A Research On Chain Furnace Combustion Control Based On BP Fuzzy Neural Network

In the combustion control of chain furnace,non-linear,time-varying and multivariable phenomenon are common to see.They make the combustion control of chain furnace become more complex and difficult,while they could lead to combustion instability.Therefore,a research on t he optimization of combustion control of chain furnace is of both great theoretical value and practical value.To achieve the highest efficiency of chain furnace,the coal in furnace shall fully burn through regulating the amounts of airflow and coal supply to reach the best air-coal ratio gradually according to the differences between real-time power and target power,adjacent efficiency and the airflow.A control program that adopts methods of coarse adjustment,fine tuning and superior adjustment is proposed in this research to coordinate combustion optimization of chain furnace with BP neural networks,as well as adaptive and self-optimizing control.In order to achieve combustion of fuel and minimal heat loss,the time optimum efficiency is compared with former optimum efficiency,and the program will choose the more optimum efficiency to implement the best air-coal ratio.With these methods functioning,the list of self-learning will be updated continuously and real-time optimization of the chain furnace will be achieved,so as optimum combustion efficiency..The results of simulation experiment show that it has solved the non-linear,timevarying and multi-variable problems with BP neural networks and adaptive selfoptimizing control;the experiment gets remarkable energy-saving effect and verifies that the proposed control program is effective and practical.Figure 34;Table 9;Reference 45...