<p>The rapid advancement of artificial intelligence has spurred the widespread application of autonomous multi-robot systems across numerous domains, which is particularly beneficial for alleviating the increasingly severe labor shortage in agriculture. However, designing efficient routing and task scheduling strategies for these robotic systems to achieve optimized collaborative operations has become a critical challenge. This study investigates the agricultural multi-objective multi-robot routing and scheduling problem, formulating it as a domain-specific variant of the multi-trip vehicle routing problem with strict agricultural constraints. Its core objectives are to simultaneously minimize the total energy consumption of the system and the task completion time (i.e., the makespan). Given the inherent conflict between these two objectives, adopting a multi-objective optimization method is the necessary choice for solving this problem. Therefore, we propose the multi-objective hierarchical route optimization algorithm (MHROA). This algorithm employs a hierarchical encoding method to clearly define the operational routes undertaken by each robot. On this basis, the algorithm optimizes both the task allocation between routes and the task sequence within individual routes to ensure the rationality of the route structure and task execution order. Through comparative experiments conducted on test instances simulating orchard environments, the effectiveness and superiority of the MHROA have been validated, demonstrating its potential in solving complex multi-robot routing and scheduling problems.</p>