A hypervolume-based evolutionary algorithm for rescue robot assignment problem of nuclear accident

Robots usually carry out rescue tasks after nuclear accidents due to harsh environmental conditions such as radiation, high temperatures, and pressure. The assignment of rescue robots is a crucial aspect of this study, serving as a preliminary step in multi-robot task allocation. Its primary objective is to allocate robots to different groups based on the distribution of tasks. This paper models robot assignment as a multi-objective optimization problem by considering total execution time, regional load balance degree, and total transfer time. An improved hypervolume estimated multi-objective evolutionary algorithm (IHypE) is proposed. An encoding method is devised to represent the assignment of the robots. A corner-point-based hypervolume approximation method is proposed to efficiently measure the quality of each solution. An environmental selection that is more applicable to this problem is developed. In the experimental section, we conduct a comparative analysis between the proposed method and four state-of-the-art algorithms, namely MMODE-ICD, MOEA/D-2TH, NSGA-II, and SPEA2SDE. This comparison is performed on nine instances with varying scales. By evaluating the algorithm’s performance using five evaluation indicators, including hypervolume, inverse generational distance, C-Metric, Spacing, and Spread, we demonstrate the competitiveness of the proposed algorithm in terms of convergence and diversity.