Robot Joint Control Architecture

Here's an image that we've seen many times before. We want this robot to do something useful, so we're interested in where the end effector is positioned in 3 dimensional space. We require the robot tool moves to a particular pose.

We've talked about how we can use inverse kinematics to work out what are the joint angles required in order for the robots end effector to be at the pose we require to do the useful task. What we haven't discussed so far is how we control the robot so that the joints move to the desired joint angles.

The control of any one joint of a robot arm, typically has a structure that looks something like this. In this diagram we can see that there's a loop and that's a very important part of the architecture. We're going to walk through this system and we're going to start with the actuator.

The actuator is the device actually makes the robot joint move, it's the motor. The output of the motor is some position and we have a mechanical transmission which connects the output of the motor to the robot joint itself. The mechanical transmission itself is commonly a gearbox and some types of robots, it might be a cable drive.

The position of the motor is monitored by a position sense and that gives a signal that says what is the actual position of the robot joint motor. An input to the system is the desired position and what we want to do is to move the motor until the actual position is equal to the desired position and that's the job of the joint controller and typically that's an imbedded micro controller. It's a piece of code running on a micro processor that provides the appropriate commands to the motor so that the actual position matches the desired position.

The output of the joint controller is typically some electronic signal. It might be a voltage or it might be a pulse with modulated waveform. This is input to the motor drive which is a lot of power electronics which controls the current or the voltage that's applied to the actuator. 

The most important part of this robot joint control system is this feedback loop. It's where we compare the actual position of the motor with the desired position and compute a control signal to the actuator to make those two things equal.
In the rest of this lecture, we're going to look at each of the individual elements in this diagram one by one. If you're going to get the most out of this lecture, then it's important that you've got some necessary pre-requisites. It will be most useful if you've completed a first undergraduate course on control theory.

If you've done that, then you'll be familiar with concepts such as the Laplace transform and you'd be familiar with concepts such as the poles of a dynamic system.

It would also be really useful if you are familiar with some fundamentals of mechatronic systems that is you know what is a motor, you know what a gearbox is, you know what an encoder is and some familiarity with the idea of embedded controllers.