【摘要】 为了科学地控制温室温度环境,提升温室温度的预测精度,提出一种改进收缩因子粒子群优化的径向基函数(RBF)神经网络预测模型。利用最大最小距离算法确定RBF神经网络的隐层节点个数;应用改进收缩因子粒子群优化RBF神经网络的隐层基函数中心和场域宽度;与RBF神经网络算法、PSO-RBF神经网络算法、CFA PSO-RBF神经网络算法的预测精度进行比较,分析预测模型性能。实验证明,在神经网络参数选择合理的情况下,与其他神经网络算法相比,改进CFA PSO-RBF神经网络算法具有更好的预测效果。更多还原

【Abstract】 In order to scientifically control the greenhouse temperature environment and improve the prediction accuracy of greenhouse temperature, we propose a prediction model of radial basis function(RBF) neural network with improved shrinkage factor particle swarm optimization. It used the maximum and minimum distance algorithm to determine the number of nodes in the hidden layer of the RBF neural network. Then, the improved shrinkage factor PSO was applied to optimize the center of the hidden layer basis function and the field width of the RBF neural network. Finally, compared with the prediction accuracy of the RBF neural network algorithm, the PSO-RBF neural network algorithm, the CFA PSO-RBF neural network algorithm, the performance of the prediction model was analyzed. The experimental results show that compared with other neural network algorithms, the improved CFA PSO-RBF neural network algorithm has better prediction effect when the neural network parameter selection is reasonable.更多还原