Hybrid Optimization Based On Bat Algorithm And Application Research

With the increasing prominence of digital globalization,artificial intelligence based on big data will be the main thrust of future industrial and commercial update progress,and the corresponding group intelligence algorithm with high accuracy,fast convergence and simple structure will be an important research direction for future high-tech development.The bat algorithm is a new population intelligence algorithm proposed by Professor Yang Xinshe of Cambridge University,UK,based on the principle of bats using ultrasonic waves for obstacle avoidance and predation,and its most prominent advantages are fewer parameters and high robustness.The classical bat algorithm has strong computing power in dealing with some problems with few parameters and low dimensionality,but the shortcomings of poor population diversity and low convergence efficiency become obvious when dealing with data processing,vehicle scheduling and wireless network sensors with multiparameter and high dimensionality characteristics.In order to enhance the classical bat algorithm,this study analyzed and compared the existing studies,proposed an improvement strategy,and successfully applied the improved algorithm to Wireless Sensor Networks(WSNs)and vehicle scheduling problems,and finally proved its applicability and superiority with extensive simulation experiments.The specific research work of this study is organized as follows.(1)The common deficiencies in the search and convergence strategies of different metaheuristic algorithms are pointed out,and the advantages and application areas of different improvement strategies are analyzed in detail by listing the current improvement and application cases of related algorithms.The development history and improvement strategy of the bat algorithm are highlighted,and the implementation flow and pseudo-code of the classical bat algorithm are given,and the optimization search effect of the algorithm is shown visually in the function 3D diagram to pave the way for the next improvement and application.(2)A multi-evolutionary strategy bat algorithm(differential sine cosine bat algorithm,DSBA)for node 3D localization is designed to address the problem that the bat algorithm runs smoothly but the evolutionary operator is weak.The bat algorithm is characterized by smooth iterations and fast approach to the optimal value in the early stage,but the search mechanism is weak in the later stage and cannot reach the global optimum,while the differential evolution algorithm and the sine cosine algorithm have the characteristics of large fluctuation in the early iteration and high convergence accuracy in the late stage.And in DSBA,differential evolution operator and sine cosine evolution operator are introduced into the update mechanism of the bat algorithm to enhance the update evolution capability of the algorithm,and dynamic weights are used to coordinate the local and global overall cooperation with the merit-seeking capability.In the end,the test function results in many different dimensions confirmed that DSBA not only inherits the characteristics of stable operation and fast convergence of the bat algorithm,but also has the advantages of high efficiency and solution quality of the other two algorithms.In addition,DSBA is also tested in node 3D localization,and the3 D localization experiments with different disturbance factors prove that DSBA not only has obvious advantages in function finding,but also applies to multi-constrained3 D localization problems.(3)Node localization in two-dimensional(2D)and three-dimensional(3D)space for wireless sensor networks(WSNs)remains a hot research topic.To improve the localization accuracy and applicability,this study first design a quantum annealing bat algorithm(QABA)for node localization in WSNs.QABA incorporated quantum evolution and annealing strategy into the framework of the bat algorithm to improve local and global search capabilities,achieved search balance with the aid of tournament and natural selection,and finally converged to the best optimized value.Additionally,this study used trilateral localization and geometric feature principles to design 2D(QABA-2D)and 3D(QABA-3D)node localization algorithms optimized with QABA,respectively.The experimental results indicated that,compared with other heuristic algorithms,the overall search and convergence performance of QABA is substantially improved,with the highest average error of QABA-2D reduced by 90.35% and the lowest by 17.22%,and the highest average error of QABA-3D reduced by 75.26% and the lowest by 7.79%.(4)Aiming at the problems of low efficiency and high cost in logistics transportation,this paper designed a double-layer collaborative vehicle route optimization and scheduling method.The upper algorithm introduced the search mechanism of bat algorithm into the simulated annealing route optimization algorithm,combined the variable neighborhood search with natural selection and tournament strategy,and optimized the vehicle route under the constraints of vehicle capacity and time window.The lower algorithm used an improved bat algorithm,which can schedule vehicles under the constraints of customer demand and the number of different vehicles.It was experimentally concluded that,compared with the comparison algorithm,the average cost and path length of the upper algorithm were reduced by 7.05% and 2.79%.The cost of the double-layer algorithm is 13.59% lower than that of the upper algorithm.Finally,the work on the improvement and application of the bat algorithm is carefully summarized,and an outlook on future research work is presented.