The comparison between single BPNN-based method and data-based HDP method for decomposition furnace optimal control

Cement decomposition furnace is a complex physics equipment of multi-variable, multi-disturbance nonlinear systems involving mass transfer, heat transfer and chemical reaction. It is difficult to establish a precise mathematical model of rotary kiln, and to achieve its optimal control. Data-based HDP(Heuristic Dynamic Programming) combines HDP method and data-based control method, in the case of difficult to establish accurate mathematical model, it makes full use of the data that generated in the production process, achieve optimal control of complex control systems. Single BP neural network control was also used for temperature control of the cement decomposition furnace, then, compared the data-based heuristic dynamic programming control method with neural network direct control method, the data-based heuristic dynamic programming control method had better control result than neural network direct control method.