In this study, an optimization model for task allocation in robotic clusters of smart warehouses based on mivar logic is developed, aiming to minimize total energy consumption. A mathematical framework is proposed that formalizes the multi-objective optimization problem, considering constraints on robot battery levels and the uniqueness of task assignment. The choice of an optimization criterion is justified, accounting for energy costs for movement to both target locations and charging stations. The effectiveness of the proposed approach is revealed through comparative experiments in the WiMi simulation environment for scenarios of varying scale (2, 3, and 5 robots and tasks). It is established that the use of the mivar decision-making system reduces the total energy consumption of the cluster by 3-10% compared to random allocation, with savings increasing as the system scales. Key elements of the model's implementation, including logical rules and dynamic adaptation algorithms, are presented. Thus, the research proves the practical significance of applying the mivar approach for enhancing the energy efficiency and operational performance of autonomous warehouse systems.