Heuristic Algorithms And Their Applications In Engineering Optimization

A large number of optimization problems have existed in modernization of national defense, control engineering, economic dispatch, machining and so on, and many optimization algorithms have been proposed to solve these practical engineering optimization problems. Some practical engineering optimization problems are usually complex optimization problems with multiple extreme points, and they are hard to solve for classical optimization algorithms. Thus, the heuristic algorithms used to obtain the global optimum have become the focus. The heuristic algorithms and their applications in engineering optimization problems have been extensively studied and analyzed in this paper.Reliability problem involves selecting the optimal combination of components and redundancy levels to meet resource constraints while maximizing system reliability. A modified particle swarm optimization (MPSO) algorithm is proposed to solve reliability problems in this paper. The MPSO updates each particle’s velocity by using its personal best particle and the global best particle separately, which is determined by a dynamic probability. In addition, a new inertia weight is introduced into the velocity updating so as to balance the global search and the local search. Based on a large number of experiments, the MPSO has demonstrated stronger convergence and stability than the other two PSO algorithms on solving reliability problems.Task assignment problem (TAP) involves assigning a number of tasks to a number of processors in a distributed system. The objective of TAP is to minimize the total execution and communication costs incurred by the task assignment, and it is a constrained optimization problem. An improved differential evolution algorithm (IDE) is proposed in this paper, and it is applied to task assignment problem. The IDE adaptively adjust scale factor according to the objective function values of all candidate solutions, and dynamically adjust crossover rate according to the number of iterations, which can improve the diversity of candidate solutions. Experimental results demonstrate that the optimal solutions obtained by the IDE algorithm are all better than those obtained by the other two DE algorithms on solving some task assignment problems.Harmony search (HS) is a new meta-heuristic optimization method inspired by the music improvisation process. The HS lacks the capability of global search, so it is easy to get trapped into the local optimum. A novel global harmony search algorithm (NGHS) is proposed in this paper so as to effectively avoid premature convergence of the HS. The NGHS uses a novel position updating and a common genetic mutation method. The former can enhance accuracy and convergence rate of HS algorithm, and the latter can diversify the harmony memory (population), which allows harmony vectors to escape from the local minima easily.The optimization performance of the NGHS is investigated, and the effect of parameters on the NGHS is analyzed. The NGHS updates the worst harmony of harmony memory in each iteration, which improves the quality of the whole harmony memory. The NGHS is further utilized to effectively solve three types of engineering optimization problems:0-1knapsack problem which belongs to a special linear integer programming problem; complex chemical equation balancing; robust PID tuning in Hoo control. A large number of simulation experiments are conducted in this paper, and experimental results show that the NGHS performs better than the HS and its several improved algorithms on solving the above three engineering optimization problems. In the mean time, the NGHS has strong convergence, stability and the development capability of solution space. It provides an efficient alternative for complex engineering optimization problems.