure data, and send the basic model to the edge data processing terminal 611; the edge data processing terminal 611 is further configured to acquire, in real time, state parameters of the stacker 61 during operation, input the state parameters into the basic model, analyze, based on the basic model, the state parameters of the stacker acquired in real time, and output a stacker fault value; and when the failure rate value of the stacker exceeds a set threshold value, a failure alarm is sent out.
In this embodiment, each stacker 61 may be provided with the edge data processing terminal 611, and the cloud server 62 may be connected to a plurality of edge data processing terminals 611, and monitor each stacker 61 at the same time, which is convenient for management.
The edge data processing terminal 611 collects historical fault data of the pilers 61, performs preprocessing on the historical fault data to obtain analysis data, sends the analysis data to the cloud server 62, and the cloud server 62 performs deep learning training on the analysis data, establishes a basic model for analyzing the fault rate of the pilers, and sends the basic model to the edge data processing terminal 611 of each piler 61.
The edge data processing terminal 611 collects the state parameters of the stacker 61 during working in real time and inputs the state parameters into the basic model, then analyzes the state parameters based on the basic model and outputs a stacker fault rate value, and if the fault rate value exceeds a threshold value, a fault alarm is sent out.
According to the embodiment of the invention, the fault of the stacker 61 can be found before the stacker 61 breaks down, and the maintenance personnel is informed to carry out maintenance, so that inconvenience caused by the stacker 61 after the fault is avoided, the potential safety hazard in a circuit can be avoided during early maintenance, and the investment cost is reduced.
In an embodiment, the edge data processing terminal 611 is further configured to send the fault alarm to the cloud server 62, and the cloud server 62 is further configured to notify a worker to perform maintenance through the mobile terminal 63.
In this embodiment, the fault alarm is sent to the cloud server 62 in real time, and the maintenance personnel acquire the fault information from the cloud server 62 through the mobile terminal 63, so that fault maintenance can be performed at the first time or possible faults can be overhauled in advance, and therefore economic loss and potential safety hazards are reduced. The embodiment of the invention has obvious effect and function and high timeliness for sudden failure alarm and early warning of the stacker.
The mobile terminal 63 may be a mobile phone, a tablet, or other paging device, and is connected to the cloud server through a wireless network.
In an embodiment, the edge data processing terminal 611 includes: a signal acquisition module 6111 for acquiring status parameters when the stacker works, a main controller 6112 for receiving the status parameters, a communication module 6113 for communicating with a cloud server, and a power supply module 6114 for providing power.
In this embodiment, the state parameter may reflect a working state of the stacker 61, and is an important parameter for determining whether the stacker 61 fails or has a failure, the signal acquisition module 6111 acquires the working state parameter of the stacker 61 in real time, and sends the state parameter to the main controller 6112, the main controller 6112 receives the state parameter and inputs the state parameter into the basic model, compares the input state parameter with the state parameter in the basic model, and outputs a failure rate value, and when the failure rate value exceeds a threshold value, the main controller 6112 sends a failure alarm to the cloud server 62 through the communication module 6113.
The power module 6114 includes a rectifier bridge and an LDC voltage stabilization chip, the rectifier bridge is used to rectify the ac power into dc power, and the LDC voltage stabilization chip can transform 220V of living voltage to 5V, which is the working voltage of the main controller.
In an embodiment, the signal collecting module 6111 includes a current sensor, a voltage sensor, a temperature sensor and a vibration sensor, which are respectively used for collecting a current parameter, a voltage parameter, a temperature parameter and a vibration parameter in the working state of the stacker 61.
In this embodiment, the state parameters include a current parameter, a voltage parameter, a temperature parameter, and a vibration parameter, and the current sensor, the voltage sensor, the temperature sensor, and the vibration sensor are respectively used to acquire the current parameter, the voltage parameter, the temperature parameter, and the vibration parameter in the working state of the stacker crane. Of course, the state parameters are not limited to the current parameters, the voltage parameters, the temperature parameters and the vibration parameters, and the state parameters may be included in the state parameters as long as the parameters are changed to cause the stacker crane to malfunction or to possibly malfunction.
The fault state parameters such as the phenomena of voltage increase, current change to zero, vibration to zero and the like, or the conditions of voltage change to zero, current increase, vibration to zero and the like can all judge that the stacker has faults.
In one embodiment, the cloud server 62 sends a fault alarm to the mobile terminals 63 in the predetermined area. The cloud server 62 monitors each stacker 61 in real time, when a stacker 61 fails, the position of the stacker 61 is located, and a failure alarm is sent to the mobile terminal 63 in a preset area, so that the stacker 61 can be maintained timely. The preset area may be an area within a certain range of the position of the stacker 61, for example, the stacker 61 is an area within 20 km of a square circle at the center, so that nearby personnel can be notified to go to the place of affairs for maintenance. Of course, the preset area may also be an administrative area or a user-defined area that is preset to be responsible for the maintenance of the stacker, for example, a certain stationer in a city is responsible for maintaining the stacker uniformly, so that the stacker can be maintained in time. Of course, the mobile terminals 63 in the two preset areas may be notified at the same time, so that even if the maintenance person in charge of the preset area is not near the maintenance location when receiving the alarm information, the maintenance person can seek help from the nearby maintenance person.
While the invention has been described with reference to specific embodiments, the invention is not limited thereto, and various equivalent modifications and substitutions can be easily made by those skilled in the art within the technical scope of the invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (10)
1. A stacker predictive maintenance method based on deep learning is characterized by comprising the following steps:
collecting historical fault data of the stacker, wherein the historical fault data comprises a fault state parameter data set in a period of time before a fault occurs;
establishing a basic model for analyzing the fault rate of the stacker according to the historical fault data;
collecting the state parameters of the stacker during working in real time, inputting the state parameters into the basic model, analyzing the state parameters of the stacker collected in real time based on the basic model, and outputting the failure rate value of the stacker;
and when the failure rate value of the stacker exceeds a set threshold value, a failure alarm is sent out.
2. The deep learning based predictive stacker maintenance method according to claim 1, wherein the establishing a base model for analyzing the failure rate of the stacker according to the historical failure data comprises:
preprocessing the historical fault data to obtain analysis data;
and importing the analysis data into a three-layer neural network based on open source framework Tensorflow to generate a basic model for analyzing the failure rate of the stacker.
3. The deep learning based stacker predictive maintenance method according to claim 2, wherein the importing the analysis data into a three-layer neural network based on open source framework Tensflow to generate a basic model for analyzing the failure rate of the stacker comprises:
when the data volume of the analysis data reaches a limit value, importing the analysis data into a three-layer neural network based on open source framework Tensorflow;
setting parameters of the three-layer neural network;
and generating basic models of various fault problems, and numbering each basic model.
4. The deep learning based predictive stacker maintenance method according to claim 3, wherein the collecting the working state parameters of the stacker in real time and inputting the collected state parameters into the basic model, analyzing the collected state parameters of the stacker in real time based on the basic model, and outputting the stacker fault value comprises:
collecting the state parameters of the stacker during working in real time, inputting the state parameters into each basic model, and outputting a plurality of similarity values;
and sequencing the similarity values to obtain the maximum similarity value, and taking the maximum similarity value as a fault value of the stacker.
5. The deep learning based stacker predictive maintenance method according to claim 1, further comprising:
when the failure rate value of the stacker does not exceed a set threshold value and the stacker judges that a failure occurs, collecting failure data of the current failure of the stacker, wherein the failure data of the current failure comprises a failure state parameter data set in a period of time before the failure occurs;
and inputting the fault data of the fault into the basic model, training and updating the basic model to obtain an updated basic model.
6. A deep learning based stacker predictive maintenance system, comprising: the system comprises an edge data processing terminal arranged on the stacker and a cloud server in communication connection with the edge data processing terminal;
the edge data processing terminal is used for collecting historical fault data of the stacker, wherein the historical fault data comprise a fault state parameter data set in a period of time before a fault occurs and are sent to the cloud server;
the cloud server is used for establishing a basic model for analyzing the failure rate of the stacker according to the historical failure data and sending the basic model to the edge data processing terminal;
the edge data processing terminal is also used for acquiring the state parameters of the stacker during working in real time, inputting the state parameters into the basic model, analyzing the state parameters of the stacker acquired in real time based on the basic model, and outputting the failure rate value of the stacker; and when the failure rate value of the stacker exceeds a set threshold value, a failure alarm is sent out.
7. The deep learning based stacker predictive maintenance system of claim 6, wherein the edge data processing terminal is further configured to send the failure alarm to a cloud server, and the cloud server is further configured to notify a worker for maintenance through a mobile terminal.
8. The deep learning based stacker predictive maintenance system of claim 6 wherein said edge data processing terminal comprises: the system comprises a signal acquisition module for acquiring state parameters of the stacker during working, a main controller for receiving the state parameters, a communication module for communicating with a cloud server and a power supply module for providing power supply.
9. The deep learning based predictive stacker maintenance system according to claim 8, wherein the signal acquisition module comprises a current sensor, a voltage sensor, a temperature sensor and a vibration sensor for acquiring a current parameter, a voltage parameter, a temperature parameter and a vibration parameter of the operating state of the stacker, respectively.
10. The deep learning based stacker predictive maintenance system of claim 7 wherein the cloud server sends a failure alert to mobile terminals in a preset area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010143230.9A CN111210092A (en) | 2020-03-04 | 2020-03-04 | Stacking machine predictive maintenance method and system based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010143230.9A CN111210092A (en) | 2020-03-04 | 2020-03-04 | Stacking machine predictive maintenance method and system based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111210092A true CN111210092A (en) | 2020-05-29 |
Family
ID=70790000
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010143230.9A Pending CN111210092A (en) | 2020-03-04 | 2020-03-04 | Stacking machine predictive maintenance method and system based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111210092A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112699104A (en) * | 2020-12-11 | 2021-04-23 | 国网宁夏电力有限公司吴忠供电公司 | Data processing method of intelligent sensor of power equipment and intelligent sensor |
| CN113651060A (en) * | 2021-08-17 | 2021-11-16 | 河南中烟工业有限责任公司 | Online detection system for operation state of stacker |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040233041A1 (en) * | 2001-03-27 | 2004-11-25 | Karl Bohman | Container surveillance system and related method |
| DE102018203816A1 (en) * | 2018-03-13 | 2019-09-19 | Gebhardt Fördertechnik GmbH | Method for, preferably anticipatory, maintenance of an automated warehouse system |
| CN110766590A (en) * | 2019-11-06 | 2020-02-07 | 山东浪潮人工智能研究院有限公司 | Street lamp predictive maintenance system and method based on deep learning |
-
2020
- 2020-03-04 CN CN202010143230.9A patent/CN111210092A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040233041A1 (en) * | 2001-03-27 | 2004-11-25 | Karl Bohman | Container surveillance system and related method |
| DE102018203816A1 (en) * | 2018-03-13 | 2019-09-19 | Gebhardt Fördertechnik GmbH | Method for, preferably anticipatory, maintenance of an automated warehouse system |
| CN110766590A (en) * | 2019-11-06 | 2020-02-07 | 山东浪潮人工智能研究院有限公司 | Street lamp predictive maintenance system and method based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| 李慧民 等著: "《城市地下工程施工安全预警***构建指南》", 冶金工业出版社, pages: 30 - 33 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112699104A (en) * | 2020-12-11 | 2021-04-23 | 国网宁夏电力有限公司吴忠供电公司 | Data processing method of intelligent sensor of power equipment and intelligent sensor |
| CN112699104B (en) * | 2020-12-11 | 2023-04-07 | 国网宁夏电力有限公司吴忠供电公司 | Data processing method of intelligent sensor of power equipment and intelligent sensor |
| CN113651060A (en) * | 2021-08-17 | 2021-11-16 | 河南中烟工业有限责任公司 | Online detection system for operation state of stacker |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109001649B (en) | Intelligent power supply diagnosis system and protection method | |
| CN111983383B (en) | Power system fault first-aid repair method and system | |
| CN107369303B (en) | Factory's intelligent diagnosing method, apparatus and system | |
| CN110707817B (en) | Intelligent power grid monitoring control system and method based on geographic information technology | |
| CN111850609B (en) | Aluminum electrolysis control system based on digital twinning | |
| CN110988559A (en) | Online monitoring method for full life cycle of transformer substation direct current system based on Internet of things | |
| CN111210092A (en) | Stacking machine predictive maintenance method and system based on deep learning | |
| CN102937802A (en) | System and method for monitoring operating state of device | |
| CN112562277A (en) | Equipment fault early warning method and system | |
| CN113627656A (en) | Electric power safety management and control method based on block chain | |
| CN110942221A (en) | Transformer substation fault rapid repairing method based on Internet of things | |
| CN111612337A (en) | Automatic issuing method and automatic issuing system for heat supply network scheduling | |
| CN101425940B (en) | Method, apparatus and system for alarm processing | |
| CN113433914B (en) | Intelligent construction site management and control platform and monitoring method thereof | |
| CN113159503B (en) | Remote control intelligent safety evaluation system and method | |
| CN103616877A (en) | Monitoring diagnostic method and system for energy pipe network | |
| CN117221145A (en) | Equipment fault predictive maintenance system based on Internet of things platform | |
| CN115238925B (en) | Motor equipment supervision method and system | |
| CN104467174A (en) | Transformer station monitoring alarm information auxiliary processing system | |
| CN114884800A (en) | Equipment maintenance operation management platform and method based on mobile terminal | |
| CN115441581A (en) | Large-scale power grid safety and intelligent scheduling system based on big data | |
| CN112346415B (en) | Workshop information physical monitoring management system | |
| CN106776193A (en) | The Virtual test equipment and method of testing of apparatus for monitoring power supply slave failure | |
| CN112433786B (en) | Information processing method, terminal and computer readable storage medium | |
| CN112073475A (en) | Real-time mold operation monitoring system based on Internet of things and working method thereof |
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 |