Study On Two-stage Adaptive Matching Of Logistics Tasks And Resources

With the constant development of information technology in the field of logistics,logistics demand is changing toward the trend of diversified,dynamic and complicated characteristics.At the same time,along with the development of the logistics resources integration and configuration platform,in the face of vast,multi-granularity characterization information of logistics resources,demand information of logistics tasks,dynamic matching information and so on,the scale of logistics matching system is getting bigger and bigger,and more and more complex,rendering characteristics of nonlinear,high dimension and information ambiguity for the matching system.Therefore,how to make the matching between logistics resources and tasks more intelligent,flexible and efficient has become an urgent problem to be solved,meanwhile,solving this problem has also become the research goal of this paper.In this paper,a two-stage matching method combining fuzzy and fine matching is proposed to improve the matching efficiency and adapt to the demand changes in different stages,so as to realize the one-to-one matching between logistics resources and tasks.In the first stage,the fuzzy matching model based on the distance space function was proposed.The fuzzy matching was carried out according to the quantized logistics capability,and the one-to-many matching results were obtained to improve the matching efficiency.In the second stage,a fine-matching model of multi-objective combinatorial optimization was proposed,and the ant colony algorithm was designed to solve the model,so as to adapt to complex and diverse logistics needs and obtain one-to-one matching results.Then,an adaptive feedback method based on multi-layer feed-forward neural network is proposed to form a result feedback control and a matching closed loop to realize the cyclic optimization of the matching process.The main research work is as follows:1)Study the fuzzy matching of logistics resources and tasks in the first stage.Firstly,the logistics capability quantization model based on fuzzy matter-element TOPSIS relative closeness degree is adopted,and the weight of quantization index is determined by projection method.According to the adaptive formalized description of logistics resources and tasks,and the quantized logistics capability,a model based on the distance space function matching is proposed,which is conducive to the rapid formation of candidate resource pool and the efficient matching of logistics tasks andresources 1 to many.2)Study the fine matching of logistics resources and tasks in the second stage.Firstly,according to the basic idea of fine matching,a multi-objective combinatorial optimization matching model was constructed.Then,based on the improved TSP problem,the ant colony algorithm was designed for combinatorial optimization matching.Adapting to the changing and complex logistics task requirements and obtaining the one-to-one matching results of logistics task and resource.3)Study adaptive matching control.Analysis of each link in two-stage matching,based on multilayer feedforward neural network adaptive feedback control algorithm is proposed,Through adaptive learning and feedback of the operation strategy and intelligent decisions of two-stages matching,improving the predictability of logistics demand,forming a matching optimization loop,shrinking matching range,and improving the matching intelligent.4)Example analysis.The validity and rationality of the two-stage matching method are verified by an example,the one-to-one matching results are formed,and the feasibility and effectiveness of the multi-layer feedforward neural network adaptive feedback algorithm for the resource recommendation of logistics tasks are verified.