Short-term urban road congestion prediction considering temporal-spatial correlation

In order to address the gaps in study of short-term urban road congestion prediction based on Baidu map real-time condition data, a model for Pearson Correlation Coefficient (PCC) and Weighted Markov Chains (WMC) is constructed by combining historical temporal correlation data with spatial between sections. The use PCC method filter out spatially significantly related sections from upstream downstream target section add them set as input WMC achieve congestion. performances proposed models are validated using manually collecting map. research results show that integrate correlations data. Compared other three models, accuracy improved 3.096% average, error reduced 0.135 average.