A software-defined vehicle (SDV) is one with a centralized digital architecture that allows software updates and upgrades throughout its lifetime. An SDV’s operation, functionality, and features are controllable and can be changed through software.
Functions like over-the-air updates and customization of the infotainment and driver-assistance systems are common opportunities for SDV functionality in internal combustion engine (IEC) vehicles and electric vehicles (EVs). This article explores additional opportunities for SDV features in EVs, including customization of the drivetrain, regenerative braking, and battery pack reconfigurability.
Functions in an SDV can be personalized explicitly or implicitly. In many EVs, regenerative braking is an example of explicit personalization. The driver can turn the system on and off and manage the amount of regenerative braking used.
The operation of the traction motor in an EV can also be explicitly or implicitly customized. For example, the operation of the motor drive inverter can be modified by the driver to control the maximum torque output of the motor, as well as acceleration and deceleration rates, maximum speed, and other characteristics. Explicit customization has its limits.
With so many variables under software control, EV operation can be extensively customized, making it increasingly difficult for drivers to optimize overall system operation. Development efforts are underway to automate the EV drivetrain optimization process.
In one case, a prototype system has been developed to maximize motor torque based on driver-specific data. The powertrain control module algorithm determines the target motor torque using a combination of a lookup table (LUT) of the initial target torque and a rate limiter (Figure 1).
The LUT inputs the vehicle state and acceleration pedal sensor (APS). The APS output reflects the driver’s actions and expectations. The algorithm includes a set-point scaling factor and a feedback controller. The set-point scaling factor reflects driver intention as measured by the APS, while the feedback controller adjusts the dynamic implementation of the desired torque.
Figure 1. A block diagram of a torque-adjustment algorithm that optimizes EV performance based on a driver’s expectations. (Image: MDPI sensors)
Reconfigurable battery packs
EV battery packs can be dynamically adjusted based on multiple variables, such as the instantaneous loading and the condition and charge states of individual cells. Adaptive battery pack reconfiguration can improve the efficiency of the power converters in the drivetrain and maximize the battery pack’s lifetime.
For instance, the effective capacity of individual cells is impacted by their discharge rate. According to the rate-capacity effect, more capacity can realized if a cell is more slowly discharged. A 2,800 mAh Li-ion cell delivered about 200 mAh less capacity when discharged at 1,000 mA compared with a slower discharge rate of 200 mA. Connecting cell strings in parallel reduces the discharge current of each cell, increasing overall pack capacity.
At the same time, the strings should be configured to deliver the optimal voltage based on the EV drivetrain’s current load requirements.
Optimizing battery pack performance based on individual cell electrochemical and thermodynamic properties is theoretically possible. Still, it requires much computing power to be implemented in real-time. Instead, an adaptive algorithm has been developed that optimizes the combination of parallel cells to minimize the discharge currents and the length of the battery strings to match the pack voltage for optimal power converter efficiency. (Figure 2).
Figure 2. The system model of a reconfigurable battery pack integrated in an EV. (Image: ACM Transactions on Sensor Networks)
Another way that an SDV function can also improve EV battery pack performance and lifetime by disabling the heating, ventilation, and air conditioning (HVAC) compressor during acceleration to minimize discharge current. Properly implemented, vehicle occupants should not experience any decrease in comfort.
For example, the blower can remain on, maintaining air circulation when the compressor is off. This function can combine implicit implementation with explicit limitations set by the driver, like the rate of acceleration when the function is initiated and the length of time it can be implemented.
Summary
EVs present several opportunities to implement SDV functions. They include implicit optimization of the drivetrain based on individual driver behaviors and reconfiguring the battery pack to maximize performance and battery lifetime. Some functions like regenerative braking and HVAC operation can benefit from a combination of implicit and explicit SDV implementations.
References
- Extending Battery System Operation via Adaptive Reconfiguration, ACM Transactions on Sensor Networks
- Personalization of Electric Vehicle Accelerating Behavior Based on Motor Torque Adjustment to Improve Individual Driving Satisfaction, MDPI sensors
- Smart Electric Vehicle Charging in the Era of Internet of Vehicles, Emerging Trends, and Open Issues, MDPI energies
Images
- Figure 1, MDPI sensors, Page 12, Figure 10
- Figure 2, ACM Transactions on Sensor Networks, Page 3, Figure 1
Filed Under: FAQs, Software