Modeling And Control Of Circulating Fluidized Bed Boiler Combustion System Based On Support Vector Machine

Circulating fluidized bed (CFB) boiler combustion technology is a clean coal combustion technology. CFB boiler has the characteristic of high combustion efficiency, fuel flexibility, high desulfurization efficiency and low-pollution combustion and it has been widely used. The special way of combustion in circulating fluidized bed boiler also determines the complexity of its combustion system. The combustion system is a kind of system features severe interference, time varying, multi-parameter and heavy coupling which brought difficulty to control system. It is difficult to achieve satisfactory control effect by using conventional control strategy and it is necessary to find new methods.This paper introduces the basic idea of support vector machine (SVM) and describes the principles of choosing the main parameters of SVM. A model based on SVM is established and this paper also verifies the accuracy and rationality of the model.On the basis of understanding the idea of predictive control, this paper sets up models of bed temperature- return valve and bed pressure-primary air based on SVM and the models are used in predictive control. The subsequent simulation shows that this method has a good control effect. The method can be generalized to multi-input system and has a broad application prospect in the control of CFB boiler combustion system in the future.