基于Isomap-BP的寒地老年住宅热舒适度模型

Thermal Comfort Model of Cold Residential Houses Based on Isomap-BP

针对寒地老年住宅的热舒适度指标偏离问题,以寒地老年住宅作为研究目标,提出了一种热舒适度预测模型,模型更接近老人个体感受,以起到对建筑设计的参考作用。首先,通过小区入户调查、问卷调研得到78个寒地老年住宅的样本数据,以及住宅环境参数、实际热感觉评价;然后,利用Isomap模型进行特征变换,以表达有效特征;最后,选取70%训练样本作为训练集,选取30%样本作为测试集,将特征样本作为BP神经网络的输入,将计算得到的PMV舒适度作为输出,进行训练,再使用老人评价热感觉进行模型修正,建立Isomap-BP预测模型。实验结果表明,所建立的模型室内舒适度预测的均方误差为0.042,说明所建立的模型预测效果较好,模型可提供寒地地区老年住宅冬季热环境量化指标,为冬季住宅供暖进行评价,合理控制供暖能耗。

Aiming at the deviation of thermal comfort index of old residential houses in cold regions,this paper proposes a thermal comfort prediction model based on the cold residential houses.The model is closer to the individual’s personal feelings and serves as a reference for architectural design.Firstly,through the household survey and questionnaire survey,the sample data of 78 cold residential houses were obtained,and the residential environmental parameters and actual thermal sensory evaluation were obtained.Then,the Isomap model was used to transform the features to express the effective features.Finally,70%of the training samples were selected as the training set,30%of the samples were selected as the test set,the feature samples are used as the input of the BP neural network,and the calculated PMV comfort is used as the output for training.After that the model was corrected by using the elderly to evaluate the thermal sensation,and Isomap-BP forecast model was established.The experimental results show that the mean square error of the model indoor comfort prediction is 0.042,which indicates that the model has better prediction effect.The model can provide quantified indicators of the thermal environment of the elderly dwellings in the winter in the cold regions,evaluate the heating of the dwellings in the winter,and reasonably control the heating energy consumption.