Improved Dynamic Programming Based on BPNN for Combustion Control of Boiler

Abstract

As a complex system boiler combustion process has the characteristics--high non-linearity, strong jamming, strong coupling and large lagging, meanwhile its internal and external disturbance are very frequent, and fuel quantity can not be exactly measured, so using conventional control scheme is hard to solve the boiler combustion control problem with non-linearity and long time lags. Improved Dynamic Programming is the method of interacting with the system (environment) and improving control effect. This method mainly consisting of three modules: model, critic and action. This paper uses dual-heuristic dynamic programming to consider the solution of boiler combustion multivariable control system, realize the emulated control of boiler combustion process, and analyze learning capacity, controlling effect and adaptive capability of the approach. Finally all of three states reach the expected control objectives.