'VIRTUAL -ACTUAL' LABORATORY FOR POWER ELECTRONICS AND MOTOR DRIVES
M. Ehsani & N. Shidore
Power Electronics and Motor Drivers Laboratory
Advanced Vehicle Systems Research Program
Texas A&M University
College Station Texas 77843
979-845-7582
Ehsani@ee.tamu.edu
Laboratory experiments are an indispensable part of undergraduate teaching; more so in the case of power electronics and motor drives, where the student runs the motors in the lab and verifies their behavior as predicted by theory taught in class, under different conditions.
Conventional labs of motor drives and power electronics have three or four workstations; with each workstation made up of a motor and a load. The student performs simple experiments on these motors, for example:
1. Open loop speed control in the case of a separately exited DC machine,
2. No load and blocked rotor test for Induction machines.
These hardware experiments are important for the students because they provide them with a sense of actually working with a motor or give them a 'feel' of the machine. But, these hardware laboratories suffer from certain inherent disadvantages:
1. Hardware is limited in number as well as variety because of cost concerns; the later restricts the nature of questions that can be asked, while the earlier may result in over crowding of students on a particular work- station. The only possible way of overcoming the above problem is to buy more and expensive hardware, which is not always feasible.
2. Limited access to motor parameters - Most of the motor /load parameters cannot be changed; the student cannot study the effect of change in parameters.
3. Safety of students, damage to hardware and similar issues; this also restricts the nature of questions that can be asked of the students.
The above three points can be summed up by the phrase 'Lack of flexibility', especially when it comes to the nature of experiments that are performed on the hardware. The student cannot satisfy his or her own curiosity by performing anything 'out of the way' because of lack of flexibility or fear of injury/damage. Similarly, intuitive questions, which can lead to better understanding of the motor behavior, cannot be asked. This is a major drawback as it drastically reduces the effectiveness of the lab.
Normal Simulations (using Simulink/ Pspice /Psim) cannot be an effective substitute for hardware, because these simulations are generally too fast or two slow (scaled time simulations). Although better (intuitive) questions can be asked of the student because of more flexibility. However, the sense of working with actual hardware is absent. Also, being scaled time, the sense of time, which is very important when it comes to electro-mechanical systems, with time constants affected by parameters like inertia, friction etc, is also not available.
The 'Actual - Virtual - Remote access lab' laboratory at Texas A&M University attempts to address all the above issues. This Lab consists of actual hardware experiments along with ' Real time - Virtual machine' experiments.
The students initially perform the conventional experiment on the hardware; as in the case of any conventional lab. Then, they perform the same experiment on a 'virtual- machine' with infinitely more freedom and much more intuitive questions asked. 'Virtual Machines' are 'Real time' Simulation environments, i.e. the actual hardware set- up is exactly duplicated in the form of a real time simulation. The real time simulation model of the motor has the exact parameters of the hardware. It comes with a lab-view interface so that the student changes parameters by turning knobs or moving sliders, thus giving him the feel of running a real motor.
As this simulation runs in real time, i.e. the simulation time shown matches exactly the actual time for which the experiment has been performed, the student gets a feel for the time constants of the electro-mechanical systems, and he /she can study the effect of change in inertia, inductance, friction etc on the time constants of the system. Therefore, simulation with these 'virtual machines' is much better than normal simulations.
Also, since the motor has been modeled inside the computer, it enables the student to change parameters of the model, be it electrical parameters like inductance, resistance or mechanical parameters like coefficient of friction, inertia etc, which are seldom possible with actual hardware. The nature of the load can be varied, from constant torque load to velocity dependent loads etc. The student can also be made to simulate faults inside the machine or control, and learn the behavior of the drive under such different conditions.
An example of a question which could be asked of the student on this virtual machine would be "What happens if the field circuit of a separately exited DC machine suddenly opens?" or " What happens if you start a the machine without field excitation?" It is impossible for the student to perform such potentially damaging tasks on actual hardware. The student observes the behavior of the virtual motor under such circumstances and then goes back to the theory taught in class to reason out such behavior. This exercise is performed for changing parameters inaccessible in the real hardware, changing the nature of loads etc. When the student observes things happening on the virtual machine and goes back to theory to figure things out, it gives him/her tremendous physical insight into the machine operation, and helps him/her to correlate the mathematical equations to the physics of the machine.
It is also planned to make this Virtual Lab a remote access laboratory.
All the universities have very high-speed Internet connection. It is proposed to utilize this facility in order to work on a lab in a novel manner - through remoter access i.e. Internet connection.
The student is able to observe the user interface of the virtual machine from his computer, and also make changes to the inputs and parameters displayed on the user interface through his computer, which is at a remote location with respect to the workstation. In short, he is able to use the virtual machine through remote access, in real time.
This sort of a lab has many advantages:
1. It gives the students the freedom to do the laboratory at a time, which they find convenient.
2. Indirectly encourages the students to work alone on the 'virtual machine' rather than being forced into groups as it happens in the normal lab.
3. The students can go over the lab again and again till they fully understand the concepts of the system.
4. Students can satisfy their own curiosity by asking themselves questions which were not a part of the original lab package, and playing with the virtual machine to get answers, without having to worry about time restrictions.
5. There can be such a thing called ' Lab HW' where in the students have to try out additional experiments on the virtual machine, having understood the concepts in the regular lab.
This Concept of Virtual labs has been implemented successfully before in the field of robotics and Controls, where in the student controls the robot remotely. Delay in signal due to Internet traffic is not as important a consideration for such applications as it is for motor drives.
For example, because of the delay in Internet traffic, the start up time of the virtual motor might appear to be more than it actually is, thus affecting results and confusing the student.
But, this technology can be implemented in experiments where the time -delay is not as critical. References [1], [2], and [3] explain different methods by which remote access is possible.
References:
[1] Mauricio F.Magalahes et.al, 'REAL-A Virtual Laboratory for mobile robot Experiments', IEEE transactions on Education, Vol.46, NO.1, February 2003.
[2] Nitin Swamy, Ognjen Kulijaca and Frank Lewis, 'Internet based Educational Control Systems Lab using NetMeeting', IEEE transactions on Education, Vol.45, NO.2, May 2003.
[3] Jamahl W.Overstreet and Anthony Tzes, 'Internet based Client/Server Virtual Instrument Designs for real time remote access Control Engineering Laboratory', Proceedings of the American Control Conference San Diego, California, June 1999
