Research On Improved Genetic Algorithm For Reheating Furnace Temperature Control

Reheating furnace is key equipment in steel rolling production line. The stability of furnace temperature is essential to follow-up processes and ensures product quality. But reheating furnace is an object of large inertia, large time-delay and time varying, been recognized as difficult to control. It's mathematical model relatively difficult to establish, conventional theories and methods used to control effects not ideal. Therefore the introduction of new methods to improve the control of the reheating furnace temperature control performance has great theoretical and practical significance.Genetic algorithm is by simulating natural genetic mechanisms and biological evolution and the formation of a process of the search algorithm for the optimal solution. It has parallel, robustness, independent of problem, adaptive self-learning, and the high rate of obtain the optimal solution characteristics, solution the nonlinear, multi-peak optimization problems show great superiority. However, the standard genetic algorithm slow convergence and easy to fall into the local optimal solution. To overcome these drawbacks, this paper attempts to improve the genetic algorithm. Improved genetic algorithms using real-coded, improving select and crossover strategy, and other accelerate convergence and improve operational efficiency strategy. After testing, improved genetic algorithm in speed of convergence and the rate of obtain the optimal solution improved significantly.In this paper, the theoretical analysis of the reheating furnace temperature object the establishment of the mathematical model, and clearly prove that this model can be used first-order plus dead time(FOPDT) to expression. Reheating furnace temperature is a large time-delay and slow time-variant object, in view of this character is not conducive to control, design of a new strategy apply genetic algorithm online identify object parameters, real time optimize PID-Smith controller and simulation proved the effectiveness of this strategy.