Hybrid neural network in circulating fluidized bed boiler based on information fusion clustering control

Circulating fluidized bed boiler (CFBB) is regarded as a clean and highly efficient coal combustion technology, and recently, this combustion technology is rapidly developing. However, CFBB is a complicated control object with distributing parameters, nonlinearity, time varying, large time-delay, and strong coupling. Traditional control methods cannot obtain good control effect, which becomes one of the bottlenecks for its extending application. In this paper, we design a three-level ART2-BP–BP of data fusion—cluster control system based on methods of multi-sensor information fusion and cluster analysis. We complete data fusion from the data level, feature level, to the decision level. By this method, more complete description of the conditions is obtained, and appropriate control strategies according to each category are taken, describing behavioral characteristics in the process. Then, the potential situation threat in working boiler is estimated, and the concept of situation threat space is proposed. Results of simulation show that the control system in this paper is feasible and effective; in particular, the control system still has more satisfactory control effect in the case of a variety of sensor failures.