SLOTSA: A Multi-Strategy Improved Tunicate Swarm Algorithm for Engineering Constrained Optimization Problems

With the increasing demands on industrial products, the design and manufacturing problems of industrial products are becoming more and more complex. Many industrial design problems belong to nonlinear optimization problems and NP-hard problems, such as shop scheduling, path planning, and industrial part design problems. In the last decade, more and more researchers have developed heuristic- based algorithms to deal with industrial design problems. Tunicate swarm algorithm (TSA) is a newly proposed high performance heuristic optimization algorithm. The TSA algorithm has the problems of low solution accuracy, slow convergence and easy attraction by local extrema. In order to solve some drawbacks of TSA, this paper improves TSA and proposes the Sine cosine -Levy-Opposition-based Tunicate swarm algorithm (SLOTSA) by combining opposition-based learning, Levy flight and positive cosine operator. Comparative experiments on SLOTSA in 10 benchmark functions fully demonstrate the rationality of the three improvement strategies. In addition, this paper applies SLOTSA to two practical industrial design problems, and the optimization results demonstrate the wide applicability of SLOTSA to industrial design problems.