Nanchang University

Multi-objective optimization struggles to balance convergence and spread on irregular or disconnected Pareto fronts. IMOCCO adds manifold-feature collision, mass-restitution collision and angular-sparsity guidance to the centered-collision search, acting on the search operator, individual interaction and environmental selection. Across 31 standard test functions it holds a Friedman mean rank of 1.73, first among seven algorithms, and scores 2/0/0 on the Wilcoxon test for a four-bar truss and a disc brake.

Multi-objective optimization struggles to balance convergence and spread on irregular or disconnected Pareto fronts. IMOCCO adds manifold-feature collision, mass-restitution collision and angular-sparsity guidance to the centered-collision search, acting on the search operator, individual interaction and environmental selection. Across 31 standard test functions it holds a Friedman mean rank of 1.73, first among seven algorithms, and scores 2/0/0 on the Wilcoxon test for a four-bar truss and a disc brake.

On the Baijiabao slope in the Three Gorges Reservoir Area, a surrogate that learns random-field spatial features together with water-level and rainfall time series cut a full-year time-dependent probabilistic stability analysis from 24,230 seconds to 15 seconds, with a coefficient of determination of 0.975.

On the Baijiabao slope in the Three Gorges Reservoir Area, a surrogate that learns random-field spatial features together with water-level and rainfall time series cut a full-year time-dependent probabilistic stability analysis from 24,230 seconds to 15 seconds, with a coefficient of determination of 0.975.