Abstract: Taking the steam drum water level as the controlled object, a cascaded three-impulse feedwater control system was adopted. A genetic algorithm was used to tune the system parameters and obtain the optimal control parameters. Simulation results in SIMULINK show that, compared with traditional tuning methods, the feedwater control system based on genetic algorithm optimization can better and faster overcome the influence of various external and internal disturbances, and has better water level setpoint tracking performance.
Keywords: Genetic algorithm, optimization, cascaded three-impulse, water supply control system
Abstract: Cascade three-element feedwater control system is adapted to regulate the drum water level. Genetic algorithm (GA) is introduced to tune the system in order to obtain the optimal parameters. SIMULINK simulation results show that the GA-optimized feedwater control system is able to reject endogenous and exogenous disturbances more effectively and rapidly, compared to the conventional tuning method.
Keywords: Genetic algorithm, Optimization, Cascade three-elements, Feedwater control system.
1 Introduction
The boiler drum water level is an important monitoring parameter in boiler operation. It indirectly reflects the balance between boiler steam load and feedwater flow. Maintaining the drum water level within the normal range is a necessary condition for ensuring the safe operation of the boiler and turbine. If the drum water level is too high, it will affect the normal operation of the steam-water separation device in the drum and easily burn out the superheater; while if the drum water level is too low, it may disrupt the boiler water circulation and cause the water-cooled wall tubes to burn out and rupture. Since the cascade three-impulse control system is technically mature and has good reliability, it is currently widely used in the feedwater control of drum boilers [1].
However, the parameter tuning of cascade three-impulse feedwater control systems is currently quite primitive and outdated. It primarily relies on semi-empirical formulas to calculate very rough parameter values, followed by on-site testing and adjustments. This primitive method is time-consuming and labor-intensive, and it doesn't consider any overall optimization index, often resulting in suboptimal control performance. Therefore, there is an urgent need for a fast, effective, time-saving, and labor-saving parameter tuning method for feedwater control systems that can achieve optimized control performance.
For details, please click: Genetic Optimization-Based Boiler Feedwater Control System
