Enhanced Beetle Whisker Search Algorithm Based On Multi-directional Perceptio

Meta-heuristic algorithms are methods for solving complex optimization problems,which find optimal or satisfactory solutions based on mechanisms of computational intelligence,and are sometimes referred to as intelligent optimization algorithms.Optimization problems involve all kinds of engineering problems and are also closely related to life.The research and application of intelligent optimization algorithms have been widely concerned by researchers,because intelligent optimization algorithms are suitable for solving complex optimization problems.The Beetle Antennae Search(BAS)algorithm,as a bio-inspired optimization algorithm,is used to solve optimization problems by studying the beetle’s process of foraging and mate seeking in nature.Due to the single individual and small size of the beetle,the algorithm has the reduced amount of code,increases the running speed,and is easier to implement.However,the basic beetle antennae search algorithm is less effective in finding the optimal solution in complex high-dimensional problems.To address the problem of poor optimization of high-dimensional functions,most of the existing improvements to the beetle antennae search algorithm have used multiple beetles(i.e.,beetle swarm)to search or have fused BAS with other algorithms to find the optimal solution.The thesis improves the algorithm in three aspects while using only a single beetle.The specific improvements are:(1)The ease of use and stability of the algorithm are improved by using the step size adaptive update method: three step size reduction methods are compared,and the optimal step size formula and the optimal accuracy factor are selected to control the step size.The step size is automatically generated based on the number of iterations according to the formula,which avoids the uncertainty of parameter settings and improves the ease of use of the beetle antennae search algorithm.(2)The efficiency of algorithm search is improved by contemporary optimal update method: Under the proposed contemporary optimal update strategy,the optimal solution among the left-right solutions and the next solution is selected for the update of the beetle position in each iteration to avoid the algorithm falling into local optimum.(3)A multi-directional sensing method is used to improve the efficiency and accuracy of the algorithm for finding the optimum: in calculating the left-right solutions and the next step solution,multiple random directions are used to increase the exploration range of the algorithm for the solution space at the current step.The performance of the algorithm is tested in the field of function optimization as well as in real engineering optimization problems,specifically:(1)F1,F3 to F30 in the CEC2017 test set are selected to test the performance of the algorithm.It is effectively verified and compared with many other intelligence algorithms,including classic,uptodate,and improved BAS algorithms.(2)The WSN coverage optimization problem is selected for testing.The performance as well as the practicality of the algorithm is tested and proved.Through the above tests,the algorithm proposed in the thesis effectively improves the efficiency as well as stability of the beetle antennae search algorithm in complex and high-dimensional function optimization,which has certain advantages.