Measuring the effect of congestion on traffic accidents in Chengdu involves a multi-faceted approach utilizing big data analytics, econometric modeling, and advanced monitoring systems. Researchers typically leverage high-resolution spatial and temporal data, such as GPS trajectories from Didi ride-hailing services and taxi fleets, to extract speed and vehicle flow variables. Statistical techniques like K-Means clustering are used to categorize traffic into levels ranging from very smooth to very slow, while the Hidden Markov Model, often combined with dissipative structure and entropy theories, helps predict congestion patterns across the city's ring roads. To isolate the impact on safety, analysts employ econometric models—such as the Negative Binomial or Bayesian spatial models for accident frequency and ordered logit models for accident severity—to investigate how congestion indices like the volume-to-capacity (v/c) ratio or traffic delay correlate with incident rates. Furthermore, the Chengdu Intelligent Transport Command Center uses a real-time smart monitoring system to detect anomalies in traffic flow and coordinate with digital mapping platforms like Baidu and Gaode, providing a comprehensive framework to quantify how congestion-driven factors like speed variance and abrupt lane changes contribute to accidents.