In the time of informationized war, the logistics system is usually the target of enemy attacks. Choosing a reasonable transportation scheme can prevent enemy attacks and reduce transportation costs so as to refine the transportation support. Based on wars’ requisition towards transportation and the battlefield environments’influence on efficiency of transportation, the paper establishes a performance assessment target system of military container transportation and puts forward a method to get weight values in complex environment, then uses the theory of Fuzzy Neural Network to build a method of best transportation programmers’ choice. Target parameters obfuscated into feature vector as the input, and comprehensive evaluation values as the expected output, the network is finally used after training and well resulted to evaluate multiple route selection schemes.