基于优化 DEA 模型对公路养护投资决策的有效性研究

为制定科学有效的养护决策，基于高速公路路面性能数据进行神经网络模拟预测了未来 5 个运营年限的路面性能，提出一种新的路面养护决策优化数据包络分析（DEA）模型。该模型以养护里程量、交通量和养护投入资金为输入指标，以路面使用性能恢复值为输出指标，对所提出的 13 种养护方案进行有效性评价，选出投入少、产出高的高效率方案。结果表明:综合考虑交通量、投入资金、产出效率和养护工作量等因素能够客观地筛选出高效率的养护方案。方案结果可为公路管理者做出合理养护决策提供有效依据。

Research on the Effectiveness of Highway Maintenance InvestmentDecision-making Based on Optimized DEA Model

To make a scientific and effective maintenance decision, the highway pavement performance in the next five operatingyears is predicted by neural network simulation based on highway pavement performance data. Additionally, a new data envelopmentanalysis (DEA) model for highway maintenance decision optimization is proposed. This model takes maintenance mileage, trafficvolume, and maintenance investment as input indexes, and pavement performance recovery value as output index. This modelevaluates the effectiveness of the proposed 13 maintenance schemes, and selects the high-efficiency scheme with less investment andhigh output. The results show that the efficient maintenance scheme can be selected objectively by considering traffic volume,investment, output efficiency and maintenance workload. The outcomes of the scheme can provide an effective basis for highwaymanagers to make reasonable maintenance decisions.