Servo Tracking System MATLAB

[Mithrandir2008]

Wall of text Warning!

Ok guys,

Just finished the control loop for the antenna tracking system. I'm uploading the .mdl file for those who are interested in looking through MATLAB.

I just have one last query I have to ask.

I need to know the parameters of the DC servo motor according to the design specs. The original motor is a 2-axis servo, but the control loop remains the same.

I need a torque of about 2 N-m and the power rating must be 300W running with a 24V DC supply.

The parameters that I need are the Torque Constant(Kt), Back emf Constant(Kf), Motor Resistance and Inductance(R and L).

The moment of Inertia I've calculated already and it comes to about 0.3375

The coefficient of viscous friction, I'm assuming to be about 0.2(which may be a bit largr though)

The model which I've implemented is COMPLETELY wrong in terms of motor parameters. I actually had to keep a 8000 amplifier(192KV!) to make it move fast enough

Oh and the speed of antenna movement must be about 80deg/sec. Motor speeds are generally much higher so I can use a gear box so no biggie there.

MODEL WORKING:

The model basically has two stages,

1) First, a manual input is given and the motor first moves/slews there.

2) Once stability is achived(error<0.5deg and oscillations<0.05deg), It has to switch to autotrack mode i.e, It has to align itself based on the signals recieved from the antenna.

I've already designed every block.

For those who need a breif idea on how auto tracking works,

A radar system has 4 points on it which recieve power(Imagine 4 points on a circle which form an X if joined). If the target is perfectly aligned with the bore(center), the points on the opposite ends will recieve the same power. So, the model is designed such that if the voltages recieved by the opposite ends are same, the target is reached.

I decided to simulate the antenna signals by simply applying the difference of the target and the posn of motor and splitting it into two.


All thats left now is to get the proper parameters to get the model working!

Any help/comments will be appreciated.

Thanks a lot.

 

[Mithrandir2008]

P.S

If anyone wants to experiment with the model,

The parameters are there in File>Model Properties>Callbacks>Model Initialization.

Still have to implement a moving target model though

 

[Mithrandir2008]

No one?

:sniff:

Finished the implementation, now the model is perfect!

Ah well, I suppose it's a bit much to expect MATLAB with people...

Feel free to comment(again).

Removing the previous one.

 

[saeedazari]

hi Mithrandir2008,
i just read your post and i'm working on a same project as yours. i've found an article that contains all the thing you need.
send me your email so i can send it for you.

 

[Mithrandir2008]

This is technically necroposting.

I've finished this project a long time back saeedazri, it was part of my project during my bachelors. Even went so far as to implement a crude form of fuzzy logic. I should go through it again because I suspect a few mistakes might have crawled in. I should be able to figure out the boo boo's now as I am pursuing my Masters .

Thanks though!

 

[saeedazari]

so glad to hear you've done the project. one question i have is that was your radar system a monopulse radar or not? because my project is to design a monopulse tracking radar servosystem in simulink.
unfortunately i have just one patent that has some equations about the servosystem of the tracking radar. Do you have any article or patent that has anything about my project?
thanks a lot for your answer.

 

[Mithrandir2008]

I think it was. Basically this project was only a part of some military project(Proprietary stuff so I can't give all the details).

Basically, the error is inferred by finding the difference in signal strengths received by the radar at the 4 corners. I was given the basic idea and asked to improvise. I assumed that all the data received would be in the form of pulses(synchronized), subtracted the two opposite ends of radar signal strength and put an assumed multiplier(which varies depending on the system) to infer the error. It was rudimentary but they seemed okay with it. Most of my work was focused on the modelling(frequency domain) and stability analysis of the motor i.e., calculating motor loading parameters depending on the size and weight of the antenna, wind speed(maximum possible) and so on. I cannot exactly remember which papers I referred, but there was one very useful for the modelling specifically. Let me see if I can find that link. I'll try and reply later today.

Hope that helped.