Multiobjective Task Allocation for Electric Harvesting Robots: A Hierarchical Route Reconstruction Approach

Area of Science:

• Agricultural Robotics
• Operations Research
• Artificial Intelligence

Background:

• Rising labor costs drive the adoption of multirobot systems in agriculture.
• Coordinating multirobot systems for orchard harvesting faces challenges in balancing makespan and energy consumption.
• Real-world constraints like variable speeds and battery limits complicate robot task allocation.

Purpose of the Study:

• To define the multiobjective agricultural multielectrical-robot task allocation (AMERTA) problem, incorporating practical constraints.
• To develop an effective algorithm for optimizing multirobot coordination in agricultural harvesting.
• To improve the efficiency and energy management of automated harvesting systems.

Main Methods:

• Formulation of the agricultural multielectrical-robot task allocation (AMERTA) problem.
• Development of a hybrid hierarchical route reconstruction algorithm (HRRA).
• Integration of hierarchical encoding, dual-phase initialization, task-sequence optimization, and specialized route reconstruction operators within HRRA.

Main Results:

• HRRA demonstrated superior performance compared to seven state-of-the-art algorithms across 45 test instances.
• Statistical analysis confirmed HRRA's competitiveness and its ability to explore novel solution spaces.
• The algorithm effectively addresses complex constraints in multirobot task allocation.

Conclusions:

• The proposed HRRA offers an effective solution for the AMERTA problem in agricultural automation.
• This research advances the theoretical understanding of multirobot coordination.
• The findings provide practical insights for enhancing the efficiency of automated orchard harvesting.