Numerical Optimization of 2026 Formula 1 Hybrid Power Unit Models for Competitor Analysis

This thesis develops a methodology to reconstruct the deployment strategy of the MGU-K and estimate the maximum thermal power of the internal combustion engine (ICE) in Formula 1 cars, using exclusively publicly available GPS telemetry data. The work is motivated by the 2026 FIA regulations, which substantially increase the role of the electric powertrain while imposing strict energy management constraints. A physics-based, parametric vehicle model is formulated to capture the longitudinal dynamics, accounting for inertial power, aerodynamic drag, rolling resistance, track slope, and lateral acceleration effects. Particular attention is given to the representation of ICE power, assumed constant in power-limited sections based on engine operating principles and experimental evidence. Model parameters are identified through nonlinear optimization, employing techniques such as variable normalization, smooth reformulation of non-differentiable operators, and multiple shooting to improve convergence and robustness. The framework is implemented with CasADi and IPOPT, enabling accurate reconstruction of both the MGU-K power profile and aerodynamic drag characteristics, even under conditions of uncertainty or incomplete reference data. The methodology is validated across multiple Formula 1 circuits, demonstrating high accuracy in reproducing resistive power profiles. The results provide novel insights into the energy deployment strategies adopted by different teams and establish a reliable tool for competitor analysis under the upcoming 2026 regulatory landscape.