Research On Optimization Technology For Energy Saving Operation Of Air Conditioning Water System

This paper takes the running condition of air conditioning water system in a shopping mall in Shenzhen as study.Because of the total energy consumption of the air conditioning water system is not only related with cold load,but also connected with the running condition of water system equipment,so the condition optimization method that achieves the minimum total energy consumption in water system which in different cold loads is proposed,and through control algorithm to obtain the minimum energy consumption for the total equipment of the system.To determine the required cold load of air conditioning water system in the next time,in consideration of large lag,non-linearity,uncertainty,and multiple influencing factors for air conditioning system load,a load prediction method that include three dimensions as input value data range,load measurement method and load prediction model is proposed.The statistical methods of correlation analysis and partial correlation analysis is used in load prediction method,and build prediction model that use Elman neural network to prediction the load of air conditioning system in the future.For the energy consumption mathematical model of air conditioning water system including discrete variables(chiller,chilled water pump and cooling water pump equipment start-stop number)and continuous variables(operation values as temperature,flux,pressure,etc.),the paper based on the law of conservation of energy and the study on the optimal start-stop number of chiller and pump operation,evolves the energy consumption model of air conditioning water system into the model which takes chilled water supply and return temperature as well as cooling water inlet and outlet temperature as input,chiller start-stop number as well as start-stop number and running frequency of cooling water pump and chilled water pump as intermediate variables,and the energy consumption of various equipment as output,to solve the technical problem that too much variables in the system of optimizing conditions.According to the optimal running conditions of air conditioning water system require high real-time performance,improved particle swarm optimization(PSO)algorithm to obtain the optimal running condition is proposed.The improved PSO algorithm not only can obtain the accurate corresponding temperature for set point and equipment start-stop number of the optimal running conditions,but also the time of searching optimal running condition is short time.Results of engineering experiments show that use improved PSO algorithm can ensure the air conditioning water system always operating under the most efficient conditions,so can energy saving for total energy consumption of the air conditioning water system.Considering the characteristics for control structure of water system is 2 input-2 output,the paper using invariance principle of feed-forward decoupling compensation,and builds the PID control algorithm that based on the Elman neural network,realizing the accurate control of the differential pressure and differential temperature of central water system.The results of the simulation show that the algorithm not only can response quickly,also has high control accuracy and stable operation results.