基于IPSO-SVM的大气候室相对湿度预测

Prediction of the relative humidity for large climate chamber based on IPSO-SVM

针对大气候室相对湿度控制效果存在明显滞后的问题,建立改进粒子群算法(IPSO)-支持向量机(SVM)的相对湿度预测模型。首先引入Tent混沌映射初始化种群,使初代粒子均匀分布于搜索空间,增加种群多样性;其次采用新的惯性权重非线性调整策略,平衡粒子的全局搜索与局部搜索能力;最后引入随机蛙跳算法(SFLA)的跳跃机制,一定程度上避免了标准PSO算法过早收敛,陷入局部最优的问题。实验结果表明:在三组数据集中,相较于PSO-SVM和GA-SVM算法,本模型具有最优的预测精度,决定系数均在0.97以上,该模型可为优化大气候室相对湿度控制策略提供参考。

Aiming at the lag problem of relative humidity control of a large climate chamber, a relative humidity prediction model based on IPSO-SVM is established. Firstly, Tent chaos mapping is introduced to initialize the population, so that the primary particles are evenly distributed in the search space, which is beneficial to increase the diversity of the population. Secondly, a new inertial weight nonlinear adjustment strategy is adopted to balance the global and local search capabilities of particles effectively.Finally, the jumping mechanism of shuffled frog leaping algorithm(SFLA) is introduced to avoid premature convergence and local optimization of standard PSO algorithm to a certain extent. Experimental results show that in three data sets, compared with PSOSVM and GA-SVM algorithm, this model has the optimal prediction accuracy, and the determination coefficients are all above0.97, which can provide a reference for optimizing the relative humidity control strategy of the large climate chamber.