Critical Design Review: Dead Reckoning System for Mobile Robots Lee FithianSteven Parkinson Ajay JosephSaba Rizvi
Problem Statement Use a mobile robot and develop a synthesized navigation algorithm. We will integrate various sensors.
Modules I. Sensor Interface I. Produces output from the electrical inputs it receives; x-, y-distance traveled & heading II. Navigation I. Allows for different methods to be used regardless of input and output needs III. Movement I. Controls motors; can be modified if motors are changed
Accelerometer
Created an algorithm that changes electrical output into position data. A(g) = (T1/T2 – 0.5)/12.5% Pos = (A(g) * t^2)/2 Pos = Pos Start + Pos New
Shaft encoder
Shaft Encoder D = (Left D + Right D) / 2 Θ = (Left D – Right D) / b X = D * cos(Θ) Y = D * sin(Θ)
Gyroscope
Compass
Merging Data Average positions calculated from sensors Weighted average of positions calculated from sensors Use sensors calculations for certain tasks and scale the results
Problems Basic Misunderstanding of problem statement OOPic limitations Counter usage, frequency generator Delay of parts Building robot vs. assembly of robot
Changes Added sensors Compass, gyroscope Upgraded OOPic Memory, Faster uC Added protoboard Switched to rechargeable batteries Integrated sensor for position calculation
Circuit Diagram
Conclusion Construction Mark III based robot with shaft encoders, accelerometers, compass, gyroscope Validation to ensure systems work at a basic level Experimentation Use dead reckoning navigation in trials. Record trials on butcher paper Analysis Numerical analysis of accuracy of navigation method.