A novel adaptive genetic algorithm for global optimization of mathematical test functions and real-world problems

AbstractGenetic algorithm (GA) is a population-based stochastic optimization technique that has two major problems, i.e. low convergence speed and falling down in local optimum points. This paper introduces an adaptive genetic algorithm (AGA) consisting of new crossover and mutation operators to handle these drawbacks. The crossover operator is based on a combination of the traditional crossover mechanism and the particle swarm optimization (PSO) operator. The proposed mutation operator intelligently uses sliding mode control (SMC) to escape from local minimums and converges to the global optimum. The performance of the proposed genetic algorithm is challenged by using twenty well-known test functions. The comparison of the obtained numerical results with those of the other optimization algorithms reported in literature demonstrates the superiority of the proposed algorithm in finding the global optimum points. At the end, the proposed method is employed to estimate the oil demand in Iran based on socio-economic indicators and using linear and exponential forms as a real-world problem that shows the AGA’s effectiveness