Research On Indoor Thermal Comfort Prediction And Control Strategy Based On PSOGSA-FNN

Nowadays,with the continuous improvement of living standards,the indoor environment quality and human comfort are demanded more and more strict for people,and the adjustment of indoor environment is mainly realized by air conditioning control.The traditional air conditioning control system takes temperature as control parameter,and ignores the influence of other indoor environmental and non-environmental factors,so it can't satisfy people's comfort requirement to a great extent.Therefore,it is an urgent problem to study the effective control of indoor thermal comfort.In this thesis,the Predicted Mean Vote(PMV)which is widely recognized as an evaluation index is taken as the control variable,and the interval PMV control strategy of indoor parameter optimization based on improved PMV prediction model is designed.Firstly,the relevant theoretical knowledge of machine learning,artificial neural network and data analysis theory is introduced.Then based on the concept of indoor environmental thermal comfort,the mathematical expression for calculating PMV is given.Secondly,an improved particle swarm gravity search algorithm is proposed to train and optimize the feedforward neural network,and called PSOGSA-FNN prediction model.Through the design of comparative experiments,it is verified that the PSOGSA-FNN has a good advantages in data prediction,and further illustrates that the improved algorithm model in this thesis has practical application value.Afterwards,a real-time PMV prediction model of indoor environment is established by using the PSOGSA-FNN prediction model,and a lot of model training and testing experiments are done according to the simulation data obtained in the experiment.The results of the experiment are analyzed,which confirms the feasibility of the PMV prediction model.Finally,the interval PMV control strategy of indoor parameter optimization based on improved PMV prediction model is designed,and relevant analysis experiments are carried out.The research shows that the interval control strategy of indoor environment comfort based on the improved PMV prediction model is excellent in the experimental environment,which not only meets the requirements of human body for indoor comfort,but also provides a dynamic and cyclical healthy living environment.