ABSTRACT
A multi-objective optimisation design method for composite tunnel linings is proposed in this paper. The geometry and material parameters of the inner and outer components of tunnel linings are treated as design variables, while the lining safety and cost value are considered as design objectives. Such a multi-objective optimisation problem is solved by integrating a computational model into the non-dominated sorting genetic algorithm. The solutions are obtained as a set of optimised design points, named the Pareto front, and all the designs from this set provide a trade-off relationship between the design objectives. In addition, the concept of a knee point is employed to provide the most recommended design which can be located on the Pareto front using the bending angle approach. Subsequently, a detailed case study is presented to demonstrate the application and advantages of the proposed method. It is shown that the proposed optimisation approach can help to achieve simultaneously a higher safety level and better cost efficiency for the design of composite tunnel linings.
