Application of PSO-BP neural network in main steam temperature control

The main steam temperature of power plant boiler has an important effect on the service life, security of the whole unit. As the load changes, the dynamics of the main steam temperature of power plant boiler changes. One of the difficulties in the heat-engine plant process control is about how to get the main steam temperature near the set point quickly. Focusing on the control of main steam temperature of the boiler unite under different loads, the writer proposes PID cascade control scheme based on BP neural network. Neural network has slow convergence speed and often fall into local extremum, therefore, this paper optimizes the initial weighting coefficient of neural network with particle swarm optimization algorithm. Then neural network obtains the best combination of PID controller's parameters through learning on system and the adjustment of weighting coefficient, and the simulation example verifies that the proposed method has better control quality and high anti-disturbing ability.