Study On Optimal Control Of District Cooling System Based On Ice Storage

In order to achieve higher comfort,the global energy use is accelerating.In China,with the development of recent decades,the urban scale is increasing day by day,the urban population is also growing rapidly,the total energy consumption of building operation and the power consumption in it are increasing greatly,in which the air conditioning energy consumption accounts for a large proportion of the building energy consumption,about 30%-50% of which is air conditioning energy consumption,and the ice storage district cooling technology has the characteristics of environmental protection,high efficiency and "peak cutting and valley filling" In order to solve the problem of peak load of power grid,the government has given policy support.Although the ice storage district cooling technology has better economic and social benefits,but because of its large-scale system and complex operation control,in order to give full play to the advantages of ice storage district cooling technology,its operation needs to be fully optimized.In this paper,an actual project is taken as the research object.In the actual operation process of the project,the operating personnel decide the operation mode according to the weather and experience.The ice storage capacity is too small.In the peak period of daytime electricity price,the refrigeration host still needs to be started,resulting in the increase of operation cost.Therefore,in view of its existing operation problems,the optimization operation research is carried out.In this paper,the main influencing factors of load forecasting accuracy of ice storage district cooling system are analyzed.Through the comparison of load forecasting models,it is found that when the input parameters are input under the condition of disorder sequence,the accuracy of the model is higher than that of the model input according to time sequence.At the same time,the input parameters select the time,the atmospheric temperature(including the atmospheric temperature at the last time,the big temperature at the last two times Degree,the atmospheric temperature at the same time of the previous day);relative humidity of the atmosphere;outdoor wind speed and direction;daily load of air conditioning operation(including the air conditioning operation load at the previous time,the air conditioning operation load at the last two times,and the air conditioning operation load at the same time of the previous day);date type.Then,different types of load sample data are used to train and predict the two load forecasting models,which are BP neural network and LSSVM least square method.The results show that LSSVM least square method is more accurate than BP neural network,but the calculation time is longer.Taking the operation cost and energy consumption as the optimization objective,this paper establishes the daily operation energy consumption model and ice tank operation model of ice storage air conditioning,including the energy consumption model of refrigeration unit,cooling tower,cooling water pump,glycol circulation pump and ice tank storage and release model.Then,according to the constraints and objective function,using the multi-objective nonlinear programming method,the optimal control strategy under 25% load and 50% load is calculated to guide the operation of the system.Through the optimal control strategy,when the load is 25%,the daily power consumption is reduced by 16.3% compared with the actual situation,and the operation cost is reduced by 28.8%.Under 50% load,the daily power consumption increased by 2.5% and the operation cost decreased by 14.3%.From the perspective of load forecasting and optimal control,this paper studies the operation strategy of the cooling system in the ice storage area.On the one hand,it can be used to formulate the optimal operation strategy of the cooling system in the ice storage area,on the other hand,it can optimize the design scheme from the perspective of annual operation.