CN-122018420-A - Energy consumption optimization system and method for dust removal fan of steel plant

Abstract

The invention belongs to the technical field of steelmaking dust removal, and particularly relates to an energy consumption optimization system and method for dust removal fans of a steel plant, wherein the method comprises the following steps that S1, an intelligent ammeter and a PLC (programmable logic controller) collect operation data and energy consumption data of a plurality of dust removal fans; the method comprises the steps of S2, performing frame cutting processing on video streams by a visual AI+ application server, S3, calling an energy consumption optimizing AI model module by the application server, inputting preprocessed historical data and real-time data, S4, pushing a set value to a PLC by an intelligent regulation module, adjusting the rotating speeds of a plurality of dust removal fans, S5, evaluating the performance of the model every 24 hours by a model updating and version management module, triggering updating and recording version information as required, and S6, displaying core indexes such as total power consumption, power saving and the like in real time by a factory-level energy consumption large screen and a system/process-level energy consumption billboard. The system and the method collect the running state, the energy consumption data and the output data of the equipment in real time, provide the dynamic monitoring and analyzing functions for the equipment, realize the control mode intellectualization and avoid the electric energy waste.

Inventors

• LIAN YUNBIN
• Hao Yuli
• ZHANG YONGHAI
• JIA YUCHAO
• BO XIAOYAN
• LI HONGYUN
• XUE GUOHUA
• CHEN JIA
• YANG HONGFEI

Assignees

• 山西太钢不锈钢股份有限公司

Dates

Publication Date

20260512

Application Date

20260116

Claims (7)

1. 1. The energy consumption optimizing system of the dust removing fan of the steel plant is characterized by comprising an equipment layer, an infrastructure layer, a data layer, a platform layer, an application layer and a display layer, wherein the layers are sequentially connected in a cooperative communication manner; The equipment layer comprises a plurality of dust removing fans, a PLC (programmable logic controller), a smart electric meter and an AI gun camera, wherein the dust removing fans are connected with the PLC and the smart electric meter, and the PLC is connected with the smart electric meter to collect operation data, energy consumption data and dust image data; the infrastructure layer comprises an ICG gateway and an industrial optical fiber switch, the PLC is connected with the ICG gateway, the ICG gateway is connected with the industrial optical fiber switch and an AI gun camera, the data layer comprises a vision AI+ application server, an application server and an Internet of things/storage server, the industrial optical fiber switch is connected with the vision AI+ application server and the application server, the vision AI+ application server and the application server are connected, data are transmitted through the ICG gateway and the industrial optical fiber switch, the vision AI+ application server and the application server are both connected with the Internet of things/storage server, and data processing and storage are carried out through the vision AI+ application server, the application server and the Internet of things/storage server; the platform layer comprises an energy consumption optimizing AI model module and a model updating and version management module, the Internet of things/storage server is connected with the energy consumption optimizing AI model module and the model updating and version management module to provide modeling and iterative support, the application layer comprises an intelligent regulation and control module, the Internet of things/storage server is connected with the intelligent regulation and control module, the energy consumption optimizing AI model module, the model updating and version management module and the intelligent regulation and control module are mutually connected, model output parameters are received and pushed to a PLC to realize fan regulation and control, the display layer comprises a factory-level energy consumption large screen and a system/procedure-level energy consumption billboard, the Internet of things/storage server is connected with the factory-level energy consumption large screen, the plant-level energy consumption large screen is connected with a system/process-level energy consumption signboard, and the operation state and the optimization effect are displayed.
2. 2. The energy consumption optimizing system of the dust removing fan of the steel plant according to claim 1, wherein the application server adopts a CPU4410Y (24 cores 2.0 GHz), a memory 64GB, a deployment energy consumption optimizing application service, an AI model training and issuing service, a relational database, a Redis memory database and a time sequence database.
3. 3. The energy consumption optimizing system of the dust removing fan of the steel plant according to claim 1, wherein the AI camera adopts 800-ten-thousand cameras, and the focal length mode is fixed focus for on-site environment monitoring and detection.
4. 4. The energy consumption optimizing system of the dust removing fan of the steel plant according to claim 1, wherein the intelligent regulation and control module supports automatic/manual mode switching and ash removing tasks according to peak-valley electricity time periods, and the period is set to be 1-3 times/day or 10-15 times/week.
5. 5. A method for optimizing the energy consumption of a dust-removing fan of a steel plant based on the system of claim 1, which is characterized by comprising the following specific steps: s1, an intelligent ammeter and a PLC (programmable logic controller) collect operation data and energy consumption data of a plurality of dust removal fans at 1-5S sampling intervals through an ICG (information and communication gateway), and an AI gun camera collects dust video streams and transmits the dust video streams to a vision AI+ application server; S2, performing 1-3 frames/S frame cutting processing on the video stream by a visual AI+ application server, identifying a smoke region by a deep learning target detection algorithm, analyzing dynamic changes by combining a background difference method, calculating dust concentration based on an image ambiguity algorithm, and calculating the error of less than or equal to 5%; S3, an application server calls an energy consumption optimization AI model module, inputs the preprocessed historical data and real-time data, identifies working conditions through a K-means clustering algorithm, outputs an optimal frequency set value in a 20-50Hz interval through a particle swarm optimization algorithm, and outputs delay less than or equal to 10S; S4, pushing the set value to a PLC (programmable logic controller) by the intelligent regulation module, adjusting the rotating speeds of a plurality of dust removal fans, and simultaneously collecting and controlling data and feeding back the data to the model; S5, the model updating and version management module evaluates the performance of the model every 24 hours, triggers updating as required and records version information; And S6, displaying core indexes such as total power consumption, power saving and the like in real time by the factory-level energy consumption large screen and the system/process-level energy consumption signboard.
6. 6. The method for optimizing energy consumption of a dust removal fan of a steel plant according to claim 5, wherein the energy consumption optimization AI model module in the step S3 adopts a random forest regression model, and the super parameters are optimized through grid search, so that the prediction accuracy of a verification set is more than or equal to 95%, the mean square error is less than or equal to 0.02, and the generalization capability is ensured.
7. 7. The method for optimizing energy consumption of dust-removing fan in steel plant according to claim 5, wherein in step S4, the intelligent control module pushes 1 set value every 3-5 minutes in automatic mode, and the manual mode only outputs recommended value.

Description

Energy consumption optimization system and method for dust removal fan of steel plant Technical Field The invention belongs to the technical field of steelmaking dust removal, and particularly relates to an energy consumption optimization system and method for a dust removal fan of a steel plant. Background In the production process of steel plants, a large amount of exhaust gas such as smoke dust, dust and the like is generated, and the exhaust gas contains harmful substances such as sulfur dioxide, nitrogen oxides and the like, so that the exhaust gas has great harm to human bodies and the environment. The dust removing system effectively reduces the concentration of smoke dust in the air by collecting and treating the smoke dust, ensures the health of workers and reduces environmental pollution. The operation of the current dust removal system has the following prominent problems: (1) The power of the motors of the fans exceeds 1000kW, the motors run for a long time under full load or half load, the annual electricity consumption is huge, the electricity cost is high, and the energy conservation and cost reduction demands are urgent; (2) The control mode is behind, most fans adopt a fixed-frequency operation or a high-speed/low-speed two-stage switching mode, stepless speed regulation cannot be realized, actual dust production change is difficult to match, the air quantity is excessive in a non-operation period, and the electric energy waste is serious; (3) The system is regulated by manual experience, such as newly turning dust removal and the like, is still regulated and controlled by post personnel according to experience, has delayed response and untimely regulation, not only influences the dust removal effect, but also cannot ensure the optimal energy conservation. Disclosure of Invention The invention aims to provide an energy consumption optimizing system and method for a dust removing fan of a steel plant, which solve the problems of higher energy consumption level, lagging control mode and dependence on manual experience adjustment in the operation of the current dust removing system. In order to achieve the above purpose, the invention adopts the following technical scheme: The energy consumption optimizing system for the dust removing fan of the steel plant comprises an equipment layer, an infrastructure layer, a data layer, a platform layer, an application layer and a display layer, wherein the layers are sequentially connected in a cooperative communication manner; The equipment layer comprises a plurality of dust removing fans, a PLC (programmable logic controller), a smart electric meter and an AI gun camera, wherein the dust removing fans are connected with the PLC and the smart electric meter, and the PLC is connected with the smart electric meter to collect operation data, energy consumption data and dust image data; the infrastructure layer comprises an ICG gateway and an industrial optical fiber switch, the PLC is connected with the ICG gateway, the ICG gateway is connected with the industrial optical fiber switch and an AI gun camera, the data layer comprises a vision AI+ application server, an application server and an Internet of things/storage server, the industrial optical fiber switch is connected with the vision AI+ application server and the application server, the vision AI+ application server and the application server are connected, data are transmitted through the ICG gateway and the industrial optical fiber switch, the vision AI+ application server and the application server are both connected with the Internet of things/storage server, and data processing and storage are carried out through the vision AI+ application server, the application server and the Internet of things/storage server; the platform layer comprises an energy consumption optimizing AI model module and a model updating and version management module, the Internet of things/storage server is connected with the energy consumption optimizing AI model module and the model updating and version management module to provide modeling and iterative support, the application layer comprises an intelligent regulation and control module, the Internet of things/storage server is connected with the intelligent regulation and control module, the energy consumption optimizing AI model module, the model updating and version management module and the intelligent regulation and control module are mutually connected, model output parameters are received and pushed to a PLC to realize fan regulation and control, the display layer comprises a factory-level energy consumption large screen and a system/procedure-level energy consumption billboard, the Internet of things/storage server is connected with the factory-level energy consumption large screen, the plant-level energy consumption large screen is connected with a system/process-level energy consumption signboard, and the operation state and the optimization effect are disp