Posted by AF [AF] on November 07, 1999 at 18:05:37 {zExURjw8QA1EwaxRomAMdaOt1gg/Zk}:
In Reply to: ****A Question for Rick posted by 5GJW on November 07, 1999 at 17:43:28:
: A project that I did last Spring was to write a program to control an inverted pendulum that was mounted on a cart. I ended up using a PID controller that actually worked! The hard part wasn't the control equations, it was the characterizations of the components.
I remember some of that stuff! One of my buddies in the Real Time Computing and Control class did the same thing, back around 1982 using a DEC LSI-11 computer. It took him several weeks. It was the computerized version of a similar demo controlled by an op-amp circuit and presented to beginning EE classes as "cool stuff". Later he decided to go one step bigger for his Bachelor's Thesis and program the cart to start from a dead position where the pendulum hung straight down, then swing it into an upright balanced position. It took him months of late nights. Many times the lab would echo to the sound of the cart smacking hard up against the stops as it failed to complete the upswing. The day he got the upswing to work right was an event. Everybody went over to the bench and cheered!
AF