Is it possible to calculate the position, with a computer algoritm, xyz coordinates, of a point in a rectangular square room, if a few, 3+, of the distances to corners a known? These corners xyz-position internally, are assumed to be known.
Yes, it is. What kind of motion application do have where you need to do this?
Thanks for the reply.
There is no specific application. I am just an “amateur” generally interested in closed loop motion control.
Since most robot arms seems to have their sensors/transducers in either the hydraulic cylinders or the arm joints. And I assume their reference is from an Origo point in an imaginary xyz-coordinate system. I was wondering if there are any closed loop motion control systems that use the same technique as the GPS system, but locally, in a room or a limited area outdoors. We could call it Local Positioning System, LPS. lf the receiver was sitting out on the end of a robotic arm, the positioning accuracy would not be dependent on the stiffness of the arm and the mass of the load.
Lets say two big telescopic cranes where to lift and position a 150ft long 30ton beam straight up, in a 100% horisontal position. Instead of cranes we could just play with 2 or 4 hydraulic cylinders. Could a LPS receiver, that detects the zyz-position, in each end of the beam make a closed loop control of the operation to work?
And what response time accuracy and could be expected with use of different types of detecting system, like mechanical pull wire transducer, laser, radio etc…
made a little video animation that shows what I mean…
I am curious whether a workable LPS system exists. We have found that wire-draw encoders (string pots) are typically not real good feedback devices. Machine vibrations can cause the wire to oscillate, inducing noise in the feedback. The laser systems I have looked at are either too slow or to coarse. I have heard tell that good ones exists, but I have not personaly seen them. And I haven’t ever seen a device that uses radio waves for accurate high-speed measurement.
Construction equipment uses GPS, or even some more local version for grading and leveling, but usually construction equipment has slower updates rates than what we normally deal with. Of course, if you are happy with slow lifting, it could certainly work.
Regardless, for precise control, I would think you still need feedback on the actuators themselves, then use cascaded PID loops to get great control. The outer loop woudl be the xyz position, and the innner loop(s) would be the actuator position(s). Depending on the geometry and degress of freedom, the math could becaome rather involved, but should be doable. Multi-dimensional vector control like this could require is not something we normally do with our motion controller.
Are you Swedish? I see your youtube channel is an Americanized form of “jajamensan”, and I assume your Delta forum user name comes from “Huvva, vad Delta är kung på hydraulisk positionstyrning!”
hej Paso! please shoot me an pm! hugaligen
I did send a PM, but apparently it isn’t working. It stays in my Outbox and doesn’t appear in my sent list. I’ll have to have our IT look into it. If you want, you can email email@example.com support with attention to Jacob.