Ocean Modeling for UUV Path Planning Peter C. Chu Department of Oceanography Naval Postgraduate School Sponsored Jointly by CRUSER ($85,344) and Naval Oceanographic Office ($120k)
Ocean Modeling for UUV Path Planning
Peter C. Chu Department of Oceanography Naval Postgraduate School
Sponsored Jointly by
CRUSER ($85,344) and Naval Oceanographic Office ($120k)
Mul?-‐Disciplinary/ Mul?-‐Ins?tu?onal Project with Par?cipa?on of 4 US Navy
Students Peter Chu, Kwang Song Chenwu Fan
Oceanography NPS
Tom WeKergren Applied Mathema?cs
NUWC-‐ Newport
Ron Betsch Mine Warfare NAVO Peter Fleischer Sedimentology NAVO Frank Bub Ocean Modeling NAVO
Four US Navy METOC Theses • LCDR Paul Ku?a, “Intelligence fused Oceanography for ASW using Unmanned Underwater Vehicles (UUV)” (Secret). MS in Meteorology and Oceanography, March 2013.
• LT Thai Phung, “Analysis of Bioluminescence and Op?cal Variability in the Arabian Gulf and Gulf of Oman for Naval Opera?ons” (Restricted). MS in Meteorology and Oceanography, June 2013.
• LT James Fritz, “Computer Aided Mine Detec?on Algorithm for Tac?cal Unmanned Aerial Vehicle (TUAV)”, MS in Meteorology and Oceanography, December 2013
• LT Mary Doty, “Analysis of Ocean Variability in the South China Sea for Naval Opera?ons” MS in Meteorology and Oceanography, December 2013.
Journal Publica?ons • Chang, Y.-‐C., R.-‐S. Tseng, G.-‐Y. Chen, P. C. Chu, and Y.-‐T. Shen, 2013: Ship rou?ng u?lizing strong ocean currents. Journal of Naviga.on, 60, doi:10.1017/ S0373463313000441
• Chu, P.C., S.E. Miller, and J.A. Hansen, 2013: Fuel-‐Saving Ship Route Using the Navy’s Ensemble Meteorological and Oceanic Forecasts. Journal of Defense Modeling and Simula.on, in press.
ArJficial PotenJal Field Method
PotenJal Field showing two closely space obstacles create an local minima
Stream Line Method
UV-22
" Static Obstacle Avoidance Solutions " Circular Obstacle, Convex hull type " Assume speed and position of Obstacle
∑ =
−−−
⎪⎪
⎭
⎪⎪
⎬
⎫
⎪⎪
⎩
⎪⎪
⎨
⎧
⎟⎟⎠
⎞⎜⎜⎝
⎛+
−+−−
⎟⎟⎠
⎞⎜⎜⎝
⎛+
−+−
−
−⎟⎟⎠
⎞⎜⎜⎝
⎛
−
−+−=
m
k
xkykxk
x
ykykxk
yk
g
g
bbybx
bxa
bbybx
bxa
Cxxyy
mCxymC
1
22
2
22
2
12
11
11
)()()(
)()()(
tantan)(tanψ- Stream function for static obstacles:
UV-22
" Moving Obstacle Avoidance " Moving Obstacle Solutions " Assume that we know speed and position of moving obstacles
∑ =
−−
⎪⎭
⎪⎬⎫
⎪⎩
⎪⎨⎧
−+−
−+−+−+
−+−
−+−+−−⎟⎟⎠
⎞⎜⎜⎝
⎛
−
−+−=
m
k ykykxk
ykxkxkxkxk
ykxk
ykxkyiyk
g
g vbybx
bybxbbyav
bybxbybxbbya
Cxxyy
mCxymC
1 22
222
22
222
21
11
1 )()(})(){()(
)()(})(){()(
tan)(tanψ
- Stream Function for Moving Obstacle
The potential and streamfunction methods are not for the real ocean
11 1
1
2
2 21
2 122 2
tan ( ) tan
( )( ) ( )
tan( )
( ) ( )
gg
ykyk
xk ykmk
xxk
xk yk
ymC mCxy yx x
a x bb
x b y bC
a x b bx b y b
ψ⎛ ⎞⎜ ⎟⎜ ⎟⎜ ⎟⎝ ⎠
⎧ ⎫⎛ ⎞⎪ ⎪⎜ ⎟⎪ ⎪⎜ ⎟⎪ ⎪⎜ ⎟⎪ ⎪⎜ ⎟⎪ ⎪⎝ ⎠⎨ ⎬⎛ ⎞⎪ ⎪⎜ ⎟⎪ ⎪⎜ ⎟⎪ ⎪⎜ ⎟⎪ ⎪⎜ ⎟⎪ ⎪⎝ ⎠⎩ ⎭
− −
−=
=− +−−
−+
− + −−
− +− + −
∑
Analytical Stream Functions:
1 11 1
2 2 2
2 2
2 1 2 2 2
2 2
tan ( ) tan
( ) {( ) ( ) }( ) ( )
( ) {( ) ( ) }( ) ( )
g
g
yiyk xk ykxk
m xk ykk
xk xk xk ykyk
xk yk
y yymC mCx x x
a y b b x b y bv
x b y bC
a y b b x b y bv
x b y b
ψ − −
=
⎛ ⎞⎜ ⎟⎜ ⎟⎝ ⎠
⎧ ⎫⎪ ⎪⎪ ⎪⎪ ⎪⎨ ⎬⎪ ⎪⎪ ⎪⎪ ⎪⎩ ⎭
−= − +
−
− + − + −
− + −−
− + − + −+
− + −
∑
Sta?c Obstacles Moving Obstacles
Gap between Ocean Modeling and UUV Path Planning
Ocean Modeling
Near Real-‐Time Ocean Data
UUV Path Planning
Analy?cal (ar?ficial): Poten?al (φ) Stream Func?on (ψ)
Poten?al and Streamfunc?on from Navy Ocean Model
2 2
, u vy x z x y zψ ϕ ψ ϕ∂ ∂ ∂ ∂
= − + = +∂ ∂ ∂ ∂ ∂ ∂
2 2, v u wx y
ψ ϕ∂ ∂
∇ = − ∇ = −∂ ∂
Ocean Modeling
Near Real-‐Time (u, v, w)
Near Real-‐Time (φ, ψ)
UUV Path Planning
3D Flow Decomposi?on
1 2
3 4
5
3D, Full Physics, Data Assimila.ng, Dynamic, Forecast Models
NOGAPS
Global NCOM
US-‐East NCOM
US-‐East NCOM Global NCOM
Groton DelU3D
Groton NCOM
Page 14
COAMPS
ObservaJons Global Regional Coastal Ocean Modeling
American Seas R-NCOM 96-Hr Series
Surface Temperature
Surface Currents,
Surface Salinity
Surface Elevation
• Same structure / algorithms as GNCOM • Boundary Conditions provided by
GNCOM • FNMOC COAMPS forcing • 3D Forecasts
– T, S, Currents, Elevation – Resolution varies (~1 / 36 deg) – 55 vertical layers – Forecast to 72/96 hr @ 3hr increments
• Assimilates data from – Satellites (SST, SSH) – insitu obs (XBTs, CTDs, floats, buoys)
• First - East China Sea (ECS) NCOM operational MAR08
• Replaced MODAS for ASW support • Implement 3 - 4 regions/year • Eventual transition to COAMPS-OS
(coupled atmosphere—ocean—waves)
Regional: 1/36 deg (3 km / 1.7 nm) Coastal: 50 to 300 m
Navy Coastal Ocean Models (NCOM)
Lagrangian Drift Forecast Examples from AMSEAS-NCOM 30 Day Period
Start with OR&R surface oil coverage in the northern Gulf on 1 June. 30-day Drift uses daily 00Z analyses. No dissipation or wind effects (other than on currents).
Material object drift starting with a line of particles from MS mouth to center of the Loop Current. 30-day persistence forecast using mean currents from 29June10 forecast series.
Material object drift at 1000m (3300ft) starting with a 1-degree disk at the wellhead. 30-day currents from daily 00Z analyses.
Example – Hampton Roads Inlet • Within 3nm of Norfolk Naval Base
• Hampton Roads-‐ world’s leading bulk cargo harbor.
• 150,000 – 500,000 barrels of petroleum in and around Hampton Roads biweekly
• Six military tankers (7.5 million gallon capaciity) homeported in Norfolk
Bathymetry for Hampton Roads Inlet
High-‐Resolu?on NAVO Ocean Model (Delj 3D)
Summary
• Navy tac?cal ocean environmental models and data are very important and useful for the CRUSER program. We will con?nue our efforts to effec?vely incorporated the ocean models in the UUV opera?ons.