Research On Temperature Control System Of Continuous Annealing Furnace

Continuous annealing is one of the major equipments of continuous annealing production line in steel rolling industry and its temperature stability can affect the quality, quantity and cost of cold strip steel directly. The research on temperature control system plays a decisive in increasing productivity, improving the quality of the product and saving the energy resources. The controlled object has the characteristic containing pure lag, great inertia and nonlinearity in the temperature control system of continuous annealing furnace and it is hard to reach the satisfied control goal adopting the traditional PID control. So researching a new control strategy in improving the temperature control system is not only having the theoretical significance but also the practice significance. The thesis studies temperature control problems of annealing furnace, according to the annealing furnace of Ben-Steel Corporation.The thesis focuses on studying the temperature control system of annealing furnace which includes furnace temperature control loop, gas flow control loop and air flow control loop. The thesis analyses the three loops deliberately and builds the control models. Then the thesis has a deeper look at live control strategy that it is double cross limit control. At last, the thesis analyses the practice running effect of Ben-Steel by means of live interface screenshot. The result is that the traditional PID control furnace has quite good performance. But the temperature system is fluctuated badly and has bad hysteretic property in ascending process of furnace temperature.In order to solve the problem of pure lag, Smith predictor is bought in. The fundamental theory is analyzed deliberately and its control model is built by the thesis. The simulation of the tradition PID control furnace and Smith predictor control furnace are made and compared. The great inertia caused by the pure lag was solved to a certain degree. However, the regulation time cannot be improved. So it is necessary to take a new control strategy.The tradition PID control parameters are optimized by BP neutral network on the basis of Smith predictor. Its control model is simulated. It is concluded that the optimized system has higher system response, shorter regulation time and eliminates the great inertia by the pure lag by means of comparing with the tradition PID, Smith prediction.