The Forklift, Model-Based Design And Rapid Control Prototyping

Updated: Jun 1

Forklifters in Sweden designed two different controllers to lower the forks of the forklift. They then went on to test them in a simulation environment, with the goal to use Rapid Control Prototyping (RCP) when implementing the control systems. Let me first explain RCP, then outline the Swedish study on forklifts, and end this blog post with yet another intriguing question about, well, cars.


In my last post I asked you guys what forklifts, model-based design and RCP have in common. So here it is: engineers use RCP to quickly test and iterate control strategies (such as lowering the fork of a forklift). Consequently, engineers deploy the designed models onto a real-time target machine with real input and output interfaces, and finally connect the real-time target machine to the real-world system. In this scenario, the real-time target machine acts as a controller, which enables engineers to get started right away with tuning the software designs. In HIL simulation it's the plant model which is deployed on to a real-time machine. In RCP however, it's the controller model which is deployed on the real-time target machine. Basically, this means that the simulated control strategy produces actual control inputs that are applied to the physical plant under control (the forklift), finally its response is measured and fed back to the controller simulation.



Finding the Best Possible Control Strategy


This also explains why engineers in Sweden chose to use the RCP workflow: Their study focused on figuring out which control strategy to use for the lowering function of the forklift. By using RCP, they were able to test, fine tune, and compare several control strategies in a very short time, without having to worry about any implementation details or spend hours in debugging hand-written code. If you'd like to know more about the controllers they designed, namely a gain-scheduled P controller and a linear Model Predictive Controller (MPC), and why they think the P controller has the better performance, you can find the study here.


How To Tell Which is Which


Usually, I provide you with a listicle with other sources, videos and white papers for you to read. This time though, I've decided not to do it, because I actually think this study is very well structured, explains workflows and thought processes thoroughly. You should definitively check it out! Before you leave the blog to read this enticing study on forklifts, let me remind you of the overall concept of real-time simulation quickly. There are two ways how it's done depending on whether the prototype or the design is being modeled:

Now I am finally ready to ask you the question I was pitching at the beginning of this blog post: Do you know what the terms "on the fly" and in-vehicle testing mean?


As always stay tuned, curious, cute and hooked!

©2020 by The Unlikely Techie. Proudly created with Wix.com