Find new location -- 2-wheeled car/robot

[TimurPlusPlus]

Hello, everyone!
Now I'm trying to develop 2 wheels robot, which travels along the line simulator.
The robot can turn only by increasing velocities of each wheel.

The conditions are below:
- I know velocities of 2 wheels
- I know radius between wheels

The task is next:
How to find new location of the robot, if it has L_wheel_v = 3 km/h and R_wheel_v = 10 km/h, radius = 3 m, for example.

 

[berkeman]

Hello, everyone!
Now I'm trying to develop 2 wheels robot, which travels along the line simulator.
The robot can turn only by increasing velocities of each wheel.

The conditions are below:
- I know velocities of 2 wheels
- I know radius between wheels

The task is next:
How to find new location of the robot, if it has L_wheel_v = 3 km/h and R_wheel_v = 10 km/h, radius = 3 m, for example.

Welcome to the PF.

Can you show us your work on these equations? They are pretty straightforward, so you should be able to show us your sketches and your work so far.

Also, how are you going to stabilize your 2-wheeled robot? Will you use some gyroscopic stabilization scheme like for the Segway?

https://selfbalancingscooters.files.wordpress.com/2015/08/41.jpg

 

[Baluncore]

The conditions are below:
- I know velocities of 2 wheels
- I know radius between wheels

Using velocity will get confusing. Instead put a shaft encoder on each wheel.
Incremental-Rotary-Encoders with 360 or 600 cycles of two phase AB code per revolution are cheap.
Keep track of wheel rotation by counting pulses up and down as the wheels rotate forward and backward.
The sum of the two wheel counters will give you the distance travelled.
The difference of the wheel counts will give you the direction it is now facing.
You can track it and navigate by summing the step movement vectors as it travels.

 

[JBA]

If you are trying to turn the robot while maintaining a forward rotation on both wheels then a virtual center of rotation can be determined by using the same data; however, in this case there is going to be a calculated vs. actual track issue that will affect the accuracy of the robot's actual subsequent position vs the calculated one. This type of error is going to be accumulative through multiple turns and therefore can result in a significant accumulated robot location error if not verified at some points by an independent position reference input and corrected.
If you are designing a toy to run around a relatively large space this may not be a issue but if the robot's application is a more functional one then it could be an issue.