Optimizing Mobile Robot Fleet Distributions in Solar Farms

The maintenance and cleaning of solar farms make up a significant portion of their operational costs. Recently, the use of robots to perform these tasks has gained popularity, requiring efficient routing for their effective deployment. To further complement their missions, charging stations can be strategically placed throughout the site which enable robots to recharge mid-operation. This work addresses this challenge by modelling the problem as a multiple travelling salesmen problem and developing a mixed integer linear programming formulation. Four different solution methods were implemented: a brute force method, a random method, a genetic algorithm based method, and an exact approach. In addition, a search strategy was developed that adds charging stops to existing tours, extending existing MTSP algorithms for problems with battery or fuel constraints.