关键词: 粒子群优化算法 组合预测 加权系数

Abstract:

Several traditional prediction methods are analyzed and a combination prediction model is proposed in this paper. In the proposed model,weight-coeffic   ients are used to combine various prediction methods,and integrate the prediction results with different sources, so as to reflect the whole prediction    process from different aspects and to make the prediction results more accurate. PSO, which has the characteristics of fast global optimization, is used  to determine the weight-coefficients for various prediction methods. This approach can reduce the blindness of search and increase the prediction preci sion of the model.

Key words: particle swarm optimization algorithm, combination prediction, weight-coefficient