I am new to Simulink and try to achieve the following:
I have a signal which simulates output power of an engine. I now want to be able to change this power output to a new value.
My question: How do I implement a linear adjustment from the current output to the newly requested output? Linear in the sense of a constant rate of change, e.g. x Watt/second.
Thanks!
The simplest way to handle this is probably the Rate Limiter block.
If you cause a step-change in the demand, the rate limiter will take the demand as an input and produce an output with the rate of change limited to the rate specified in its dialog parameters.
There is a dynamic version if you want the maximum slew rates to be specified via signals.
Related
I have a system, described by a black-box, that takes as input a signal in time (let's say something similar to a sine wave) and returns a single scalar as output.
My goal is to find the optimum signal that maximizes/minimizes the output. As constraints, the time average of the signal must be kept constant and the minimum and maximum value of the signal must be within a specific range.
The parameters to be optimized are all the points describing my signal in time.
I want to use the scipy.optimize library to minimize the output of my black-box system however, everytime the parameters are modified by the optimizer, I need to smooth the input signal in order to avoid discontinuity.
My question is: is it possible to access the input variables after they are modified by the algorithm, smooth them, and substitute them at every iteration of the algorithm?
An outside Library (from PreScan) requests 200 Hz while my control plant model needs to run at 100 Hz. Therefore, my question is that how can I coordinate these two activities? My concern is that if I use 200Hz in Simulink, it may compromise my control plant’s fidelity.
Is it possible to set simulink time step as 1/100 while keep the outside library to run at 200Hz?
Simulink works perfectly happily with multi-rate models. The thing (it appears) that you don't understand is the difference between the overall model sample rate - i.e. the settings of your solver - and the sample rate of individual blocks within your model.
It's very typical to have some blocks in your model sampled at say 100Hz, while other parts of your model sampled at 200Hz. In this case you would choose a discrete solver and give it a sample time of 200Hz. The 200Hz blocks would get executed at every solver time step, while the 100Hz blocks would get executed every second solver time step.
You should look at the Sample Times in Systems section of the documentation.
You can use both explicit and implicit rate control in Simulink.
Sample Time
To set the fundamental sample time go to: Configuration Parameter>Solver>Fixed-step size. You can also use the Simulink API:
get_param(bdroot, 'FixedStep')
set_param(bdroot, 'FixedStep', '0.005') % 200Hz
Colors
To activate the Sample Time colors go to: Display>Sample Time>All. The Sample Time Legend will help you understanding how the implicit rate control works.
Sample time Option
You can control the tasking and sample time options via: Configuration Parameter>Solver>Tasking and sample time options.
At the beginning you can activate the automatic handling of rate transition for data transfer. Then you shall analyse what colors your model elements are and place the Rate-Transition blocks on the data signal lines between the model elements with different sample rates.
Now the rate control is implicit. If you use the function calls to explicitly call your subsystems at a required rate by involving a predefined scheduler, than the rate control is explicit.
You can open build in Simulink examples to see how it works:
sf_ladder_logic_scheduler
sf_loop_scheduler
I am trying to find a transfer function of bldc motor speed over duty cycle percent. I made two measurements for different duty cycle percentages in order to both estimate transfer function and its validation.
For the first one (%65 duty cycle step input) I got below measurement and its transfer function estimation.
For the second one (%70 duty cycle step input) I got below measurement and it transfer function estimation.
The problem is that my transfer functions are not validating each other as shown below. They do not give the same response for the same input. Can anyone explain the reason?
It looks like the two measurements are very different. One has a maximum of 220, the other has a maximum of 350. This means the data acquisition is at fault, or the motor is itself variable.
Why don't you try measuring 20 times, and see if the raw data look similar?
Otherwise would need more info about your recording setup and the protocol for testing the duty cycles. It doesn't sound like a matlab or programming problem.
-- edit
Transfer functions are usually the output as a function of the input. Not functions of time.
Transfer function estimation assumes that the system is linear and time-invariant.
Most likely the system exhibits a nonlinear response characteristic which causes a very large change in output amplitude when input is increased from 65% to 70%, so a transfer function obtained at one operating point is not valid for the other.
I would like to make a discrete filter, where the sampling rate can be controlled by an input. I am trying to understand how the discrete filter block looks, "under its own mask." Is there anyway to retrieve the code behind this block so it can be modified for my use?
You can use an user-defined function as a filter, pick the filter transfer function, translate it into a difference equation (the discrete-time equivalant to the diferrential equation), implement that difference equation in a function and feed the sampling rate as an input (the sampling rate will appear as a constant in your difference equation).
The block is too complex and has too many options to simply be able to look under the mask. Your best option is to have a look at the documentation, which does show some detailed implementation of the block in some articular cases to get an idea and then try to recreate the discrete filter you want from basic building blocks, using a constant sample time at first, until you can validate your own implementation against the Simulink library block. Only then, start considering how you will change the sample time. Your main problem though is that the filter coefficients will change with the sample time, so you need to be able to re-calculate them on the fly. It's not an easy problem, I don't even know if it's possible.
How to use rate limiter with square signal and variable step size in Simulink?
Here's a screenshot of the model I'd like to set up:
model:
I feed a customized rectangular signal to a rate limiter to avoid vertical slopes.
Unfortunately that doesn't seem to work. I'm using ode15s, it's a requirement. Here's the error message Simulink throws:
Error: Input signals to Rate Limiter '.../Rate Limiter' are neither
discrete nor continuous sample time signals. Only discrete or
continuous input signals are supported
Quite surprisingly I found a workaroud by adding an integrator directly followed by a derivative. This works:
workaroud:
But it's ugly and I'm getting some very annoying stability issues in some cases. And I doubt very much that it is considered "good practice".
So how is one supposed to use this rate limiter block in such a situation?
John
Thank you both for your answers.
I forgot to say I had already checked the sample time with the colour display. It was "Fixed-in-Minor-Step".
Actually it was quite simple. If I get it right, the sample time was not specified or specified in a wrong way in my subsystem. Specifying Continuous in the rate limiter dialog box solved the problem!
thewaywewalk, I'll keep your suggestion in mind. Since I'm using steps a lot it might be useful.
Try displaying the sample time colours in your model to check what sample times your signals are using.
Introducing an integrator block will force the signal to become continuous, hence why it works. Maybe using a Signal Specification block or a Rate Transition block with a sample time of [0, 0] (for continuous signal, see Specify Sample Time in the documentation) will achieve the same thing and be slightly more elegant (using the derivative block is not considered good practice).