Spearman-Coefficient-Based Improved D-S Evidence Theory Algorithm Considering the Initial Importance of Evidence Body and Its Application on Multi-Source Information Fusion

Decision-making technology based on multi-source information fusion plays an increasingly important role in the field of R&D, design and manufacturing in digital-intelligence industry at the present time. Dempster-Shafer (D-S) evidence theory is a method that can effectively integrate multiple sources of information, conduct uncertainty reasoning and support decision making. How to carry out information fusion under the condition of evidence body conflict is the key to the effective application of D-S evidence theory. Moreover, due to the different sources of evidence, the initial importance of each evidence body is also a necessary concern prerequisite for the effective application of D-S evidence theory in real-world decision-making problems. In this paper, we propose an improved D-S evidence theory algorithm based on Spearman correlation analysis. According to the importance of each evidence body, the basic probability assignment (BPA) of D-S evidence theory is revised firstly. Then, based on the correlation of the evidence bodies in the same recognition framework, Spearman correlation coefficient is used to further modify the weight of evidence bodies. Finally, the modified evidence bodies are fused based on the improved evidence combination rule, thus the inference result is obtained. The method proposed in this paper is compared with several existing improved D-S evidence theory algorithms, and the numerical case results show that the proposed method is more robust in obtaining results aligning with common sense. The application results of a reliability prediction case for a new type of armored vehicle show that, for taking the initial importance of each evidence body into account, the proposed method can lead us to a decision that more in line with intuition and field test results compared with other existing improved methods.