Design of Dynamic Load Balancing Optimization Model Based on Improved Weights in Dempster–Shafer Evidence Algorithm for Cloud Computing Environments

This paper improves the Dempster–Shafter (D-S) evidence algorithm, proposes a dynamic load balancing model based on the improved weights of the D-S evidence algorithm, combines with the cloud computing environment, proposes a cloud computing resource load balancing model, trains the model and finally compares the centralized balancing algorithms model. The specific conclusions are as follows: (1) This paper can correctly identify the target when two pieces of evidence are collected and the target can be correctly and efficiently identified for three, four and five pieces of evidence, especially for five pieces of evidence m(A)[Formula: see text] [Formula: see text] [Formula: see text]0.907, 7 which is higher than 0.795 8 of Murphy method. The target can be identified correctly and efficiently with 3, 4 and 5 evidence, especially with five evidences m(A) = 0.907, 7 which is higher than 0.795 8 in Murphy’s method, 0.890 9 in Deng Yong’s method, 0.893 7 in Cao Jie’s method and 0.896 1 in Yuan Xiao Ke’s method. (2) The decision-making method is based on the trust function and can be used to avoid serious conflict of evidence. The trust results are continuously adjusted according to the weights and finally, the trust results of multiple factors are fused and the fusion results are increased, then the point is judged to be overloaded. Otherwise, it is lightly loaded. (3) Construct the cloud computing resource load balancing model to achieve the balance of the overall resource nodes and improve the utilization of resources. (4) By thoroughly examining the D-S evidence algorithm with enhanced weights, this model of resource load balancing incorporates the D-S evidence algorithm with distinctly improved weights, refining it to identify the most effective load-balancing approach.