1 Energy Model and Measurement for DC Motors in PPRK Yongguo Mei Sep. 18, 2003
2 Outline PPPK Energy model Experiment design Future work
3 PPRK Palm Pilot Robot Kit Cheap, easy to build, and fun Created by two Carnegie Mellon Robotics Institute research groups Desktop Robotics, fully autonomous robot controlled by a Palm handheld computer Open resource available from their website
4 PPPK (continued) Central control Unit: BrainStem, General Purpose Module 3 MX-400 Servo Motors, driving three wheels 3 GP2D12 infrared sensors Connect to PDA through RS-232
5 Power source and consumers Source: Hobbico HydriMax 6.0V Battery, or 4 regular AA alkaline batteries, precious Consumers: GP module, sensors, DC motors Not consider PDA here Concentrated on the speed control of motors to save energy
6 DC Motor
7 Circuit Model V = iR + Ldi/dt + V emf T = K t i V emf = K t ω T– torque ω– angular velocity K t – torque constant P=V*i Convenient, but enough?
8 Other factors Saturation caused by high current Non-linear commutation effect Eddy current losses Frictions No general accepted model How do we handle all of these?
9 From another direction Real measurement to get the energy model for our Robot? Few people did this before Avoid the complicated theoretical analysis A viable way for our problem
10 Possible format For specific Motors and working environment, Power is a function of the velocity and acceleration What are possible formats? Polynomials with highest exponent 2 P(v,a) = c(1)v*v+c(2)v*a+c(3)a*a+c(4)v+c(5)a+c(6)
11 Possible format Second-degree polynomial is enough It can incorporate most of the factors mentioned previously Compare this with other models Can this model be extended to other Robots?
12 Experiment design MX-400 Pro Servo motor Three lines: ground, power and control Measure voltage and current by Data Acquisition, DAQ Program by Labview to process the data
13 Experiment Design Accerelation, constant speed and Deceleration Rotation Speed control by c program, hope not too difficult!
14 Future work Real measurement Energy optimal control Motion planning Multimover problems