An improved simulated annealing for solving the linear constrained optimization problems

In this paper, we propose an improved simulated annealing (ISA), a global optimization algorithm for solving the linear constrained optimization problems, of which the main characteristics are that only one component of current solution is changed based on the Gaussian distribution in each iteration and ISA can directly solve the linear constrained optimization problems. By solving 6 benchmark functions with the lower and upper bounds constraints and 6 functions with linear constraints, ISA is superior to the classical techniques for solving the problems with lower and upper bounds and reduces greatly the number of function evolutions compared with GENOCOP with the same precision condition.