Social spider optimization(SSO)is a heuristic optimization algorithm designed to simulate a series of behaviors of social spiders such as cooperative predation,public web information exchange and reproduction between male and female spiders.In this algorithm model,the population is divided into female population and male population according to a certain proportion,and the individuals collaborate depending on their gender differences.Such mechanism not only reflects the biological behavior of the social spider population,but also balances the local and global search ability of the algorithm to a certain extent.Social spider optimization algorithm has the advantages of clear structure,easy to understand and strong search ability,so it has been widely concerned by scholars at home and abroad.With the further study of social spider optimization algorithm,scholars have found some shortcomings of the algorithm,such as slow convergence speed,low accuracy,easy to fall into local optimization,etc.,which seriously restrains the application of social spider optimization algorithm in areas of science and engineering.Regarding the several shortcomings of social spider optimization algorithm,we first improve the algorithm from two aspects: introduction of the new strategy and combination with good mechanism of other algorithms,and then the improved algorithm is applied to a series of function optimization problems and the wireless sensor network coverage optimization problem,the main research work is as follows:(1)By combining the differential evolution algorithm and the social spider optimization algorithm,an improved differential-evolution-based social spider optimization algorithm(DESSOcw)is proposed.This algorithm uses the differential mutation mechanism of differential evolution algorithm to mutate some randomly selected female spiders,so as to improve the diversity of the population,and thus to improve the shortcomings of slow convergence speed and low convergence progress of the original swarm spider optimization algorithm.In addition,the inertia weight and learning factor mechanism of particle swarm optimization algorithm are introduced into the position updating equation of the original social spider algorithm to accomplish the dynamic adjustment of the learning ability of particles from the global optimal individuals and local optimal individuals,and therefore to effectively balance the local and global search ability of the algorithm.(2)An improved social spider algorithm(CGSSO)is proposed,which combines chaos optimization strategy and Gaussian disturbance.The chaos optimization strategy is introduced into the initialization of the algorithm,so that the initial solution of the algorithm is uniformly distributed in the solution space,thus expanding the search scope of the algorithm.At the same time,during each iteration of the population,some randomly selected female spiders are adaptively disturbed by Gaussian disturbance to increase the diversity of the population and thus to improve the ability of the algorithm to jump out of the local extremum.Experimental results show that CGSSO algorithm has better convergence speed and solution accuracy.(3)A DESSOcw-based method is provided to solve the wireless sensor network coverage problem.In this method,each spider is encoded as a node deployment mode,the probability perception model is utilized to calculate the sensor node coverage,and the resulted coverage function is selected as the fitness function of the DESSOcw algorithm.The simulation results show that our DESSOcw-based algorithm can obtain better the sensor network coverage compared with differential evolution algorithm and the original social spider optimization algorithm. |