A Genetic Algorithm with Feasibility-Preserving Representation for the Tripper Car Movement Problem

The Tripper Car Movement Problem presents a critical challenge in optimizing material transport within mineral processing systems. This paper addresses the problem and proposes a tailored Genetic Algorithm. The algorithm uses a ternary representation based on relative movements, a context-aware mutation operator, and an adaptive immigration mechanism.
The representation preserves movement feasibility by construction, so repair procedures for movement-related constraints are not required.
 Computational experiments on benchmark instances show that the method reproduces optimal solutions for small instances and achieves highly competitive results for larger ones. For several large-scale instances with longer planning horizons, the proposed GA outperformed the best solutions obtained by the MILP model within the imposed time limit, while requiring significantly less computational effort. The results highlight the method's scalability on benchmark instances covering problem dimensions representative of real-world industrial applications.