Multi-objective optimization for battery electric vehicle power train topologies

Get this Project:

Fields with * are mandatory

ABSTRACT

Electric vehicles are becoming more popular in the market. To be competitive, manufacturers need to produce vehicles with a low energy consumption, a good range and an acceptable driving performance. These are dependent on the choice of components and the topology in which they are used. In a conventional gasoline vehicle, the powertrain topology is constrained to a few well-understood layouts; these typically consist of a single engine driving one axle or both axles through a multi-ratio gearbox.

With electric vehicles, there is more flexibility, and the design space is relatively unexplored. In this paper, we evaluate several different topologies as follows: a traditional topology using a single electric motor driving a single axle with a fixed gear ratio; a topology using separate motors for the front axle and the rear axle, each with its own fixed gear ratio; a topology using in-wheel motors on a single axle; a four-wheel-drive topology using in-wheel motors on both axes.

Multi-objective optimization techniques are used to find the optimal component sizing for a given requirement set and to investigate the trade-offs between the energy consumption, the powertrain cost and the acceleration performance. The paper concludes with a discussion of the relative merits of the different topologies and their applicability to real-world passenger cars.

BEV POWERTRAIN TOPOLOGIES

Figure 1. BEV powertrain topologies considered in this study: (a) SB-SA; (b) DM-DA; (c) IWM-SA; (d) IWM-DA. diff: differential

Figure 1. BEV powertrain topologies considered in this study: (a) SB-SA; (b) DM-DA; (c) IWM-SA; (d) IWM-DA. diff: differential

When two traction sources are combined in a single vehicle in this way with an appropriate control strategy, the designer has more degrees of freedom and the flexibility to investigate more effective and efficient operation methods. This powertrain topology is presented in Figure 1(b). This topology has been sufficiently developed for use in a production vehicle, namely the Tesla Model X, introduced in late 2015, which is configured this way, with separate front and rear motors, each with single-ratio gearboxes.

VEHICLE MODEL AND SIMULATION TECHNIQUE

Figure 2. Backward-facing model of the vehicle showing the trajectory as the input

Figure 2. Backward-facing model of the vehicle showing the trajectory as the input

Figure 2 illustrates the structure of generic forward-facing simulations. Starting from the driver model, the reference speed and the actual vehicle speed are compared, giving an ‘error’ signal, and this is used to generate a control signal to control the torque used to power the vehicle. (Driver models normally use proportional–integral controllers). In practice, the source of power of the BEV is an electric motor, and so an appropriate model is used to translate the throttle and brake commands from the driver into the torque and mechanical braking signals.

Figure 5. Inputs and outputs of backward-facing simulations of the motor–machine

Figure 5. Inputs and outputs of backward-facing simulations of the motor–machine

The electric machine acts as a motor–generator, providing a traction torque during acceleration and steady-state driving and then recovering the kinetic energy that is otherwise lost during braking. A schematic diagram for a backward-facing model is shown in Figure 5. The relationship between the motor input power and the output torque–speed pair is given later in equation (6).

VEHICLE SIMULATIONS WITHOUT OPTIMIZATION

In order to understand the implications of each chosen topology, two studies were carried out. In the first study, the properties of each topology with an 80 kW electric machine were considered. The battery back was kept at the case-study size, and there were no attempts to optimize the powertrain. The aim of this part of the study was to demonstrate that the basic methodology was sensible and that the simulations gave reasonable results.

MULTI-OBJECTIVE OPTIMIZATION FOR DIFFERENT BEV TOPOLOGIES

Figure 13. Pareto front of the acceleration time and the energy consumption

Figure 13. Pareto front of the acceleration time and the energy consumption

Figure 14. Pareto front of the energy consumption and the powertrain cost

Figure 14. Pareto front of the energy consumption and the powertrain cost

The results of the multi-objective optimization are shown in Figures 13 and 14. Figure 13 shows the trade-offs of between the energy consumption and the acceleration time for each topology. The vertical solid line shows as a benchmark the acceleration of the Nissan LEAF, 9.9 s for 0–100 km/h. For each topology, it is possible to match or improve the original acceleration performance relative to the benchmark vehicle; this is most noticeable for the topologies using in-wheel motors. Better acceleration naturally requires a large motor with a higher mass, which therefore consumes more energy.

CONCLUSIONS

In this paper a study was presented in which the proper- ties of four possible BEV topologies were studied. The topologies used different combinations of in body and in-wheel motors, driving axles and powertrain arrangements. The background to the topologies was presented, and four topologies selected for detailed analysis were described: SM-SA, DM-DA, IWM-SA and IWM-DA. These were modelled, and the calculations and the assumptions that we used were presented. The results of the models were presented, first showing how each topology performed with a fixed 80 kW total motor power, and then after performing multi-objective optimization.

The relative benefits of each topology were evaluated and presented, with guidelines on the choice of topology for given objectives. For pure energy efficiency or a good compromise between the energy efficiency and the driving performance, the IWM-DA topology was the best; for pure driving performance, the DM-DA topology was the best; for a cheap vehicle or a good compromise between the cost and the efficiency, the SM-S Atopology was the best; for a cheap vehicle with a good performance, the IWM-SA topology was the best.

Also, there was a significant sensitivity of the result to the problem formulation, as shown by the addition of a full payload to the vehicle. It is hoped that these guidelines and the method used to obtain them are useful for vehicle manufacturers in determining the best topologies to consider in the design of future BEVs.

Source: University of California
Authors: Pongpun Othaganont | Francis Assadian | Daniel J Auger

Download Project

Get this Project:

Fields with * are mandatory

Leave a Comment

Your email address will not be published. Required fields are marked *