LAUSR

Massive data sets and complex scheduling processes have high-dimensional and non-convex features bringing challenges on various applications. With deep insight into the bio-heuristic opinion, we propose a novel Beetle Colony Optimization (BCO) being able to adapt NP-hard issues to meet growing application demands. Two important mechanisms are introduced into the proposed BCO algorithm. The first one is Beetle Antennae Search (BAS), which is a mechanism of random search along the gradient direction but not use gradient information at all. The second one is swarm intelligence, which is a collective mechanism of decentralized and self-organized agents. Both of them have reached a performance balance to elevate the proposed algorithm to maintain a wide search horizon and high search efficiency. Finally, our algorithm is applied to traveling salesman problem, and quadratic assignment problem and possesses excellent performance, which also shows that the algorithm has good applicability from the side. The effectiveness of the algorithm is also substantiated by comparing the results with the original ant colony optimization (ACO) algorithm in 3D simulation model experimental path planning.