A Multi-strategy Improved Fireworks Optimization Algorithm

To solve the shortcomings of traditional Fireworks Algorithm (FWA), such as slow convergence, being easy to fall into local optimum and low precision, a multi-operator improved Multi-strategy Fireworks Algorithm (MSFWA) was proposed. For initialization, the position of individual fireworks is initialized by chaos. As for the explosion operator, the explosion range is reduced nonlinearly and the explosion range of each fireworks particle is divided according to the level of fitness. It is beneficial to improve the development and exploration of the algorithm. For mutation operator, this paper adds mutation information on the basis of retaining the original information, and adopts adaptive strategy to select different mutation modes to further improve the ability to jump out of local optimum. For the selection operator, a brand-new strategy of multi-elite reservation + random / elite reservation is adopted, improving the global and local searching ability of the algorithm. Combining various strategies improves the global and local searching ability of the algorithm, and accelerates the convergence speed. Finally, 8 benchmark test functions and optimization problems of Design of Reducer are tested. The experimental results show that MSFWA has better optimization accuracy and performance than FWA and other heuristic intelligent algorithms.