Hello,
I am working on algorithm which would control a robot exploring unknown random maze. Maze in unknown. Robot may acquire information about what it sees using sensors in 4 directions. Size of maze may be e.g. 25x25 oraz 20x20. There are no open spaces in maze (in each cell, robot is close to at least one wall). Each step of robot costs energy point. When I discover new cell (it is not demand to move robot to that cell; it is enough that I see that cell using sensors) I earn energy point. I need to discover whole maze in as small number of moves as possible and earn as much points as possible. I developed my own algorithm which makes list of "oppenings" (cells which are connected to unknown regions of maze) indicating how much unknown "walls" each cell has (1, 2 or 3). There is also information, how much steps robot must do to move from actual position to each "opening". After small calculations, my algorithm chooses which oppening is the best and moves robot to it. My question is: is there any known algorithm which could I use in described exercise?
Thanks in advance for help and suggestions.
Best regards.