CN118277956B - Dust removal, mildew removal and acid removal control method - Google Patents
Dust removal, mildew removal and acid removal control method Download PDFInfo
- Publication number
- CN118277956B CN118277956B CN202410424424.4A CN202410424424A CN118277956B CN 118277956 B CN118277956 B CN 118277956B CN 202410424424 A CN202410424424 A CN 202410424424A CN 118277956 B CN118277956 B CN 118277956B
- Authority
- CN
- China
- Prior art keywords
- removal
- layer
- treatment
- convolution
- cabinet
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/254—Fusion techniques of classification results, e.g. of results related to same input data
- G06F18/256—Fusion techniques of classification results, e.g. of results related to same input data of results relating to different input data, e.g. multimodal recognition
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2431—Multiple classes
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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 OR CALCULATING; 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
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
- G06N3/082—Learning methods modifying the architecture, e.g. adding, deleting or silencing nodes or connections
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Software Systems (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Image Processing (AREA)
Abstract
The invention relates to the field of data processing, in particular to a dust, mildew and acid removal control method, which is characterized in that an environment evaluation model is established, the environment is evaluated according to environment data, whether dust, mildew and acid removal treatment is needed or not is judged on a large scale, the calculated amount is reduced, and the calculated cost is reduced; an image processing neural network model based on combination optimization of a convolutional neural network and a long-short-term neural network is constructed, structured data and unstructured data are processed by using a convolutional memory module and a full-connection module, effective information is extracted from an image, and fusion of multi-source heterogeneous data is achieved. The historical data of dust removal, mould removal and acid removal treatment are obtained, daily treatment frequency and instant treatment judgment are obtained through calculation, the treatment mode is set in the situation, the actual application is more fitted, resources are saved, and the treatment cost is reduced.
Description
Technical Field
The invention relates to the field of data processing, in particular to a dust removal, mildew removal and acid removal control method.
Background
Dust removal and cleaning of various cabinets, such as dense cabinets, file cabinets, communication cabinets, electrical equipment cabinets, elevator control cabinets and other various cabinet bodies, and particularly devices in the cabinet are always dirty and similar. Because some of the charged devices in the screen cabinet are powered devices, a water-flush cleaning mode cannot be used to avoid damaging the devices. However, by adopting the manual cleaning mode, the problems of low working efficiency, high strength, incapability of cleaning up and the like exist. Especially, when an outdoor sweeping mode is adopted, dust in the cabinet overflows to pollute the environment. Because the operation intensity is big, the operation time is long, operational environment is abominable, and dust removal effect is poor, consequently the dust removal cleanness to cabinet word, especially electrified device in it is a difficult problem that is difficult to solve always. Adopt radiating rack of new trend, its dust keeper is placed in the rack in the air intake inboard, forms the plane wind channel almost with the air intake, easily blocks up, difficult clearance, and can only manual cleaning. Especially in northern areas, the wind and sand is large, the period is long, and frequent cleaning is needed.
Some cabinets adopt dustproof nets, but only the cabinet doors can be opened for manual replacement or cleaning, so that the maintenance efficiency is low and the maintenance cost is high. In addition, acid removal control is required for some articles stored in the environment where the cleanliness is highest, such as documents and the like. At present, manual modes are mainly adopted to place independent cleaning work of the plates on each layer, so that cleaning efficiency of staff on the device is low.
The prior art mostly adopts the manual cleaning mode, is not intelligent enough and wastes human resources, can not know the internal condition of the cabinet in real time, easily has the problem of untimely cleaning, and the calculation complexity of the existing intelligent cleaning control method is higher, the realization difficulty is high, and the processing cost is high.
Disclosure of Invention
The embodiment of the application solves the problems that the prior art mostly adopts a manual cleaning mode, is not intelligent enough and wastes human resources, cannot know the internal condition of a cabinet in real time, and is easy to clean in time, and the traditional intelligent cleaning control method has higher calculation complexity, large realization difficulty and high treatment cost, realizes intelligent cleaning, timely and efficient cabinet cleaning, reduces calculation complexity and cleaning cost.
The application provides a dust removal, mildew removal and acid removal control method, which specifically comprises the following technical scheme:
The dust and mildew removal and acid removal control method comprises the following steps:
S1, acquiring environment data and image information by a perception monitoring component, setting up an environment evaluation model, evaluating the environment according to the environment data, and confirming whether dust removal, mildew removal and acid removal are needed; acquiring historical data of dust removal, mildew removal and acid removal treatment of a cabinet, and calculating to obtain daily treatment frequency and immediate treatment judgment;
S2, the specific method for judging the immediate treatment of dust removal, mould removal and acid removal comprises the following steps: constructing an image processing neural network model, extracting effective information from image information, and forming influence factor information; and constructing a deep learning neural network model, judging the degree level of the current cabinet needing dust removal, mildew removal and acid removal treatment according to the influence factor information, and forming a control strategy.
Preferably, the step S1 specifically includes:
Setting up an environment evaluation model, and evaluating the environment according to the environment data of the cabinet.
Preferably, the step S1 specifically includes:
and acquiring historical data of the dust removal, mildew removal and acid removal treatment of the cabinet, processing and analyzing the historical data, and calculating to obtain daily treatment frequency and instant treatment judgment according to the historical data of the dust removal, mildew removal and acid removal treatment of the cabinet. Preferably, the step S2 specifically includes:
An image processing neural network model is constructed, and the image processing neural network model comprises three convolution memory modules, wherein each convolution memory module consists of two convolution layers, a fusion layer, a long-term memory layer and a pooling layer.
Preferably, the step S2 specifically includes:
The fused image features are input into a long-period memory layer, the long-period memory layer is provided with an input gate, a forgetting gate and an output gate, each gate is composed of a Sigmoid neural network layer and a dot multiplication operation and is used for protecting and controlling the states of neurons, and finally dimension reduction operation is carried out on the image features through a pooling layer.
Preferably, the step S2 specifically includes:
After three convolution memory modules, the output of the three convolution memory modules enters a full-connection module, the full-connection module comprises a dimension reduction layer, a full-connection layer and an output layer, the dimension reduction layer continues to carry out dimension reduction operation on image features to obtain a one-dimensional vector, the full-connection layer outputs each vector element, namely effective information extracted from image information, and the effective information and environment information jointly form influence factor information.
Preferably, the step S2 specifically includes:
and constructing a deep learning neural network model, and judging the degree level of the current cabinet needing dust removal, mildew removal and acid removal treatment according to the influence factor information.
Preferably, the step S2 specifically includes:
the convolution layer comprises three convolution units with the same structure, each convolution unit comprises convolution operation, normalization operation and activation operation, the normalization operation and the activation operation are carried out once after each convolution operation, the convergence speed can be increased, independent calculation spaces are set for elements with different attributes, and after three convolution operations, influence factor feature vectors are obtained and input to the correlation layer.
Preferably, the step S2 specifically includes:
The association layer calculates an association function between the influence factor feature vectors, the classification layer obtains classification values according to the attention kernel function, the classification layer inputs the classification values to the output layer, the output layer outputs corresponding processing degree grades according to the classification values, namely, each processing degree grade corresponds to different classification value ranges, and the classification values are compared with each range, so that the corresponding processing degree grade is obtained through matching.
The beneficial effects are that:
the technical schemes provided by the embodiment of the application have at least the following technical effects or advantages:
1. According to the application, an environment evaluation model is established, the environment is evaluated according to the environment data of the cabinet, and whether dust removal, mildew removal and acid removal treatment are needed or not is judged on a large scale, so that the calculated amount is reduced, and the calculation cost is reduced; acquiring historical data of dust removal, mould removal and acid removal treatment, calculating to obtain daily treatment frequency and immediate treatment judgment, and setting a treatment mode in the situation, so that the method is more suitable for practical application, resources are saved, and treatment cost is reduced;
2. an image processing neural network model based on combination optimization of a convolutional neural network and a long-short-term neural network is constructed, structured data and unstructured data are processed by using a convolutional memory module and a full-connection module, effective information is extracted from an image, and fusion of multi-source heterogeneous data is realized; the deep learning neural network model is constructed, the degree level of the current cabinet needing dust removal, mildew removal and acid removal treatment is judged according to the influence factor information, a control strategy is formed, the problem of model overfitting is avoided, the level judgment accuracy of the model is improved, the convergence speed is accelerated, and the high-precision dust removal, mildew removal and acid removal control method is provided.
3. The technical scheme of the application can effectively solve the problems that the prior art mostly adopts a manual cleaning mode, is not intelligent enough and wastes human resources, cannot know the internal condition of the cabinet in real time, and is easy to clean, but the existing intelligent cleaning control method has higher calculation complexity, high realization difficulty and high treatment cost, and the system or the method is subjected to a series of effect investigation, and can finally realize intelligent cleaning through verification, timely and efficient cabinet cleaning, reduce calculation complexity and cleaning cost.
Drawings
FIG. 1 is a flow chart of a dust removal, mildew removal and acid removal control method according to the application;
Fig. 2 is a diagram of a model structure of an image processing neural network according to the present application.
Detailed Description
The embodiment of the application solves the problems that the prior art mostly adopts a manual cleaning mode, is not intelligent enough and wastes human resources, cannot know the internal conditions in real time, and is easy to cause untimely cleaning, and the traditional intelligent cleaning control method has higher calculation complexity, large realization difficulty and high treatment cost.
The technical scheme in the embodiment of the application aims to solve the problems, and the overall thought is as follows:
According to the application, an environment evaluation model is established, the environment is evaluated according to the environment data of the cabinet, and whether dust removal, mildew removal and acid removal treatment are needed or not is judged on a large scale, so that the calculated amount is reduced, and the calculation cost is reduced; acquiring historical data of dust removal, mildew removal and acid removal treatment of a cabinet, and calculating to obtain daily treatment frequency and immediate treatment judgment, wherein the treatment mode is set in the situation, so that the method is more suitable for practical application, resources are saved, and treatment cost is reduced; an image processing neural network model based on combination optimization of a convolutional neural network and a long-short-term neural network is constructed, structured data and unstructured data are processed by using a convolutional memory module and a full-connection module, effective information is extracted from an image, and fusion of multi-source heterogeneous data is realized; the deep learning neural network model is constructed, the degree level of the current cabinet needing dust removal, mildew removal and acid removal treatment is judged according to the influence factor information, a control strategy is formed, the problem of model overfitting is avoided, the level judgment accuracy of the model is improved, the convergence speed is accelerated, and the high-precision dust removal, mildew removal and acid removal control method is provided.
In order to better understand the above technical solutions, the following detailed description will refer to the accompanying drawings and specific embodiments.
Referring to fig. 1, the dust removing, mildew removing and acid removing control method of the application comprises the following steps:
s1, acquiring environmental data and image information in a cabinet by a perception monitoring component, setting up an environmental evaluation model, evaluating the environment according to the environmental data of the cabinet, and confirming whether dust removal, mildew removal and acid removal are needed; acquiring historical data of dust removal, mildew removal and acid removal treatment of a cabinet, and calculating to obtain daily treatment frequency and immediate treatment judgment;
The intelligent control system is used for intelligently controlling a cabinet with dust removing, mildew removing and acid removing functions, wherein dust removing, mildew removing and acid removing components are arranged on the cabinet, the dust removing, mildew removing and acid removing functions of the cabinet are respectively realized by corresponding dust removing, mildew removing and acid removing components, and the dust removing components can be equipment such as a negative pressure dust collector, a fan and the like; the mildew removing component can be a purple light lamp, an activated carbon filter and other devices; the acid removal component can be a filtration device such as an activated carbon filter.
The cabinet is provided with a perception monitoring component which is used for acquiring the environmental conditions and visual change conditions in the cabinet, wherein the environmental conditions comprise temperature, humidity, PM value, carbon dioxide concentration and other data; the visual change condition is image information obtained according to video image change in the cabinet, which is acquired by the perception monitoring component.
The cabinet is provided with a control panel, and the control panel comprises a sensing module, an analysis module and a control module. The sensing and monitoring component transmits the acquired environmental conditions and visual change conditions to the control panel, the sensing module receives data, and the sensing module converts the environmental conditions and visual change condition information into a data format which can be processed by the control panel to obtain environmental data and image information. The sensing module is in communication connection with the analysis module, the analysis module is in communication connection with the control module, the control module is in communication connection with the dust removal, mildew removal and acid removal assembly switch on the cabinet, and the dust removal, mildew removal and acid removal assembly operation is controlled.
The analysis module is used for respectively analyzing and processing the environmental data and the image information, and determining the control strategies of dust removal, mildew removal and acid removal of the cabinet according to the analysis results. Setting up an environment evaluation model, and evaluating the environment according to the environment data of the cabinet, wherein the specific formula is as follows:
Wherein EA represents an environmental evaluation value, alpha 1,α2,α3,α4 is a weight coefficient, QT represents a difference between the environmental temperature of the cabinet and the optimal temperature, QH represents a difference between the environmental humidity of the cabinet and the optimal humidity, QPM represents a PM value of the environment of the cabinet, and QC represents the carbon dioxide concentration in the environment of the cabinet. Setting an environment threshold through experiments, if the environment evaluation value is larger than the environment threshold, indicating that the whole environment of the cabinet is not up to the standard, and carrying out dust removal, mildew removal and acid removal treatment; otherwise, the cabinet environment reaches the standard, and the subsequent dust removal, mildew removal and acid removal treatment are not needed.
And respectively confirming influence factors of dust removal, mildew removal and acid removal treatment and treatment standards, wherein the influence factors can be obtained according to an expert experience method, and the treatment standards are classified into three grades according to the treatment degree: the degree of dust removal, mildew removal and acid removal treatment is determined according to the values of all the influencing factors, and different control instructions are issued according to different treatment degrees. The division of the processing criteria may be set by itself.
And acquiring historical data of dust removal, mildew removal and acid removal treatment of the cabinet, wherein the historical data comprises treatment data, environment data and image information of the cabinet in a certain period before. And processing and analyzing the historical data, and calculating to obtain daily processing frequency and instant processing judgment according to the historical data of the dust removal, mildew removal and acid removal processes of the cabinet. The daily treatment refers to the time interval of dust removal, mildew removal and acid removal treatment under the condition of no external interference, so that the dust removal, mildew removal and acid removal assembly is started periodically to perform daily treatment; the instant processing judgment refers to judging whether dust removal, mildew removal and acid removal are needed or not according to the environmental data and the image information acquired at the current moment.
As a specific embodiment, for the daily treatment frequency, in the dust removal treatment history data, if the cabinet monitoring shows that the date of the environmental data and the image information reaching the treatment standard is 7 days under the condition that the cabinet body does not have any external interference, the dust removal daily treatment frequency is that dust removal treatment is carried out once every 7 days.
S2, the specific method for judging the immediate treatment of dust removal, mould removal and acid removal comprises the following steps: constructing an image processing neural network model, extracting effective information from image information, and forming influence factor information; and constructing a deep learning neural network model, judging the degree level of the current cabinet needing dust removal, mildew removal and acid removal treatment according to the influence factor information, and forming a control strategy.
The historical data of dust removal, mildew removal and acid removal treatment is processed, firstly effective information is required to be extracted from image information, the effective information refers to characteristic differences between current image information and standard images, and whether mildewing, deformation and the like exist can be found through comparison of the current image and the standard images. In order to reduce the computational complexity, the image is scaled, and the specific formula is as follows:
Where (x, y) is the coordinates of the image pixel after scaling, (x ', y') is the coordinates of the image pixel before scaling, H 0×W0 is the original image size, and H 1×W1 is the scaled image size.
The image is subjected to gray processing, an image dividing template is set, namely, an image dividing mode is set up in advance for the image shot at each position, and files in a cabinet are divided independently, so that the image is divided into a plurality of small blocks, and the calculation complexity is reduced.
The method comprises the steps of constructing an image processing neural network model by adopting a mode of combining and optimizing a convolutional neural network and a long-short-period neural network, wherein the image processing neural network model comprises three convolutional memory modules, and each convolutional memory module consists of two convolutional layers, a fusion layer, a long-short-period memory layer and a pooling layer.
Firstly, inputting image information into a first convolution memory module, obtaining image features of different scales through convolution layers of convolution kernels of two different sizes, and fusing the image features of different scales by a fusion layer, wherein the specific formula is as follows:
wherein s k denotes the kth image feature after fusion, And θ k respectively represent two image features of different scales, δ is a scale factor, and μ k represents a featureIs used as a means for controlling the speed of the vehicle,Representing characteristicsIs a variance of (c). The fused image features are input into a long-period memory layer, the long-period memory layer is provided with an input gate, a forgetting gate and an output gate, each gate is composed of a Sigmoid neural network layer and a dot multiplication operation and is used for protecting and controlling the states of neurons, and finally dimension reduction operation is carried out on the image features through a pooling layer.
After three convolution memory modules, the output of the three convolution memory modules enters a full-connection module, the full-connection module comprises a dimension reduction layer, a full-connection layer and an output layer, the dimension reduction layer continues to carry out dimension reduction operation on image features to obtain a one-dimensional vector, the full-connection layer outputs each vector element, namely effective information extracted from image information, and the effective information and environment information jointly form influence factor information.
And constructing a deep learning neural network model, judging the degree level of the current cabinet needing dust removal, mildew removal and acid removal treatment according to influence factor information, and expressing the influence factor information as X= { c 1,c2,...,cn},cn for an nth influence factor information element, wherein any one influence factor information element is used for c i. M groups of influence factor information and corresponding processing degrees are selected as training samples, wherein the processing degrees are represented as { Y 1,Y2,Y3 }, and the processing degrees are respectively three levels of processing degrees. Inputting a group of influence factor information into a deep learning neural network for training, wherein the method comprises the following specific steps of:
The { c 1,c2,...,cn } is input into an input layer, n neurons are arranged in the input layer, the input layer and the convolution layer are in full connection, the convolution layer comprises three convolution units with the same structure, each convolution unit comprises convolution operation, normalization operation and activation operation, the normalization operation and the activation operation are carried out once after each convolution operation, the convergence speed can be increased, and independent calculation spaces are arranged for elements with different attributes. And after three convolution operations, obtaining an influence factor feature vector, and inputting the influence factor feature vector to the association layer.
The association layer calculates an association function between the influence factor feature vectors, and the specific formula is as follows:
Wherein l (c i,cj) represents the association degree between the influence factor feature vectors c i and c j, I epsilon [1, n ], J epsilon [1, n ], I not equal to J, I represents the class I influence factor feature set, J represents the class J influence factor feature set, dust removal, mildew removal and acid removal represent different treatment items, and the influence factor features affecting the same treatment item are called similar influence factor features. The association layer inputs the calculation result to the classification layer.
The classification layer obtains classification values according to the attention kernel function, and a specific calculation formula is as follows:
the classification layer inputs the classification value into the output layer, the output layer outputs the corresponding processing degree grade according to the classification value, namely, each processing degree grade corresponds to different classification value ranges, and the classification value is compared with each range, so that the corresponding processing degree grade is obtained through matching.
The control module issues corresponding control instructions according to different treatment degree grades, and the switch of the dust removing, mildew removing and acid removing assembly performs dust removing, mildew removing and acid removing treatments of corresponding degrees according to the received control instructions.
In conclusion, the dust removal, mildew removal and acid removal control method is completed.
The technical scheme provided by the embodiment of the application at least has the following technical effects or advantages:
1. According to the application, an environment evaluation model is established, the environment is evaluated according to the environment data of the cabinet, and whether dust removal, mildew removal and acid removal treatment are needed or not is judged on a large scale, so that the calculated amount is reduced, and the calculation cost is reduced; acquiring historical data of dust removal, mould removal and acid removal treatment, calculating to obtain daily treatment frequency and immediate treatment judgment, and setting a treatment mode in the situation, so that the method is more suitable for practical application, resources are saved, and treatment cost is reduced;
2. an image processing neural network model based on combination optimization of a convolutional neural network and a long-short-term neural network is constructed, structured data and unstructured data are processed by using a convolutional memory module and a full-connection module, effective information is extracted from an image, and fusion of multi-source heterogeneous data is realized; the deep learning neural network model is constructed, the degree level of the current cabinet needing dust removal, mildew removal and acid removal treatment is judged according to the influence factor information, a control strategy is formed, the problem of model overfitting is avoided, the level judgment accuracy of the model is improved, the convergence speed is accelerated, and the high-precision dust removal, mildew removal and acid removal control method is provided.
Effect investigation:
The technical scheme of the application can effectively solve the problems that the prior art mostly adopts a manual cleaning mode, is not intelligent enough and wastes human resources, cannot know the internal condition of the cabinet in real time, and is easy to clean, but the existing intelligent cleaning control method has higher calculation complexity, high realization difficulty and high treatment cost, and the system or the method is subjected to a series of effect investigation, and can finally realize intelligent cleaning through verification, timely and efficient cabinet cleaning, reduce calculation complexity and cleaning cost.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
Claims (2)
1. The dust and mildew removal and acid removal control method is characterized by comprising the following steps of:
S1, acquiring environment data and image information by a perception monitoring component, setting up an environment evaluation model, and evaluating the environment according to the environment data of a cabinet, wherein the specific formula is as follows:
，
wherein, Representing an environmental evaluation value; are all weight coefficients; representing the difference between the ambient temperature of the cabinet and the optimal temperature; representing the difference between the cabinet environment humidity and the optimal humidity; Representing a cabinet environment PM value; representing the carbon dioxide concentration in the cabinet environment; confirming whether dust removal, mildew removal and acid removal treatment are needed; acquiring historical data of dust removal, mildew removal and acid removal treatment of a cabinet, and calculating to obtain daily treatment frequency and immediate treatment judgment;
S2, the specific method for judging the immediate treatment of dust removal, mould removal and acid removal comprises the following steps: constructing an image processing neural network model, wherein the image processing neural network model comprises three convolution memory modules, and each convolution memory module consists of two convolution layers, a fusion layer, a long-term memory layer and a pooling layer;
Firstly, inputting image information into a first convolution memory module, obtaining image features with different scales through convolution layers of convolution kernels with different sizes, and fusing the image features with different scales by a fusion layer to obtain fused image features; the fused image features are input into a long-period memory layer, the long-period memory layer is provided with an input gate, a forgetting gate and an output gate, each gate is composed of a Sigmoid neural network layer and a dot multiplication operation, and finally the dimension reduction operation is carried out on the image features through a pooling layer;
After three convolution memory modules, the output of the three convolution memory modules enters a full-connection module, the full-connection module comprises a dimension reduction layer, a full-connection layer and an output layer, the dimension reduction layer continues to carry out dimension reduction operation on image features to obtain a one-dimensional vector, the full-connection layer outputs each vector element, namely effective information extracted from image information, and the effective information and environment information form influence factor information together;
Constructing a deep learning neural network model, and judging the degree level of the current cabinet needing dust removal, mildew removal and acid removal treatment according to the influence factor information; inputting a group of influence factor information into a deep learning neural network for training, wherein the method comprises the following specific steps of: the method comprises the steps that influence factor information is input into an input layer, the input layer and a convolution layer are in a full connection relationship, the convolution layer comprises three convolution units with the same structure, each convolution unit comprises convolution operation, normalization operation and activation operation, the normalization operation and the activation operation are carried out once after each convolution operation, influence factor feature vectors are obtained after three convolution operations, and the influence factor feature vectors are input into an association layer;
the association layer calculates an association function between feature vectors of the influence factors, the classification layer obtains classification values according to the attention kernel function, the classification layer inputs the classification values to the output layer, the output layer outputs corresponding processing degree grades according to the classification values, namely, each processing degree grade corresponds to different classification value ranges, the classification values are compared with each range, and accordingly, the corresponding processing degree grades are obtained through matching, and a control strategy is formed.
2. The method for controlling dust removal, mildew removal and acid removal according to claim 1, wherein the step S1 specifically comprises:
And acquiring historical data of the dust removal, mildew removal and acid removal treatment of the cabinet, processing and analyzing the historical data, and calculating to obtain daily treatment frequency and instant treatment judgment according to the historical data of the dust removal, mildew removal and acid removal treatment of the cabinet.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410424424.4A CN118277956B (en) | 2024-04-09 | 2024-04-09 | Dust removal, mildew removal and acid removal control method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410424424.4A CN118277956B (en) | 2024-04-09 | 2024-04-09 | Dust removal, mildew removal and acid removal control method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118277956A CN118277956A (en) | 2024-07-02 |
| CN118277956B true CN118277956B (en) | 2024-10-25 |
Family
ID=91646654
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410424424.4A Active CN118277956B (en) | 2024-04-09 | 2024-04-09 | Dust removal, mildew removal and acid removal control method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118277956B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118913506A (en) * | 2024-07-26 | 2024-11-08 | 绍兴市上虞春晖风冷设备有限公司 | Corrosion prevention method, corrosion prevention system, corrosion prevention terminal and corrosion prevention storage medium for ship shaft power instrument |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006015061A2 (en) * | 2004-07-29 | 2006-02-09 | Can Technologies, Inc. | System and method for optimizing animal production based on environmental nutrient inputs |
| CN113642627A (en) * | 2021-08-09 | 2021-11-12 | 中国人民解放军海军航空大学航空作战勤务学院 | Image and decision multi-source heterogeneous information fusion identification method and device based on deep learning |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007229554A (en) * | 2006-02-28 | 2007-09-13 | Casio Comput Co Ltd | Performance evaluation method of dust cleaner |
| WO2023275897A1 (en) * | 2021-06-30 | 2023-01-05 | Infilect Technologies Private Limited | System and method for preventing shrinkage in a retail environment using real-time camera feeds |
| KR20230086233A (en) * | 2021-12-08 | 2023-06-15 | 순천대학교 산학협력단 | A Reconstructed Convolutional Neural Network Using Conditional Min Pooling and Method thereof |
| CN116592993A (en) * | 2023-04-11 | 2023-08-15 | 辽宁科技大学 | Mechanical vibration fault diagnosis method based on deep learning |
-
2024
- 2024-04-09 CN CN202410424424.4A patent/CN118277956B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006015061A2 (en) * | 2004-07-29 | 2006-02-09 | Can Technologies, Inc. | System and method for optimizing animal production based on environmental nutrient inputs |
| CN113642627A (en) * | 2021-08-09 | 2021-11-12 | 中国人民解放军海军航空大学航空作战勤务学院 | Image and decision multi-source heterogeneous information fusion identification method and device based on deep learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118277956A (en) | 2024-07-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109305534B (en) | Self-adaptive control method of belt conveyor of coal wharf based on computer vision | |
| CN109858569A (en) | Multi-tag object detecting method, system, device based on target detection network | |
| CN110009095B (en) | Efficient segmentation method of road driving area based on deep feature compression convolutional network | |
| CN118277956B (en) | Dust removal, mildew removal and acid removal control method | |
| CN108710919A (en) | A kind of crack automation delineation method based on multi-scale feature fusion deep learning | |
| CN114255357B (en) | A method for group-rearing pig identity recognition and health monitoring based on computer vision | |
| CN112132784A (en) | Method for classifying and segmenting industrial magnetic tile defect image based on small sample deep convolution neural network | |
| CN109978835A (en) | A kind of online assembly defect recognition system and method | |
| CN117253192A (en) | Intelligent system and method for silkworm breeding | |
| CN110599459A (en) | Underground pipe network risk assessment cloud system based on deep learning | |
| CN112802006A (en) | Deep learning-based edge calculation motor oil stain identification method | |
| CN118559940B (en) | Self-adaptive adjusting system and method based on automatic burr removing equipment | |
| CN109444604A (en) | A kind of DC/DC converter method for diagnosing faults based on convolutional neural networks | |
| CN116675277B (en) | Control method of low-energy-consumption wastewater concentration system based on Internet of things | |
| CN118230045A (en) | A garbage classification system and method based on visual recognition technology | |
| CN115762169B (en) | Unmanned intelligent control system and method for sanitation vehicle | |
| CN118690276B (en) | Recognition method for sensing countermeasure learning damage of high-speed railway wheel rail | |
| CN119443704A (en) | A method and system for dynamic optimization control of smart apartment resources | |
| CN118675112A (en) | Water ecology intelligence accuse algae equipment and early warning system | |
| CN111383219B (en) | Method and system for intelligently detecting cleanliness of aerial work platform equipment | |
| CN119206460A (en) | A method and system for identifying benthic animal community structure based on image recognition | |
| CN117269168B (en) | New energy automobile precision part surface defect detection device and detection method | |
| CN118552578A (en) | A recycling and identification method for waste lithium batteries based on image analysis | |
| CN118372853A (en) | A car automatic driving system | |
| CN117876353A (en) | Electronic equipment surface scratch detecting system based on visual recognition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |