ME 475 is the highest undergraduate-level control course at Purdue ME. A classical control problem, inverted pendulum, was our final project. 

In the final project, a rotating rod was attached on a cart sliding horizontally on a track, where an encoder measured the angular position and angular speed of the rod and only the cart is controlled by an input motor. We conducted the system identification of the system by the Equation of Motion, kinematics and balance equations and performed calculations in MATLAB. Then we used the state space feedback method to determine the appropriate gains, K1, K2, K3, K4 for car position, car velocity, rod angle, rod angular velocity, respectively, for the input voltage to the motor and designed our controller. Therefore, when the pendulum, or rod, was inverted, the input voltage to the motor should control the motor such that the cart moved, corresponding to the direction of the falling rod, and kept the pendulum vertically inverted in balance for more than 20 seconds (Stability Status). The control process was completed on a LabVIEW algorithm with a myRIO.

In the competition, the movement algorithm would need to swing the rod back and forth to the top and then use the stability algorithm to stabilize the pendulum. The team spending the least time to stabilize the pendulum on the top of the cart will win the championship in the competition. In the competition, we spent around 2.8 seconds and won the second place throughout the whole class.

Phase 1: The cart is moved back and forth by a pre-determined frequency and amplitude in order to swing the rod up to its top.

Phase 2: Once the rod is within certain range of angle, the algorithm will try to keep the rod straight on the cart by slightly adjusting the cart position. The balance should be kept for more than 30 seconds, and the cart position steady error should be less than 5 cm.

Wiring Diagram of Main VI in LabVIEW