The steam generator, an essential component of a pressurized water reactor nuclear power plant, has level parameters that directly influence the safety and stability of the nuclear power unit. The nonlinearity of the steam generator, coupled with the ‘false level’ issue, complicates its level control. Therefore, research on steam generator level control is considered to be of great significance. Traditional level control in steam generators employs a fixed-parameter Proportional-Integral-Derivative (PID) scheme, a widely used control strategy, which can struggle to adapt to changes in operating conditions and system characteristics. By integrating the adaptive learning capability of the Backpropagation (BP) neural network into the PID framework, a BP-PID control scheme is developed, where the BP network dynamically adjusts the PID parameters in real-time based on system feedback, enabling adaptive adjustment to varying operating conditions and enhancing control effectiveness. The proposed control algorithm was tested on a hardware-in-the-loop simulation platform. Results indicate that the BP-PID control significantly outperforms the original fixed-parameter PID control, demonstrating its potential for engineering implementation.
扫码关注我们
求助内容:
应助结果提醒方式:
