Simulation Dataset for Forward Modeling and Inverse Design of Lumped-Element Nonlinear Transmission Lines
Description
This dataset contains simulation-derived data used for the forward modeling and inverse design of compact lumped-element nonlinear transmission lines (NLTLs). The data were generated using a nonlinear RLC time-domain simulation framework implemented in MATLAB. The design variables include the series inductance L, zero-bias capacitance C0, capacitance reference voltage Vr, and the number of NLTL sections N. The extracted response quantities include output amplitude Aout, dominant oscillation frequency fout, transient energy Etr, and maximum current slew rate SImax. Three dataset sizes are provided, containing 1,960, 5,040, and 9,680 samples, respectively. These datasets were used to evaluate RBF, SVR, MLP, and MVPC-MTNet models for forward modeling and inverse parameter identification.
Files
Steps to reproduce
Define the NLTL design variables L, C0, Vr, and N within the predefined feasible parameter ranges. 1. Simulate each NLTL configuration using the nonlinear lumped-element RLC time-domain model implemented in MATLAB. The nonlinear capacitance is evaluated as a voltage-dependent element during the transient simulation. 2. Obtain the output voltage waveform and the corresponding current response for each simulated NLTL configuration. 3. Extract the response features from the simulated signals: output amplitude (Aout), dominant oscillation frequency (fout), transient energy (Etr), and maximum current slew rate (SI,max). 4. Store the corresponding design parameters and extracted response quantities for each simulated sample. 5. The F2, F3, and F4 inverse-design feature configurations are obtained using Aout and fout for F2, Aout, fout, and Etr for F3, and Aout, fout, Etr, and SI,max for F4. 6. The resulting datasets containing 1,960, 5,040, and 9,680 samples are those provided in this repository.
Institutions
- Malek Ashtar University of TechnologyTehran, Tehran