The most powerful feature of my free model airplane power system calculator is the EZ Optimizer. Here’s how to use it.
Review Online Tutorials
What I’m about to say won’t make much sense until you have gone through the online interactive calculator tutorials. I wrote them based on the questions the early calculator users asked me. Since I wrote the tutorials, I get very few questions on how to use the calculator. They do a really good job of explaining how to use it, from the basics to more advanced uses.
Just a Pretty Face
The regular optimizer came first. It is very flexible and extremely powerful. The EZ (“easy”) Optimizer is just a pretty face on top of the regular optimizer. It makes running the optimizer a lot easier, but keep in mind that it is entirely optional. In practice, I normally run the EZ Optimizer once to get a general feel for what the power system solution looks like. Then I iteratively edit the filters and run the regular optimizer until I get the result that I’m looking for.
Filters are Your Friends
Each tab in the calculator is called an editor. The first one is for airplane components. The next to last tab is for the scale editor. Each editor has a filter for filtering or limiting the components displayed. The filters also limit the components that the optimizer will consider when looking for the best power system. The best power system is the one that provides the longest endurance and meets all of the imposed constraints. The filters start out turned off.
EZ Optimizer Sets the Filters
The EZ Optimizer creates filters for you based on the choices that you make on its dialog box. It also sets the knobs on the airplane editor to match your selections on its dialog box. Then it runs the regular optimizer.
It may not sound like much when I describe it like that, but it’s a great convenience. On each editor, it creates at a minimum a filter to omit from the results the components that are considered incomplete. This is always a good idea.
The EZ Optimizer will also compute a reasonable cruise speed for the model. This is based on the flying weight, wing type, and wing area. We tend to fly our model airplanes faster than we need to, so feel free to increase the recommendation shown in the dialog.
If the filters are too restrictive, then the optimizer won’t be able to find a combination of power system components that satisfy them all. This is what’s called an over-constrained problem. I get an email once in a while from somebody that ran the optimizer and it “did nothing”. This is what happened.
The solution is to ease up on the constraints. Remove some filters or at least broaden the list of components that they include.
The optimization algorithm is extremely fast. I only have to wait for it to finish if I haven’t filtered out any components. Then the optimizer has to try every single component combination, which can take a couple of minutes.
It is prudent to create filters first that eliminate components that you know cannot possibly be a part of the best power system. Once I do that, the optimizer always finishes in a second or two. That is when the true power of the calculator comes into play. With near instant “what if” scenarios, I can try out all sorts of power system component combinations very quickly.
Secret to Speed
Despite the pretty face on the calculator, it uses some really advanced programming techniques under the covers. The optimizer is programmed as a constraint satisfaction problem with intelligent heuristics that dramatically speed up the search. The power system equation solver uses a neural network at its core. I kid you not.
Squeezing More Performance Out
It can sometimes be really tricky to improve the overall power system efficiency. For example, sometimes improving the efficiency of the motor decreases the efficiency of the propeller. The good news here is that it gets easier with practice. Remember that the larger the propeller and the higher the motor input voltage, the higher the overall system efficiency will be.
One Optimization Goal at a Time
The optimizer will only optimize for one scenario at a time. What if you want to optimize for shortest take off distance and most efficient cruise at the same time? Well, you cannot, at least not directly.
Here’s one possible solution. Most airplanes spend far more time cruising around than they do taking off. Optimize for the cruise first, and then tweak the components until you get an acceptable take off distance.
Like I said, the optimization algorithm always tries to maximize the endurance of the model. It is really good at doing that. Left to its own devices, it tends to come up with very long values like 30 minute flight times or even longer. Ten to fifteen minutes is plenty for most of our flying needs. Examine carefully the power system it came up with, and edit the filters to shape the results of the next optimizer run so that it better matches your flying goals.