Design And Implementation Of Logistics System Based On Heuristic Algorithm

The rapid development of the logistics industry benefits from the continuous improvement of internet technology,and the optimization of transportation routes in logistics distribution has become a hot issue in the logistics industry.How to effectively reduce the cost of logistics distribution,reduce the time of logistics distribution,and form an efficient logistics distribution path solution is a problem that the current logistics industry urgently needs to solve and optimize.Among the algorithms for solving logistics route optimization problems,heuristic algorithms are the most widely used.The purpose of this article is to combine heuristic algorithms to design and implement a logistics system to provide efficient logistics solutions for the logistics industry.Among the heuristic algorithms,the spider monkey algorithm uses fission fusion mechanism to make the algorithm more and more popular,especially in solving high-dimensional optimization problems,it has certain advantages,and it surpasses other algorithms in reliability,effectiveness and accuracy.Therefore,this paper proposes a LASMO algorithm,which uses Laplace distribution instead of uniform distribution and applies it to the logistics location problem.In the initial stage of the population,the random number generated by the Laplace distribution is used to initialize the entire population,so that the initial position distribution of the entire spider monkey is more uniform,and the quality of the entire population initialization is optimized;in order to further accelerate the convergence speed of the algorithm,improve the algorithm In the local leader stage of the algorithm,the method of exponential decline and random logarithmic decline is adopted to make the step size adaptive.The step size in the early stage of the iteration is larger,and the step size in the later stage of the iteration is smaller;for further to balance the search speed and convergence accuracy of the algorithm,an acceleration factor is added in the global leader stage.Through eight very classic test functions,the performance and application scheme of the LASMO algorithm are simulated and tested,it is applied to the analysis of logistics site selection.The analysis shows that LASMO has ideal optimization and superiority.In order to improve the problem of time and cost in the logistics process,this paper uses the twolevel planning method,combined with the characteristics of logistics,and proposes a logistics transportation model based on the two-level planning method.The two-level programming method takes advantage of the independence and mutual influence between the upper and lower decision makers to solve the problem.First,the upper decision makers make decisions,and then the lower decision makers optimize their goals and make decisions based on the upper decision information.Finally,the upper-level decision-makers make the final decision using the optimized decisions of the lower-level decision-makers.Using the characteristics of mutual independence and mutual influence between the upper and lower decision makers in the two-level programming method,the cost and time in the logistics process are divided into the upper and lower objective functions to construct the logistics transportation model.The upper objective is the minimum total cost,and the lower objective for the shortest total time.According to the characteristics of mutual independence and mutual influence between the upper and lower target decision-makers,by fusing genetic algorithm and tabu search algorithm to improve,a HTSGA algorithm is proposed,which is used to solve the logistics transportation model,and the simulation verification is carried out through a design example,which proves the effectiveness of the algorithm and model in solving the logistics distribution problem.Finally,on the basis of the proposed algorithm and model,this paper implements several major functional modules of the logistics system based on heuristic algorithm,such as transportation management and distribution management,and tests several modules to achieve the expected results.It further verifies that the algorithm and model proposed in this article have strong practicability.