I am currently working on implementing a closed-loop control system for anti-sway using the MTGntCraneController library functions. The control schematic is attached for your reference. I have been using a joystick to input the velocity, which is designed to start from zero, accelerate uniformly, then decelerate uniformly back to zero. The input velocity is used to generate a trajectory with MTGntCraneManualTrajectory, which is subsequently fed into the controller.
In the closed-loop control, I have experimented with different feedback sources: one using MapleSim simulation for input and output, and the other using the built-in MTGntCraneSimulationModel.
During my tests, I observed that the parameters ModeX, ModeZ, and FeedForwardMode significantly impact the control’s performance. Here are my specific findings:
When FeedForwardMode is set to 0, I am unable to control the direction of the rope’s movement at all.
When FeedForwardMode is set to 1, the system operates as expected.
With ModeX and ModeZ both set to 1,and FeedForwardMode is set to 1, the trolley and gantry move in the direction of the applied velocity during the acceleration phase. However, during the deceleration phase, they move in the opposite direction and eventually return to the starting point.
With ModeX and ModeZ both set to 2,and FeedForwardMode is set to 1,trolley and gantry move as expected
I would like to understand if this behavior is considered normal and what is the fundamental principle behind this control system.
