Real-time torque distribution simulation of parallel hybrid vehicle engine
IntroductionParallel hybrid vehicles face challenges in real-time torque distribution, including slow feedback speeds and suboptimal energy allocation, which constrain overall energy efficiency. This study aims to develop a high-precision, robust torque distribution model to enhance energy utilizati...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-08-01
|
| Series: | Frontiers in Mechanical Engineering |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fmech.2025.1647691/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | IntroductionParallel hybrid vehicles face challenges in real-time torque distribution, including slow feedback speeds and suboptimal energy allocation, which constrain overall energy efficiency. This study aims to develop a high-precision, robust torque distribution model to enhance energy utilization while addressing interference from environmental noise and extreme temperatures.MethodsA real-time torque distribution model integrates three core components: a Markov Decision Process framework transforms torque allocation into a mathematical optimization problem; the Proximal Policy Optimization algorithm enhanced with Prioritized Experience Replay dynamically generates control strategies; and Fiber Bragg Grating sensors achieve millisecond-level torque measurement by correlating shaft strain forces with wavelength shifts. Validation employed the Gamma Technologies Suite simulation platform and the Next Generation Simulation dataset, with benchmark comparisons against Equivalent Consumption Minimization Strategy, Fuzzy Logic Control, and Thermostat Strategy models.ResultsThe optimized Proximal Policy Optimization algorithm achieved 93.2% accuracy and 1.0% loss rate upon convergence, with an average feedback time of 32 milliseconds. In simulated vehicle operations, torque distribution was completed within 70 milliseconds, while energy utilization rates reached 75.5% during startup, 42.3% in normal driving, 41.5% under acceleration, 22.5% during deceleration braking, and 50.0% in high-speed driving. Robustness testing demonstrated 82.3% accuracy under 300-decibel noise interference and 83.1% accuracy at 180-degree Celsius temperatures.DiscussionThe model establishes a closed-loop system that synergizes rapid Fiber Bragg Grating sensing with Markov Decision Process-driven decision-making, enabling efficient torque distribution under extreme operating conditions. While energy utilization during deceleration braking remains suboptimal, future work will optimize regenerative braking strategies through road condition prediction and advanced power devices. This approach provides a viable pathway to improve energy sustainability in hybrid transportation systems. |
|---|---|
| ISSN: | 2297-3079 |