Modeling and Control for Emission Management in Hybrid Electric Commercial Vehicles

Abstract

Electrification of powertrains is a major trend in the vehicle industry. The reason behind this is mainly that electrification of a powertrain generally results in better fuel economy, by eliminating inefficient, low load, operation of the engine. This can be done in two ways: load shifting to shift the operation point of the engine to a more efficient one, or by turning off the engine completely. When it comes to emissions, load shifting generally have positive effect since it usually result in higher exhaust temperatures which are beneficial for the aftertreatment system. The effect from turning off the engine completely is more complicated. When the engine is turned off the aftertreatment system will start to cool down and will eventually lose its effectiveness, resulting in higher emissions when the engine is restarted. So-called green zones, zones established by legislation or demand of costumers, where the use of combustion engines is prohibited, are a good example of where this can be expected and is therefore a focus of this thesis. The applications are not limited to hybrids but also useful for all vehicles that make stops, e.g., commercial vehicles that make regulated 45 minutes breaks and loading/off-loading cargo.A model of a complete hybrid electric heavy-duty vehicle is developed and validated. The model is a compilation of several submodels of the different components in the vehicle. To correctly estimate the pollutive emissions, the components in the aftertreatment system are the most important components and emphasis is put on how the concentrations in them are calculated. It is shown that a quasi-static model for the concentrations gives the best balance in terms of accuracy and simulation time for the application. The aftertreatment system submodels are validated against data from a high-fidelity model and the complete powertrain is validated against experimental data from a powertrain in a test stand, all with satisfactory results. The model is used to create a virtual environment where the effect different control strategies have on the emissions around green zones can be studied and optimized.A control strategy based on pre-heating of the aftertreatment system is developed. The strategy heats the aftertreatment before turning off the engine in an optimal way to reduce NO x . This strategy is shown to be effective for engine-off times up to a few hours. However, for longer engine-off times, pre-heating of the aftertreatment system induces a limitation on the amount of stored ammonia, making the strategy ineffective or even bad. The strategy is extended to handle scenarios with multiple engine-off events using an algorithm that finds the engine-off events and handle them separately, but with a common equivalence factor between fuel and NO x to link them. The strategy is shown to handle scenarios with multiple engine-off events well, and the resulting distribution of fuel between the events is close to optimal. vii viii