Work schedule optimization for electric agricultural robots in orchards

• Published in: Computers and electronics in agriculture 2023-07, Vol.210, p.107889, Article 107889
• Main Authors: Hizatate, Tomoaki, Noguchi, Noboru
• Format: Article
• Language: English
• Subjects: agriculture; algorithms; batteries; Electric vehicle; electric vehicles; electronics; Genetic algorithm; mountains; Optimization; pesticides; Work schedule
• ISSN: 0168-1699
• DOI: 10.1016/j.compag.2023.107889

•We present an algorithm to optimize electric agricultural robots' work schedules.•Our algorithm is effective in the scheduling the robots working on the sloping fields.•The potential for applications to robot operational and design problems was confirmed. We developed an optimal work-scheduling algorithm for autonomous spraying electric robots in orchards. Orchards are often located in mountainous areas, where the slopes are greater than those in plain areas. The work efficiency of agricultural operations in orchards thus depends on the quality of the work schedule, which in turn requires that the work schedule be optimized. We developed an algorithm that considers the returns of an agricultural robot electric vehicle to a replenishment base for pesticide refills and battery recharging. Our proposed method consists of three steps: the reconstruction of individual robots' work schedules, optimization of the overall work schedule, and optimization of the direction of entry for each driving lane in a schedule. We used a multi-layered genetic algorithm (LGA) to solve these problems. A comparison of the work schedules created by the LGA with those generated by other algorithms confirmed that the LGA can create an optimum or a suboptimum work schedule. We also created work schedules for multiple robots, which demonstrated that the proposed method can be applied to robot-operation problems such as the determination of the required number of robots within the constraints of total work time and robot specifications. Our results confirmed that the quality of a work schedule for robot EVs improves when the parameters related to robot modeling and field modeling are optimized, suggesting the possibility of applying the proposed method to robot and field design problems.