一种基于改进TD3算法的超临界机组协调控制方法OA

The coupling between the steam turbine and boiler in supercritical units is severe,and the dynamic response varies greatly.The traditional proportional integral differential(PID)coordinated control is difficult to balance the stability of the main steam pressure and rapid load response.To this end,a control strategy based on the twin delayed deep deterministic policy gradient(TD3)algorithm is proposed.Through the introduction of the double Critic evaluation network and the delayed policy update,The training instability issue of the deep deterministic policy gradient(DDPG)algorithm has been effectively addressed,which enhances the control quality of the coordinated control system.Firstly,the improved seagull optimization algorithm is applied to conduct model identification of the unit coordination system,and establishes mathematical models of the controlled object;Secondly,the initial exploration mechanism of TD3 algorithm is improved based on action-enhancing strategies(AES)to speed up the convergence speed of the algorithm.The multi-dimensional state information and multi-objective reward function are designed,and a control framework centered on the TD3 agent is constructed.The optimal fuel control command is generated through TD3 algorithm to improve the response speed and control accuracy of the main steam pressure on the boiler side.Finally,the simulink simulation based on 600 MW supercritical unit shows that this strategy can effectively improve the control quality.