Assuming I understood what you meant, I wrote a solution notebook in Wolfram Language. I don't think you can solve it in WolframAlpha (maybe with the desktop version, which I don't have).

You may try intersecting three annuli, each given by inequalities.

Something like

minimize the sum of the distance from p to p1 and the distance from p to p2 and the distance from p to p3 with p1 = (1,0) and p2 = (0,2) and p3 = (-1,0) and p = (x,y)

should do. But in my WolframAlpha Android App this query is not understood.

**Postscript: ** Sorry. This is not what you need. It will give the center of gravity, i.e. mean, of the three sensor locations. Anyway, you'd better handle that problem directly within Wolfram Language, not WolframAlpha. The latter's natural language queries are a nightmare concerning mathematics. Most of the time they are not understood and it's and endless trial and error game to achieve results.

Minimize the sum of the squared error? Can you think how to write an expression for that?

For one sensor that will be the square of the distance from the unknown x,y point to the circle given by the center and the sensor's estimated radius.

Sum the three squared errors and ask WolframAlpha to find the x,y that minimizes that sum.

Do you have an example of the three sensor positions and three distances estimated?