基于神经网络的智能家居热舒适度研究

Research on Thermal Comfort of Smart Home Based on Neural Network

本文主要研究智能家居的热舒适度,采用PMV热舒适性指标方程进行计算。针对BP神经网络容易陷入局部最优、收敛速度慢等相关问题,本文引入了遗传算法,利用遗传算法对BP神经网络中的参数进行了系统性的分析与优化,然后将GA-BP算法运用于智能家居室内环境舒适度的预测中,通过MATLAB的仿真验证,该算法大大提高了预测PMV值的精确度。

This paper mainly studies the thermal comfort of smart home,and uses the PMV thermal comfort index equation for calculation.For BP neural network is prone to local optimal,slow convergence problems,introduced the genetic algorithm,using the genetic algorithm of BP neural network parameters in the systematic analysis and optimization,then apply GA-BP algorithm in intelligent home indoor environment comfort prediction,through the MATLAB simulation validation,the algorithm greatly improves the accuracy of predicting PMV value.