Improvement And Application Research Of Bat Algorithm

Bat algorithm(BA) is a novel swarm intelligence algorithm, inspired by the echolocation behavior of the bats with varying pulse rates of emission and loudness. BA increases the pulse rates of emission and decreases the loudness to balance the global search and local search process of the algorithm. The of algorithm has simple structure and less parameter, so it is easy to understand and implement. What’s more, it has good search performance and strong robustness. Therefore, it has received widespread attention and research at home and abroad, and has been successfully applied to solving complex combinatorial optimization problem. But at the same time the drawback of easily being trapped in local optimum and lower optimization accuracy has greatly limited the application range of BA.In this paper, aiming to overcome the shortcomings of the bat algorithm, we improved the bat algorithm from aspects of the coding scheme and new evolutionary strategy, and the improved algorithm was applied to some complex combinatorial optimization problems. The main purpose is to improve the bat algorithm performance, perfect its theoretical basis, to expand its range of application.In summary, the works in this paper are as follows:(1) Improve the bat algorithm by adopting the idea of complex-valued encoding scheme. The diversity of population was increased. Consequently, this strategy can enhance the global exploring ability of the algorithm, which avoid algorithm trapped in local optimum prematurely and overcome the shortcoming of the lower searching precision.(2) The computational cost of traditional algorithm in solving 0-1 knapsack problem is too large. So in order to solve 0-1 knapsack problems effectively, based on the complex-encoding bat algorithm, we introduced the greedy search strategy in the algorithm to provide a moving direction for bat individuals, and to modify the abnormal solution in the solving process timely.(3) To further optimize the bat algorithm encoding scheme, a novel quantum inspired meta-heuristic algorithm namely quantum bat algorithm(QBA) is proposed in this paper. QBA uses a novel representation that is based on the concept of qubits. Quantum bits are updated by quantum rotation gates. And quantum NOT gate is used to realize quantum mutation to avoid premature convergence. The search performance of the bat algorithm is improved further. Finally, this article also applied quantum bat algorithm to solve the UCAV path planning problem, and the precision and solution efficiency are improved.(4) Introducing modified random localization(MRL) strategy, the bat population is encouraged as much as possible to cover the entire search space. The global exploring ab ility of the algorithm is enhanced and the balance between global search and local search become better. Experiment results illustrate that the improved bat algorithm has a good performance for image multilevel threshold segmentation problem.