<![CDATA[This systematic review aims to synthesize and critically evaluate the application of ordinary differential equations (ODEs) in modeling and controlling autonomous vehicle dynamics from 2020 to 2025. Following the PRISMA framework, 36 peer-reviewed journal articles and conference proceedings were selected and analyzed to map model types, control strategies, and their practical trade-offs. The review identifies six primary ODE-based modeling approaches kinematic, dynamic, linear, nonlinear, single-track, and two-track each exhibiting distinct balances among accuracy, computational demand, and real-time feasibility. Integration with control methods such as PID, LQR, Model Predictive Control (MPC), and nonlinear techniques reveals context-dependent performance: while MPC and sliding mode control offer high accuracy and robustness, they impose significant computational burdens; in contrast, PID and LQR are lightweight but limited in nonlinear or high-disturbance scenarios. Critical gaps persist, including insufficient real-world validation, limited fusion of physics-based ODEs with data-driven learning, scarcity of open dynamic datasets, and unresolved real-time implementation challenges. The study concludes that model–control selection must be mission- and hardware-aware, and calls for stronger empirical validation, standardized benchmarks, and hybrid architectures combining ODE rigor with machine learning adaptability.]]>
Copyrights © 2026
