Modeling And Optimization Control Of Main Steam Temperature System Based On Harmony Search Algorithm

As thermal power units develop in high-parameter, high-capacity direction, thecontrol requirements of the thermal system become more stringent, and the control ofmain steam temperature, which related to the safe and economic operation of the unit,has been an important and difficult task. In this paper, IMC-PI cascade control strategyfor main steam temperature control is discussed. Outer loop uses advanced internalmodel control, while inner loop using classic PI control. Then use a kind of improvedharmony search algorithm (PSO-IHS) for main steam temperature model identificationand primary and secondary loop controller’s parameters optimization to achieveintelligent optimization control of main steam temperature system.Study found that, stochastic optimization results satisfy the normal distribution, andthe expected value is the optimal solution. Therefore, the secondary optimization methodbased on statistical theory and Newton method, is proposed to improve the optimizationresults of stochastic optimization algorithms. Experiments show that, the secondaryoptimization method can effectively improve the accuracy of stochastic optimizationalgorithms.Some problems are encountered in main steam temperature identification which isbased on PSO-IHS algorithm and using field data. Empirical mode decomposition is usedfor filtering of noised data according to the characteristics that field data contains manydisturbance signals especially high frequency disturbance. Identification strategies areanalyzed and compared, which is aiming at the computer implementation and numericalcalculation about the identification of transfer function model. The identification result ofmain steam temperature system of a circulating fluidized bed unit shows that theobtained model has high precision and can reflect the dynamic and static characteristicsof the actual main steam temperature system.PSO-IHS algorithm is used to optimize the parameters of the outer and inner loopcontroller of IMC-PI cascade control system respectively. The simulation experiments ofmodel mismatch cases of internal model controller show that the IMC-PI cascade controlstrategy is robust enough to meet the requirements of controlling main steam temperaturesystem with multi-interference, time-varying, large inertia and large delay characteristics.