Minimum Lap Time Simulator

<- All projects

Aprilia Racing | University of Trento
2021 - present

Overview

An offline, optimal-control minimum lap time (MLT) simulator developed with Aprilia Racing for MotoGP. Given a track and a vehicle model, the simulator computes the fastest theoretically achievable lap and the full state and control trajectory that achieves it, rather than integrating a driver model forward in time like a traditional lap-time simulator.

Approach

The motorcycle is modeled as a full multibody system (around 30 states), including a Magic-Formula-based tyre model (MF-MCTyre) and CFD-derived aerodynamic maps, and the minimum-time problem is solved as an indirect optimal control problem. Because the resulting boundary-value problem is numerically stiff and sensitive to the initial guess, the solver uses homotopy continuation - it solves a sequence of progressively harder problems, using each solution as the initial guess for the next, rather than attempting to solve the full nonlinear problem from a cold start.

Output

Beyond the lap time itself, the simulator produces the full optimal state and control history - speed, lean angle, throttle, brake, tyre loads - which is compared against real telemetry to validate the vehicle model and probe where a rider is leaving performance on the table.

Status

Ongoing, 2021 - present, developed with Aprilia Racing at the University of Trento.