Research On Monitoring Of Key Variables And Operation Optimization In Boiler Side Of Coal - Fired Units

In the foreseeable future,thermal power plant,especially the coal-fired units,will remain the dominant electricity supply in China.However,due to the complex mechanism,the incomplete measurements,and the inherent disturbance features of the coal-fired unit such as time-varying,nonlinear,great time constant and lag time,multiple variables,various interferences,etc.,there are a lot of challenging problems to be investigated in the process monitoring,operation,and optimization.Under the ever increasing pressure on environment protection,the demand for the key variable monitoring and operation optimization becomes more and more urgent.In this study,the mechanism modeling and the key issues in the operation optimization for the coal-fired power plant are investigated.The main innovations of this paper include:(1)The boiler side mechanism model of coal-fired units is established.The model covers the heat exchanger system,metal wall energy compensation,and heat loss.(2)The approach of flue gas temperature estimation in the high temperature region is presented based on the boiler side mechanism model.The flue gas temperature in the furnace exit is calculated by the particle swarm optimization algorithm.Then the quasi-distribution parameter model is established for the semi-radiative heating surface along the flue gas flow direction.By calculating the radiation and convection part of energy seperately,the flue gas temperature of some key locations are obtained.The calculation value is validated using the original assays in the spatial scale and in time scales of 24 hours,respectively.(3)Based on the quasi distribution parameter model,ash pollution of the heat exchanger is evaluated.The ash pollution coefficient is taken as the characteristic ash deposition parameter of the heat exchanger.The coefficient is calculated based on the energy balance of the flue gas/working fluid.Then,ash pollution coefficient prediction model is established.By solving the optimization problem of soot-blowing economic benefit,the soot-blowing time and soot-blowing interval are determined,according to these,the strategy of optimizing soot-blowing is formulated.(4)The boiler efficiency model based on least squares support vector regression(LSSVR)is established.The model is validated by several sets of historical data,and compared with another model based on the artificial neural network(ANN)algorithm.The results show that both of them can simulate the real boiler operation process,and the proposed model has higher precision and reliability.The influence of primary air flow on boiler efficiency is studied based on the boiler efficiency prediction model proposed in this paper.It is found that there are potential optimization space for boiler efficiency,by optimizing the primary air volume.