利用遗传算法具有的全局寻优能力来优化差分格式的多变量灰色模型MGM(1,n,q)。详细介绍了遗传算法优化参数q的过程,并用此模型对大坝变形监测数据进行建模分析及预测。结果表明,经遗传算法优化的MGM(1,n,q)模型比MGM(1,n)模型具有更高的拟合精度和预测精度。
Difference scheme of MGM( 1,n,q) model could be optimized through using the genetic algorithm because genetic algorithm possesses the characteristics of full optimal. The process of optimizing the parameter q by genetic algorithm is introduced in details and applied to analyze and predict the monitoring data of dam deformation. The results show that MGM( 1,n,q) model optimizing by genetic algorithm not only has higher fitting precision,but also improve the prediction accuracy.