Research On Method Of Fluoride Detection And Energy Saving Control Optimization For HVAC Equipment

With the improvement of living standards,the application of HVAC(Heating,ventilation,and air conditioning)equipment has become more and more widespread,but it also brings huge energy consumption problems.In terms of energy saving of HVAC equipment,fault diagnosis and energy-saving control optimization are two important research directions.Therefore,this paper takes the household dehumidifier and split air conditioner in HVAC equipment as research objects,and conducts research on the refrigerant capacity detection and energy-saving control optimization.Firstly,due to the accurate model of the dehumidifier is difficult to establish and the expert experience is limited,we adopt the neural network to construct the relationship between the dehumidifier operation data and the refrigerant content,and a large amount of data under complex conditions is used to ensure the generalization of the model.At the same time,we compare the detection effects of several different neural network topologies in this paper,and related techniques of deep learning are used to accelerate the training of the network and improve the performance of the model.In addition,the time series features are used to characterize the operation of the dehumidifier to improve the performance of fluorine deficiency detection.Secondly,in the optimization of energy-saving control of air-conditioning,because the existing methods are more focused on versatility,and it is impossible to adjust according to the specific environment.The deep reinforcement learning is used to optimize the existing air-conditioning energy-saving control strategy without establishing models of airconditioning and environment.Through the continuous trial and exploration of air conditioning in the environment,the adaptive optimization of air conditioning control in the environment is realized.The experimental results in this paper show that the refrigerant capacity detection method based on neural network can effectively detect whether the dehumidifier refrigerant quantity is in the state of lack of fluoride,and has a high detection rate and detection accuracy.The air conditioning energy-saving control strategy optimized by deep reinforcement learning can achieve the same control effect with lower energy consumption than the un-optimized control strategy,and realize the adaptive adjustment of the air conditioner output to the environment.