利用支持向量机回归算法(SVM)结合粒子群优化算法(PSO)建立了用于蒸发预测的PSO_SVM模型,用和田地区实测蒸发量对其进行拟合与预测,并与传统的最小二乘支持向量机(LS—SVM)的预测结果进行了对比,结果表明PSO_SVM预测蒸发量的精度要高于LS_SVM,说明该模型可以用于蒸发预测。
This paper has used Support Vector Machine(SVM) combined with Particle Swarm Optimization (PSO) to establish PSO SVM model in predicting evaporation. The model was applied to fit and predict evaporation in Hetian, and comparing with the traditional Least Squares Support Vector Machine (LS_SVM), the results have shown that PSO_SVM has better forecast accuracy than LS_SVM, so this model could be used in predicting evaporation.