A hybrid RS-SVM dynamic prediction approach to rotary kiln sintering process

Based on the idea of the attributes reduction of the rough sets theory (RS) and the support vector machine regression (SVM), a kind of RS-SVM dynamic prediction approach is presented and applied to predict the temperature of the rotary kiln sintering process. Firstly, the new approach refines the sensor signals closely associated with the sintering temperature using the attribute reduction theory. Then, it constructs a nonlinear predictive model between those sensor signals and sintering temperature utilizing SVM, and dynamically corrects the SVM predictive model via continuous tracing of the predictive error. Thereby, the anti-interference and the fault-tolerant performances have been improved. Through the comparative experiments between the direct SVM approach and the RS-SVM approach proposed The results show that the RS-SVM approach has superiority in the temperature predictive task of rotary kiln sintering process.