Temperature Control Algorithm Design For Steel Rolling Furnace

Furnace is important rolling production line equipment. Furnace temperature control directly influences the quality of billet. The furnace is a typical non-linear, large inertia, strong coupling, time delay systems, coupled with the furnace temperature distribution can not be measured directly, while it be restricted by a variety of factors. So the temperature control is a hot issue in the study of control.Usually, steel rolling furnace divided into three sections– preheating section, heating section, and soaking period section. Soaking period section is the last section before billets leave furnace, but also the most stringent control, during this billet rolling line needs to be heated to the temperature and its cross-section temperature control within a certain range. This paper focuses on the furnace temperature control for soaking period section. Currently, the furnace temperature control algorithm based PID algorithm. PID algorithm with a simple, stable, and reliable, but the parameter tuning is more difficult. In the nonlinear time-varying large delay system control result is not satisfactory. Fuzzy control algorithm is an intelligent control algorithm, through mimicking ideas and experiences of people to adjust the object. Do not need to establish an accurate mathematical model during the design, so it is very appropriately used to some fields, such as difficultly obtain accurate mathematical model and difficultly grasp the dynamic characteristics of the furnace process.Taking into account the PID algorithm and fuzzy algorithm with different advantages, this paper integrates the different algorithms. Through fuzzy PID algorithm control furnace temperature, and used PID algorithm, fuzzy algorithms, fuzzy PID control algorithm to simulation furnace temperature, comparative and analysis the control effects. The results show that PID algorithm is faster in response, but there are large oscillations and overshoot, settling time is longer. Fuzzy control oscillation is small, has smaller overshoot, faster response, the control effect curve smooth, smooth arrival for a given value, but the transition process is longer, take longer to reach steady state. Fuzzy PID algorithm combines the advantages of fuzzy algorithm and PID algorithm to avoid the establishment of precise mathematical model, and play the simple and stable properties of PID algorithm. It computation is small, and easy to implement. Its adaptability is better than PID algorithm, with better control effect, small overshoot and fast response. Its transition time is short; steady-state error is small, and with practical.