Abstract: In order to obtain the best emergency rescue path on time when an emergency occurs,an emergency rescue path selection model was built according to the random consistency condition after using the Zadeh fuzzy operators to quantize the fuzziness of links' travel speed and introducing a random dynamic variable to characterize the randomness of link's travel time,which could depict a road network under poor information condition with fuzziness and randomness.The model was transformed into a robust optimization model by mathematical induction,which could be solved by the modified Dijkstra algorithm.The KAUFMAN road network was applied to validate model's effectiveness.The results show that the mean absolute error of links' travel speed based on 2 different fuzzy operators is 6.271 km/h,that compared with traditional methods,a shorter time emergency rescue path can be obtained by using the robust optimal model,and that the model's algorithm complexity is expressed as a polynomial,which makes the model suitable for large gauge road networks.

Key words: poor information, fuzziness, random consistency condition, robust optimization, mathematical induction