Raw Data_Mountainous Logistics Park Simulation Data

The dataset, "Raw Data_Mountainous Logistics Park Simulation Data," was generated using a custom-developed digital twin simulation platform built with Python. The simulation was based on Geographic Information System (GIS) data and operational patterns of a real mountainous logistics park (Park G) in Guizhou Province, China. It models one week (168 hours) of continuous park operations under two scenarios: a Control Group (traditional rule-based scheduling) and an Experimental Group (a novel "Prediction-Simulation-Control" integrated framework). A 3-hour heavy rainfall event was artificially introduced to test system robustness, and additional sensitivity tests (e.g., adding 10% Gaussian noise, extending rainfall to 6 hours) were performed. After data cleaning and removal of transient startup/shutdown anomalies, the final dataset comprises two tables: the Simulation Base Data Table (aggregated by operation period and work area) and the KPI Summary Statistics Table (containing averages, minimums, maximums, and standard deviations across multiple simulation runs). The dataset provides key performance indicators such as vehicle turnaround time, queue length, response time to abnormal events, and dock utilization. It offers important insights into the operational efficiency and resilience of mountainous logistics parks, supporting the core conclusions of the related research, including documented improvements in efficiency (e.g., an 18% reduction in turnaround time), congestion mitigation (e.g., a 32.1% reduction in queue length), and model robustness (e.g., maintaining 88.3% scheduling effectiveness during a simulated 6-hour rainstorm).