Multi-objective particle swarm optimization for mechanical harvester route planning of sugarcane field operations

One of the important aspects to increasing sugarcane mechanical harvesting efficiency is the path planning of the harvester, involving direction and field accessibility constraints. Moreover, in real-life applications, the two objective functions pertaining to minimization of harvested distance and maximization of sugarcane yield are conflicting and must be considered simultaneously. This paper presents a multi-objective with the variant of the particle swarm optimization combined gbest, lbest and nbest social structures (MO-GLNPSO), to solve sugarcane mechanical harvester route planning (MHRP). A new particle encoding/decoding scheme has been devised for combining the path planning with the accessibility and split harvesting constraints. Numerical computation results on several networks with sugarcane field topologies illustrate the efficiency of the proposed MO-GLNPSO method for computation of MHRP, which is compared with other methods such as the traditional particle swarm optimization (PSO) and Non-dominated Sorting Genetic Algorithm-II (NSGAII) by the values of C˜ metric indicator. The solutions found in this work can offer a decision maker a choice of trade-off solutions, providing sufficient options to give planners the power to make an informed choice that balances the important objectives.