Research On Model Identification Algorithm For Typical Thermal Process Of Gas-Steam Combined Cycle Unit

Chinese thermal power generating units are mainly coal-fired generating units,which cause serious pollution and high consumption of coal for power supply.It cannot meet the development needs of the power industry in the new century.Gas-steam combined cycle unit has high power generation efficiency and low environmental pollution,which is a good choice to replace thermal power unit as the main energy development in the future.In order to better study the typical thermal process of gas-steam combined cycle unit and lay a good foundation for the design of its control strategy,it is necessary to identify its thermal process.In order to solve this problem,the actual thermal process data of a gas-steam combined cycle unit thermal power plant in Beijing were used as samples for analysis and processing in this paper.The process model based on historical data was obtained by offline identification method.Then,on the basis of offline identification,the feasibility of online identification was explored,and the process model was identified online in real time using the fast online algebraic parameter identification algorithm.When the parameters of the process model changed,the parameters of the model could be quickly tracked through online identification.Firstly,this paper introduces the working principle and process flow of gas-steam combined cycle unit.After having a deeper understanding of gas-steam combined cycle unit,four most typical thermal processes are selected from its various thermal processes to confirm its input and output data.After that,through zero initial value processing and rough value processing,the actual operation data of the gas-steam combined cycle unit collected from the actual site are preprocessed to prepare for the subsequent offline identification.Then using the cuckoo search(CS)and cloud theory optimize the weight of particle swarm optimization(PSO)iterative formula and location parameters respectively,conducive to jump out of local optimum and get the best parameter identification results.Thus,particle swarm cuckoo search fusion algorithm based on cloud theory(CPSO-CS)is obtained.Apply it to the actual operation of the processed data offline identification,get a combined cycle unit is a typical thermal transfer function of the process.Using the Graphical User Interface(GUI)of MATLAB design function,a typical thermal process model automatic identification software is designed,which can identify the process model of thermal process in power generation according to various actual field data.Finally,the feasibility of online identification is explored by using the transfer function of typical thermal process obtained by offline identification.Through experimental simulation,the change of model parameters due to steady-state changes in the field is simulated,and the fast online algebraic parameter identification algorithm is used for online identification.When the process model parameters change,the online identification algorithm can quickly track the change of parameter values,thus verifying the feasibility of online identification.This paper aims to discover the potential information contained in the massive actual field data,the intelligent identification algorithm is used to extract the content which is beneficial to automatic control and to study and control the typical thermal process of gas-steam combined cycle unit and other units.