This project simulates a simple robot that decides how to move based on sensor input using a basic artificial intelligence model.
-
Initialization:
- The robot starts with a random set of weights. These weights help the robot decide its actions based on sensor data.
-
Sensor Input:
- The robot receives input from its environment, such as the distance to an obstacle. In this example, we use a fixed value of
0.3
.
- The robot receives input from its environment, such as the distance to an obstacle. In this example, we use a fixed value of
-
Forward Propagation:
- The sensor input and a bias value (always
1.0
) are combined into a matrix. The robot uses the random weights to perform a calculation called "forward propagation" to determine a score.
- The sensor input and a bias value (always
-
Decision Making:
- The score is passed through a function called sigmoid, which transforms it into a value between
0
and1
. - If this value is greater than
0.5
, the robot decides to move forward. Otherwise, it decides to turn left.
- The score is passed through a function called sigmoid, which transforms it into a value between
-
Robot Movement:
- Based on the decision, the robot either moves forward a set distance or turns left by a specified angle.
-
Output:
- Finally, the program prints the robot's action and its new position after moving.
When you run the program:
- It initializes the weights.
- The robot checks its sensor input.
- It decides whether to move forward or turn left.
- It updates its position accordingly.
- Nvidia support (Robotic)