CN113469343A - Industrial time sequence data processing method and system - Google Patents
Industrial time sequence data processing method and system Download PDFInfo
- Publication number
- CN113469343A CN113469343A CN202110774963.7A CN202110774963A CN113469343A CN 113469343 A CN113469343 A CN 113469343A CN 202110774963 A CN202110774963 A CN 202110774963A CN 113469343 A CN113469343 A CN 113469343A
- Authority
- CN
- China
- Prior art keywords
- model
- module
- data
- industrial
- data processing
- 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.)
- Withdrawn
Links
Images
Classifications
-
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/04—Inference or reasoning models
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Fuzzy Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides an industrial time sequence data processing method based on an Encoder-Decoder neural network structure, which comprises the following steps: collecting data and extracting edge features; acquiring edge feature data, and establishing a first model based on an end-to-end deep learning algorithm; training the first model and outputting a second model; and continuously inputting the time sequence data into the second model for analysis and then outputting a processing result. The method can support data input of a time sequence, obtain and output a whole string of abnormal sequences, identify all abnormal points and corresponding time coordinates thereof, and quickly obtain accurate results and position the abnormal points; the processing result can more fully meet the requirements of guiding production and risk management of enterprises. The industrial time sequence data processing system provided by the invention is easy to configure on the basis of the existing hardware, and is beneficial to the realization of data processing and collaborative analysis of enterprises in an industrial internet environment, thereby promoting the progress of the production level of the enterprises.
Description
Technical Field
The invention belongs to the technical field of industrial big data processing, and particularly relates to a time-series-based industrial big data processing method and a corresponding data processing system.
Background
With the development of industrialization, the production process is more and more intensive, the connection of each link is more and more compact, and the occurrence of a fault or error can cause considerable loss of the whole factory production. At present, enterprises invest a large amount of operation and maintenance cost in order to ensure the normal operation of production processes. From some statistical data, 15% to 70% of the total cost of industrial production needs to be paid as the maintenance cost for repairing the fault, so that the abnormality in the industrial production is detected in time, the processing is responded in time according to the abnormality type, the smooth proceeding of the industrial production process is ensured, and the industrial production cost can be greatly saved. The big data from links such as real-time monitoring of equipment, data acquisition, product quality detection, remote maintenance of products and the like and the conventional data such as design, process, production, logistics, operation and the like form industrial big data together. Through industrial big data analysis, the abnormal conditions in the production process can be found in advance, the source of quality problems can be analyzed, and the factors restricting the improvement of the production level can be found, so that scientific decision support can be provided for process optimization, quality improvement, preventive maintenance of equipment, even the improvement and design of products and the like. Particularly under the promotion of intelligent manufacturing and industrial internet, the intelligent production can integrate the advantageous resources of different enterprises in a range far beyond the traditional factory, and the cross-region dispersion cooperative operation is realized. Any equipment, a station, a workshop or even a factory can participate in task nodes of networked manufacturing as long as the equipment is within the resource configuration authority range, so that the complex business cooperation is realized. In production, collected time sequence data are processed, the state of industrial equipment is judged in a classified mode, the method is important for timely adjusting and eliminating faults, and the importance of the method is increasingly strengthened particularly in a new collaborative production mode. The improvement of industrial time series data processing capability is also particularly critical to advance the development of intelligent manufacturing.
In modern industrial equipment, hundreds of millions of sensors are often installed to detect various types of data such as temperature, pressure, vibration, noise, etc., the data generated by the sensors are decoded and converted to form a one-dimensional or high-dimensional time series, and the time series data is a data type which changes with time and generally comprises two basic factors, namely, time (time coordinate) to which a detected phenomenon belongs and an index value reflecting the quantitative characteristics of the phenomenon under a certain time condition, for example, the temperature within 24 hours. The processing and analysis of time series data is highly required for accuracy and timeliness. In the current industrial big data system, simple mathematical analysis statistics or traditional deep learning is usually adopted for processing and analyzing time series data, only one output can be obtained for one input series data, and only whether an abnormality exists in the series can be indicated. But the time coordinate corresponding to the abnormal point cannot be accurately analyzed; moreover, if a plurality of abnormal points exist in the sequence, the existing system is difficult to detect simultaneously, and even the abnormal points cannot be output simultaneously. And the existing analysis statistics or deep learning algorithm can not completely meet the requirements of precision and efficiency of industrial big data processing.
Therefore, at present, it is very necessary to research an industrial time series data processing method and a corresponding system, which can support the large data processing and analysis of the industrial time series, so as to further promote the deep development of intelligent manufacturing, and promote the high-quality realization of the intelligent manufacturing in a wider industrial field.
Disclosure of Invention
The invention aims to solve all or part of problems in the prior art, and provides an industrial time sequence data processing method which is suitable for analyzing and processing industrial big data of a time sequence, finds an abnormality and a time point (time coordinate) of the abnormality and assists in diagnosing an industrial fault. And. Another aspect of the present invention provides a data processing system capable of implementing the data processing method of the present invention.
The industrial time sequence data processing method provided by the invention is based on an Encoder-Decoder neural network structure and comprises the following steps: s1, collecting data and extracting edge characteristics; s2, acquiring edge feature data, and establishing a first model based on an end-to-end deep learning algorithm; the first model is formed by splicing an Encoder, a Decoder and a loss function unit; s3, training the first model and outputting a second model; and S4, continuously inputting the time sequence data into the second model for analysis and then outputting a processing result, wherein the processing result comprises an abnormal sequence formed by abnormal data and corresponding time coordinates.
For the neural network model, two problems of pattern classification and pattern regression can be divided according to different output tasks. When training the neural network, the gradient descent-based method continuously reduces the difference between the predicted value and the true value, and the difference is called Loss (Loss), and the function for calculating the Loss is called Loss function (Loss function). Such as a two-class cross-entropy loss function, a multi-class cross-entropy loss function, and so forth.
The penalty function of the penalty function unit is a Transducer Loss. The Transducer Loss is a Loss function proposed by Alex Graves, and the specific implementation can be referred to the paper: A. sequences transformation with recovery Neural networks, retrieval Learning workshop wortcll 2012, Edinburgh, Scotland. The Loss function adopts Transducer Loss to effectively solve the problem of sequence probability path modeling.
In step S2, the building a first model includes: splicing the LSTM layer with a preset first layer number, the Transformer unit and the TDNN with a preset second layer number to obtain the Encoder; splicing the LSTM layers with a preset third layer number to obtain the Decoder; and splicing the Encoder, the Decoder and the loss function unit to complete modeling. The Encoder and Decoder structures which are many in theory can be built in the modeling process, the model can be built in a targeted mode aiming at the specific time sequence modeling problem, the number of model layers and the model structure can be selected by a user at will, the model can be built specifically according to the specific problem, and the first model can be applied flexibly, effectively and reliably.
TDNN (Time-Delay Neural Network) is a common module in deep learning modeling, and is generally used to solve phoneme recognition. The LSTM (Long Short-Term Memory) is a time cycle neural network and is a deep learning modeling unit for time sequence modeling; the Transformer is a model based on an encoder-decoder structure, abandons RNN in the prior seq2seq model, and adopts Self-attention or Mulit-head-Self-attention to enable input data to be processed in parallel, thereby improving the operating efficiency; the Transformer can well solve the NLP task of the text in the field of Natural Language Processing (NLP).
The first layer number or the second layer number is greater than 1; the second number of layers is greater than 2.
In step S3, the first model is trained in parallel. The training of the deep learning model is an iterative process. In each iteration, the forward propagation algorithm calculates the predicted value on a small part of training data according to the value of the current parameter, and then the backward propagation algorithm calculates the gradient of the parameter according to the loss function and updates the parameter. In training the deep learning model in parallel, different devices (GPU or CPU) can run this iterative process on different training data, while different parallel modes differ in different parameter update modes.
In step S3, the first model is trained using a PyTorch-based training platform. PyTorch is an open-source Python machine learning library, based on Torch, commonly used in Natural Language Processing (NLP) applications. PyTorch provides two advanced functions: tensor calculations with powerful GPU acceleration (e.g., NumPy); a deep neural network comprising an automatic derivation system. The PyTorch is a relatively compact, efficient and fast framework, the design pursues a minimum of packaging, and the required basic environment is generally a PC device and Ubuntu system, which is easy to configure.
The time sequence data processing method of the invention also comprises the steps of deploying the second model on line; in step S4, the data collection stream is analyzed in real time. The deployment of the model mainly refers to the deployment of the trained algorithm model into the production environment, generally in two ways, namely online (online) and offline (offline), and the online deployment generally includes: and performing primary preprocessing at the mobile terminal, transmitting data to the server for prediction, and returning the data to the mobile terminal. On-line deployment is relatively simple, and the ready-made frame (such as TensorFlow, PyTorch, Keras, Caffe, Theano, MxNet) can be directly used for packaging; the server is used for calculation, the performance is high, and the algorithm model with large calculation amount can be processed.
The industrial time sequence data processing system provided by the invention comprises a data acquisition module, an edge layer processing module, a deep learning module, a model training module and an analysis module; the data acquisition module is used for acquiring data and transmitting the acquired data to the edge layer processing module; the edge layer processing module is used for receiving the data sent by the data acquisition module and performing feature extraction (edge data preprocessing) on the data; the deep learning module receives the data sent from the edge layer processing module and establishes an algorithm model, namely a first model, based on an end-to-end deep learning algorithm; the model training module acquires the first model, performs parallel training on the first model by using the acquired data, and outputs an algorithm model, namely a second model after the training is finished; and the analysis module acquires the second model and performs real-time analysis on the data acquisition flow based on the second model.
The system comprises a deep learning module and/or a model training module, and is characterized by also comprising a server, wherein the server is a network server or a cloud server and is integrated with the deep learning module and/or the model training module; the model training module is based on PyTorch; the server is in communication connection with the data acquisition module and the edge layer processing module through the internet. And transmitting the data acquired by the data acquisition module to the server through a network to establish an algorithm model and train the algorithm model.
The server is also integrated with the analysis module and used for completing the real-time analysis of the data acquisition stream on line.
The device also comprises a display module and a storage module, wherein the display module is used for displaying the processing result, and the storage module is used for storing the processing result.
Compared with the prior art, the invention has the main beneficial effects that:
1. the industrial time sequence data processing method is based on an Encoder-Decoder neural network structure, an end-to-end deep learning algorithm model is established and trained to process industrial big data of a time sequence, data input of the time sequence can be supported, a whole string of output abnormal sequences can be obtained after the second model analysis, all abnormal points and corresponding time coordinates can be identified, the traditional analysis effect of simple mathematical analysis statistics or traditional deep learning is effectively improved, the service calculation amount of the whole time sequence data processing is low, and accurate results can be quickly obtained; the processing result can more fully meet the requirements of guiding production and risk management of enterprises. In addition, the Encoder provides an abstract mode for model building, theoretically any number of Encoder and Decoder structures can be built according to specific application by using the abstract mode, so that a specific model can be built for specific time sequence modeling problems by presetting the number of model layers and the model structure by a user, the specific problem building is realized, the judgment capability and the prediction accuracy of the model are high, and the application is flexible.
2. The industrial time sequence data processing system can realize the industrial time sequence data processing method on one hand, is easy to configure on the basis of the existing hardware, and saves the hardware cost; the method can be combined with the flexible spatial arrangement of the existing equipment of the enterprise, is favorable for the enterprise to realize data processing and collaborative analysis under the industrial internet environment, can effectively improve the management capability and risk control capability of the enterprise on the production process and the production equipment data, and promotes the progress of the production level of the enterprise.
Drawings
Fig. 1 is a schematic diagram of an industrial time series data processing system according to a first embodiment of the invention.
Fig. 2 is a schematic diagram of an industrial time series data processing method according to a first embodiment of the invention.
Fig. 3 is a schematic diagram of an industrial time series data processing process according to a second embodiment of the invention.
Fig. 4 is a schematic structural diagram of an Encoder-Decoder neural network according to a second embodiment of the present invention.
Detailed Description
The technical solutions in the specific embodiments of the present invention will be clearly and completely described below, and it should be understood that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. 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.
The above and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings. In the figures, parts of the same structure or function are denoted by the same reference numerals, and not all parts shown are denoted by the associated reference numerals in all figures for reasons of clarity of presentation.
The operations of the embodiments are depicted in the following embodiments in a particular order, which is provided for better understanding of the details of the embodiments and to provide a thorough understanding of the present invention, but the order is not necessarily one-to-one correspondence with the methods of the present invention, and is not intended to limit the scope of the present invention.
It is to be noted that the flow charts and block diagrams in the figures illustrate the operational procedures which may be implemented by the methods according to the embodiments of the present invention. It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the alternative, depending upon the functionality involved. It is also noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and manual acts.
Example one
The industrial time series data processing system disclosed by the first embodiment of the invention comprises a data acquisition module, an edge layer processing module, a deep learning module, a model training module and an analysis module, as shown in fig. 1; the data acquisition module is responsible for collecting and summarizing data based on time series from a large number of devices and sensors arranged by enterprises and transmitting the collected data to the edge layer processing module. In this embodiment, access and data acquisition to devices in a factory are mainly realized through an industrial communication network such as a field bus, an industrial ethernet, an industrial optical fiber network, and the like, and mainly data acquisition to special acquisition devices such as a sensor, a transmitter, an acquisition device, and the like. In some implementation cases, the method also comprises data acquisition of general control equipment such as a PLC, an RTU, an embedded system, an IPC and the like; the data acquisition and the like of special intelligent equipment such as robots, numerical control machines, AGV and the like are not limited, and the data acquisition is carried out according to the actual application needs of enterprises. The data acquisition module in the embodiment accesses various industrial field devices and intelligent products/equipment through various wired and wireless communication technologies such as industrial ethernet, industrial optical fiber network, industrial bus, 3G/4G, NB-IoT and the like, and acquires industrial big data. The edge layer processing module is used for receiving the data sent by the data acquisition module and extracting the features of the data. In this embodiment, the edge layer processing module includes edge computing software, a supporting database, a related module, and the like, and performs data preprocessing, storage, and intelligent analysis application on the side of the network edge near the device or the data source based on technical supports such as high-performance computing, a real-time operating system, and an edge analysis algorithm, so as to improve the response sensitivity of operation, eliminate network congestion, and form cooperation with the deep learning module, the model training module, and the analysis module. The edge computing software selected by the embodiment can realize massive and heterogeneous connection, meet the real-time requirement of the service, realize data optimization, pay attention to the intelligence of application, and protect safety and privacy. The deep learning module receives data sent from the edge layer processing module and establishes an algorithm model, namely a first model, based on an end-to-end deep learning algorithm; the model training module acquires the first model, performs parallel training on the first model by using the acquired data, and outputs an algorithm model, namely a second model after the training is finished; and the analysis module acquires the second model and performs real-time analysis on the data acquisition flow based on the second model.
In this embodiment, still include display module and storage module, display module is used for showing the processing result, and display module can include a plurality of display screens that distribute at the specific monitoring post of enterprise, handheld mobile terminal's display screen, is convenient for need acquire the processing result and carry out the nimble convenient acquisition processing result of personnel that follow-up analysis was carried out to the processing result, also is convenient for many people's cooperative operation. The storage module is used for storing the processing results, and the memories of the plurality of PCs in the embodiment can be used for reserving and regularly summarizing and analyzing historical data of the processing results and guiding improvement of actual production.
As shown in fig. 2, the industrial time series data processing method of the embodiment includes: s1, collecting data and extracting edge characteristics; s2, acquiring edge feature data, and establishing a first model based on an end-to-end deep learning algorithm; the first model is formed by splicing an Encoder, a Decode and a loss function unit; s3, training the first model and outputting a second model; and S4, continuously inputting the time sequence data into the second model for analysis and then outputting a processing result, wherein the processing result comprises an abnormal sequence formed by abnormal data and corresponding time coordinates. In step S3, the first model is trained in parallel. In step S4, the data collection stream is analyzed in real time. The specific Loss function of the Loss function unit in this embodiment is a transmitter Loss.
Example two
As shown in fig. 3, the difference between the second embodiment and the first embodiment is mainly that, in the time series data processing method, the second model is deployed online. And deploying the algorithm model on a line, and carrying out real-time analysis on the data acquisition flow. In step S3 of this embodiment, the first model is trained by using a PyTorch-based training platform. The industrial time sequence data processing system of the embodiment also comprises a plurality of servers, wherein the servers are integrated with an analysis module, a deep learning module and a model training module; the server is in communication connection with the data acquisition module and the edge layer processing module through the internet and is in cooperation with the cloud data analysis.
The following description is provided for the process of creating the first model in step S2 in this embodiment: splicing the LSTM layer with a preset first layer number, the Transformer unit and the TDNN with a preset second layer number to obtain the Encoder; splicing the LSTM layers with a preset third number m to obtain the Decoder; and splicing the Encoder, the Decoder and the Transducer unit to complete modeling. As shown in fig. 4 in the present embodiment, the exemplary first layer number is preset to 1 layer; the second number of layers is preset to 3 layers. The Transducer Unit is composed of a Concat & Gated Linear Unit and project Softmax, and can build an arbitrary number of (n-layer) Encoder and (m-layer) Decoder structures.
In the training process of this embodiment, the training of the whole model and the inference online are completed by using the training platform based on PyTorch, which is not further described.
For clarity of description, the use of certain conventional and specific terms and phrases is intended to be illustrative and not restrictive, but rather to limit the scope of the invention to the particular letter and translation thereof.
It is further noted that, herein, relational terms such as "first" and "second," and the like, may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
The present invention has been described in detail, and the structure and operation principle of the present invention are explained by applying specific embodiments, and the above description of the embodiments is only used to help understanding the method and core idea of the present invention. It should be noted that, for those skilled in the art, it is possible to make various improvements and modifications to the present invention without departing from the principle of the present invention, and those improvements and modifications also fall within the scope of the claims of the present invention.
Claims (10)
1. The industrial time series data processing method is characterized by comprising the following steps: based on Encoder-Decoder neural network structure, include:
s1, collecting data and extracting edge characteristics;
s2, acquiring edge feature data, and establishing a first model based on an end-to-end deep learning algorithm; the first model is formed by splicing an Encoder, a Decode and a loss function unit;
s3, training the first model and outputting a second model;
and S4, continuously inputting the time sequence data into the second model for analysis and then outputting a processing result, wherein the processing result comprises an abnormal sequence formed by abnormal data and corresponding time coordinates.
2. The industrial time series data processing method according to claim 1, characterized in that: the penalty function of the penalty function unit is a Transducer Loss.
3. The industrial time series data processing method according to claim 1, characterized in that: in step S2, the building a first model includes:
splicing the LSTM layer with a preset first layer number, the Transformer unit and the TDNN with a preset second layer number to obtain the Encoder;
splicing the LSTM layers with a preset third layer number to obtain the Decoder;
and splicing the Encoder, the Decoder and the loss function unit to complete modeling.
4. The industrial time series data processing method according to claim 3, characterized in that: the first layer number or the second layer number is greater than 1; the second number of layers is greater than 2.
5. The industrial time series data processing method according to claim 1, characterized in that: in step S3, the first model is trained in parallel.
6. The industrial time series data processing method according to claim 1, characterized in that: in step S3, the first model is trained using a PyTorch-based training platform.
7. The industrial time series data processing method according to any one of claims 1 to 6, characterized in that: further comprising deploying the second model online; in step S4, the data collection stream is analyzed in real time.
8. An industrial time series data processing system, characterized by: the system comprises a data acquisition module, an edge layer processing module, a deep learning module, a model training module and an analysis module;
the data acquisition module is used for acquiring data and transmitting the acquired data to the edge layer processing module;
the edge layer processing module is used for receiving the data sent by the data acquisition module and extracting the characteristics of the data;
the deep learning module receives the data sent from the edge layer processing module and establishes an algorithm model, namely a first model, based on an end-to-end deep learning algorithm;
the model training module acquires the first model, performs parallel training on the first model by using the acquired data, and outputs an algorithm model, namely a second model after the training is finished;
and the analysis module acquires the second model and performs real-time analysis on the data acquisition flow based on the second model.
9. The industrial time series data processing system of claim 8, wherein: the system comprises a deep learning module and/or a model training module, and is characterized by also comprising a server, wherein the server is a network server or a cloud server and is integrated with the deep learning module and/or the model training module; the model training module is based on PyTorch; the server is in communication connection with the data acquisition module and the edge layer processing module through the internet.
10. The industrial time series data processing system of claim 9, wherein: the server is also integrated with the analysis module and used for completing the real-time analysis of the data acquisition stream on line.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110774963.7A CN113469343A (en) | 2021-07-08 | 2021-07-08 | Industrial time sequence data processing method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110774963.7A CN113469343A (en) | 2021-07-08 | 2021-07-08 | Industrial time sequence data processing method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113469343A true CN113469343A (en) | 2021-10-01 |
Family
ID=77879241
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110774963.7A Withdrawn CN113469343A (en) | 2021-07-08 | 2021-07-08 | Industrial time sequence data processing method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113469343A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114253242A (en) * | 2021-12-21 | 2022-03-29 | 上海纽酷信息科技有限公司 | VPN-based Internet of things cloud equipment data acquisition system |
-
2021
- 2021-07-08 CN CN202110774963.7A patent/CN113469343A/en not_active Withdrawn
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114253242A (en) * | 2021-12-21 | 2022-03-29 | 上海纽酷信息科技有限公司 | VPN-based Internet of things cloud equipment data acquisition system |
| CN114253242B (en) * | 2021-12-21 | 2023-12-26 | 上海纽酷信息科技有限公司 | VPN-based cloud equipment data acquisition system for Internet of things |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lade et al. | Manufacturing analytics and industrial internet of things | |
| CN110148285B (en) | Intelligent oil well parameter early warning system based on big data technology and early warning method thereof | |
| US10902368B2 (en) | Intelligent decision synchronization in real time for both discrete and continuous process industries | |
| CN110221558B (en) | Electro-hydraulic servo valve online fault diagnosis gateway based on edge computing technology | |
| KR102086451B1 (en) | Smart Factory remote collaboration CMS system in augmented reality and Drive method of the Same | |
| Alves et al. | Deployment of a smart and predictive maintenance system in an industrial case study | |
| CN104679828A (en) | Rules-based intelligent system for grid fault diagnosis | |
| CN112041867A (en) | Intelligent factory data acquisition platform and implementation method thereof | |
| CN104390657A (en) | Generator set operating parameter measuring sensor fault diagnosis method and system | |
| CN112926257A (en) | Reciprocating natural gas compressor fault diagnosis system and diagnosis method | |
| CN115097788A (en) | Intelligent management and control platform based on digital twin factory | |
| CN104412190A (en) | Systems and methods for health assessment of a human-machine interface (HMI) device | |
| CN116485576A (en) | Intelligent manufacturing management platform for brain data with known source in aviation manufacturing industry | |
| KR20230017556A (en) | System and operational methods for manufacturing execution based on artificial intelligence and bigdata | |
| CN117055502A (en) | Intelligent control system based on Internet of things and big data analysis | |
| CN113469343A (en) | Industrial time sequence data processing method and system | |
| Singh et al. | EYE-on-HMI: A Framework for monitoring human machine interfaces in control rooms | |
| CN117696224A (en) | Ore grinding optimizing treatment system based on large model | |
| CN109732603A (en) | Method for controlling robot based on the fusion of multi-source heterogeneous information | |
| CN114819350A (en) | RUL prediction method under sensor fault condition based on multiple tasks | |
| CN112364088A (en) | Visual configuration system based on factory digital manufacturing resources | |
| Fu et al. | Environmental Intelligent Perception in the Industrial Internet of Things: A Case Study Analysis of a Multi-crane Visual Sorting System | |
| CN111832832B (en) | District self-inspection system based on thing networking | |
| KR102429393B1 (en) | Behavior abnormal detection intelligent sharing system and method in smart factory based on profiling | |
| US20240160198A1 (en) | Industrial data source writeback |
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 | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20211001 |
|
| WW01 | Invention patent application withdrawn after publication |