Abstract: The evaluation of decoy lure effectiveness in electromagnetic countermeasure environment is influenced by multidimensional factors, and there are complex coupling and causal relationships between different indicators. Traditional linear weighting methods are difficult to fully reflect their inherent mechanisms. A performance evaluation method combining grey decision laboratory analysis and improved genetic algorithm is proposed to enhance modeling rationality and optimization performance. Establish a causal relationship network for performance indicators using grey decision laboratory analysis, extract key factors, and determine indicator weights; construct a comprehensive efficiency objective function and introduce adaptive crossover and mutation probabilities, as well as grey prediction operators, to improve the genetic algorithm and achieve efficiency configuration optimization. The experimental results show that during the iteration process, the loss value of the proposed method rapidly decreases to within 0.05 and tends to stabilize around 40 generations, while the differential evolution method takes 80 iterations to converge to around 0.1, and the weighted average rule always maintains above 0.2. In terms of lure success rate, it reached 92% when 10 decoys were deployed, and approached 97% when 50 decoys were deployed. The research conclusion indicates that this method can maintain high efficiency and convergence stability under limited resource conditions, providing effective support for decoy deployment optimization in complex electromagnetic environments.