Difference between revisions of "HANtune/MATLAB Interface"

From OpenMBD
Jump to: navigation, search
(Reading data from the model)
(Creating a daq list)
Line 69: Line 69:
  
 
====Creating a daq list====
 
====Creating a daq list====
Use the MATLAB engine to call createDaqList.m with your modelName as an argument. The last argument ('signals') determines what items will be in the daq list. If 'signals' is used, a daq list will be created of all signals in your model. If 'blocks' is used, a daq list will be created of all non-virtual blocks in your model.
+
Use the MATLAB engine to call createDaqList.m with your modelName as an argument. The last argument ('signals') determines what items will be in the daq list. If 'signals' is used, a daq list will be created of all signals in your model. If 'blocks' is used, a daq list will be created of all non-virtual blocks in your model. The createDaqList function returns two arrays. The first contains all signal names, used to create signals in HANtune. The second one contains the daq items used to retrieve the data from the model. A daq item has the following format: [blockHandle, portHandle, dataIndex].  
 
  <nowiki>#Create a daqlist from signals in the model
 
  <nowiki>#Create a daqlist from signals in the model
 
result = engine.feval(2, 'createDaqList', modelName, 'signals')
 
result = engine.feval(2, 'createDaqList', modelName, 'signals')

Revision as of 07:25, 24 October 2019

HANtune can be interfaced with MATLAB® (and by extension Simulink®). We can do this by using the MATLAB engine and calling it from a script.

The scripts needed to run this example can be found in HANtune\scripts\examples\MATLAB_Interface. These are:

  • ConnectToMatlab.py
  • ReadSimulinkModel.py
  • createDaqList.m
  • getDaqListData.m


Note: You will need to have installed a version of MATLAB® with the Java MATLAB engine.


Connecting to the MATLAB Engine

The first thing we need to do is tell HANtune where to find the MATLAB engine.

Open ConnectToMatlab.py and change the following line to the location of your MATLAB engine:

#Change this to your location of the MATLAB engine
enginePath = "C:\Program Files\MATLAB\R2017b\extern\engines\java\jar\engine.jar"

Add the engine to the java classpath so we can call it from a script.

#Add matlab engine to the classpath
from datahandling import CurrentConfig
CurrentConfig.getInstance().addSoftwareLibrary(enginePath)

Import the engine and connect to a running instance of MATLAB.

Note: The scripts assumes there is shared instance of MATLAB currently running. To to share an instance of MATLAB, go to the command window of MATLAB and enter: Matlab.Engine.shareEngine. If you don't want to connect to a running instance, but start a new instance of MATLAB instead, remove the engines[0] argument from the call to MatlabEngine.connectMatlab().

#Import matlab engine
from com.mathworks.engine import MatlabEngine

#Initialize matlab engine
engines = MatlabEngine.findMatlab()
engine = MatlabEngine.connectMatlab(engines[0])

Optionally you can set the engine as a global variable across all interpreters. This allows you to initialize the engine once and then call it from another script.

#Add engine as a global variable to all interpreters
from nl.han.hantune.scripting import ScriptingManager
ScriptingManager.getInstance().setGlobal("engine", engine)

Optionally we can redirect the output (e.g. print statements, warnings, errors) to the console of HANtune.

#Redirect matlab output to HANtune console
from java.lang import System
from nl.han.hantune.scripting import Console
System.setOut(Console.getGUI().getOut())
System.setErr(Console.getGUI().getErr())

Reading data from a running Simulink model

Set your model

Open ReadSimulinkModel.py and change the modelName and modelPath variables to that of your model.

#Change these to the name and location of your model
modelName = 'MyModel'
modelPath = 'C:\Users\Michiel Klifman\Desktop'

The path to the example scripts and the path to your model must be added to the MATLAB search path so that MATLAB knows where to find them.

#Add folders to MATLAB path
from java.lang import System
scriptPath = System.getProperty("user.dir") + "\scripts\examples\MATLAB_Interface"
engine.eval("addpath('" + scriptPath + "')")
engine.eval("addpath('" + modelPath + "')")

Creating a daq list

Use the MATLAB engine to call createDaqList.m with your modelName as an argument. The last argument ('signals') determines what items will be in the daq list. If 'signals' is used, a daq list will be created of all signals in your model. If 'blocks' is used, a daq list will be created of all non-virtual blocks in your model. The createDaqList function returns two arrays. The first contains all signal names, used to create signals in HANtune. The second one contains the daq items used to retrieve the data from the model. A daq item has the following format: [blockHandle, portHandle, dataIndex].

#Create a daqlist from signals in the model
result = engine.feval(2, 'createDaqList', modelName, 'signals')
daqItemNames = result[0]
daqList = result[1]
daqSize = len(daqList)
#Create a signal for each item in the daq list
signals = []
removeAllSignals()
for name in daqItemNames:
	signal = createSignal(name)
	signals.append(signal)
updateLayout()
print 'DAQ list created with ' + str(daqSize) + ' items'

Reading data from the model

First we will start the simulation:

#Start the simulation
engine.eval("set_param('" + modelName +"','SimulationCommand','start');")
print 'Simulation started...'

And then start reading the model by calling getDaqListData with the daq list that was created earlier as an argument.

#Read data from the model in Simulink and add it to the signals in HANtune
while (True):
	data = engine.feval('getDaqListData', modelName, daqList, daqSize)
	if not data:
		break

	i = 0
	for value in data:
		signals[i].setValueAndTime(value, 0)
		i += 1

print 'Simulation completed!'