07 ni from modeling to system testing with single hardware and software platform as 2010 slo

8/3/2019 07 Ni From Modeling to System Testing With Single Hardware and Software Platform as 2010 Slo

http://slidepdf.com/reader/full/07-ni-from-modeling-to-system-testing-with-single-hardware-and-software-platform 1/19

1

From Modeling to System Testing

with

Single Software-Hardware Platform

Agenda

• Graphical System Design

• -

Model andDesign

SystemTesting

Unified Tool Chain Modeling

System Identification

Control Design

Simulation

• FPGA As a Protot in Technolo

RapidPrototyping

Targeting

Hardware inthe Loop

2

• System Testing Data Management Challenges



2

Customer-DefinedPC-Based Measurement and

Automation Solutions

Traditional Vendor-DefinedInstruments

Origins of Virtual Instrumentation

Processor

RAM

Display

Power Supply

3

ROM Hard Disk

Virtual Instrumentation Architecture

4

• Software Defined Measurement Hardware

• Rapid Development Software

• PC-based Platform With Timing and Synchronization



3

LabVIEW Graphical Programming

5

•

• Development Time Reduction of Four to Ten Times

• Tools to Acquire, Analyze, and Present Your Data

• Built‐in Engineering Libraries

• Easily Connects to Hardware I/O

Speed Development with Graphical Programming

LabVIEW C*LabVIEW C*

6

* does not include code to generate UI



4

Automotive Design – Closed Loop Control

System Test

Rapid

Prototyping

Hardware-in-the-

Loop Validation

Design

Kc Kp

Kc KpKc Kc KpKc

7

Targeting

What Is a Software Model?• Software model is a dynamic representation of

8

Model and Design



5

Demonstration: Mathematical Modeling

Transfer Function (State Space, Zero-Pole-Gain)

Time Domain Differential Equation

• Root Locus

•

9

• MIMO/SISO

System Identification and Control Design in LabVIEW

System Identification

“MIMO s stem identification w/ control desi n in LabVIEW allowed me to ski

Stimulus ResponseSystem

10

first principles modeling. The approach saved so much time that I completed

the entire project in 1/3 the time.”

Dr. George Anwar, Integrated Motions, Inc., NI Alliance Member



6

Simulation Model Conversion

Convert your plant and controller models developed in The MathWorks,

Inc. Simulink® environment into LabVIEW Simulation Module code

11

Simulink® is a registered trademark of The MathWorks, Inc. All other trademarks are the property of their respective owners.

Litt le or No Learning Curve for The MathWorks, Inc. Simulink ®

Software Users

The Simulink Software

Environment

LabVIEW

Simulation Module

12

Simulink® is a registered trademark of The MathWorks, Inc. All other trademarks are the property of their respective owners.



7

Plant Analysis Requirements

• Model construction, conversion,and reduction

• Time and frequency response

• Dynamic characteristics

• Continuous, discrete, stochastic

13

LabVIEW MathScript

• Powerful textual programming fo r signalprocessing, analysis, and math More than 650 built-in functions

Reuse many of your m-file scripts created withThe MathWorks, Inc. MATLAB® software and

others Partially based on original math from

NI MATRIXx

• A native LabVIEW solut ion

14

Interactive and programmatic interfaces

Does not require third-party software

MATLAB® is a registered trademark of The MathWorks, Inc. All other trademarks are the property of their respective owners.



8

Little or No Learning Curve for Customers Familiar

with The MathWorks Inc. MATLAB® Language Syntax

LabVIEW MathScript Syntax MATLAB ® syntax

15

MATLAB® is a registered trademark of The MathWorks, Inc. All other trademarks are the property of their respective owners.

Control Design Requirements

• Classical control design- root locus, PID, lead/lag ...

• -- LQR, LQG, pole placement,Kalman filter ...

• PID Synthesis• Model Predictive Control (MPC),

Adaptive Controlers, non-linear control

16



9

Demo: DC Motor Control NI 9472

DO Module

Ethernet

Crossover

• 9472 Digital Output Module

• QuadratureEncoder

17

Brushed DC Motor

Simulation Requirements

• Linear systems –continuous anddiscrete time

tables

• Fixed-step, variable step, and stiff

solvers• Trimming and linearization

• Model hierarchy

• 3D picture control for system

18

• Graphical and Textual Math



10

Demonstration: LabVIEW Simulation ModuleDemo

a ev ys

LabVIEW System ID

Toolkit

LabVIEW CD&Sim

a mua on o ue

DC Motor

Model

Controller

ModelSpeed

Set oint

Actual

Speed

19

Case Study - Visteon

• TICTV Engine Control

Reference trackin of engine Torque (Tq)

Minimization of Brake

Specific Fuel Consumption(BSFC)

Minimization of CombustionInstabilit CI

20

• NI Software Benefits

Front panel UI

Libraries



11

Algorithm Deployment RapidPrototyping

Demontration: Real-Time Prototyping

• Simulation Module and LabVIEW Real-Time

mpemen conro er on rea- me ar ware

LabVIEW Dev Sys

LabVIEW CD&Sim

LabVIEW Simulation Module

DC Motor

ModelController

Model

Speed

Setpoint ActualSpeed

22



12

Demontration: Real-Time Prototyping

• Simulation Module and LabVIEW Real-Time

mpemen conro er on rea- me ar ware

LabVIEW Dev Sys

LabVIEW CD&Sim

LabVIEW Simulation Module

Controller

Model

Speed

Setpoint

Actual

Speed

AO

Update

AI

Scan

23

Standard Embedded Architecture,

Standard Design ToolLabVIEW Real-Time

target

LabVIEW FPGACOTS I/O

COTS I/O

COTS I/O

CompactRIO

PXI RIO target

LabVIEW uP SDK

target

Custom I/O

24

Single-Board

CompactRIO

Third-

Party

SBC Custom

DesignTargeting



13

FPGA Technology and Logic ImplementationImplementing Logic on FPGA: F = {(A+B)CD}⊕ E

LabVIEW FPGA Code

A

B

C

D

25

FPGA Technology and Logic Implementation

FE

Implementing Logic on FPGA: F = {(A+B)CD}⊕ E

LabVIEW FPGA Code

A

B

C

D

26



14

FE

True Parallelism

A

B

CD

27

FE

True Parallelism

A

B

C

D

28

YW X



15

Analog Control Over 40 kHzControl in Software

• Single PID ~40 kHz

Control in FPGA

• Single PID ~200 kHz

29

Analog Control Over 40 kHz

Control in Software

• Four PID Loops ~25 kHz

Control in FPGA

• Four PID Loops ~200 kHz

30



16

Importance of FPGA in Automotive Systems

31


• High Reliability –Designs become a custom circuit

• High Determinism –Runs algorithms at deterministic rates

down to 25 ns (faster in many cases)

• True Parallelism –Enables parallel tasks

32

• Reconfigurable –Create new and alter existing task-specific

personalities



17

System Testing

Designed Component Needs to be

Tested in Real World:

• In-vehicle Datalogging

• Noise Harshness Vibration Anal sis

33

• Compliance with Safety Requirements

• Crash Tests

System Testing

System Testing - Data Acquisit ion (DAQ)

• Plug-in DAQ

PCI/e

PXI/e

SCXI

SC Express (New!)• Intelligent DAQ

Standalone cRIO

34

System Testing

Ethernet/USB Compact DAQ

Multifunction USB X Series (New!)

Wi-fi



18

The Data Problem

Keep graphics belowthe

text and to the right.

35

System Testing

NI DIAdem

Revolutionary software for

organizing, finding andreporting technical data

36



Summary

• From Modeling to System Testing

• V-Diagram Covered with Unified Toolset:

LabVIEW with Control Design Toolkits

-

37

PXI with 1300+ Modules

DIAdem to Keep Track of Data


• High Reliability –Designs become a custom circuit

• High Determinism –Runs algorithms at deterministic rates

down to 25 ns (faster in many cases)

• True Parallelism Enables arallel tasks

44

• Reconfigurable –Create new and alter existing task-specific

personalities

07 ni from modeling to system testing with single hardware and software platform as 2010 slo

Documents