API Techniques Using ActiveX for Cosimulation Purposes

1
API Techniques Using ActiveX for Co-simulation
Purposes

• Aalhad A. Deshmukh
• Dr. A. Monti, Dr. F. Ponci
• VTB Conference 2004

2
Overview

• Objectives
• Existing Techniques
• Shortcomings
• Introduction to ActiveX
• COM concepts
• Automation
• Types of Access Mechanisms
• ActiveX in VTB

3
Overview

• Implementation of ActiveX
• Network Based Monitoring System
• Advantages
• Conclusion

4
Aim

• To investigate methods and develop Active-X
  interface between VTB Matlab and VTB LAB View

5
Earlier Methods

• API, What is it? Why do we need it?
• Matlab Engine, example FFT
• LABView Wrapper

6
Disadvantages

• Distribution of DLLs
• More development time
• License issues

7
Introduction to ActiveX

• OLE
• - Definition
• - Origin
• Faster software development

8
Component Object Model

• COM Components
• COM objects
• Program identifier
• COM interface

9
COM Interface

• A COM interface is not an object
• It is strongly typed
• They are immutable

10
IUnknown Interface

• QueryInterface( )
• AddRef( )
• Release( )

11
Summary of COM Development

• Interface definition
• Interface implementation
• The Client

12
Automation

• Uses the concept of COM
• Simplifies coding
• Similar Client-Server Relationship

13
Client-Server Interaction

• Server
• -Component class
• -Objects
• -Dispatch Interface
• -Type of Access Mechanism
• Client

14
Type of Access Mechanism

• IDispatch Interface
• VTBL
• Dual

15
Member Functions

• Obtain the dispID of the member functions
• IDispatchGetIDsOfNames ( )
• IDispatchInvoke ( )
• Backbone of OLE Automation

16
Overview of OLE Automation Interface
17
Steps for OLE Automation

• Initialize COM library
• Obtain Class ID from Program ID
• Interface Identifier for IDispatch/Get the Vtbl
  pointer.

18
Steps for OLE Automation

• Obtain the DispID for members of the
  dispinterface
• Invoke methods of the component
• Pass the necessary arguments

19
Implementation

• COM and DDE support in Matlab
• COM Support in LabView

20
Software Interaction

• Signal Coupling
• Natural Coupling
• Time Steps

21
Network Based Monitoring
22
Network Based Monitoring

• A VI collects actual data from a scope
• This data is then pulled into VTB through another
  VI
• Results of the simulation are passed back to a
  third VI
• These results are compared with the actual
  readings taken from the scope

23
Conclusion

• Faster development time
• Distribution of DLLs not required