Answer the question
In order to leave comments, you need to log in
How to implement a mobile robot navigation system?
There is a mobile robot that is stuffed with all sorts of sensors. It moves inside an empty field with dimensions of 3 by 3 m.
How can I make the robot move to the point of the field that I specified?
And if I move it with my hands, then the robot should return back to the point I specified.
What if there are 10 robots?
Answer the question
In order to leave comments, you need to log in
He needs to know the current coordinate, you can put some sensors or the like in each corner of the field, to which he will be guided (ie, north, south, west, east).
One way or another, the robot needs to highlight the features of its particular location and associate them with coordinates. This requires additional information about the field.
Additional information is entered either from an existing field ("fence", "pillars", "beacons", a pattern on a field or fence),
or additionally created specially (a special pattern on a field or fence, active ultrasonic beacons, some kind of interference ultrasonic field special for each point).
In general, historical and everyday methods of geopositioning are a storehouse of ideas for your question. =)
If there is a compass, install a servo on the robot (360 degrees) adjust the servo so that the "table" on the servo always looks in 1 direction. On the "table" place 2 or more laser or some other rangefinders that will measure the distance to the walls. These distances will be your coordinates.
wedge.
the task is divided into at least 2 :
1. positioning.
2. building a route.
The first one is the most difficult and the most interesting.
in order not to wipe my fingers into the blood, I will refer you to a good article here .
in fact, under these conditions, only a local positioning system will suit us - only they will provide the necessary accuracy. if there is no restriction on the equipment of the field (for oneself, not for competitions) - then imho it will be more efficient to use stationary emitting beacons of the field and the receiver (s) on the robot.
this is both a variant with ultrasound, which has already been proposed, and a variant of the classic "beacon" - visible / infrared light (plus light - no dependence on humidity / temperature). "lighthouses" of different colors on the corners of the field. Place receivers with color filters on a rotating platform / servo. Knowing the angles between the received "readouts" from the "beacons", it is easy to calculate the coordinates on the field.
everything is worse if beacons cannot be set. Here perhaps only CV will save. look at an example from the competition on YouTube - a field, 2 robots, positioning mainly by CV.
2. building a route - knowing the coordinates of the starting point and the end point of the vector for 2 currents .
the presence of other robots does not change the essence - except that they can start blocking each other's beacons (decided by raising the beacons) and start "pushing" - well, for this it is worth considering this during the movement cycle.
I propose a scheme with my system of beacons. Lay antenna wires along two perpendicular walls. Each wire has its own generator with its own frequency. Two amplitude detectors determine the induced voltage in the receiving antenna. Calculate the removal by the power of each amplitude.
Or solve the reverse problem. The camera from above takes pictures of all robots and transmits (IR, radio, sound) a packet with the coordinates of all robots.
Once they discussed such a task, the only decision accepted by all the participants was "To come up with a sausage smell sensor and put a piece of this food product in the right place" :)
High-precision (+-2cm) indoor navigation system for autonomous robots and copters: marvelmind.com
Demo video on your topic:
- https://youtu.be/TBTw1njQ7QY
- https://youtu.be/sqUcEjRizbc
Didn't find what you were looking for?
Ask your questionAsk a Question
731 491 924 answers to any question