control-toolbox

control-toolbox is a small Python library for LQR-based state-space controller design, simulation, preprocessing, and visualisation. The package wraps python-control models with convenience utilities for tracking references, handling delays, and plotting closed-loop responses.

Features

• Predefined plant models (inverted pendulum, mass‑spring, motor, etc.) inheriting from control.StateSpace.
• build_controller — synthesise LQR/DLQR state-feedback plus pre- compensator (Nbar) for reference tracking. Works with continuous and discrete plants, and accounts for input delays.
• simulate — run a closed-loop simulation from a ControllerResult.
• add_delay / discretize utilities for preprocessing plants.
• Plotting helpers and simple performance metrics.

Installation

python -m venv .venv
source .venv/bin/activate
pip install -e .  # installs this package in editable mode

Quick start

from control_toolbox import *

plant = InvertedPendulum()
plant = add_delay(plant, T_input=0.1, T_output=0.05, pade_order=3)
plant = discretize(plant, dt=0.01)

ctrl = build_controller(plant, track={"x":0.5, "phi":0.0}, p=100)
resp = simulate(ctrl, t_end=5.0)

fig = plot_response(resp, metrics=compute_metrics(resp))
fig.savefig("response.png")