How is the Kalman filter estimator implemented in an MPC?

I have a physical process with 2 inputs, 4 states and 2 outputs. Two of the states are
measured while the other two needs to be estimated using kalman filter.I have
developed the controller (2 i/p-2 o/p) with an MPC. I can see the 4 states in the implemented
MPC, but how do I implement a K.filter required to estimate the states of the 2 unmeasured
states in the physical process. ...help help help...
Attached is the implemented MPC.vi for comments.
Attachments:
MPC controller.zip ‏116 KB

Hi Vicky,
I dont have Labview installed on this laptop, but I have implemented the Labview MPC on a distillation column. You have 2 blocks that are necessary for the MPC, one outside the loop and one inside. You can connect all the necesary settings, like the model, the control horizon, the kalman filter settings,... to the block outside the loop. I think the control for Kalman is called estimation parameters. There you select between no estimation, pole placement and Kalman filter. For the Kalman filter you have to give the controller 3 matrices, that define your Kalman filter, N, Q and R. These matrices are related to your noise model, I don't know the exact theory unfortunately.
regards
KF

Similar Messages

How come the FTTC speed estimation has gone down.

I'm confused as to how the FTTC speed estimation for my phone number has gone down, it's now showing as 33.8 when it was 39.9. Currently I'm sync at 68 and 20 up. The only thing that I can think that possibly has caused the estimation to go down is the fact that I had line issues for a year or so and always got sub 25meg speeds?

Most of the checkers results are now lower it looks like they have updated the checker data from more up to date line data now that there are more fibre users
If you want to say thanks for a helpful answer,please click on the Ratings star on the left-hand side If the reply answers your question then please mark as ’Mark as Accepted Solution’

Custom task (uc_task_custom)? how is the best way to implement intention

Hello all,
i would like to tell you our intention. We want to calculate Capital Charge. Thius means we want to read a couple of records from transactional cube (differing e.g. on item) then do a calculation, then write it back as one record (new item) into the transactional cube. The Custom method is not enough to ful fill the requirements because it can be used just for status monitor updating. Does anyone have a suggestion how to implemet this. Thanks in advance,
Murat

I did not use any BPS or IP planning functions; that might be an approach, but I've never worked with either application. The calculation, in my case, was done in ABAP.
I don't have access to a BCS system at the moment, but I believe there is documentation and an example of how to implment the UC_TASK_CUSTOM BAdI delivered with the system. Check the BAdI in SE18 and see if there is an example class listed. I seem to remember CL_IM_EX_UC_TASK_CUSTOM or something similar.
There is also the documentation in the BAdI which explains exactly which importing/exporting parameters exist. IT_SEL will include things like group currency, period, year, FYV, cons group, company, etc. ET_MESSAGE will determine the cons monitor status when you are finished. You can show additional results on the screen (like a regular BCS task would produce), but you need to write your own ALV. I don't believe it is possible for this list to be available with the "last log" functionality though (I could never find a way, at least).
uc_task_custom method is only to change status monitor
I would disagree. The exporting parameters, as mentioned above, determine the monitor status. But it's also an entry point for you to develop whatever you need to meet your requirements. Need to have a document approval process for manual documents? You can write a custom task for that. Need to post equity pickups back to the GL? You could write a custom task to call an RFC or BAPI in ECC and do that. Need to export results from from the BCS virtual cube to a flat file, basic cube, or an end user's PC? I'd write a custom task. Custom Document? Custom Task.
The issue becomes that greater development effort is required to create a custom task, than to customize a reclassification or an open hub. If the standard functionality does not meet your requirements though, you can either change the requirements or hire a developer to implement them.
- Chris

How works the analogo filter of the 4472?

I need understand how filter the signals the 4472 board.
Thanks

Hello!
That's a pretty broad question, the best thing to do would be to take a look at the 4472 manual on pages 3-4 through 3-9 at the following URL:
http://www.ni.com/pdf/manuals/322940c.pdf
This is an extensive explanation of how the antialiasing filters work, including several graphs of outputs.
Hope this helps!
Brian Spears
Applications Engineer
National Instruments

How does the Java Gernerics being implemented?

Generic programming is supported in Java, and I want to know how does this feature being considered and implemented? Any one can give me some introduction or resources link? Thank you.
Edited by: AaronZhang on Aug 25, 2008 9:37 PM

Thank you ! But I've gotten this tutorial and actaully want to known the implementation mechanism of java generic.

How is the 'code fold' functionality implemented?

Hi all,
I'm thinking that I'd like to have a form, which is broken into several categories, that has 'code fold' functionality ala NetBeans. I'm thinking the way to do this would be to simply place all of these JPanels in the form in the nodes of a tree, when collapsed, just the title is displayed, when they expand, the JPanel appears. I admit I've not yet tried this approach yet , I first wanted to find out if anyone has done something like this and what approach they might use. Any feedback is appreciated.
Thanks
Mike

I read this and thought it would be great to have in my own app.
So, after getting some code from these forums I have cobbled together the following classes.
This one creates a collapsible panel:
import java.awt.BorderLayout;
import java.awt.Color;
import java.awt.Component;
import java.awt.GradientPaint;
import java.awt.Graphics;
import java.awt.Graphics2D;
import java.awt.GridLayout;
import java.awt.LayoutManager;
import java.awt.Rectangle;
import java.awt.event.MouseAdapter;
import java.awt.event.MouseEvent;
import javax.swing.ImageIcon;
import javax.swing.JLabel;
import javax.swing.JPanel;
import javax.swing.border.EmptyBorder;
import javax.swing.border.LineBorder;
* CollapsiblePanel.java Created on 22-Apr-2005
public class CollapsiblePanel extends JPanel
    private class TitlePanel extends JPanel
        private Color fadeto = new Color(200, 212, 247);
        public TitlePanel(LayoutManager layout)
            super(layout);
        protected void paintComponent(Graphics g)
            GradientPaint gradient = new GradientPaint(0, 0, Color.white,
                    getSize().width - 10, getSize().height, fadeto, true);
            Graphics2D g2d = (Graphics2D) g;
            g2d.setPaint(gradient);
            g2d.fill(new Rectangle(getSize().width, getSize().height));
    private JPanel mPanel = new JPanel(new GridLayout(0, 1, 2, 2));
    private JLabel mTitle;
    private JLabel mButton;
    public CollapsiblePanel(String title)
        setLayout(new BorderLayout());
        final Color active = new Color(66, 142, 255);
        final Color inactive = new Color(33, 93, 198);
        final ImageIcon up = new ImageIcon("Up.gif");
        final ImageIcon down = new ImageIcon("Down.gif");
        mTitle = new JLabel(title);
        mTitle.setForeground(inactive);
        mButton = new JLabel(up);
        TitlePanel panel = new TitlePanel(new BorderLayout());
        panel.setBorder(new LineBorder(Color.BLACK));//EmptyBorder(2, 2, 2, 2));
        panel.add(mTitle, BorderLayout.CENTER);
        panel.add(mButton, BorderLayout.EAST);
        panel.addMouseListener(new MouseAdapter()
            public void mouseClicked(MouseEvent e)
                mPanel.setVisible(!mPanel.isVisible());
                mButton.setIcon(mPanel.isVisible() ? up : down);
            public void mouseEntered(MouseEvent e)
                mTitle.setForeground(active);
            public void mouseExited(MouseEvent e)
                mTitle.setForeground(inactive);
        panel.setBackground(Color.WHITE);
        mPanel.setBorder(new EmptyBorder(2, 2, 2, 2));
        //mPanel.setBackground(new Color(214, 223, 247));
        add(panel, BorderLayout.NORTH);
        add(mPanel, BorderLayout.CENTER);
    public Component add(Component comp)
        return mPanel.add(comp);
}This one creates a panel to hold many collapsible panels.
NOTE: This uses the TableLayout manager which can be found on the web. I am sure that all of you here could change it to use GridbagLayout.
import javax.swing.JPanel;
import info.clearthought.layout.TableLayout;
* Panel to hold many CollapsiblePanels
* @author Garry Lovesey
* @version 1.0 22-Apr-2005
public class CollapsiblePanelBar2 extends JPanel
    public static double PREFERRED = TableLayout.PREFERRED;
    public static double FILL = TableLayout.FILL;
    private TableLayout layout;
    public CollapsiblePanelBar2(CollapsiblePanel panel, double layoutPreference)
        super();
         // create the GUI
          // b - border
          // f - FILL
          // p - PREFERRED
          // vs - vertical space between labels and text fields
          // vg - vertical gap between form elements
          // hg - horizontal gap between form elements
          double b= 10;
          double f= TableLayout.FILL;
          double p= TableLayout.PREFERRED;
          double vs= 5;
          double vg= 10;
          double hg= 10;
          double size[][]= { { f }, { // our column sizes
               layoutPreference }}; // our row sizes
          layout= new TableLayout(size);
          setLayout(layout);
          // now we add in our starter panel
          add(panel, "0,0");
    public void add(CollapsiblePanel panel, double layoutPreference)
          // we need to add our panel into our layout
        int rowToAdd = layout.getNumRow();
        layout.insertRow(rowToAdd, layoutPreference);
        add(panel, "0," + rowToAdd);
}

How is the licesning of XI implementation done?

I heard that SAP XI licesning is based on the message volume.
Does any body have any info or pointers on this?

It all depends.....there is tool for sizing in SAP....based on that we can check if XI can handle the load...if not go with 2 servers.....
It does depend on Message volume...when there is lots of load....but normally i never heard any scenario where there were 2 servers....
SAP People will come to your place and analyze the req and give you the solution...........
Coming to licensing part, its not a big deal...
if you have the useroid of market place,u can do the sizing by urself....
https://websmp102.sap-ag.de/SIZING
Edited by: PS Reddy on May 15, 2008 5:05 PM

How to implement State of charge kalman filter algorithm in C code

hi,
I am going to implement kamlan filter algorithm in C code. Is anyone here did this before. Please share your ideas and tell me how i can implement this. Give me some ideas. It would be highly apreciate. thanks in advance
here i attached the kalman filter algorithm file.
regards,
usman
Attachments:
1.docx ‏74 KB

Hi,
did you already have a look at some implementations of the Kalman filter ? For example, that one :
Kalman filter c code implementation. How to tune required variables?
http://forums.udacity.com/questions/1021647/kalman-filter-c-code-implementation-how-to-tune-required...
or that one :
http://alumni.media.mit.edu/~wad/mas864/psrc/kalman.c.txt
Hope it helps!
Aurelie

How to use the size filter in Liquid?

How does the "size" filter work in Liquid if the size of an array is zero. I would like to check if an array's size == 0 or if it is empty. Does the size filter output anything if the array is empty? If not, is there a better way to check if an array is empty? Thanks.
{% if {{lastMinuteOpeningsApp | size}} == 0 %}

{% endif %}
I really wish there were more/better examples on the BC documentation for Liquid.

Hi Dean,
Firstly - You do not need to do the braces inside the if:
{% if {{lastMinuteOpeningsApp | size}} == 0 %}
{% if lastMinuteOpeningsApp | size == 0 -%}
Secondly, If you are accessing web app items, you do not access the webapp as you have.
<pre>{{lastMinuteOpeningsApp | json }}</pre> Will get you the output of the web app liquid object. You will see the items actually inside "items"
So your web app items would but items are not returned if there are no web app items - Have a look. With that you can just say if lastMinuteOpeningsApp.items. Size will only return the maximum number of items you have per pagination limit.

How does the servlet implement the multi-thread feature?

There is only one instance of every servlet in one wep application.
When several clients invoke the doPost() method of the same servlet,the servlet can process their request respectively.
So there much multi threads of one servelt.
But the Servlet doesn't implement the Runnable interface of extends the Thread class.
I wan't to know,how does the servlet/servlet container implement the multi-thread feature?

Hi johnnylzb
There is only one servlet instance existing.
(assuming u have <load-on-startup>1</load-on-startup>)
The server creates a thread for every request ..
and all the threads access the same servlet instance.
For this its not necessary for the Servlet to extend Thread or Runnable.
hope this helps

Needs help in using Discrete Extended Kalman Filter.vi

Dear fellow members,
Currently I am working on a research estimating tremor parameter from patients using Discrete Extended Kalman Filter.
I am from Nanyang Technological Univeristy Singapore.
Now, I face a problem in running the Kalman Filter. The error board keep saying that my E(v) has wrong dimension although I have checked my code several times but not find any mistakes in dimension.
Please advice.
Thank you.
Attachments:
kalman filter.vi ‏23 KB
SIM Discrete Nonlinear Plant Model.vi ‏35 KB
acc_emg.vi ‏41 KB

The related vi(s) were attached

Not able to get current user in PCD Filter factory implementation method

hi all,
In the PCD Filter factory implementation method . i am using the following code
public Object getObjectInstance(Object arg0, Name arg1, Context arg2, Hashtable env)
     throws Exception
          if (Constants.ASPECT_NAVIGATION.equals(env.get(Constants.REQUESTED_ASPECT)))
          IUser user = (IUser) env.get(IPcdContext.SECURITY_PRINCIPAL);
          String department="null";
          if (user != null)
             department=user.getDepartment();
               if ((department != null) && (!department.equals("")))
          filterExpression = "(" + FilterDepartmentService.Department_ATTRIBUTE_KEY + "=" + department + ")";
          return filterExpression;
but i cannot able to get the current user. how to get the current user in the above method..
Regards,
Shanthakumar.

Hi,
First print the output for env.get(IPcdContext.SECURITY_PRINCIPAL);
If it returns null, set it from the request object first (from the method where Hashtable is declared and request obj is available) and then use it in this getObjectInstance method.
env.put(IPcdContext.SECURITY_PRINCIPAL, request.getUser());
Regards,
Harini S

Output of Quad as input of Kalman filter

I am attaching 2 codes; one is a Quad Detector VI and the other a Kalman filter VI. I want to know how can I make the output of Quad Detector as the input of Kalman Filter.
Quad equations: X = [(B+D) - (A+C)] / A+B+C+D ; Y = [(A+B) - (C+D)] / A+B+C+D
Kalman filter equation: Yn = Xn + a * Yn-1 where Xn is the input.
Attachments:
quad.vi ‏66 KB
kalman2.vi ‏35 KB

On the two vi's ... go the symbol in the upper right hand part of the screen which shows an icon for the vi. Right click on it and select "show connector". On the Quad Vi, click one of the right hand connector boxes and then click the output indicator on your Quad front panel. On the other vi...do the same thing except connect one of the left hand connectors and then click on the control which inputs the input data for the vi.
Then in the Quad vi...open the block diagram, and select add a vi from the bottom of the function palet...add the Kalman vi...
right click on the Kalman vi...set the show as icon to off ... that will let you look at all of the i/o for the Kalman filter vi. You should see your Quad input ...connect the dots and run...put whatever outputs you want on the Kalman front panel...put them in loops if you wish...but basically that I think is what you are trying to do.
You could also add the Quad vi as a part of the Kalman filter block diagram...which might actually be a better idea.
Hope that helps.

Kalman filter in Labview

I would like to know if anyone has developped a labview function for adaptive Kalman filtering.
I would not wish to use the MATLAB m-file function given the strict timing-requirements of my application.

The Kalman Filter is available in the NI System Identification toolkit.
I couldn't get my hands on the toolkit so don't know how it works.
Should be intuitive I hope.
Sankar.

How well does website filter work in parental controls?

Hello there,
I'm curious to know how well the adult filter works ok the ios7 built in parental controls. I noticed they have a websites option now to filter different levels of the internet. Does this also work across the entire device in 3 party in app browsers?
Thanks,
LJ

They work pretty good, & are across the device. Read here:
http://support.apple.com/kb/HT4213
Just be aware, crafty kids can figure out a way to bypass pretty much anything they want. Just get in the habit of regularly checking the device.

How is the Kalman filter estimator implemented in an MPC?

Similar Messages

Maybe you are looking for