基于改进RBF算法的建筑能效预测模型的构建

Construction of Building Energy Efficiency Prediction Model Based on Improved RBF Algorithm

采用PSO-LM算法对RBF神经网络进行优化,并构建出基于PSO-LM-RBF神经网络算法的建筑能效预测模型,并以某办公建筑在某段时间内的能耗数据为例,最后验证了本研究提出的优化模型。结果显示,PSO-LM-RBF神经网络算法的收敛速度显著高于改进之前,且相对误差低于2.3%,平均、最大相对误差均显著低于改进前。PSO-LM-RBF神经网络模型可以更好的预测数据变化过程,改善了RBF神经网络的预测能力。

With the rapid development of Chinese economy,the energy conservation of office buildings has gradually attracted the attention of all sectors of society.In this study,PSO-LM algorithm is used to optimize RBF neural network,and a building energy efficiency prediction model based on PSO-LM-RBF neural network algorithm is constructed.Taking the energy consumption data of an office building in a certain period of time as an example,the optimization model proposed in this study is finally verified.The results show that the convergence speed of PSO-LM-RBF neural network algorithm is significantly higher than that before improvement,and the relative error is less than 2.3%,and the average and maximum relative errors are significantly lower than that before improvement.PSO-LM-RBF neural network model can better predict the data change process and improve the prediction ability of RBF neural network.