FADE-RIPE: A Robust Quantum Optimal Control Algorithm for Superconducting Transmons with Realistic Hardware Constraints (reference implementation)

FADE-RIPE: Transmon Quantum-Gate Optimizer FADE-RIPE (Fidelity Aware, Distortion-aware, Environment-conscious Robust Iterative Pulse Engineering) is a gradient-based quantum optimal control method for a single-qubit gate on a 4-level (N=4) transmon, benchmarked against GRAPE and CRAB (via qutip_qtrl). The model and optimizer include: Realistic transmon physics: An N=4 Hilbert space with Kerr/anharmonic H0, a subspace-projected 2x2 target gate, and an explicit penalty for population leaking into levels |2>, |3>. Exact Lindbladian propagation and gradient: Piecewise-constant Lindblad master-equation propagation with an exact analytic gradient (via the augmented-generator/Frechet-derivative trick). AWG bandwidth filter: A Butterworth low-pass filter on the control pulses, represented as an exact linear operator so the gradient chain rule through it is exact. Filter-function noise penalty: A numerical filter function F(omega) of the optimized pulse, penalized against a 1/f (flux) or thermal-photon power spectral density via an overlap integral. Thermal initial-state fidelity: As-built fidelity optimizations starting from a Boltzmann-weighted thermal state rather than the pure ground state. This upload contains the reference implementation fade_realistic.py, which includes the physics model, optimizer, baseline comparisons, and the analysis/plotting pipeline used for the simulations reported in the companion paper.