+ All Categories
Home > Documents > GPS/INS Computing System

GPS/INS Computing System

Date post: 23-Feb-2016
Category:
Upload: duane
View: 27 times
Download: 0 times
Share this document with a friend
Description:
GPS/INS Computing System. Performed by: Alexander PavlovDavid Domb Supervisor: Mony Orbach. Project Characterization Spring 2008/9. Agenda. 1. General overview. 2. Our Project. 3. Working environment. 4. Design Solution. 5. Timeline. General overview. - PowerPoint PPT Presentation
29
Company LOGO Project Characterization Performed by: Alexander Pavlov David Domb Supervisor: Mony Orbach GPS/INS Computing System
Transcript
Page 1: GPS/INS Computing System

Company

LOGO

Project CharacterizationSpring 2008/9

Performed by:Alexander Pavlov David Domb

Supervisor:Mony Orbach

GPS/INS Computing

System

Page 2: GPS/INS Computing System

Agenda

1. General overview

2. Our Project

3. Working environment

5. Timeline

GPS/INS Computing System

4. Design Solution

Page 3: GPS/INS Computing System

GPS/INS Computing System

General overview

“Even Noah got no salary for the first six months partly on account of the weather and partly because he was learning navigation.”

Mark Twain

Page 4: GPS/INS Computing System

GPS/INS navigation

INS Characteristics

• Self-contained• Provides accurate

position and velocity over short time periods but slowly drifts over time

• More expensive and heavier than GPS

GPS Characteristics

• Relies on GPS satellites: susceptible to jamming, RF interference, multipath and integrity problems

• Provides accurate position and velocity over longer time periods but has high frequency noise

By combining the outputs of a GPS and an INS, the performance issues of both systems can be remedied.

Both Global Positioning System (GPS) and Inertial Navigation System (INS) has its advantages and disadvantages.

GPS/INS Computing System

Page 5: GPS/INS Computing System

Tightly Coupled INS/GPS

GPS/INS Computing System

IMU

Strap down Inertial

Navigation Algorithm

State Update

FilterInnovation CalculationGPS

∆Vm

∆ θm

PI

VI

qI

PINS/GPS

VINS/GPS

qINS/GPS

PSAT

meas

P V q

b

SDINS

.

Page 6: GPS/INS Computing System

Kalman & Particle filter• Standard filter used in navigation systems is

extended Kalman filter (EKF)• Disadvantages of the extended Kalman filter:

– EKF is not an optimal estimator for non-linear systems

– Optimized for statistical noise only• Particle filter can be used as an alternative to EKF,

these to improve estimation's accuracy.• Particle filters is a sophisticated model estimation

technique based on simulation using sufficient number of samples

GPS/INS Computing System

Page 7: GPS/INS Computing System

Theoretical Solution

0 • Initialization1 • Particle Propagation2

• Particle Update & Normalization

3 • State Estimation4 • Effective N calculation5 • D computation6 • Re-sampling7 • Regularization8 • Weight Re-computation

GPS/INS Computing System

Implementing the tightly coupled INS/GPS navigation unit with the particle filter, according to algorithm developed in Technion.

The theoretical algorithm stages:

Page 8: GPS/INS Computing System

Project Goals

Establishing the efficiency of the particle filter based, tightly

coupled INS/GPS navigation unit realization.

Designing an efficient real-time particle filter based, tightly

coupled INS/GPS navigation unit.

GPS/INS Computing System

Page 9: GPS/INS Computing System

GPS Computing System

Our Project

Page 10: GPS/INS Computing System

General• Project will be performed in 2

stages. First part in this semester.• Project will be performed by

several work groups• Our group will implement Particle

Propagation and State Estimation stages in this first part.

• Both stages need to be performed each 0.01 sec, regardless of other stages performance.

GPS Computing System

Page 11: GPS/INS Computing System

Group Project Goals – PART 1Learning GPS/INS navigation using Particle

Filter algorithm

Learning VHDL language

Learning FPGA environment

Implementation of Particle Propagation and State Estimation stages of algorithm

GPS/INS Computing System

Page 12: GPS/INS Computing System

GPS/INS Computing System

WorkingEnvironment

Page 13: GPS/INS Computing System

Gidel PROCStar II

GPS/INS Computing System

• Up to 4 ALTERA Stratix II 60 to 180 FPGAs• Five level memory structure (over 2.5GB)• Typical system frequencies: 100-300 Mhz. • Flexible clocking system. • Up to 695 available I/Os.• Up to 5 PSDBs (ProcStar II Daughter Boards):

  Camera Links, machine I/Os and other interfaces. • Expandable system: up to 96 DDR II I/Os between

  PROCStar II boards. • Up to 660 Gbits per second connectivity between FPGAs.

Page 14: GPS/INS Computing System

Altera Stratix II

GPS/INS Computing System

15,600 to 179,400 equivalent Logic elements Adaptive logic module (ALM), maximizes performance

and resource usage efficiency Up to 9,383,040 RAM bits High-speed DSP blocks provide dedicated

implementation of multiply-accumulate functions. Up to 12 PLLs (four enhanced

PLLs and eight fast PLLs) per device.

Support for high-speedexternal memory

Megafunctions support

Page 15: GPS/INS Computing System

Altera Quartus IIProvides a multiplatform design environment for all phases of FPGA design.

GPS/INS Computing System

Page 16: GPS/INS Computing System

GPS/INS Computing System

Design solutions

Page 17: GPS/INS Computing System

State Vector - X[1..18]PositionX[1..3]

VelocityX[4..6]

QuaternionX[7..10]

Accelerometers offset X[11..13]

Gyroscope driftX[14..16]

GPS clock errorX[17..18]

GPS/INS Computing System

, ,h

, ,N E DV V V

1 2 3 4, , ,q q q q

, ,X Y Z

, ,X Y Z

,b b

Page 18: GPS/INS Computing System

Design guidelinesConstrains: large amount of calculations Limited hardware real-time resultsPossible solutions:PipeliningLarge amount of parallel calculation unitsSelected solution:Max. Parallel processingWhat can’t be parallel – will be Pipelined.

GPS/INS Computing System

Page 19: GPS/INS Computing System

Solution – Top design

GPS/INS Computing System

Weight vector

Particles propagation

unit

State estimation

unit

Estimated State Vector

]1..18[

xN Extended State Vector

]1..18[

Extended State Vector

]1..18[

Extended State Vector

]1..18[

Controller

Page 20: GPS/INS Computing System

Parallel VS. Pipeline

GPS/INS Computing System

Considerations: Max. parallel processes will result in Min.

calculation time. Number of parallel processes is limited by

the hardware. Not all calculations have to be parallel in

order to comply with 100 Hz.

Page 21: GPS/INS Computing System

Parallel VS. Pipeline

GPS/INS Computing System

Conclusions: The number of L.E.’s (available on the

FPGA) will determine the number of parallel processes in the “Particle propagation unit” and the “State estimation unit”.

To complete this number to N, we will pipeline the processes in those units.

Page 22: GPS/INS Computing System

Particles Propagation block

GPS/NS Computing System

Single particle

Propagationunit

Single particle

Propagationunit

Single particle

Propagationunit

1 particle

1 particle

1 particle

1 particle

1 particle

1 particle

After Propagation

After Propagation

After Propagation

# units

Page 23: GPS/INS Computing System

1 Particle Propagation block

GPS/INS Computing System

Propagationunit After

Propagation

Propagation

unit After Propagation

Propagationunit After

Propagation

hh

Propagationunit After

Propagation

NVNV

h

NV

Propagationunit After

Propagation

Propagationunit After

Propagation

Propagationunit After

Propagation

q

Propagationunit After

Propagation

CLK

EV EVEV

DVDV DV

qq

CLKCLK

Page 24: GPS/INS Computing System

State estimation block

GPS/INS Computing System

Estimationunit

Estimated # Weightesq

# q

#q#q

# other componentsEstimation

unit

# other Estimated

components# other

Weightes

# other components

Page 25: GPS/INS Computing System

q estimation block – N units

GPS/INS Computing System

# units

computationunit After

estimation

1q

1qW1qM

1qM

1q

estimationunit

1q

1q

computationunit After

estimation

2q

2qW2qM

2qM

2q

estimationunit

2q

2q

computationunit After

estimation

iq

qiWqiM

qiM

iq

estimationunit

iq

iq

#q

Page 26: GPS/INS Computing System

Not q components estimation block

GPS/INS Computing System

Multiplyunit

1

1W1 1W

Multiplyunit

2

2W2 2W

# units

Multiplyunit

#

#W# #W

Adder

X14 – for every component other then q

Estimated

Page 27: GPS/INS Computing System

Growth capability  

GPS/INS Computing System

Each design unit, deals with a number of particles.

The basic calculations in each unit, are designed for one particle and is then multiplied.

The same design can be implemented with “bigger” FPGAs, by increasing the number of multiplications and parallel processes.

This can result in lesser pipelines which means faster realization.

It can also implement bigger N.

Page 28: GPS/INS Computing System

GPS/INS Computing System

Timeline

Page 29: GPS/INS Computing System

GPS/INS Computing System

GANTT – PART A

ID Task Name 2009 2009 2009 2009 2009

15/3 22/3 29/3 5/4 12/4 19/4 26/4 3/5 10/5 17/5 24/5 31/5 7/6 14/6 21/6 28/6 5/7 12/7 2/826/7

1 Theoretical background learning

2 Working environment learning

3 Initial Characterization

4 Characterization presentation

5 DESIGN

19/7

6

7

9

Middle report

More DESIGN

Tests and Simulations

10

11

Final report preperation

Final report

8 Realization


Recommended