Bayesian reasoning approach based recursive algorithm for online updating belief rule based expert system of pipeline leak detection

In this paper a recursive algorithm based on the Bayesian reasoning approach is proposed to update a belief rule based (BRB) expert system for pipeline leak detection and leak size estimation. In addition to using available real time data, expert knowledge on the relationships of the parameters among different rules is incorporated into the updating process so that the performance of the expert system can be improved. Experiments are carried out to compare the newly proposed algorithm with the previously published algorithms, and results show that the proposed algorithm can update the BRB expert system faster and more accurately, which is important for real-time applications. The BRB expert systems can be automatically tuned to represent complex real world systems, and applied widely in engineering.