A hybrid least square support vector machine for boiler efficiency prediction

Abstract

A hybrid least square support vector machine (LSSVM) is proposed to predict the boiler combustion efficiency. In this approach, a principal component analysis (PCA) is employed to reconstruct new variables as the input of the predictive model. Then, a particle swarm optimization (PSO) algorithm optimized LSSVM is proposed. The parameters of LSSVM are optimized dynamically by PSO and the output value of the model is corrected to improve the prediction accuracy. The experimental results based on practical data set illustrate that the proposed hybrid LSSVM obtains better accuracy compared with other data-driven approaches, such as the multi-layer perceptron (MLP) and Elman neural network. The proposed boiler combustion efficiency model can meet the requirements of boiler control and optimization.