Autopilot — Page 358, Lesson 421

Right, let's get into the autopilot chapter. This is where we start building the mental model of how an autopilot actually thinks and acts. I want to walk you through the foundational concept first, because everything else in this chapter hangs off it. An autopilot is fundamentally a control system, and it works by using what we call control loops. Now, the very first thing to understand is the difference between the two types of loops, because they're the skeleton of the whole system. Let's start with the inner loop. The inner loop is a classic example of a closed loop control system. In fact, that's one of its common titles — a closed loop. The key word here is "closed," and I'll explain why in a moment. Then we have an outer loop. The outer loop interacts with the inner loop to give the aeroplane guidance, so that it can achieve the required flight path. So think of the inner loop as the muscle that actually moves the controls, and the outer loop as the brain that tells the muscle where to go. Now, let's break down the basic elements of a closed loop control system. There are five of them, and I want you to hold onto each one. First, we have the Input. That's the desired value — what we want the aircraft to do. Second, we have the Error Detector, or Signal Processor. This is the component that compares what we want with what's actually happening, and it produces a signal based on the difference. Third, we have the Output. That's the actual result — the real state of the aircraft. Fourth, we have the Control Element. That's the thing that acts on the output to change it, like the actuator moving a control surface. And fifth, we have Feedback. This is the crucial one. Feedback is the path that takes a measure of the output and sends it back to the error detector, so the system can compare it against the input. That's what "closes" the loop — the output feeds back and influences the next input. That continuous cycle of compare, act, measure, and feed back is what makes it a closed loop. Now, contrast that with Figure 26.2, which shows an open loop control system. The defining feature of an open loop is that it does not have feedback. There's no path from the output back to the input. The controller in an open loop system may consist of a pre-determined programme, or it could be a human operator. But here's a subtle and important point: if a human operator is used, then the system, in effect, becomes a closed loop system. Why? Because the human is closing the loop — the human is watching the output and feeding that information back into their decisions. So the human becomes the feedback path. Let me give you a really clear illustration of the difference, using something you know from everyday life: the domestic central heating system. Imagine a system with a timing controller but no thermostat. That's an open loop system. The pump would continue to send hot water round the house regardless of the room temperature, for the whole duration of the period set on the timer control. It doesn't care what the actual temperature is — there's no feedback. Now, on the other hand, a system with a thermostat would interrupt the circulation when the room temperature reaches the preselected level. That's feedback control. The thermostat measures the output — the room temperature — and feeds it back to stop the pump. That's the closed loop in action. Now, there's one more piece of terminology I want to give you, because it's specific to aviation. Feedback control systems that are used for positional control — and aircraft flying controls are a perfect example — are usually referred to as servo systems, or servomechanisms. So when you hear "servo" in an aircraft context, think positional control with feedback. And here's the precise definition of a servomechanism: it's a closed loop control system where a small input is converted into a larger output in a strictly proportionate manner. So a tiny movement or signal at the input gets amplified into a much larger movement at the output, but the relationship stays strictly proportional — the output is always a faithful, scaled-up version of the input. That proportionality is what makes it a servo, not just any amplifier. So, to tie it all together: the autopilot is built on these control loops. The inner loop is the closed loop that does the actual controlling of the aircraft's position, and the outer loop guides it to the flight path. The whole thing works because of feedback, and when we're talking about moving the flying controls, we're talking about servomechanisms. That's the foundation.