An improved differential-based harmony search algorithm with linear dynamic domain

As a relatively new optimization algorithm, Harmony Search (HS) has been widely used to solve global optimization tasks in various fields due to its simplicity of operation and good performance. However, the basic HS has low fine-tuning ability, easy trapping into local optimum and premature convergence. To overcome the drawbacks and further enhance the precision of calculation results, an improved differential-based harmony search algorithm with linear dynamic domain (ID-HS-LDD) is proposed. In the ID-HS-LDD, two main innovative strategies are adopted: Firstly, inspired by one mutation in the Differential Evolution (DE) algorithm, an improved differential-based method is used as a new pitch adjuster. Secondly, for the search domain of optimal values, introducing a linear dynamic change model is considered. In addition, a parameter is also introduced to modify the new vectors generation formula for updating the harmony memory (HM) in the process of computation. A series of comparative experiments is carried out to verify the performance of the ID-HS-LDD using twenty-four typical benchmark functions. The experimental results show that, for most cases, the ID-HS-LDD has superior performance compared with other HS variants and advanced nature-inspired optimizations. Therefore, the proposed ID-HS-LDD is successfully implemented as a novel optimization method.