Georgia Tech researchers are using an electric-powered autonomous vehicle to help driverless vehicles maintain control at the edge of their handling limits. (Photo: Rob Felt)

Researchers have devised a novel way to help keep a driverless vehicle under control as it maneuvers at the edge of its handling limits. The new technology is being tested by racing, sliding, and jumping one-fifth-scale, fully autonomous auto-rally cars at the equivalent of 90 mph. The technique – model predictive path integral control (MPPI) – was developed to address the nonlinear dynamics involved in controlling a vehicle near its friction limits.

The control algorithm continuously samples data coming from GPS hardware, inertial motion sensors, and other sensors. The onboard hardware-software system performs real-time analysis of a vast number of possible trajectories, and relays optimal handling decisions to the vehicle moment by moment. The auto-rally vehicles use special electric motors to achieve the right balance between weight and power.

Source 

Topics:
Automation Automotive Motion Control Motors & Drives Positioning Equipment Power Transmission Sensors Software

More From SAE Media Group

    Transcript

    00:00:00 The exciting thing about this technology that we have is the ability for the car to perform what are known as progressive maneuvers. So during turns, for example, the car is literally sliding around the turn. It's not maintaining its complete contact with the road surface. And yet in spite of that sliding maneuver, which gives us a lot of speed in that turns, the car

    00:00:26 is actually stable and safe in navigating this particular feature of our track. The algorithms that we have developed are able to project into the future what the vehicle is going to do in the next three or four or five seconds. And generate approximately 2,000 or 3,000 possibilities of what's going to happen. And based on these possibilities, it chooses the best one.

    00:00:49 And this can be done very, very fast. I think we're calculating about 2000 different possibilities every 50 milliseconds. This algorithm will have a broad impact. It doesn't have to be within aggressive driving. It's can be within any domain, with autonomy. We can think about tasks that relate to locomotion. Manipulation. You have a robot, and you want to be able to manipulate objects.

    00:01:14 Everything that involves, essentially, controlling of a system. That is very non-linear. That is very uncertain. You don't know exactly how it's going to react.