This study investigates- the spectrum allocation problem for multiple unmanned aerial vehicles(multi-UAVs) systems operating under uncertain environmental interference, aiming to balance mission effectiveness and energy efficiency. An enhanced NSGA-II optimization framework is proposed, integrating a Gaussian incomplete interference model with an innovative partially mapped crossover(PMX) interval crossover mechanism. The methodology establishes a dual-objective optimization paradigm that simultaneously maximizes task utility and minimizes communication power consumption through Pareto optimality analysis. Experimental results demonstrate that the proposed approach significantly outperforms conventional methods in interference mitigation and energy conservation, achieving substantial improvements in spectrum utilization efficiency under dynamic conditions. The core contributions lie in two aspects: (1) a probabilistic interference prediction model addressing environmental uncertainty, and (2) a diversity-preserving evolutionary operator design that enhances solution convergence. This work provides a theoretically grounded framework for resource-constrained UAV network deployments in complex electromagnetic environments.