A belief-rule-based model for information fusion with insufficient multi-sensor data and domain knowledge using evolutionary algorithms with operator recommendations

Multi-sensor information fusion (IF) has attracted the attention of many researchers in different fields because it can improve modeling accuracy by integrating information gathered from multiple sensors. Traditionally, the inputs of the multi-sensor IF problem are mostly quantitative data. However, to provide a more comprehensive decision support, both quantitative data and experts’ domain knowledge should be synthesized in the IF process. Moreover, there may be insufficient data in many practical conditions, which would make many conventional approaches inapplicable. Because the belief rule base (BRB) has shown advantages in nonlinear modeling with insufficient data and experts’ domain knowledge, a BRB-IF model is proposed for the multi-sensor IF problem. To improve its efficiency, an optimization model and the corresponding optimization algorithm for BRB-IF are proposed. The particle swarm optimization algorithm and differential evolutionary algorithm are tested as the optimization engine to solve the optimization model with the operator recommendation strategy. The efficiency of the proposed BRB-IF is validated by a practical case study of threat level assessment, where a comparison between BRB-IF and the neural network is conducted.