【摘要】 在混沌时间序列研究中,相空间重构和预测模型参数优化是影响预测性能的关键步骤,利用两者之间的相互联系来提高混沌时间序列预测模型的整体性能,提出一种基于遗传算法的混沌时间序列参数同步优化方法。同步优化方法将相空间重构和最小二乘支持向量机参数作为遗传算法的染色体,预测精度作为遗传算法的适应度函数值,通过遗传算法对参数同步优化问题进行求解。通过混沌时间数据对同步优化方法进行了验证性实验。实验结果表明,相对于单独参数优化方法,同步优化方法不仅提高了混沌时间序列的预测精度,同时降低了计算时间的复杂度。更多还原

【Abstract】 In chaotic time series prediction,phase space reconstruction and prediction model parameters optimization were two key steps,used the relationship between phase space reconstruction and prediction model parameters to improve the model performance,proposed a synchronous optimization method of chaotic time series parameters based on genetic algorithm.In the synchronous optimization method,used PSR and least square support vector machine parameters as genetic algorithm chromosomes while used prediction accuracy as the evaluation function of genetic algorithm,solved the synchronous optimization problem by genetic algorithm.Tested the proposed method by the chaotic time series.The experiment results show that the proposed algorithm improves the prediction accurate and reduces the computational complexity compared with the separate optimization methods.更多还原