Research On The Thermodynamic Modeling Of Clean Circulating Cooling Water System

With industrial energy-saving and water-saving awareness being gradually recognized,most industrial cooling water systems are using circulating water.However,there are many problems during the operation of existing circulating cooling water systems,such as improper choice of the operating point of pump or cooling tower,unreasonable allocation of cooling water of heat exchangers and so on.Thus,the purpose of saving energy and water can not be realized.Therefore,with the clean circulating cooling water system of iron and steel enterprise as the background,thermodynamic models of the main heat transfer equipment,including plate-type heat exchanger,mechanically ventilated reverse cooling tower and pipe in the system,are built.The models lay foundation for operation optimization aiming at energy reduction of the system.In this thesis,based on the study of basic compositions of clean circulating cooling water system and operation mechanism of main equipment in the system,the thermodynamic mechanism models of plate-type heat exchanger,mechanically ventilated reverse cooling tower and pipe are established.The parameters in the mechanism model of heat exchanger and cooling tower are usually obtained by experiences or table looking-up.The parameters can not accurately reflect changes in device characteristics.So the key parameters in the mechanism model of heat exchanger and cooling tower are identified using the sequential quadratic programming(SQP)algorithm.The simulation results show that the output of the mechanism model can reflect the trend of the actual output.The heat exchanger and cooling tower have the characteristics of complex nonlinearity and coupling,and the mechanism model is established under certain assumptions and simplifying conditions,so the accuracy and reliability of the model are not enough to meet the requirement of operation optimization of the system.RBF neural network with the nonlinear approximation ability is used to compensate the bias between the mechanism model and the process output,and the parallel hybrid model of heat exchanger and cooling tower is established,thus models with higher accuary are obtained in this thesis.During the process of training the neural network,a dynamic adaptive learning algorithm is adopted to determine the network centers and the output weights are calculated using least square method.The simulation results of hybrid model show that the hybrid model has higher accuracy than the mechanism model.Finally,based on the model of each equipment,the thermodynamic solution of the entire pipe networks is carried out,and the simulation software for clean circulating cooling water system is developed.The software provides simulation platform for running simulation,energy consumption assessment and operation optimization of the system.