论文部分内容阅读
在分析城市用水特点、筛选相关影响因素的基础上建立城市生活需水量预测模型,并研究了模型求解过程中智能算法的应用。采用改进的粒子群优化(PSO)算法对反向传播(BP)神经网络的初始设置进行智能优化,避免了传统BP神经网络模型在训练过程中容易陷入局部极小值的缺点。应用该粒子群优化神经网络(PSO-BP)算法求解需水量预测模型,其实例结果表明,该算法提高了神经网络的训练效率,基于该算法的预测模型具有较理想的可靠性和精度。
Based on the analysis of the characteristics of urban water use and the selection of relevant factors, a prediction model of urban living water demand is established, and the application of the intelligent algorithm in solving the model is also studied. The improved Particle Swarm Optimization (PSO) algorithm is used to optimize the initial setting of backpropagation (BP) neural network, which avoids the disadvantage of traditional BP neural network model that it is easy to fall into local minimum during training. The particle swarm optimization neural network (PSO-BP) algorithm is used to solve water demand forecasting model. The experimental results show that this algorithm improves the training efficiency of neural network. The prediction model based on this algorithm has better reliability and accuracy.