Improved Cuckoo Search Algorithm for Engineering Optimization Problems

This paper proposes a grouped dynamic adaptive cuckoo search algorithm (GDACS) to overcome the shortcomings of the standard cuckoo search algorithm (CS), such as poor local search ability, slower search speed in the later stage and low convergence accuracy. The modifications of GDACS could be summarized into two aspects. On the one hand, a chaotic map of the chaotic algorithm is used to generate a chaotic sequence for ensuring the distribution of the initial solution more uniform and increasing the diversity of the solution because of the randomness of initialization. On the other hand, a modified update strategy is discussed to replace the random walk strategy of standard CS by evaluating the fitness value of the bird's nest. According the proposed update strategy, the populations will be grouped due to the pros and cons of fitness, and the different steps strategies are adopted for different groups to realize an adaptive update mechanism. In the experiment part, the optimization of six selected functions is executed to reveal the performance of the proposed GDACS among standard CS, the GDACS, and the other two CS variants. Experiential results and analysis demonstrate that the proposed GDACS has more efficient search performance and more accurate optimization results.