Abstract: In the existing studies on the layout scheme of ventilation shafts in the underground caves of hydropower stations, usually only a single factor is selected as an evaluation basis, and the specification of index weights often ignores the relevance of factors or relies too much on expert opinions. To improve decision making, this paper first constructs a comprehensive evaluation index system for layout schemes which includes technical feasibility, economic rationality, and safe reliability, etc. Then, for scheme selection, we develop an improved decision model of fuzzy cognitive map (FCM)-entropy weight method (EWM)-multi-attribute boundary area comparison (MABAC). This model calculates the index combining weights by adopting an improved FCM that calculates factor correlation and has strong autonomous learning ability, and combines it with EWM. We improve the MABAC model by taking into the decision basis the relative distance between the selected scheme and the positive and negative ideal schemes, so as to overcome the problem with the traditional MABAC in distinguishing between a standard-distance scheme and its similar schemes. Finally, this new method is applied to optimizing the ventilation shaft layout scheme for an underground cavern at the Luoning pumped storage power station. The results show that the optimized scheme can achieve satisfactory effects in technology, economy and safety, demonstrating the advantages of the improved FCM and improved MABAC and a new idea for further study on multi-attribute decision making for cave ventilation shafts.

Key words: underground cave, ventilation shaft layout scheme, improved FCM, entropy method, multi-attribute decision making, MABAC