A Generative Physics-Informed Reinforcement Learning-Based Approach for Construction of Representative Drive Cycle

2025-06-09

AI Summary

To address low fidelity and poor physical consistency in driving cycle modeling for vehicle design and energy consumption evaluation, this paper proposes a Physics-Informed Embedded Expected SARSA–Monte Carlo (PIESMC) generative framework. The method integrates physics-based modeling, generative reinforcement learning, Monte Carlo sampling, wavelet transform, and power distribution analysis—achieving, for the first time, synergistic optimization of dynamical constraints and data-driven modeling. Evaluated on real-world datasets, PIESMC reduces cumulative kinematic segment error by 57.3% and 10.5% compared to MTB and MCB, respectively, while improving computational efficiency by nearly an order of magnitude. Moreover, it accurately reproduces statistical distributions and frequency-domain characteristics of target driving cycles. These advances significantly enhance representativeness and engineering applicability of synthesized driving cycles.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Planning, Routing, and Scheduling: Model-Based Reasoning
• Constraint Satisfaction and Optimization: Constraint Learning and Acquisition

Application Category

• Web Mining and Content Analysis: Web data generation and simulation
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems

Abstract

Accurate driving cycle construction is crucial for vehicle design, fuel economy analysis, and environmental impact assessments. A generative Physics-Informed Expected SARSA-Monte Carlo (PIESMC) approach that constructs representative driving cycles by capturing transient dynamics, acceleration, deceleration, idling, and road grade transitions while ensuring model fidelity is introduced. Leveraging a physics-informed reinforcement learning framework with Monte Carlo sampling, PIESMC delivers efficient cycle construction with reduced computational cost. Experimental evaluations on two real-world datasets demonstrate that PIESMC replicates key kinematic and energy metrics, achieving up to a 57.3% reduction in cumulative kinematic fragment errors compared to the Micro-trip-based (MTB) method and a 10.5% reduction relative to the Markov-chain-based (MCB) method. Moreover, it is nearly an order of magnitude faster than conventional techniques. Analyses of vehicle-specific power distributions and wavelet-transformed frequency content further confirm its ability to reproduce experimental central tendencies and variability.

Problem

• Accurate driving cycle construction for vehicle design and analysis
• Generative approach capturing transient dynamics and energy metrics
• Reducing computational cost while improving kinematic accuracy

Innovation

• Physics-Informed Reinforcement Learning for driving cycles
• Monte Carlo sampling reduces computational cost
• Replicates kinematic and energy metrics accurately