Unmanned aerial vehicles （UAVs） are widely applied across various fields due to their high mobility and strong coupling capabilities. In the military domain， multi-objective task allocation for UAV swarms holds critical significance. Based on the current research status and latest advancements in UAV task allocation， this paper addresses the limitations of traditional algorithms， such as insufficient robustness， slow convergence， and susceptibility to local optima in complex missions—by proposing an improved wolf pack algorithm （IWPA） based on a multi-rule mechanism. An optimization model is constructed with the objective of maximizing strike effectiveness， incorporating factors such as enemy target distance， attack threat， and target value.Simulation results demonstrate that， compared to other heuristic algorithms， the proposed method achieves faster optimization speed and superior performance， indicating broad application prospects.