CN-121979138-A - Multi-objective optimization-based energy-saving adjustment control method and device for pulverizing system

CN121979138ACN 121979138 ACN121979138 ACN 121979138ACN-121979138-A

Abstract

The invention provides an energy-saving regulation control method and device for a pulverizing system based on multi-objective optimization, and the method and device comprise the steps of obtaining operation parameters of the pulverizing system, preprocessing, carrying out partition identification on current working conditions based on an expert rule base, dynamically configuring a multi-objective optimization weight matrix, carrying out multi-objective optimization by utilizing an improved particle swarm optimization algorithm, realizing the balance between global search and local development by self-adaptively adjusting inertia weight and learning factors, inputting an optimization result into an expert rule base, carrying out hard constraint verification and experience correction to generate control parameters meeting safety and energy efficiency requirements, converting the corrected control parameters into control instructions, and carrying out closed loop feedback regulation to form a periodic optimization control flow. The multi-target collaborative optimization control of the powder making system under different load working conditions can be realized, the powder making unit consumption is obviously reduced, the fineness stability of the coal powder is improved, the algorithm convergence speed and the fault early warning advance are improved, and the dynamic balance of the system safety and the energy efficiency is ensured.

Inventors

FU QI
YANG FENGSONG
ZHANG RUIMIN
CAO WEI
JU RONG
LIU DINGPO
YANG PEIJUN
YUAN PANFENG
JIANG QINGYUN
CHEN HAIYAN

Assignees

河北凯清环保科技有限公司
西安西热锅炉环保工程有限公司

Dates

Publication Date: 20260505
Application Date: 20260112

Claims (10)

1. The multi-objective optimization-based energy-saving adjustment control method for the pulverizing system is characterized by comprising the following steps of: S1, collecting and preprocessing operation parameters of a coal mill, wherein the operation parameters comprise mill outlet temperature, primary air quantity, coal supply quantity, coal mill current, outlet pressure, inlet primary air pressure and inlet primary air temperature; S2, identifying a current working condition interval based on the operation parameters, and configuring multi-objective optimization weights according to a preset expert rule base, wherein the multi-objective optimization weights correspond to priority relations of energy consumption, stability and response speed; S3, adopting an improved particle swarm optimization algorithm to dynamically optimize the operation parameters, wherein the improved particle swarm optimization algorithm generates preliminary optimization control parameters under preset constraint conditions through an inertia weight linear decreasing strategy and a learning factor self-adaptive adjustment mechanism; S4, inputting the preliminary optimization control parameters into the expert rule base to carry out hard constraint verification and experience correction, wherein the hard constraint verification comprises that the primary air quantity is not lower than the minimum fluidization air quantity and the grinding outlet temperature is not higher than a preset threshold value, and the experience correction comprises that an advanced adjustment strategy is triggered according to the load change rate, a final control instruction is output, and the final control instruction is fed back to an executing mechanism to adjust the running state of the coal mill.
2. The method of claim 1, wherein S1 comprises: S11, collecting 7 core parameters of mill outlet temperature, primary air quantity, coal supply quantity, coal mill current, outlet pressure, inlet primary air pressure and inlet primary air temperature at sampling intervals of 1 minute; s12, a filtering denoising technology is adopted in the preprocessing stage, the data length is ensured to cover a full working condition period of at least 1 week, and the data distribution characteristics of high, medium and low load typical scenes are included.
3. The method of claim 1, wherein S2 comprises: s21, dividing the unit load into a high load section, a medium load section and a low load section; S22, automatically matching a target weight matrix according to a load interval, wherein the stability weight of the high load interval is 0.6, the energy consumption weight of the medium load interval is 0.5, and the response speed weight of the low load interval is 0.5.
4. The method of claim 1, wherein S3 further comprises: s31, adopting an inertia weight linear decreasing strategy, and according to a formula Calculating a current inertial weight, wherein 、、 ; And S32, when the fitness is not improved for 10 continuous generations of particles, the cognitive factor is adjusted to be 1.5, and the social factor is adjusted to be 2.5 so as to enhance the population information sharing.
5. The method of claim 1, wherein S4 comprises: S41, ensuring that the primary air quantity is not lower than 60% rated value in hard constraint verification, and ensuring that the grinding outlet temperature is not higher than 75 ℃; s42, when the load change rate is in the experience correction And triggering an advanced regulation strategy, and presetting the coal feeding amount by +5% to avoid the risk of coal blockage.
6. The method as recited in claim 1, further comprising: S5, fault diagnosis information is extracted based on local features of the coal mill operation image, wherein the local features comprise shapes, textures and particle size distribution, 7 typical fault types such as accumulation blocking, insufficient crushing, dust backflow and the like are identified through an expert rule base, and fault early warning advance is improved to be the same as that of the fault types 。
7. The utility model provides a powder process system energy-saving regulation controlling means based on multi-objective optimization which characterized in that includes: The operation parameter acquisition and pretreatment module is used for acquiring the operation parameters of the coal mill and carrying out pretreatment, wherein the operation parameters comprise mill outlet temperature, primary air quantity, coal supply quantity, coal mill current, outlet pressure, inlet primary air pressure and inlet primary air temperature; The working condition identification and weight configuration module is used for identifying a current working condition interval based on the operation parameters, and configuring multi-objective optimization weights according to a preset expert rule base, wherein the multi-objective optimization weights correspond to the priority relation of energy consumption, stability and response speed; The dynamic optimizing module is used for dynamically optimizing the operation parameters by adopting an improved particle swarm optimization algorithm, and the improved particle swarm optimization algorithm generates preliminary optimized control parameters under preset constraint conditions through an inertia weight linear decreasing strategy and a learning factor self-adaptive adjustment mechanism; The hard constraint verification and experience correction module is used for inputting the preliminary optimization control parameters into the expert rule base to carry out hard constraint verification and experience correction, the hard constraint verification comprises that the primary air quantity is not lower than the minimum fluidization air quantity and the mill outlet temperature is not higher than a preset threshold value, and the experience correction comprises that an advanced adjustment strategy is triggered according to the load change rate, a final control instruction is output, and the final control instruction is fed back to an executing mechanism to adjust the running state of the coal mill.
8. The apparatus of claim 7, wherein the operating parameter acquisition and pre-processing module is further to: 7 core parameters including mill outlet temperature, primary air quantity, coal supply quantity, coal mill current, outlet pressure, inlet primary air pressure and inlet primary air temperature are collected at sampling intervals of 1 minute; The preprocessing stage adopts a filtering denoising technology, and ensures that the data length covers a full working condition period of at least 1 week, and the full working condition period comprises data distribution characteristics of high, medium and low load typical scenes.
9. A computer device comprising a processor and a memory; Wherein the processor executes a program corresponding to the executable program code by reading the executable program code stored in the memory, for implementing a multi-objective optimization-based pulverizing system energy saving adjustment control method according to any one of claims 1 to 6.
10. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the program when executed by a processor implements a multi-objective optimization-based pulverizing system energy saving adjustment control method according to any one of claims 1 to 6.

Description

Multi-objective optimization-based energy-saving adjustment control method and device for pulverizing system Technical Field The invention belongs to the technical field of control, and particularly relates to an energy-saving adjusting control method and device for a pulverizing system based on multi-objective optimization. Background Coal-making system as fire coal the core auxiliary machine system of the power plant, the energy consumption of the system occupies 20 percent of the total station service electricity, and directly influences the running cost and the equipment reliability. The existing control technology of the pulverizing system can be divided into traditional control, intelligent optimization and hybrid architecture. The traditional method mainly comprises a PLC central automation system and PID control, and part of enterprises adopt closed-loop control, but the coverage rate is less than 20%. The intelligent optimization technology comprises prediction control (such as gradient lifting tree working condition prediction) based on machine learning, fuzzy control (such as a multidimensional fuzzy controller for adjusting feeding quantity and throttle opening) and a multi-objective optimization algorithm. The hybrid architecture is typically such as the NSGA-III + DDPG architecture. The existing control technology of the pulverizing system has obvious defects in the aspects of energy consumption optimization, dynamic adaptability, fault early warning and the like. In terms of energy consumption, the electricity consumption of the coal pulverizing system of the medium-storage steel ball coal mill is 20% of the station service electricity, and the unit consumption reduction rate of the traditional control method is only 3.2%. On dynamic control, the traditional PID control precision is limited, the outlet temperature fluctuation of the coal mill is +/-3.5 ℃, the variable load working condition stabilization time is 32s, and the unit consumption is obviously increased in low load. The optimization algorithm has the problems of low convergence speed and low solution quality, NSGA-II has low convergence speed, NSGA-III has high complexity, and the calculation efficiency of a large-scale scene is poor. The fault early warning response is lagged, and the diagnosis lead of the vibration signal is only 10min. In modeling aspect, the prediction error of the pure data driving model is 7.8% under the working condition of coal variety change, and the interpretation is poor. The evaluation system is limited to a single energy efficiency index, the dimensionalities of reliability, environmental influence and the like are not covered, and multi-objective optimization is often simplified into a weighted sum, so that the system cannot adapt to the dynamic change of working conditions. Disclosure of Invention The present invention aims to solve at least one of the technical problems in the related art to some extent. Therefore, a first object of the present invention is to provide an energy-saving adjustment control method for a pulverizing system based on multi-objective optimization. The second aim of the invention is to provide an energy-saving adjusting and controlling device of a pulverizing system based on multi-objective optimization. A third object of the invention is to propose a computer device. A fourth object of the present invention is to propose a non-transitory computer readable storage medium. To achieve the above objective, an embodiment of a first aspect of the present invention provides an energy-saving adjustment control method for a pulverizing system based on multi-objective optimization, including: S1, collecting and preprocessing operation parameters of a coal mill, wherein the operation parameters comprise mill outlet temperature, primary air quantity, coal supply quantity, coal mill current, outlet pressure, inlet primary air pressure and inlet primary air temperature; S2, identifying a current working condition interval based on the operation parameters, and configuring multi-objective optimization weights according to a preset expert rule base, wherein the multi-objective optimization weights correspond to priority relations of energy consumption, stability and response speed; S3, adopting an improved particle swarm optimization algorithm to dynamically optimize the operation parameters, wherein the improved particle swarm optimization algorithm generates preliminary optimization control parameters under preset constraint conditions through an inertia weight linear decreasing strategy and a learning factor self-adaptive adjustment mechanism; S4, inputting the preliminary optimization control parameters into the expert rule base to carry out hard constraint verification and experience correction, wherein the hard constraint verification comprises that the primary air quantity is not lower than the minimum fluidization air quantity and the grinding outlet temperature is not higher than