Investigating the use of simulated dynamic pricing to optimize transport systems

运输系统的使用者不仅为自己承担一般成本(例如自己的旅行时间)，而且还对其他运输使用者(例如拥堵、事故)、居民(例如噪音、致癌空气污染物)和全球人口(温室气体排放)造成损害。在这个项目中，提出了一种简单的机械方法，将这些损害的成本反馈给造成损害的人。开发了一个模型，该模型计算动态和特定于用户的修正项，并将其添加到每个用户的一般旅行成本中。这种成本修正可以被解释为向运输用户收取的价格，但它也可以被解释为将模拟系统移动到更好的运行点的计算设备。这种方法将被集成到一个开源的多代理模拟框架中，在该框架中，每个运输用户都被微观地建模。该模型将用于研究如何使用模拟的动态价格来优化交通系统。因此，该模型利用了创新的基于智能体的仿真方法，特别考虑了需求的异质性、大规模网络、动态拥堵和复杂的行为决策，如模式选择、路线选择和发车时间选择。为了验证该优化方法在现实世界中的适用性，将所提出的优化方法应用于大柏林地区。

Users of transport systems do not only accept generalized costs for themselves (e.g. own travel time), but they also cause damages to other transport users (e.g. congestion, accidents), to residents (e.g. noise, carcinogenic air pollutants) and to the global population (greenhouse gas emissions). In this project, a simple mechanistic approach is proposed to to feed the costs of these damages back to the persons causing them. A model is developed which calculates dynamic and user-specific correction terms, to be added to each user's generalized travel cost. This cost correction may be interpreted as a price to be charged from the transport user, but it can as well just be interpreted as a computational device to move the simulated system to a better operating point. This approach will be integrated into an open source multi-agent simulation framework in which each transport user is modeled microscopically. The model will be used to investigate how the simulated dynamic prices can be used to optimize transport systems. Thereby, the proposed model makes use of the innovative agent-based simulation methodology which particularly allows for heterogeneity in demand, large-scale networks, dynamic congestion and complex behavioral decisions, such as mode choice, route choice and departure time choice. In order to demonstrate the real world applicability, the proposed optimization approach is applied to the Greater Berlin area.