Hybrid Genetic Algorithm with PSO Effect for Combinatorial Optimisation Problems

In engineering field, many problems are hard to solve in some definite interval of time. These problems known as “combinatorial optimisation problems” are of the category NP. These problems are easy to solve in some polynomial time when input size is small but as input size grows problems become toughest to solve in some definite interval of time. Long known conventional methods are not able to solve the problems and thus proper heuristics is necessary. Evolutionary algorithms based on behaviours of different animals and species have been invented and studied for this purpose. Genetic Algorithm is considered a powerful algorithm for solving combinatorial optimisation problems. Genetic algorithms work on these problems mimicking the human genetics. It follows principle of “survival of the fittest” kind of strategy. Particle swarm optimisation is a new evolutionary approach that copies behaviour of swarm in nature. However, neither traditional genetic algorithms nor particle swarm optimisation alone has been completely successful for solving combinatorial optimisation problems. Here a hybrid algorithm is proposed in which strengths of both algorithms are merged and performance of proposed algorithm is compared with simple genetic algorithm. Results show that proposed algorithm works definitely better than the simple genetic algorithm.