Experimental Validation of Eco-Driving and Eco-Heating Strategies for Connected and Automated HEVs

Mohammad Reza Amini, Qiuhao Hu, Hao Wang, Yiheng Feng, Ilya V. Kolmanovsky, Jing SunarXiv 2021 · OpenReview

TL;DR

This paper experimentally validates eco-driving and eco-heating strategies for hybrid electric connected and automated vehicles (CAVs) in cold weather. By integrating V2I communication for traffic signal prediction and optimizing thermal loads, they achieved significant fuel savings of up to 19% compared to a human driver with constant heating.

What Problem Does This Paper Address?

The transportation sector is a major contributor to energy consumption and emissions. While connected and automated vehicles (CAVs) offer potential for improved safety and efficiency, a significant portion of energy savings remains untapped, particularly in thermal management. Specifically, for hybrid electric vehicles (HEVs) in cold climates, cabin heating represents a substantial energy load. Traditional approaches to cabin heating often lead to inefficient engine operation, especially during low-speed driving or stops, by forcing the engine to idle to maintain cabin temperature. This paper addresses the challenge of optimizing both driving efficiency (eco-driving) and thermal management (eco-heating) in CAVs, particularly in cold ambient conditions, to maximize fuel savings without compromising passenger comfort.

Methods

The authors developed and experimentally validated two key strategies:

  • Eco-Driving: This strategy leverages Vehicle-to-Infrastructure (V2I) communication to obtain traffic signal timing and estimate queue lengths at intersections. This information is used by an eco-trajectory planning algorithm to generate optimized speed profiles that minimize energy consumption. This algorithm accounts for vehicle queuing dynamics and aims for "eco-arrival and eco-departure" at intersections.
  • Eco-Heating: This strategy builds upon the eco-driving trajectory by using it as a look-ahead preview for a Model Predictive Control (MPC) framework. The eco-heating strategy intelligently coordinates the cabin thermal load with the vehicle's speed and predicted future states. It treats the engine coolant and cabin air as thermal energy storages. The core idea is to store thermal energy in the coolant and cabin air when the vehicle is driving at higher speeds (when the engine is more efficient and generates excess heat) and then release this stored energy slowly during vehicle stops or low-speed driving, thereby avoiding unnecessary engine idling for heating. The MPC controller optimizes the HVAC blower flow rate and cabin temperature setpoint to achieve this coordinated thermal management.

Datasets

The experiments were conducted using real-world drive cycles on a 2017 Toyota Prius Four Turing HEV. The testing was performed on a road segment in Ann Arbor, MI, emulating a congested arterial corridor with virtual signalized intersections. The data collected included powertrain and HVAC system data, as well as cabin temperatures.

Key Insights and Ideas

  • Thermal Energy Storage is Key: The paper highlights the underutilized potential of using the engine coolant and cabin air as thermal energy storage systems. By strategically managing heat during different driving phases, significant energy savings can be realized.
  • V2I for Thermal Management: Beyond just optimizing driving, V2I communication is crucial for providing the necessary look-ahead information to enable predictive thermal management strategies like eco-heating.
  • Coordinated Optimization Pays Off: Combining eco-driving with eco-heating yields synergistic benefits, leading to greater overall energy savings than either strategy alone.
  • Cold Weather Amplifies Benefits: The eco-heating strategy shows more pronounced benefits in colder ambient temperatures, where cabin heating demands are higher and engine idling for heating is more likely to occur inefficiently.
  • Trade-offs Exist: While eco-heating can save fuel, there's a potential trade-off with passenger comfort, especially in milder ambient conditions where reduced heating might be acceptable for greater fuel savings.

Key Results

  • Eco-Driving: Achieved an average fuel saving of 14.5% compared to normal driving by a human driver.
  • Eco-Heating: Achieved an average fuel saving of 4.7% compared to a constant-heating strategy.
  • Combined Savings: The integrated eco-driving and eco-heating strategies resulted in a combined energy saving of over 19% compared to the baseline vehicle driven by a human driver with a constant-heating strategy.
  • Consistency: The experimental results showed good repeatability and confirmed the simulation-based predictions, with some minor differences attributed to real-world factors like driver tracking error and proprietary powertrain control.

Limitations and Future Work

  • Baseline Comparison: The eco-heating strategy was compared against a fixed HVAC input (constant heating) scenario. The authors note that comparing against the OEM's "AUTO" mode might yield different results.
  • Passenger Comfort Modeling: Future work will focus on developing a more accurate model for passenger comfort to better address the trade-offs between fuel economy and comfort.
  • Controlled Environment Testing: Performing experiments in a controlled environment is suggested to reduce uncertainties, particularly regarding ambient temperature effects on eco-heating.
  • Benchmark HVAC Controller: Further investigation is needed to fully quantify the performance of benchmark HVAC controllers and identify opportunities for energy savings through connectivity.

Summary updated

BibTeX
@misc{amini2021experimental,
  author = {Mohammad Reza Amini and Qiuhao Hu and Hao Wang and Yiheng Feng and Ilya V. Kolmanovsky and Jing Sun},
  title = {{Experimental Validation of Eco-Driving and Eco-Heating Strategies for Connected and Automated HEVs}},
  year = {2021},
  url = {https://openreview.net/forum?id=yyplEkEL7o},
}
Figure