Multi-Domain Modeling and Optimization of Integrated Renewable Energy and Urban Electric Vehicle Systems

德国和中国以及其他国家已经制定了雄心勃勃的目标，即在未来几年内大力增加电动汽车（ev）的数量和可再生能源在电力供应中的份额。如果不制定可再生能源和电动汽车协调可持续的电力系统整合战略，这种快速增长将给基础设施带来严重压力——正如今天在电力需求高峰时段所经历的那样。拟议的项目解决了基于多领域方法的必要集成阶段，该方法包括基于大数据的分析以及集成可再生能源和电动汽车车队的建模和优化。中国和德国城市的内城参考区可以作为基准。大多数相关的科学文献通过随机模型、旅行调查或模拟数据来描述驾驶员的行为。相比之下，在这项工作中，使用了北京数千辆电动汽车的运动和充电过程的真实记录数据。这是可能的，因为从中国工信部大数据监测中心获得的信息正在整合。对公交车和城市商用车的大量数据也进行了评估。目前还没有对如此庞大的电动汽车数据集进行全面的分析，有必要开发合适的大数据方法，在异构硬件上应用并行处理和机器学习。能源消耗记录将分为几种模式，以开发城市参考区的运输和能源需求预测模型。这包括拓扑结构、不同的车辆类型及其行驶行为。基于此分析，可以生成负载分布，为电动汽车车队的电网整合建模提供基础。将开发不同充电模式的电力系统参考模型，根据这些参考模型分析电能质量问题，并得出缓解问题的解决办法。通过创建能源管理系统，解决了电动汽车车队的能源市场整合问题。所谓的“纳米电网”被研究作为电网薄弱和容易产生大电压波动的地方的解决方案。电动汽车的性能在很大程度上取决于电池，电池劣化严重影响汽车的续航里程和寿命周期。当采用车联网技术时，后者也会受到影响。因此，我们将研究电池老化和健康对电动汽车车队运行和整合优化的影响。在研究的最后阶段，推导了综合考虑短期和长期影响的充电管理优化和总成本最小化的集成多域模型。本文将针对不同的运行模式提出建议，以最大限度地利用可再生能源或减少二氧化碳排放。

Germany and China, as well as other countries, have set ambitious objectives for strongly increasing the number of electric vehicles (EVs) and the share of renewable energies in the electric power supply over the next years. Without the development of strategies for a coordinated sustainable power system integration of renewables and EVs, this rapid growth is set to cause serious stress to the infrastructure – as already experienced today during peak hours of power demand. The proposed project addresses the necessary integration stages based on a multi-domain methodology comprising big data based analysis and the modelling and optimization of integrated renewable energy and electric vehicle fleets. Inner-city reference districts of Chinese and German cities serve as benchmarks.The majority of relevant scientific references describe the driver behaviour by means of stochastic models, travel surveys or simulation data. In contrast, in this work real-world recorded data of the movement and the charging process for several thousand EVs in Beijing are used. This is possible as information obtained from the Big Data Monitor Center of the Chinese MIIT is being integrated. Extensive data for buses and urban commercial vehicles are evaluated as well. A comprehensive analysis of such a large data set of EVs has not been carried out so far and the development of suitable big data methods is necessary, applying parallel processing and machine learning on heterogeneous hardware.Energy consumptions records will be classified into several patterns to develop a transport and energy demand prediction model for an urban reference district. This includes the topology, different vehicle types, and their travel behaviour. Based on this analysis, load profiles can be generated to provide a foundation for modelling the grid integration of EV fleets. Power system reference models for different charging patterns will be developed, power quality issues will be analysed based on these reference models and mitigation solutions will be derived. The energy market integration of EV fleets is addressed by the creation of an energy management system. So-called “nanogrids” are investigated as solution for locations where the electric network is weak and prone to heavy voltage fluctuations. The performance of EVs is highly determined by its batteries and battery degradation has a serious impact on the vehicle range and life cycle. The latter is also affected when vehicle-to-grid technology is applied. Therefore, the influence of battery aging and health on the optimization of EV fleet operation and integration will be examined.In a concluding stage of research an integrated multi-domain model is derived which facilitates an optimized charging management and overall cost minimization considering relevant short- and long-term effects. Recommendations for different operating modes to maximize the use of renewable energy or to minimize CO2 emissions will be given.