Prediction　of　the　city　distribution　is　an　important　part　of　urban　distribution　nodes　and　channel　setting　in　the　process　of　transportation　resources　reasonable　configuration.　This　paper　adopted　158　subdistricts＇　logistics　data　of　Wuhan　to　analyze　of　correlative　factors　influencing　the　distribution　generation　by　studying　the　generation　mechanism　of　urban　distribution.　Then　significant　factors　were　extracted　through　the　Pearson　correlation　analysis.　At　last,　a　Takagi-Sugeno(T-S)　fuzzy　inference　system　was　established,　by　using　the　Gaussian　mixture　model　to　determine　the　fuzzy　rule　base　and　former　parameters　of　fuzzy　inference　model,　and　the　consequent　parameters　was　determined　by　the　least　square　method.　The　results　of　training　and　forecasting　indicated　the　method　has　higher　accuracy,　better　stability　and　extensibility.