Research On Multi-Task Collaboration Mechanism In Distributed Intelligent Systems

In complex distributed intelligent systems(DIS),collaboration between multiple intelligent subsystems is a critical problem.This dissertation uses the cloud model theory,intelligent optimization algorithms and the multi-agent systems(MAS)theory to conduct in-depth theoretical discussions on the key issues of the resource allocation algorithm,the decision-making evaluation method and the utility allocation strategy of collaborative coalition,which is the main way to realize multi-task collaboration in DIS.It is also applied to the optimization of resource allocation and the decision-making evaluation for the new energy vehicle(NEV)coalition.Therefore,it provides an effective method and approach to solve the mechanism problems in complex DIS,such as resource allocation,decision-making evaluation and utility allocation.The main research contents and innovations in this dissertation are as follows:1)To solve the problem of resource allocation of multi-task collaborative coalition(MTCC),a MTCC resource allocation method based on the BPSO-BDE hybrid algorithm is proposed.A resource-oriented MTCC resource allocation model is constructed,and a resource conflict resolution mechanism based on two-dimensional binary coding revision is established to avoid potential resource conflicts and coalition deadlocks in MTCC.Additionally,the two-dimensional binary coding method is used to encode individual population,the total profit value of coalition is taken as the fitness value,and the particle velocity and position update of BPSO and operations such as population variation,crossover,selection of nonparametric variation BDE are integrated to obtain the optimal scheme of coalition resource allocation through continuous optimization iteration.The advantages of the BPSO-BDE hybrid algorithm are verified by the comparative analysis of the MTCC resource allocation of the BPSO algorithm and the BDE algorithm without parameter variation.2)On the basis of the optimal resource allocation of MTCC,aiming at the uncertainty problem in the process of MTCC decision-making evaluation,the concept of the evaluation cloud is introduced and a decision-making evaluation method of MTCC based on the cloud model theory is proposed.First of all,on the basis of establishing a decision-making evaluation system of multi-task coalition,the index evaluation cloud and the weight of evaluation index are determined,and the decision-making evaluation cloud of single-task coalition is obtained by applying cloud weighted arithmetic averaging(CWAA)of the cloud clustering method.Then,the MTCC decision-making evaluation cloud is obtained by combining the task weight and the single-task coalition decision-making evaluation cloud to perform two-times cloud clustering,and the cloud similarity based on distance is used to rank and select alternative collaborative coalition schemes in order to determine the optimal one.Finally,the effectiveness of the collaborative coalition decision-making evaluation method based on the cloud model theory is verified through the comparative analysis of the coalition evaluation method based on the D-S evidence theory with an example.3)On the basis of the above research work,a utility allocation strategy of MTCC based on trust and ability is proposed.Under the condition of non-reducing utility and resource nonconflict,the utility of each member of the coalition is allocated by integrating the trust degree and the evaluation information of contribution resource ability.The rationality and effectiveness of the utility allocation strategy are proved through the comparative analysis of an example.4)Considering that the big data is an important strategic resource and in view of the characteristics of the new energy vehicle industry coalition,a new energy vehicle coalition resource optimal allocation and decision-making evaluation method orienting to the big data is proposed on the basis of the aforementioned study on the resource optimal allocation and decision-making evaluation method.Firstly,based on the BPSO-BDE hybrid algorithm,the coalition resource allocation orienting to big data is optimized.Then,the coalition decision-making evaluation of the big data-oriented is carried out based on the cloud model theory.Finally,an example is given to verify the effectiveness and rationality of the big data-oriented resource allocation optimization and cloud decision-making evaluation for the new energy vehicle coalition.