“移动救护车”需求预测分析及运营模式探讨——以网约车为例

Demand Forecast Analysis and Operational Model Exploration of “Mobile Ambulances”—A Case Study of Ride-Hailing Services

摘要: 为探究网约车“移动救护车”适配性运营模式，优化中小城市非紧急就医出行运力配置，本文以娄底市为研究区域，选取2020~2024年各区县GDP与常住人口数据，采用极差标准化与熵权法构建需求强度指数，量化区域就医出行需求差异；基于GM (1, 1)灰色模型开展2025~2026年需求趋势预测，通过后验差比、小误差概率检验模型精度，并采用留一交叉法验证预测可靠性；依托静态需求分级结果与动态增长预测数据，分层设计适配高、中、低需求区县的差异化空间运营方案，结合通勤双峰出行规律搭建分时运力调度模式，配套司机激励、服务标准化等长效运营保障机制。结果表明：娄星区为高需求核心区，新化县、涟源市、双峰县为中需求区，冷水江市为低需求区；GM (1, 1)灰色模型对娄星区、双峰县、新化县、涟源市预测精度达一级，冷水江市为二级。研究形成需求量化、趋势预判与运营落地的完整分析体系，可为娄底市网约车“移动救护车”分区域、分时段运力调度与可持续运营模式落地提供量化数据支撑与实操决策参考。

Abstract: To explore the operational model of ride-hailing services as “mobile ambulances” and to optimize the allocation of transportation capacity for non-emergency medical travel in small and medium-sized cities, this study takes Loudi City as the research area, selects GDP and resident population data from various districts and counties spanning the period 2020~2024, employs extreme value normalization combined with the entropy weight method to construct a demand intensity index to quantify regional disparities in medical travel demand; utilizes the grey prediction GM (1, 1) model to forecast demand trends for the period 2025~2026, evaluates the model’s accuracy using the posterior difference ratio and small-error probability tests, and validates the predictive reliability through the leave-one-out cross-validation method; based on the static demand classification results and dynamic growth forecast data, this study designs differentiated spatial operational schemes tailored for high-, medium-, and low-demand districts and counties; it further establishes a time-based capacity dispatching model by incorporating the dual-peaked commuting pattern, along with long-term operational support mechanisms such as driver incentives and service standardization. The results indicate that Louxing District is the high-demand core zone, Xinhua County, Lianyuan City, and Shuangfeng County are medium-demand zones, while Lengshuijiang City is the low-demand zone; the GM (1, 1) model achieved first-level prediction accuracy for Louxing District, Shuangfeng County, Xinhua County, and Lianyuan City, and second-level accuracy for Lengshuijiang City. This study has established a comprehensive analytical framework encompassing demand quantification, trend forecasting, and operational implementation, providing quantitative data support and practical decision-making references for the regional and temporal allocation of ride-hailing “mobile ambulance” capacity and the implementation of sustainable operational models in Loudi City.