Research On Energy Saving Optimal Control Of Cold Station System In A Pharmaceutical Factory Based On Machine Learning

With the rapid development of economic and social,the energy mix is changing on and on.Traditional energy resources are decreasing because of inherently non-renewable.At the same time,using conventional energy releases a lot of carbon emissions,which causes environmental pollution seriously.Our country is a great manufacturing power,many places need to consume energy.There are many energy-saving policies in our country,such as enhancing energy efficiency and the change of energy structure.In daily life,the utilization rate of air conditioning is relatively frequent,and there are many problems about energy waste in the operation process.The energy consumption of water chiller occupy 60%-80%of the whole cooling system.Therefore,it is necessary to study the energy-saving optimization of water chilling unit.The operating performance of the chiller is affected by internal and external factors.The reasonable number of chiller starts and stops and load distribution will be more effective than the original method with artificial intelligence technology,which uses artificial intelligence technology.The goal of energy saving and emission reduction can be achieved.To address the problem,this article puts forward to study how to rationally optimize the distribution of cold load and the number of start and stop about two chiller units with the same cooling capacity rating.On the premise of meeting the cold load of the day,the improved grey Wolf optimization algorithm is used to optimize the objective function established in this paper.Optimize the distribution of the cold load of each chiller to find the lowest value of total energy consumption.The main research is as follows:Taking the cold station system of a large pharmaceutical factory as the research object.Meteonorm was used to obtain the meteorological report of Wuxi City and the meteorological data.According to the mathematical model of the associated equipment,the cooling system simulation platform was built in Energy simulation software,to obtain related parameter data.At the same time,according to the formula P=Q/COP of total energy consumption of chiller,the cold load and energy efficiency ratio COP were studied respectively in the third and fourth chapters,to make preparations for the reasonable load allocation of each chiller.Secondly,in order to study the daily cold load required by two chiller units,this article through Pearson correlation coefficient analysis and select several influence factors which have high correlation with cooling load,to establish the improved PCC-LSTM-GRU prediction model,and the daily cooling load Q required by the chiller was calculated.The accuracy of the model is improved,which contrasted with the traditional LSTM-GRU model,LSTM model and GRU model.Meanwhile,the study found that COP reflects the energy use rate of air conditioning.The COP is increasingly high,manifesting the energy efficiency is getting better and the energy consumption is on the decrease.Therefore,the fourth chapter of this paper mainly studies COP.The results of Pearson correlation coefficient show that COP has the highest correlation with cold load rate.On the basis of previous studies,polynomial regression algorithm is used to establish COP-PLR relationship model in this paper.The results show that the fifth order polynomial regression algorithm can improve the accuracy of the model.Based on the above analysis,it is the key to optimize the operation of chiller to find the optimal load distribution cut-off point and the minimum total energy consumption₈8)8)8)8)8))by optimization algorithm,which that the required cooling load value Q is met on the premise of the day.Therefore,the objective function of energy consumption P is established in this paper.The improved Grey Wolf algorithm is used to optimize the objective function according to the constraint conditions,which enables multiple chiller units to carry out reasonable load distribution with on the premise of meeting the required load for the day.The optimal load distribution cut-off point is obtained,and the total energy consumption reaches the lowest value.The effect of energy saving and emission reduction has been achieved.