View Full Version : Control System for Trailer Dynamometer Application

01-22-2006, 08:32 AM

My name is Clement and I am working with a rally racing team on a project to develop a trailer dynamometer.· For the trailer dynamometer (which will be towed behind the rally car), the aim is to provide a variable braking load to the rally car, in order to evaluate the rally car's cooling system while driving on a test track.· In addition, the trailer dynamometer must maintain the rally car & trailer at a pre-set constant speed.· Therefore, we need a control system to monitor the actual speed to be compared with the pre-set desired speed and then control the braking load accordingly to maintain constant speed.

The actual speed of the trailer will be measured using a GPS speed sensor.· The braking load will be provided by an eddy current retarder which is controlled by the use of an alternator; therefore, the control system will need to control the alternator in order to control the retarder and hence the speed of the car.· However, it should be noted that in order to control the alternator, a duty cycle voltage will be needed, which the control system will need to supply.

I have little experience and knowledge with controllers and so Im hoping·to get some help.· Therefore, it would be greatly appreciated·if·someone can provide a list of components that will be needed for this controller and how·I can implement it as a system.· .

Thank you very much!


David B
01-23-2006, 12:21 AM
You might need more sensor inputs than just GPS speed to be able to do a good job of vehicle speed control.

My own GPS sensor does a lot of internal averaging before it delivers a speed result to the display. Once I was riding my bicycle as I was watching the GPS speed. I had to stop fast for a red light, but even as I was stopped, my speed at reported by GPS was still my travelling speed. It took about 5 seconds for the display to taper down to zero.

It is possible that other sensors may be optimized for faster response, but a little Google searching turns up several other descriptions similar to my experience.

One website describes a rapid-response vehicle speed sensor built by combining the GPS speed with accelerometer readings. (They claim that wheel rotation sensors are inferior over time because of tire wear.)

There's a technique known as Kalman filtering that mathematically works through combining the results of several different sensors, including discarding sensor readings if they wildly disagree with the others. Something like this might be useful - you wouldn't want the trailer to slam on the brakes because your GPS receiver suddenly caught a bit of multipath signal reflection.