Improved Cooperative Search Algorithm with Multi-Strategy Fusion

Proposed in response to the slow convergence speed, limited search space, and susceptibility to local optima of the basic cooperative search algorithm when dealing with complex optimization problems, the improved cooperative search algorithm with multi-strategy fusion is introduced. By using an improved Circle chaotic mapping to initialize individual positions of the employees, the diversity of the population is enhanced, addressing the issues of uneven population. By using an improved Circle chaotic mapping to initialize individual positions of the employees, the diversity of the population is enhanced, addressing the issues of uneven population distribution and insufficient search space. By combining it with the Sparrow algorithm, global guidance is introduced to accelerate the convergence speed of the algorithm. The foraging strategy of the Artificial Rabbit algorithm is integrated to the foraging strategy of the Artificial Rabbit algorithm is integrated to balance the algorithm's local development and search capabilities. Furthermore, the simulated annealing mechanism is incorporated into the position updating process to help the algorithm escape from local optima. Through a comparison using 5 test cases, the results demonstrate that the Improved Cooperative Search Algorithm (ICSA) exhibits superior optimization performance. The algorithm, both before and after improvement, is combined with the Kernel Extreme Learning Algorithm (KEA). The algorithm, both before and after improvement, is combined with the Kernel Extreme Learning Machine to predict the total frictional force of bowstring contact, and the results confirm the feasibility and superiority of the improved algorithm.