CN114139811A - Control system optimization method and device based on deep learning and electronic device - Google Patents
Control system optimization method and device based on deep learning and electronic device Download PDFInfo
- Publication number
- CN114139811A CN114139811A CN202111469851.7A CN202111469851A CN114139811A CN 114139811 A CN114139811 A CN 114139811A CN 202111469851 A CN202111469851 A CN 202111469851A CN 114139811 A CN114139811 A CN 114139811A
- Authority
- CN
- China
- Prior art keywords
- index
- control system
- index prediction
- operation information
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000005457 optimization Methods 0.000 title claims abstract description 39
- 238000013135 deep learning Methods 0.000 title claims abstract description 33
- 230000010076 replication Effects 0.000 claims description 60
- 230000006870 function Effects 0.000 claims description 45
- 238000003062 neural network model Methods 0.000 claims description 36
- 238000012549 training Methods 0.000 claims description 34
- 238000012545 processing Methods 0.000 claims description 23
- 238000004590 computer program Methods 0.000 claims description 20
- 238000003860 storage Methods 0.000 claims description 5
- 238000012423 maintenance Methods 0.000 abstract description 10
- 238000010586 diagram Methods 0.000 description 7
- 238000007781 pre-processing Methods 0.000 description 7
- 230000005540 biological transmission Effects 0.000 description 6
- 238000004519 manufacturing process Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 5
- 238000013527 convolutional neural network Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000009499 grossing Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Molecular Biology (AREA)
- Development Economics (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Game Theory and Decision Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Feedback Control In General (AREA)
Abstract
The application relates to a control system optimization method and device based on deep learning and an electronic device, wherein the control system optimization method based on deep learning comprises the following steps: acquiring node operation information of each node of a control system; adjusting parameter content in an index prediction model according to the node operation information; inputting the node operation information into an index prediction model to obtain an index prediction result; and optimizing the control system according to the index prediction result. According to the method and the device, the index prediction model is adopted to predict the node operation information to generate the index prediction result, and the control system is optimized according to the index prediction result, so that the problem of difficulty in maintaining the control system is solved, and the timely maintenance of each node of the control system is realized.
Description
Technical Field
The present application relates to the field of deep learning technologies, and in particular, to a method and an apparatus for optimizing a control system based on deep learning, and an electronic apparatus.
Background
As the process industry is increasingly large-scale and complex, an economic growth mode which is dominated by capacity scale expansion is formed, and the industrial development mode is changed from a scale speed type to a quality benefit type. The optimization and upgrading of the industry result in high integration and high efficiency of the industrial process, and simultaneously, the complexity and the coupling degree of the system are increased. However, along with the long-term operation of the production device, the control system is often degraded gradually due to lack of timely and professional maintenance, which affects the operation stability, product quality, product yield, material consumption, energy consumption and the like of the device, some loops degraded due to the performance of the controller are directly driven to a manual state, and the number of the loops is large, which is difficult to accurately position, and the cost of field maintenance of project engineers is too high.
At present, the predictive control can achieve good effects on a production process or an object with slower control change, but the algorithm of the predictive control is proposed based on a linear object, and the industrial process in the process industry has the characteristics of time variation, nonlinearity and the like and has the influence of various uncertain factors; the traditional control mode is difficult to meet the real-time requirement of control optimization. In the face of a large number of non-linear and uncertain processes in industrial processes, the algorithm needs to be further improved.
In an actual industrial process, there are various constraint requirements on physical quantities in a system due to various reasons such as device physical characteristics and product quality indexes. However, over time, the characteristics of the multivariable circuit change, and the instrumentation valves also age during use, resulting in degraded control performance. The number of control loops under the industrial device is large and complex, and the large controller cannot achieve professional maintenance in time.
Aiming at the problem of difficult maintenance of a control system in the related art, no effective solution is provided at present.
Disclosure of Invention
The embodiment provides a control system optimization method, a control system optimization device and an electronic device based on deep learning, so as to solve the problem of difficult maintenance of a control system in the related art.
In a first aspect, in this embodiment, a method for optimizing a control system based on deep learning is provided, which includes:
acquiring node operation information of each node of a control system;
adjusting parameter content in an index prediction model according to the node operation information;
inputting the node operation information into an index prediction model to obtain an index prediction result;
and optimizing the control system according to the index prediction result.
In some embodiments, the obtaining node operation information of each node of the control system includes:
and preprocessing the node operation information, wherein the preprocessing comprises one or more of smoothing processing, noise reduction processing, fitting processing, normalization processing, up-sampling processing, down-sampling processing and enhancement processing.
In some embodiments, the inputting the node operation information into an index prediction model, and obtaining an index prediction result includes:
establishing a neural network model, training the neural network model by taking node operation information containing different types of data as a training set to obtain an index prediction model, wherein the input of the index prediction model is the training set, and the output of the index prediction model is the index prediction result.
In some embodiments, the establishing a neural network model, and training the neural network model by using node operation information containing different types of data as a training set, and the obtaining an index prediction model includes:
copying the neural network model to obtain a plurality of copied models;
respectively training the plurality of replication models according to node operation information containing different types of data to obtain a plurality of replication index models;
generating the index prediction model from a plurality of the replicated index models.
In some of these embodiments, said generating said metric predictive model from a plurality of said replicated metric models comprises:
wherein the replication index model corresponds to a loss function;
sorting the loss functions of the plurality of replication index models, and distributing corresponding weight coefficients for the replication index models;
and performing weighting processing on the plurality of copied index models according to the weighting coefficient to generate the index prediction model.
In some embodiments, the sorting the loss functions of the plurality of the replica index models, and the assigning the corresponding weight coefficients to the replica index models includes:
detecting whether a loss function of the replication index model is smaller than a preset function threshold value or not;
and if the loss function of the replication index model is greater than or equal to a preset function threshold value, the replication index model is not sequenced.
In some embodiments, the inputting the node operation information into an index prediction model, and obtaining an index prediction result further includes:
and adjusting the structure and parameters of the neural network model according to the index prediction result.
In a second aspect, in this embodiment, a deep learning based control system optimization apparatus is provided, including:
the information acquisition module is used for acquiring node operation information of each node of the control system;
the parameter adjusting module is used for adjusting parameter contents in the index prediction model according to the node operation information;
the index prediction module is used for inputting the node operation information into an index prediction model to obtain an index prediction result;
and the system optimization module is used for optimizing the control system according to the index prediction result.
In a third aspect, in the present embodiment, there is provided an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements the deep learning based control system optimization method according to the first aspect when executing the computer program.
In a fourth aspect, in the present embodiment, there is provided a storage medium having stored thereon a computer program which, when executed by a processor, implements the deep learning based control system optimization method of the first aspect.
Compared with the related art, the control system optimization method and device based on deep learning and the electronic device provided by the embodiment predict the node operation information through the index prediction model to generate the index prediction result, and optimize the control system according to the index prediction result, so that the problem of difficult maintenance of the control system is solved, and the timely maintenance of each node of the control system is realized.
The details of one or more embodiments of the application are set forth in the accompanying drawings and the description below to provide a more thorough understanding of the application.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. In the drawings:
fig. 1 is a block diagram of a hardware structure of a terminal of a deep learning-based control system optimization method according to the present embodiment;
FIG. 2 is a flowchart of a deep learning based control system optimization method according to the present embodiment;
fig. 3 is a block diagram of the control system optimization device based on deep learning according to the present embodiment.
Detailed Description
For a clearer understanding of the objects, aspects and advantages of the present application, reference is made to the following description and accompanying drawings.
Unless defined otherwise, technical or scientific terms used herein shall have the same general meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The use of the terms "a" and "an" and "the" and similar referents in the context of this application do not denote a limitation of quantity, either in the singular or the plural. The terms "comprises," "comprising," "has," "having," and any variations thereof, as referred to in this application, are intended to cover non-exclusive inclusions; for example, a process, method, and system, article, or apparatus that comprises a list of steps or modules (elements) is not limited to the listed steps or modules, but may include other steps or modules (elements) not listed or inherent to such process, method, article, or apparatus. Reference throughout this application to "connected," "coupled," and the like is not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. Reference to "a plurality" in this application means two or more. "and/or" describes an association relationship of associated objects, meaning that three relationships may exist, for example, "A and/or B" may mean: a exists alone, A and B exist simultaneously, and B exists alone. In general, the character "/" indicates a relationship in which the objects associated before and after are an "or". The terms "first," "second," "third," and the like in this application are used for distinguishing between similar items and not necessarily for describing a particular sequential or chronological order.
The method embodiments provided in the present embodiment may be executed in a terminal, a computer, or a similar computing device. For example, the method is executed on a terminal, and fig. 1 is a block diagram of a hardware structure of the terminal of the deep learning-based control system optimization method according to the embodiment. As shown in fig. 1, the terminal may include one or more processors 102 (only one shown in fig. 1) and a memory 104 for storing data, wherein the processor 102 may include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA. The terminal may also include a transmission device 106 for communication functions and an input-output device 108. It will be understood by those of ordinary skill in the art that the structure shown in fig. 1 is merely an illustration and is not intended to limit the structure of the terminal described above. For example, the terminal may also include more or fewer components than shown in FIG. 1, or have a different configuration than shown in FIG. 1.
The memory 104 can be used for storing computer programs, for example, software programs and modules of application software, such as a computer program corresponding to the deep learning based control system optimization method in the present embodiment, and the processor 102 executes various functional applications and data processing by running the computer programs stored in the memory 104, so as to implement the above-mentioned method. The memory 104 may include high speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include memory located remotely from the processor 102, which may be connected to the terminal over a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The transmission device 106 is used to receive or transmit data via a network. The network described above includes a wireless network provided by a communication provider of the terminal. In one example, the transmission device 106 includes a Network adapter (NIC) that can be connected to other Network devices through a base station to communicate with the internet. In one example, the transmission device 106 may be a Radio Frequency (RF) module, which is used to communicate with the internet in a wireless manner.
In the present embodiment, a control system optimization method based on deep learning is provided, and fig. 2 is a flowchart of the control system optimization method based on deep learning of the present embodiment, as shown in fig. 2, the flowchart includes the following steps:
step S201, obtaining node operation information of each node of the control system.
The Control System may be a Distributed Control System (Distributed Control System), which is a new generation of instrument Control System based on a microprocessor and adopting a design principle of Distributed Control functions, centralized display operation, and consideration of distribution, autonomy and comprehensive coordination.
Specifically, the node running information of the controller on each node of the control system is obtained through the monitor. The node operation information is information of variable indexes associated with the controller.
The acquiring of the node operation information of each node of the control system includes: and detecting whether the node operation information is complete, and if the node operation information is incomplete, positioning the lost node and reading the node operation information of the node again.
Specifically, each node has a unique identifier, and whether the node operation information of the corresponding node is collected is judged by detecting whether the node operation information carries the corresponding unique identifier. If the node operation information carries the unique identification of all the nodes, the node operation information is complete; and if the node operation information does not carry the unique identification of all the nodes, positioning the lost node and reading the node operation information of the node again.
The acquiring of the node operation information of each node of the control system includes: and preprocessing the node operation information.
Wherein the preprocessing comprises one or more of smoothing processing, denoising processing, fitting processing, normalization processing, upsampling processing, downsampling processing and enhancing processing.
Specifically, through the preprocessing, abnormal data in the node operation information are removed, and the accuracy of the index prediction model is improved.
And S202, adjusting parameter content in the index prediction model according to the node operation information.
And step S203, inputting the node operation information into an index prediction model to obtain an index prediction result.
Specifically, the preprocessed node operation information is input into an index prediction model to obtain an index prediction result. More specifically, the node operation information is converted into a two-dimensional image, and the two-dimensional image is input into the index prediction model. The index prediction model identifies characteristic points on the two-dimensional image and generates an index prediction result according to the characteristic points. In one embodiment, data points are plotted according to the node operation information, and the plotted data points are converted to generate a two-dimensional image.
The step of inputting the node operation information into an index prediction model to obtain an index prediction result comprises the following steps: establishing a neural network model, training the neural network model by taking node operation information containing different types of data as a training set to obtain an index prediction model, wherein the input of the index prediction model is the training set, and the output of the index prediction model is the index prediction result.
The architecture of the Neural Network model may be a Convolutional Neural Network (CNN), such as a full Convolutional Neural Network (FCN), or other deep learning artificial Neural Network architectures such as a U-Net Convolutional Neural Network based on the FCN, a self-coding Neural Network based on multiple layers of neurons, or a deep belief Neural Network.
Specifically, a neural network model is established, and node operation information containing different types of data is used as a training set to train the neural network model. The neural network model carries out operations such as feature extraction, feature mapping, sampling and fitting on a training set, and obtains an index prediction model through continuous training of the training set on the neural network model.
The establishing of the neural network model, training the neural network model by taking node operation information containing different types of data as a training set, and obtaining the index prediction model comprises the following steps: copying the neural network model to obtain a plurality of copied models; respectively training the plurality of replication models according to node operation information containing different types of data to obtain a plurality of replication index models; generating the index prediction model from a plurality of the replicated index models.
Specifically, the structure of the neural network model is copied to obtain a plurality of copy models with the same structure; and training the plurality of replication models respectively according to the node operation information containing different types of data to obtain a plurality of replication index models. And verifying the plurality of replication index models, detecting errors between output results of the plurality of replication index models and standard results, screening the replication index models with smaller errors, and distributing corresponding weight coefficients to form the index prediction model.
In one embodiment, the node operation information includes data of different data types collected by Distributed Control Systems (DCS) of different manufacturers, the data of different data types in the node operation information are respectively allocated to corresponding replication models for training, so as to obtain corresponding replication index models, and each replication index model corresponds to one Distributed Control System. Verifying a plurality of replication index models corresponding to the same type of distributed control system, and screening the replication index models with errors smaller than a preset error threshold; and distributing weight coefficients corresponding to a plurality of copy index models according to the currently acquired node operation information to form the index prediction model.
The generating the metric prediction model from the plurality of replicated metric models comprises: sorting the loss functions of the plurality of replication index models, and distributing corresponding weight coefficients for the replication index models; and performing weighting processing on the plurality of copied index models according to the weighting coefficient to generate the index prediction model.
Wherein the replication index model corresponds to a loss function.
Specifically, the index prediction models are generated by performing ranking according to the size of the loss function of the plurality of replica index models, and the smaller the loss function is, the larger the assigned weight coefficient is, and performing weighting processing on the plurality of replica index models to be fused.
More specifically, the proportion of each type of data in the node operation information is identified, and the weight coefficients of the plurality of replication index models are adjusted according to the proportion. And weighting the plurality of adjusted copy index models to generate an index prediction model. In one embodiment, the node operation information includes 30% of type a data, 30% of type B data, and 40% of type C data. Identifying the proportion of each type of data in the node operation information, adjusting the weight coefficient of the replication index model corresponding to the type A data to be 30 percent, adjusting the weight coefficient of the replication index model corresponding to the type B data to be 30 percent, adjusting the weight coefficient of the replication index model corresponding to the type C data to be 40 percent according to the proportion, and then performing weighting processing to generate an index prediction model. It is understood that the proportions of the various types of data may be other values.
The sorting the loss functions of the plurality of replication index models, and the assigning of the corresponding weight coefficients to the replication index models includes: detecting whether a loss function of the replication index model is smaller than a preset function threshold value or not; and if the loss function of the replication index model is greater than or equal to a preset function threshold value, the replication index model is not sequenced.
Specifically, whether the loss function of the replication index model is smaller than a preset function threshold value is detected. If the loss function of the replication index model is larger than or equal to the preset function threshold, the replication index model is not sorted, and the replication index model has no reference significance due to the fact that the loss function is too large. And if the loss function of the replication index model is smaller than a preset function threshold, sequencing the loss functions of the replication index models, and distributing corresponding weight coefficients for the replication index models.
The inputting the node operation information into an index prediction model and obtaining an index prediction result further comprises: and adjusting the structure and parameters of the neural network model according to the index prediction result.
Specifically, the operation information of the nodes containing the optimization results is used as a verification set to train a neural network model, so as to obtain an index prediction model, wherein the input of the index prediction model is the training set, and the output is the index prediction result. And comparing the index prediction result with the optimization result, detecting the data parameter of the index prediction result, and adjusting the structure and the parameter of the neural network model according to the data parameter, thereby improving the prediction accuracy of the subsequent index prediction model.
Wherein the data parameters include at least one of accuracy, precision, and recall.
And S204, optimizing the control system according to the index prediction result.
Specifically, according to the index prediction result, parameter adjustment is performed on the controller of each node in the control system, and various constraint requirements are met, so that the control system is maintained, and the optimization of the control system is completed.
The method further comprises the following steps: updating node operation information according to a preset time interval; inputting the updated node operation information into the index prediction model to obtain an updated index prediction result; and optimizing the control system according to the updated index prediction result.
Specifically, the node operation information of the controller on each node is collected according to a preset time interval. Replacing and updating the previous node operation information, and inputting the updated node operation information into the index prediction model to obtain an updated index prediction result; and according to the updated index prediction result, parameter adjustment is carried out on the controller of each node of the control system, the corresponding index prediction result is obtained by dynamically updating the node operation information according to the preset time interval, the control system is optimized in real time, and the stability of the control system is improved.
Through the steps, the node operation information is predicted through the index prediction model to generate an index prediction result, and the control system is optimized according to the index prediction result, so that the problem of difficult maintenance of the control system is solved, and the timely maintenance of each node of the control system is realized.
It should be noted that the steps illustrated in the above-described flow diagrams or in the flow diagrams of the figures may be performed in a computer system, such as a set of computer-executable instructions, and that, although a logical order is illustrated in the flow diagrams, in some cases, the steps illustrated or described may be performed in an order different than here.
In this embodiment, a control system optimization device based on deep learning is further provided, and the device is used to implement the foregoing embodiments and preferred embodiments, and the description of which has been already made is omitted. The terms "module," "unit," "subunit," and the like as used below may implement a combination of software and/or hardware for a predetermined function. Although the means described in the embodiments below are preferably implemented in software, an implementation in hardware, or a combination of software and hardware is also possible and contemplated.
Fig. 3 is a block diagram of a control system optimization device based on deep learning according to the present embodiment, and as shown in fig. 3, the device includes: an information acquisition module 310, a parameter adjustment module 320, an index prediction module 330, and a system optimization module 340.
An information obtaining module 310, configured to obtain node operation information of each node of the control system.
And a parameter adjusting module 320, configured to adjust parameter content in the index prediction model according to the node operation information.
And the index prediction module 330 is configured to input the node operation information into an index prediction model to obtain an index prediction result.
And the system optimization module 340 is configured to optimize the control system according to the index prediction result.
The information obtaining module 310 is further configured to perform preprocessing on the node operation information.
The index prediction module 330 is further configured to establish a neural network model, train the neural network model by using node operation information containing different types of data as a training set, obtain an index prediction model, where an input of the index prediction model is the training set, and an output is the index prediction result.
The index prediction module 330 is further configured to copy the neural network model to obtain a plurality of copy models; respectively training the plurality of replication models according to node operation information containing different types of data to obtain a plurality of replication index models; generating the index prediction model from a plurality of the replicated index models.
The index prediction module 330 is further configured to rank the loss functions of the multiple replicated index models, and assign corresponding weight coefficients to the replicated index models; and performing weighting processing on the plurality of copied index models according to the weighting coefficient to generate the index prediction model.
The index prediction module 330 is further configured to detect whether a loss function of the replicated index model is smaller than a preset function threshold; and if the loss function of the replication index model is greater than or equal to a preset function threshold value, the replication index model is not sequenced.
The index prediction module 330 is further configured to adjust the structure and parameters of the neural network model according to the index prediction result.
The above modules may be functional modules or program modules, and may be implemented by software or hardware. For a module implemented by hardware, the modules may be located in the same processor; or the modules can be respectively positioned in different processors in any combination.
There is also provided in this embodiment an electronic device comprising a memory having a computer program stored therein and a processor arranged to run the computer program to perform the steps of any of the above method embodiments.
Optionally, the electronic apparatus may further include a transmission device and an input/output device, wherein the transmission device is connected to the processor, and the input/output device is connected to the processor.
Optionally, in this embodiment, the processor may be configured to execute the following steps by a computer program:
acquiring node operation information of each node of a control system;
and adjusting the parameter content in the index prediction model according to the node operation information.
Inputting the node operation information into an index prediction model to obtain an index prediction result;
and optimizing the control system according to the index prediction result.
Optionally, in this embodiment, the processor may be configured to execute the following steps by a computer program: the acquiring of the node operation information of each node of the control system includes: and preprocessing the node operation information.
Optionally, in this embodiment, the processor may be configured to execute the following steps by a computer program: the step of inputting the node operation information into an index prediction model to obtain an index prediction result comprises the following steps: establishing a neural network model, training the neural network model by taking node operation information containing different types of data as a training set to obtain an index prediction model, wherein the input of the index prediction model is the training set, and the output of the index prediction model is the index prediction result.
Optionally, in this embodiment, the processor may be configured to execute the following steps by a computer program: the establishing of the neural network model, training the neural network model by taking node operation information containing different types of data as a training set, and obtaining the index prediction model comprises the following steps: copying the neural network model to obtain a plurality of copied models; respectively training the plurality of replication models according to node operation information containing different types of data to obtain a plurality of replication index models; generating the index prediction model from a plurality of the replicated index models.
Optionally, in this embodiment, the processor may be configured to execute the following steps by a computer program: the generating the metric prediction model from the plurality of replicated metric models comprises: sorting the loss functions of the plurality of replication index models, and distributing corresponding weight coefficients for the replication index models; and performing weighting processing on the plurality of copied index models according to the weighting coefficient to generate the index prediction model.
Optionally, in this embodiment, the processor may be configured to execute the following steps by a computer program: the sorting the loss functions of the plurality of replication index models, and the assigning of the corresponding weight coefficients to the replication index models includes: detecting whether a loss function of the replication index model is smaller than a preset function threshold value or not; and if the loss function of the replication index model is greater than or equal to a preset function threshold value, the replication index model is not sequenced.
Optionally, in this embodiment, the processor may be configured to execute the following steps by a computer program: the inputting the node operation information into an index prediction model and obtaining an index prediction result further comprises: and adjusting the structure and parameters of the neural network model according to the index prediction result.
It should be noted that, for specific examples in this embodiment, reference may be made to the examples described in the foregoing embodiments and optional implementations, and details are not described again in this embodiment.
In addition, in combination with the control system optimization method based on deep learning provided in the foregoing embodiments, a storage medium may also be provided to implement this embodiment. The storage medium having stored thereon a computer program; the computer program, when executed by a processor, implements any of the deep learning based control system optimization methods of the above embodiments.
It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to be limiting. All other embodiments, which can be derived by a person skilled in the art from the examples provided herein without any inventive step, shall fall within the scope of protection of the present application.
It is obvious that the drawings are only examples or embodiments of the present application, and it is obvious to those skilled in the art that the present application can be applied to other similar cases according to the drawings without creative efforts. Moreover, it should be appreciated that in the development of any such actual implementation, as in any engineering or design project, numerous implementation-specific decisions must be made to achieve the developers' specific goals, such as compliance with system-related and business-related constraints, which may vary from one implementation to another.
The term "embodiment" is used herein to mean that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the present application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is to be expressly or implicitly understood by one of ordinary skill in the art that the embodiments described in this application may be combined with other embodiments without conflict.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the patent protection. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.
Claims (9)
1. A control system optimization method based on deep learning is characterized by comprising the following steps:
acquiring node operation information of each node of a control system;
adjusting parameter content in an index prediction model according to the node operation information;
inputting the node operation information into an index prediction model to obtain an index prediction result;
and optimizing the control system according to the index prediction result.
2. The deep learning-based control system optimization method according to claim 1, wherein the inputting the node operation information into an index prediction model and obtaining an index prediction result comprises:
establishing a neural network model, training the neural network model by taking node operation information containing different types of data as a training set to obtain an index prediction model, wherein the input of the index prediction model is the training set, and the output of the index prediction model is the index prediction result.
3. The deep learning-based control system optimization method according to claim 2, wherein the establishing of the neural network model, the training of the neural network model using node operation information containing different types of data as a training set, and the obtaining of the index prediction model comprises:
copying the neural network model to obtain a plurality of copied models;
respectively training the plurality of replication models according to node operation information containing different types of data to obtain a plurality of replication index models;
generating the index prediction model from a plurality of the replicated index models.
4. The deep learning based control system optimization method of claim 3, wherein the generating the metric prediction model from the plurality of replicated metric models comprises:
wherein the replication index model corresponds to a loss function;
sorting the loss functions of the plurality of replication index models, and distributing corresponding weight coefficients for the replication index models;
and performing weighting processing on the plurality of copied index models according to the weighting coefficient to generate the index prediction model.
5. The deep learning based control system optimization method of claim 4, wherein the ranking the loss functions of the plurality of replicated metric models and assigning corresponding weight coefficients to the replicated metric models comprises:
detecting whether a loss function of the replication index model is smaller than a preset function threshold value or not;
and if the loss function of the replication index model is greater than or equal to a preset function threshold value, the replication index model is not sequenced.
6. The deep learning-based control system optimization method according to any one of claims 3 to 5, wherein the inputting the node operation information into an index prediction model and obtaining an index prediction result further comprises:
and adjusting the structure and parameters of the neural network model according to the index prediction result.
7. A deep learning based control system optimization apparatus, comprising:
the information acquisition module is used for acquiring node operation information of each node of the control system;
the parameter adjusting module is used for adjusting parameter contents in the index prediction model according to the node operation information;
the index prediction module is used for inputting the node operation information into an index prediction model to obtain an index prediction result;
and the system optimization module is used for optimizing the control system according to the index prediction result.
8. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the computer program to perform the deep learning based control system optimization method of any one of claims 1 to 6.
9. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the deep learning based control system optimization method of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111469851.7A CN114139811A (en) | 2021-12-04 | 2021-12-04 | Control system optimization method and device based on deep learning and electronic device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111469851.7A CN114139811A (en) | 2021-12-04 | 2021-12-04 | Control system optimization method and device based on deep learning and electronic device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114139811A true CN114139811A (en) | 2022-03-04 |
Family
ID=80387523
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111469851.7A Pending CN114139811A (en) | 2021-12-04 | 2021-12-04 | Control system optimization method and device based on deep learning and electronic device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114139811A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111132190A (en) * | 2019-11-28 | 2020-05-08 | 中通服建设有限公司 | Base station load early warning method and device |
| CN111580411A (en) * | 2020-04-27 | 2020-08-25 | 珠海格力电器股份有限公司 | Control parameter optimization method, device and system |
| CN112099343A (en) * | 2020-07-29 | 2020-12-18 | 福建龙净环保股份有限公司 | Intelligent energy-saving optimization method of electric precipitation system based on neural network |
| CN112445136A (en) * | 2020-12-16 | 2021-03-05 | 北京科技大学 | Thickener prediction control method and system based on continuous time neural network |
-
2021
- 2021-12-04 CN CN202111469851.7A patent/CN114139811A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111132190A (en) * | 2019-11-28 | 2020-05-08 | 中通服建设有限公司 | Base station load early warning method and device |
| CN111580411A (en) * | 2020-04-27 | 2020-08-25 | 珠海格力电器股份有限公司 | Control parameter optimization method, device and system |
| CN112099343A (en) * | 2020-07-29 | 2020-12-18 | 福建龙净环保股份有限公司 | Intelligent energy-saving optimization method of electric precipitation system based on neural network |
| CN112445136A (en) * | 2020-12-16 | 2021-03-05 | 北京科技大学 | Thickener prediction control method and system based on continuous time neural network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3822880A1 (en) | Load prediction method and apparatus based on neural network | |
| CN113361680B (en) | Neural network architecture searching method, device, equipment and medium | |
| CN113239639B (en) | Policy information generation method, policy information generation device, electronic device, and storage medium | |
| CN110363355B (en) | Cloud-edge collaborative forecasting system and method for alumina production index | |
| CN109344969B (en) | Neural network system, training method thereof, and computer-readable medium | |
| CN109783459A (en) | The method, apparatus and computer readable storage medium of data are extracted from log | |
| CN112559963A (en) | Power distribution network dynamic parameter identification method and device | |
| CN108921287A (en) | A kind of optimization method and system of neural network model | |
| CN113935390A (en) | Data processing method, system, device and storage medium | |
| CN114819168B (en) | Quantum comparison method and device for matrix eigenvalues | |
| CN114896216A (en) | Industrial Internet data storage method and system based on block chain and electronic equipment | |
| CN114139811A (en) | Control system optimization method and device based on deep learning and electronic device | |
| CN114678114A (en) | Big data mining evaluation method and big data mining system applied to intelligent medical treatment | |
| CN109726824A (en) | The transfer learning method and terminal device of training pattern | |
| CN110349180B (en) | Human body joint point prediction method and device and motion type identification method and device | |
| CN113962384A (en) | Automatic integrated architecture search system and method for click rate prediction model | |
| CN115169426A (en) | Anomaly detection method and system based on similarity learning fusion model | |
| CN109684471B (en) | Application method of AI intelligent text processing system in new retail field | |
| CN116157809A (en) | Method for automatically determining model compression technology based on compression of artificial intelligence model | |
| CN113887820A (en) | Method and device for predicting fault of electric power spot business system, computer equipment and storage medium | |
| CN112926952A (en) | Cloud computing-combined big data office business processing method and big data server | |
| CN117608809B (en) | Multi-task plan progress prediction system based on gradient lifting decision tree | |
| CN110084511B (en) | Unmanned aerial vehicle configuration method, device, equipment and readable storage medium | |
| CN112861951B (en) | Image neural network parameter determining method and electronic equipment | |
| CN111597415B (en) | Neural network-based distribution network account data penetration method and device |
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 |