Control And Optimization Of Air Cooled Condenser Fan Operation

Thermal power generation occupies a dominant position in the power supply structure of our country,and the traditional thermal power unit has great potential for energy saving and consumption reduction.Cold-end system,as an important auxiliary system of thermal power unit,has a very important influence on the safety and economic operation of the unit.Therefore,studying the subsystems involved in the cold-end system,optimizing and improving the overall performance of the system,is an important means of saving energy and reducing energy consumption of the power plant and improving thermal economy of the unit.In this paper,we mainly focus on modeling and optimization of cold junction system,and try our best to establish a neural network model to train each fan speed based on the theoretical analysis and operation data.At the same time,Of the back-pressure model of the unit,and energy-saving optimization of cold-side system proposed effective solutions to achieve air-cooled fan matrix control.First of all,the structure and characteristics of the direct air cooling system are briefly introduced.The problems existing in the cold end system are introduced.The mathematical model of the direct air cooled condenser is deduced through the mechanism analysis.Thermal coefficient,fan-related parameters of the calculation.On this basis,the relationship between unit back pressure increment and unit load increment is studied,and the influence of back pressure on the economic operation of steam turbine unit is described in detail.Then,the definition of the net profit of the cold-end system is proposed,and the target energy-saving optimization architecture of the air-cooled fan speed regulation system is established.Based on this,a calculation method to determine the optimal vacuum of the unit is proposed.In this paper,from the environmental factors as the starting point to analyze the impact on the unit performance,combined with the problems in the theoretical calculations,expand the system modeling based on operational data.Taking the BP neural network as the research idea,the input samples of the neural network are enriched and de-redundant based on the data processing methods such as principal component analysis and data filtering,and the mathematical relationships existing between the parameters are used for integration.As much as possible,Physical meaning,and carried out a certain model validation.On this basis,the optimization direction of the optimal speed with the back pressure of the unit as an independent variable is put forward,and the antifreeze operation solution of the unit with the fin wall temperature being controlled is proposed.In the end of this paper,the fan speed is taken as the model input and a series of other parameters are used to identify the backpressure model of the unit by genetic algorithm,and a better identification model is obtained.This paper integrates the neural network model into the multivariable predictive control system to realize the matrix control of the wind turbine.The scheme has good engineering application value.