Fuzing Validation RAeS Conference Nov 2012 Fuzing... · Fuzing Validation RAeS WS&T Conference...

LAND DEFENCE Thales Proprietary

Fuzing Validation

RAeS WS&T Conference

November 2012

Jason Cowell

2 /


Content

Introduction & BackgroundIntroduction & BackgroundIntroduction & BackgroundIntroduction & Background

Proximity Fuze DevelopmentProximity Fuze DevelopmentProximity Fuze DevelopmentProximity Fuze Development

� Challenges for proximity fuzes

� Advancement in signal process capability

� Principles of operation and applications

Cloud discrimination algorithmsCloud discrimination algorithmsCloud discrimination algorithmsCloud discrimination algorithms

� Cloud data gathering

� Algorithm development

Sea Clutter discriminationSea Clutter discriminationSea Clutter discriminationSea Clutter discrimination

� Algorithm development

� Validation through trials

SummarySummarySummarySummary

Acknowledgement:The work described in this presentation has been sponsored by the UK MoD

3 /


Previous LIDAR based fuzesPrevious LIDAR based fuzesPrevious LIDAR based fuzesPrevious LIDAR based fuzes

� Proven and reliable sensor technology� In-service with several UK missiles

� Existing products employ � Bespoke laser diode and detector components

� Special to purpose ground and polished glass optical components

� Optical alignment mechanisms and associated bonding processes

� Machined metal housings and structural components

� ASICs (Application Specific Integrated Circuits)

� Performance constrained by simple detection & trigger processing logic

� Simple received signal threshold crossing & event counting

� Limited clutter discrimination capability

Introduction

Advanced Short Range Air to Air Missile Prox

Rapier Ground to Air Missile Prox

4 /


Modular Prox Strategy

Modular StrategyModular StrategyModular StrategyModular Strategy

� In the last 5 years TME has developed a new approach to Prox fuze design

� Common Signal Processor module compatible with both future laser and radar sensor technologies

� Associated laser and radar modules designed to be compatible with Common Processor

� Modular family of Prox fuze sensor products to meet a wide range of future applications

� Improved flexibility of algorithmic approach to deal with a range of environments

Dual Mode Options

Common Signal

Processor

Radar RF Transceiver

Modules

+

Laser Transceiver

Modules

+

Radar Sensor Product Family

Laser Sensor Product Family

5 /


Common Signal Processor

Fully developed item (4Fully developed item (4Fully developed item (4Fully developed item (4thththth iteration)iteration)iteration)iteration)

� Low cost components� Dual Channel Analogue to Digital converter� Field Programmable Gate Array � Digital Signal Processor � Read Only Memory

� Timing control & address lines� Multiplexed interface / control of multiple

analogue transceiver modules

Powerful platform for algorithmsPowerful platform for algorithmsPowerful platform for algorithmsPowerful platform for algorithms

� Optimised triggering effectiveness� Enhanced clutter discrimination

� Real time analysis of ‘time history’ of received signal waveforms

� Exploit differences in ‘texture’ between background and target

Wea

pon

Memory

x5 Timing control lines

Externalinterface

Serial data link,Trigger to SAU

ADCTwo

Channel100Ms/s

2x10 Bit

Ch2

Dual Core DSPe.g. AD Black Fin

x2 Address Lines

Ch1

FPGAData Buffering,

Timing & Control

Wea

pon

Memory

x5 Timing control lines

Externalinterface

Serial data link,Trigger to SAU

ADCTwo

Channel100Ms/s

2x10 Bit

Ch2

Dual Core DSPe.g. AD Black Fin

x2 Address Lines

Ch1

FPGAData Buffering,

Timing & Control

φ60mm

φ<2½ inches

6 /


LIDAR Prox Principle of Operation

Concept of OperationConcept of OperationConcept of OperationConcept of Operation

� Multiple fan beam LIDAR sensor� Precise hollow cone overall envelope

� Pulsed laser emissions in each beam

� Detection processing in each beam� Received pulse digitisation & integration

� Pulse time of flight down beam range measurements in each beam

� Clutter discrimination processing� Sea clutter, clouds, smokes, etc

� Based on time history of pulse returns

Nose on view

1

32

45

8

7

6Side view

7 /


Demonstration Unit with Visible Laser Diodes

Red laser diodes fitted to illustrate beam geometryRed laser diodes fitted to illustrate beam geometryRed laser diodes fitted to illustrate beam geometryRed laser diodes fitted to illustrate beam geometry

� Near IR diodes used in real sensor

8 /


Performance Validation Strategy

Performance validation strategyPerformance validation strategyPerformance validation strategyPerformance validation strategy

� To avoid expensive weapon level trials the fuze per formance predictions are based on modelling results

� Fuze mathematical model is validated through cost e ffective trials� Sensor characterisation� Signatures of targets

� Signatures of background

� Signatures of clutter

� Algorithm development utilises the trials data to d evelop and then prove algorithm (detection & rejection of false ala rms) performance

� Model vast numbers of engagement scenarios in Monte -Carlo assessment

� Range of missile engagement geometries

� Range of target geometries� Range of environmental conditions

� Range of clutter or background condition

9 /


Effects of cloud on Air target Fuze

Cloud impact on Prox fuze performanceCloud impact on Prox fuze performanceCloud impact on Prox fuze performanceCloud impact on Prox fuze performance

� Historically radar fuzes have been used which are b roadly unaffected by clouds

� Simple detection processing for laser fuzes can result in false alarms

� Improvements in signal processing techniques provid e opportunityfor enhanced clutter discrimination for laser Prox fuzes

� Attenuation of signal not major issue over short ra nges

� Aerosols provide backscatter of laser� Cloud “edges” provide change in signal level that could cause false alarm

� When entering or leaving cloud (at interface)

� Robust algorithmic approach required for laser fuze s that rejectsignals from cloud to avoid false alarms

10 /


Cloud Clutter Data Gathering

Prototype Laser Prox modulePrototype Laser Prox modulePrototype Laser Prox modulePrototype Laser Prox module

� Integrated within sealed enclosure� Laptop data logging of reflected pulses

� Secured against open map window� Uninterrupted ‘view’ of cloud

Data recording in dense cloudData recording in dense cloudData recording in dense cloudData recording in dense cloud

� Cumulus, Cumulonimbus� In cloud, entering cloud, exiting cloud

� Pulse rate scaled for aircraft velocity

31

24

31

24

Experimental 4 Fan beam

Laser Module

11 /


Time during flight (s) Digitised Pulse Waveform

Am

plitu

de

Sector 4


Am

plitu

de

Sector 3


Am

plitu

de

Sector 2


Am

plitu

de

Sector 1

Cloud Backscatter Pulse Signatures Gathered

31

24

31

24

12 /


Cloud Clutter Discrimination Algorithms

Algorithms can analyse backscattered pulse waveform ‘history’Algorithms can analyse backscattered pulse waveform ‘history’Algorithms can analyse backscattered pulse waveform ‘history’Algorithms can analyse backscattered pulse waveform ‘history’

� Determination of cloud presence and point of entry/ exit� Exploitation of received signal amplitude rate of change in multiple sectors� Valid targets not falsely detected as cloud

� Cloud backscatter suppression techniques� Slowly varying cloud backscatter waveforms can effectively be ‘subtracted’� Residual signals can still occur at cloud edges

� Adaptive threshold techniques to fully suppress clo ud detections

Green – Raw Pulse AmplitudeRed – Amplitude Post Suppression

Effective cloud clutter discrimination algorithms demonstrated

Residual post suppression spike

13 /


Small targets in sea clutter

Threat proximity to sea surface a challenge for the fuzeThreat proximity to sea surface a challenge for the fuzeThreat proximity to sea surface a challenge for the fuzeThreat proximity to sea surface a challenge for the fuze

� Sea skimming missiles close to sea clutter

� Fast Inshore Attack Craft (FIACs) embedded in sea c lutter

Clutter reflections difficult to differentiate from targetClutter reflections difficult to differentiate from targetClutter reflections difficult to differentiate from targetClutter reflections difficult to differentiate from target

� Can be similar range and amplitude

Analysis of fuze performance requires representative Analysis of fuze performance requires representative Analysis of fuze performance requires representative Analysis of fuze performance requires representative models of sensor interaction with the sea surfacemodels of sensor interaction with the sea surfacemodels of sensor interaction with the sea surfacemodels of sensor interaction with the sea surface

Boston Whaler with rocket launcher

Sea Skimming

Missile

14 /


Laser Sensor Interaction with the Sea

Operating at near IR wavelength (Operating at near IR wavelength (Operating at near IR wavelength (Operating at near IR wavelength (λλλλ~0.9~0.9~0.9~0.9µµµµm)m)m)m)

� Imaginary component of refractivity (k) very small

� Bulk absorption high hence volume backscatter can b e ignored

� Real component of refractivity (n) ~1.33 can be use d to estimatesurface reflectivity ( ρ) using Fresnel

� Only incident angles close to normal are of interes t� Small sensor bistatic angle

� Fresnel equations simplify

� Reflectivity ~2%

( )( ) 02.0

11

2

≈

+−=

nnρ

Sea surface

Lidar with low bistatic angle

15 /


Laser Sensor Interaction with the Sea

Active IR (laser) sensor response to sea ‘intermittent’Active IR (laser) sensor response to sea ‘intermittent’Active IR (laser) sensor response to sea ‘intermittent’Active IR (laser) sensor response to sea ‘intermittent’

� Sea surface behaves like a rippled mirror with a 2% reflectivity� Strong reflection if surface elements intersect beam near normal

� Very low response if illuminated surface not close to normal

� Response depends upon complex geometry of beam and rippled shape of sea surface

� White caps can present a diffusely scattered signat ure� Detected over a broad range of illumination angles

Sea detection unreliable

But detection of white caps likely

But detection of white caps likely

Sea detection unreliable

Sea detection may be reliable enough to track

16 /


Modelling the Sea Surface

Sea surface modelled as an array of small 2% reflectorsSea surface modelled as an array of small 2% reflectorsSea surface modelled as an array of small 2% reflectorsSea surface modelled as an array of small 2% reflectors

� Contiguous surface comprising non planar facets� 5mm x 5mm (or smaller)

� Arranged to represent 3D geometry of sea surface

� Model shares origins with existing radar TDD intera ction model� Smaller facets due to much shorter wavelength (~1µm versus ~10cm)

� 64bit PC with large memory capacity used to run analyses (slowly)

Fuze sensor interaction modelFuze sensor interaction modelFuze sensor interaction modelFuze sensor interaction model

� Contiguous surface comprising non planar facets

� Multiple fan beam geometry modelled

� Defined engagement trajectories

� Intersection of beams with 3D sea model

� ‘Pulse by pulse’ response modelled

� Summation of reflected pulse components from multiple facets computed

17 /


Modelling the Sea Surface

Model uses wave spectrum proposed by Model uses wave spectrum proposed by Model uses wave spectrum proposed by Model uses wave spectrum proposed by ElfouhailyElfouhailyElfouhailyElfouhaily

� Both gravity & surface capillary waves modelled

� Capillary waves (e.g. λλλλ<25mm) significant at laser wavelengths

Parameters adjusted to vary Parameters adjusted to vary Parameters adjusted to vary Parameters adjusted to vary

sea conditionssea conditionssea conditionssea conditions

� Fetch

� Wind speed & Direction

� Resolution (e.g. 5mm)

� Patch Size

Wide variety of sea Wide variety of sea Wide variety of sea Wide variety of sea

conditions modelledconditions modelledconditions modelledconditions modelled

� Case shown a 80m by 80m patch, 12m/s wind, 500km fetch

18 /


Modelling Sensor Response to the Sea

Sea surface modelled as a regular grid of heightsSea surface modelled as a regular grid of heightsSea surface modelled as a regular grid of heightsSea surface modelled as a regular grid of heights

� Height at each vertex derived using the Elfouhaily spectrum

� Characteristics of each element calculated from adj acent vertices� Normal vector of each element� Radii of curvature in two orthogonal axes

Intersection of beams with gridIntersection of beams with gridIntersection of beams with gridIntersection of beams with grid

� Shot lines calculated to each element

� Occurrences of surface normals found

� Incremental contributions to pulse responses determined from;

� Sensor parameters (e.g. power, etc)

� Element radii of curvature

� Repeated at Pulse Repetition Rate

19 /


Model Validation – Sea Data Gathering

Pulsed laser sensorPulsed laser sensorPulsed laser sensorPulsed laser sensor

� Narrow beam width <1°

� Sensitivity calibrated

Mounted on bows of vesselMounted on bows of vesselMounted on bows of vesselMounted on bows of vessel

� Beam viewing sea surface ahead of wake

� Adjustable pitch & roll angles

� Adjustable height

� Vessel speed ~13 knots

� Wind speed/bearing recorded

Threshold crossings recordedThreshold crossings recordedThreshold crossings recordedThreshold crossings recorded

� Fair correlation with model

� Provided initial validation

Initial Pencil Beam Laser Sensor TrialsInitial Pencil Beam Laser Sensor TrialsInitial Pencil Beam Laser Sensor TrialsInitial Pencil Beam Laser Sensor Trials

20 /


Model Validation – Sea Data Gathering

Experimental form of future TDDExperimental form of future TDDExperimental form of future TDDExperimental form of future TDD

� Four 30°contiguous fan beams

� Partial azimuth coverage (only downward beams see reflections)

� Received pulse waveforms digitised

� Data recorded for various sensor orientations and sea conditions

Multiple Fan Beam Laser Sensor TrialsMultiple Fan Beam Laser Sensor TrialsMultiple Fan Beam Laser Sensor TrialsMultiple Fan Beam Laser Sensor Trials

31

24

Pulse count (x10 4)Time (ns)

Example of Sector 3 pulse

responses

21 /


Model Applications – Anti FIAC Algorithms

Algorithm developmentAlgorithm developmentAlgorithm developmentAlgorithm development

� Sea clutter rejection

� Reliable target detection

� Initial algorithms constructed and tested

� Initial results encouraging

� Validation in progress

FIAC targets modelledFIAC targets modelledFIAC targets modelledFIAC targets modelled

� 3D facet models� Diffuse Lambertian reflectors

� Embedded in sea clutter models

� Various dive angles modelled

� Combined response to target and clutter modelled

22 /


Sea Data Gathering Trial

Sensor deployed on boom to one side of vesselSensor deployed on boom to one side of vesselSensor deployed on boom to one side of vesselSensor deployed on boom to one side of vessel

Rib ‘target’ travelling at speed under / to one side of sensorRib ‘target’ travelling at speed under / to one side of sensorRib ‘target’ travelling at speed under / to one side of sensorRib ‘target’ travelling at speed under / to one side of sensor

� Provides representative wake data

� Data to be used for validating models and developin g algorithms

31

24

15 knots

30 knots

23 /


Recent ‘AFIAC’ Sea Data Gathering Trial

24 /


Received Pulse Signatures Gathered

31

24

31

24

Target

Sporadic Sea

reflections


Am

plitu

de

Sector 1


Am

plitu

deTime during flight (s) Digitised Pulse Waveform

Am

plitu

de

Time during flight (s) Digitised Pulse WaveformA

mpl

itude

Wake Wake Target

25 /


Laser TDD FIAC Detection Algorithms

Algorithms can analyse Algorithms can analyse Algorithms can analyse Algorithms can analyse received pulse waveformsreceived pulse waveformsreceived pulse waveformsreceived pulse waveforms

� Valid targets consistently detected in multiple fan beams

� Simultaneous sea detections very rarely occur in multiple sectors

� Reflected power variability� Sea reflections sporadic

� Pulse amplitude from sea varies significantly from pulse to pulse

� Target signatures more consistent

� Range measurements� Available from each TDD sector

� Range step onto target evident if the sea is being tracked beforehand

� Valid targets exhibit greater range variability and rate than the sea

Sensor (and camera) rolled so that sector 3 looks down

Effective sea clutter discrimination algorithms demonstrated

26 /


Model Applications – Anti FIAC Algorithms

Example Model Output Example Model Output Example Model Output Example Model Output –––– Case of Horizontal TrajectoryCase of Horizontal TrajectoryCase of Horizontal TrajectoryCase of Horizontal Trajectory

Sea Surface Only Target Present

First target facet enters

beam

Target detection

Sector 1 - Red

Sector 2 – Green

Sector 3 – Cyan

Sector 4 – Magenta

27 /


Real-time algorithm

Exploits difference in reflectivity Exploits difference in reflectivity Exploits difference in reflectivity Exploits difference in reflectivity characteristics of undisturbed sea, and characteristics of undisturbed sea, and characteristics of undisturbed sea, and characteristics of undisturbed sea, and target / waketarget / waketarget / waketarget / wake

Two parallel processing pathsTwo parallel processing pathsTwo parallel processing pathsTwo parallel processing paths

� Tracking of sea surface� Detection of “bumps”

� Range gradient estimation� Detection of sudden changes in range, which

may correspond with a target profile

Comparison with “baseline” fuze modes to Comparison with “baseline” fuze modes to Comparison with “baseline” fuze modes to Comparison with “baseline” fuze modes to determine possible benefitdetermine possible benefitdetermine possible benefitdetermine possible benefit

Wake

Target

For each Integration

Calculate range

gradient

Sea Tracking

Looking for sudden bump in the sea surfaceRange

Detection Criteria

Detection Criteria

Target Detected

Dive Indication

Dive Estimate

Missile characteristics

model Missile speed wrto time

Data from the CAN (if available)

28 /


Summary

A fuze sensor model for the response of a multiple fan beam laseA fuze sensor model for the response of a multiple fan beam laseA fuze sensor model for the response of a multiple fan beam laseA fuze sensor model for the response of a multiple fan beam laser fuze has r fuze has r fuze has r fuze has been developed along with sea surface and cloud discrimination abeen developed along with sea surface and cloud discrimination abeen developed along with sea surface and cloud discrimination abeen developed along with sea surface and cloud discrimination algorithmslgorithmslgorithmslgorithms

Cost effective data gathering trials have provided the evidence Cost effective data gathering trials have provided the evidence Cost effective data gathering trials have provided the evidence Cost effective data gathering trials have provided the evidence to validate to validate to validate to validate these models these models these models these models

� Received signal levels estimated by the model compa re favourably with those of the trials data

� The predicted variability of the signal returns fro m the sea appears to be confirmed by the trials

Facility to embed targets in sceneFacility to embed targets in sceneFacility to embed targets in sceneFacility to embed targets in scene

� e.g. FIACs and sea skimming missiles

� Supports the development of a Laser Fuze sensor for Anti FIAC and anti Sea Skimmer missile applications

Research has demonstrated the feasibility of enhanced AFIAC fuziResearch has demonstrated the feasibility of enhanced AFIAC fuziResearch has demonstrated the feasibility of enhanced AFIAC fuziResearch has demonstrated the feasibility of enhanced AFIAC fuzingngngng

� Smaller FIACs, embedded in high sea clutter, over a broader miss distribution

� Current phase is to embed algorithms into hardware for trials and then into production hardware (March 2013)

29 /


Any Questions ?Any Questions ?Any Questions ?Any Questions ?

Fuzing Validation RAeS Conference Nov 2012 Fuzing... · Fuzing Validation RAeS WS&T Conference...

Documents

Transcript of Fuzing Validation RAeS Conference Nov 2012 Fuzing... · Fuzing Validation RAeS WS&T Conference...