Engineering optimization

Engineering Optimization is the subject which uses optimization techniques to achieve design goals in engineering and applications.

It is also called design optimization. Its topics include structural design (e.g., pressure vessel design, welded beam design), shape optimization, topological optimization (e.g., airfoil), inverse optimization, processing planning, and product designs and others.

The techniques used for solving such optimization problems can be classified in three categories: traditional deterministic algorithms, evolutionary algorithms (or genetic algorithms, GA), and metahueristic algorithms. Traditional algorithms such as Hooke-Jeeves pattern search and hill-climbing are widely used for simple problems, while evolutionary algorithms/strategies are used for more complex problems. Metaheuristic algorithms are a recent trend, and are very promising. These algorithms include particle swarm optimization, simulated annealing, cuckoo search, differential evolution, genetic algorithms, harmony search and many others.

The 'simple problems' referred to above are problems which have a single minimum. For this case, when a minimum is found, it is also the global minimum. Other problems have more than one local minima. In this case, if a gradient method is used, a local minimum may be found, but the method may not find the global minimum. Methods which use a higher number of initial search points, such as genetic algorithms, particle swarm optimization and others have a higher probability of finding the global optimum.