| Abstract: | In this paper, two questions common in tunnel design are addressed: (i) how to choose an optimum solution when more than one conflicting objective must be achieved; (ii) how to deal with data affected both by imprecision and randomness. Fuzzy Set Theory and Random Set Theory are used to develop a general interactive multiobjective procedure, which is then applied to the design of tunnel support/reinforcement. A case history illustrates how the procedure was successfully used in the preliminary design of a total of 40 km of tunnels in Central Italy. It is shown that the procedure allows the designer to become a knowledgeable decision maker because his interaction is required at the key points of the process, and because the trade-offs among the objective functions can be easily assessed. The designer's personal input is valued and clearly defined in its impact on the solution. The case history demonstrates that, without an optimization procedure, it is extremely likely (probability of 99%) that a solution is chosen, which either increases the costs without increasing safety, or decreases the safety without decreasing the costs. Finally, it is shown that both imprecision and randomness can be easily taken into account in tunnel design. |
