Traffic is a major contributor to emissions in many large cities with severe air pollution. Experience in London, Milan, and Stockholm shows that charging for the use of roads can be effective in reducing emissions, as well as congestion. This study examines the use of predictive cordon tolls based on weather forecasts to reduce ambient air pollution and congestion. Travelers choose their destinations inside or outside the cordon, and whether to drive or take public transport. Passenger vehicles are divided into three classes according to their emission characteristics, and higher tolls are imposed on more polluting vehicles. The Box model of emission dispersion is used to predict air quality. A Markov decision-making process then determines daily toll levels with the objective of maximizing welfare measured by travelers’ surplus, toll revenue, and air pollution health costs. The model is applied to a hypothetical network using recorded weather data for Tehran in 2016. With base-case parameter values, predictive pricing reduces the daily average CO concentration as well as the number of days with dangerous air quality. Predictive pricing yields a higher welfare gain than a fixed toll (i.e., the same every day regardless of weather conditions). The effects of weather information, wind forecast accuracy, forecast time horizon, values of travel time, destination attractions, and road link capacity on the benefits from predictive pricing are analyzed. The performance of the model under randomized weather conditions is also assessed.

在许多空气污染严重的大城市，交通是造成排放的主要因素。伦敦，米兰和斯德哥尔摩的经验表明，对道路使用收费可以有效减少排放和交通拥堵。这项研究检查了基于天气预报的警戒线收费的使用，以减少环境空气污染和交通拥堵。旅客可以选择在警戒线内还是警戒线外选择目的地，以及选择开车还是乘坐公共交通工具。乘用车根据其排放特性分为三类，高污染车辆要加重通行费。排放扩散的Box模型用于预测空气质量。然后，马尔可夫决策过程将确定每日的通行费水平，以使通过旅客剩余，通行费收入，和空气污染健康成本。该模型使用记录的德黑兰2016年天气数据应用到假设网络中。通过使用基本参数值，预测性定价可以降低每日平均CO浓度以及危险空气质量的天数。预测性定价所产生的福利收益要高于固定通行费（即，无论天气情况如何，每天都相同）。分析了天气信息，风力预报的准确性，预报的时间范围，行驶时间的值，目的地的吸引力以及道路通行能力对预测价格收益的影响。还评估了该模型在随机天气条件下的性能。预测性定价会降低每日平均一氧化碳浓度以及危险空气质量的天数。预测性定价所产生的福利收益要高于固定通行费（即，无论天气情况如何，每天都相同）。分析了天气信息，风力预报的准确性，预报的时间范围，行驶时间的值，目的地的吸引力以及道路通行能力对预测价格收益的影响。还评估了该模型在随机天气条件下的性能。预测性定价会降低每日平均一氧化碳浓度以及危险空气质量的天数。预测性定价所产生的福利收益要高于固定通行费（即，无论天气情况如何，每天都相同）。分析了天气信息，风力预报的准确性，预报的时间范围，行驶时间的值，目的地的吸引力以及道路通行能力对预测价格收益的影响。还评估了该模型在随机天气条件下的性能。分析了预测时间范围，旅行时间，目的地景点的价值以及基于预测性价格收益的道路通行能力。还评估了该模型在随机天气条件下的性能。分析了预测时间范围，旅行时间，目的地景点的价值以及基于预测性价格收益的道路通行能力。还评估了该模型在随机天气条件下的性能。