WOLFRAM

Electrical Engineering

Use System Modeler to construct and study electrical circuits, power electronics, and electrical machinery. Combine electrical and mechanical components to build complete system models. Perform analysis tasks and measure performance.

Inverted Pendulum: Symbolic Model Linearization

Linearization makes it possible to use tools for linear systems to study nonlinear systems around an operating point. In this example, an inverted pendulum is linearized around its upright position. The goal is to analyze the behavior of the controlled inverted pendulum system for varying lengths of the pendulum.

The Model

In this example, a PID controller is used to keep the pendulum in its upright position. The pivot point of the pendulum is mounted on a cart that can move horizontally, controlled by a motor and gear arrangement.

Model of the control system.

Control Design

The inverted pendulum system is linearized around its upright position. The length of the pendulum is kept in symbolic form. The control system is then designed by tuning the PID parameters for different lengths of the pendulum.

New in SystemModeler 12.0

Obtain symbolic linearization with your selected symbolic parameters.

Linearized state-space representation with the length of the pendulum as the symbolic parameter.

The combined system consisting of the PID controller and the inverted pendulum is verified for stability using the Nyquist plot. The open-loop transfer function of the combined system has one unstable pole, and there is one encirclement around {-1,0} in the counterclockwise direction. As the unstable pole and encirclement cancel out, it can be inferred that the closed-loop system will have no unstable poles.

Check stability

Use Nyquist plot to check unstable poles of the closed-loop system.

Nyquist plot for different lengths.

After verifying the stability, the model is simulated. You can now observe the raw data for the oscillations and plot the envelope of the response. You can see that with an increase in length, the time to reach steady state increases.

Deploy controller

Observe the system behavior for different lengths.

Response of the system for an initial pendulum angle of 5°.

Control design with symbolic model linearization