基于ADE-LSSVM算法的粮食产量预测研究

Research on grain yield prediction based on ADE-LSSVM algorithm

针对粮食产量预测精度低、影响因素复杂等问题,提出了自适应差分进化(ADE)算法优化最小二乘支持向量机(LSSVM)的粮食产量预测方法。针对基本差分进化(DE)算法容易出现局部最优,陷入早熟的现象,通过引入自适应变异算子和交叉概率算子改善算法寻优性能。将改进算法用于对LSSVM的参数优化中,选择1975~2018年我国粮食产量数据进行仿真,建立ADELSSVM预测模型。结果表明,模型具有较高的预测精度,对粮食产量预测有一定的实用价值。

Aiming at the problems of low prediction accuracy and complex influencing factors,an adaptive differential evolution(ADE)algorithm for optimizing the least squares support vector machine(LSSVM)for grain yield prediction was proposed.In view of the phenomenon that the basic differential evolution(DE)algorithm was prone to appear local optimization and fall into the premature,the adaptive mutation operator and cross-probability operator were introduced to improve the optimization performance of the algorithm.The improved algorithm was used to optimize the parameters of LSSVM.The grain yield data of China from 1975 to 2018 was selected to simulate and establish the ADE-LSSVM prediction model.The results showed that the model had higher prediction accuracy and had certain practical value for grain yield prediction.