DEZIO LORENZO

Summary of PhD Research - Implementation of Genetic Algorithms for the Optimization of Ground and Rock-Mass Improvement Works

The PhD project develops a C# computational framework for the multi-objective design of ground-improvement measures against soil liquefaction. The current case study is the Christchurch Central Business District (New Zealand), where CPTu data are used to characterize the subsoil, evaluate liquefaction susceptibility, compute LPI, and identify spatially critical zones. A Genetic Algorithm based on NSGA-II explores alternative reinforcement/stiffening solutions by varying treatment geometry and mechanical parameters, while minimizing residual liquefaction severity and the amount of material required. Preliminary results show a stable workflow from CPTu processing and spatial zoning to Pareto-front generation and graphical visualization of optimized column and wall/grid layouts. The current development phase focuses on the scientific verification and refinement of the reinforcement equations, applicability domains, and numerical controls. Further work will extend the validated treatment models and improve software robustness through sensitivity analyses, verification across multiple CPTu profiles, site-specific calibration, and systematic testing of alternative ground-improvement strategies.

Key words

CPTu; Soil liquefaction; Christchurch CBD; C# software; NSGA-II; Multi-objective optimization; Pareto front; Ground improvement; Reinforcement