| The passive ultra-low energy consumption building promotes the development of integrated heat pump environment control unit with outdoor air.In order to meet the personal thermal comfort requirements of users,this paper applies the machine learning algorithm to the integrated machine environment control,constructing personalized thermal comfort model and intelligent control scheme.The specific content of the study is as follows:Firstly,the thermal comfort prediction model is established by using Phython language.Indoor air temperature,relative humidity,air speed,outdoor air temperature of environmental parameters and clothing thermal resistance,height,weight,gender,age and metabolic rate of personalized parameters are taken as input values,PMV indexes were used as output indexes,and MAE,RMSE and R~2were used as evaluation indexes to evaluate the model.Three algorithms including BP neural network,support vector machine and XGBoost are used to predict thermal comfort.It is found that the thermal comfort model based on XGBoost algorithm has the best prediction effect and the highest accuracy.Based on the thermal comfort prediction model of XGBoost algorithm,the SHAP value method is used to conclude that air temperature and air speed are used as control variables,and other parameters are used as interference variables.Secondly,TRNSYS software was used as the simulation platform to establish three simulation models of temperature PID control,PMV start-stop control and temperature start-stop control in the full heat-refrigeration heating mode of integrated heat pump environment control unit with outdoor air.Taking typical residential indoor environment as the control object,compared with temperature-based PID control mode,the indoor PMV can be controlled between[-0.5,0.5]by using temperature-based start-stop control mode and PMV-based start-stop control mode.The dual-objective evaluation method is used to evaluate the three control modes,and the evaluation results show that PMV start-stop control mode is the best control mode.Based on PMV start-stop control mode,the energy saving strategy of the unit is analyzed from three aspects:running time,window opening mode and heat recovery mode.The results show that the all-heat and heat recovery mode of heat pump fresh air environmental control integrated machine is optimal when it runs for 12h a day and closes the window.Finally,the intelligent control scheme of integrated heat pump environment control unit with outdoor air is designed.Based on the personalized thermal comfort model and PMV start-stop control mode,the intelligent control scheme framework is designed and the control strategy of the the unit under cold and hot conditions was proposed.The hardware end of the control system such as SCM,sensor design and selection.The pages of"intelligent perception","intelligent prediction"and"intelligent control"on the user software side are designed.Two cases were selected to verify the effectiveness of the control strategy. |