Research On The Superheat Control Method Of Cold Storage Based On Fuzzy Neural Network

In recent years,the demand for domestic cold storage has increased rapidly,a stable and energy-saving refrigeration control system is of great significance to ensure quality of food,reduce the operating costs,and save energy and reduce consumption.As an important part of the refrigeration control system,the quality of the evaporator outlet superheat control not only affects the refrigeration efficiency of the cold storage,but also is closely related to the safe operation of the cold storage.Therefore,the research of control of superheat has very important theoretical and practical significance.Firstly,a small cold storage building was designed as an experimental platform.based on the superheat measurement principle and control method,the scheme of electronic expansion valve controller driving electronic expansion valve to control superheat was designed.the hardware platform of the data acquisition system was built by using RS485 communication,the data acquisition system software was modularized,human-computer interaction,real-time data acquisition of the operating conditions,local database,abnormal alarm,and remote data wireless transmission,etc.the real-time data acquisition,dynamic curve display,storage,historical data query,abnormal alarm and remote monitoring and other functions of the cold storage state variables such as temperature,pressure,flow and other parameters were completed.Then,the superheat deviation and the superheat deviation change rate were taken as the input,and the opening of the electronic expansion valve was taken as the output,on the based on the design of the fuzzy controller,a fuzzy neural network controller was designed by using the Mamdani structural model.Combined with the working characteristics and temperature sequence of the cold storage under the fuzzy controller,the superheat and the opening of the electronic expansion valve were analyzed and processed,and the training samples were made.The central value and width of the membership function and control rules in the fuzzy control were learned and adjusted by the BP learning algorithm.the fuzzy neural network model was trained.Finally,the superheat control simulation and experiment were carried out.The superheat control simulation results show that the superheat degree under the fuzzy neural network controller has stronger anti-interference ability and stronger robustness than the fuzzy controller.The superheat control test shows that the fuzzy neural network controller can track the superheat set value more quickly and more stably,can better meet the requirements of variable load conditions,and the COP has increased by 9.3% on average.