Actually-measured marine environment data assimilation method based on sequence recursive filtering three-dimensional variation

Patent metadata

Field	Value
Publication number	US-10439594-B2
Application number	US-201415519823-A
Country	US
Kind code	B2
Filing date	Dec 1, 2014
Priority date	Dec 1, 2014
Publication date	Oct 8, 2019
Grant date	Oct 8, 2019

Abstract

The present invention provides an actually-measured marine environment data assimilation method based on sequence recursive filtering three-dimensional variation. The method includes: preprocessing actually-measured marine environment data; calculating a target function value; calculating a gradient value of a target function; calculating a minimum value of the target function; extracting space multi-scale information from the actually-measured data; and updating background field data to form a final data assimilation analysis field. The present invention improves the traditional recursive filtering three-dimensional variation method, and sequentially assimilates information with different scales, thereby effectively overcoming the problem that multi-scale information cannot be effectively extracted by a traditional three-dimensional variation method. A high-order recursive Gaussian filter is used, and a cascaded form of the high-order recursive filter is converted into a parallel structure, so that the recursive filtering process of the recursive Gaussian filter can be executed in parallel, and many problems caused by a cascaded filter are overcome.

First claim

The invention claimed is:

1. A system for determining an underwater marine environment based on sequence recursive filtering three-dimensional variation, comprising:
a plurality of sensors configured to measure and generate actually-measured marine environment data proximate an underwater vehicle;
a computer system for receiving the marine environment data from the plurality of sensors, the computer system comprising a processor, the processor
configured to:
carry out preprocessing and quality control on the actually-measured marine environment data by filtering out invalid data and abnormal data in the actually-measured marine environment data and then converting the actually-measured marine environment data into a format necessary for assimilation;
calculate a target function value by adopting the preprocessed actually-measured marine environment data and background field data by a parallel recursive filter by giving a filtering parameter;
calculate a gradient value of the target function based on the target function value and a source code of a gradient of the target function generated by a tangent and adjoint model compiler;
calculate a minimum value of the target function by inputting the target function value and the gradient value of the target function in a limited Broyden-Fletcher-Glodfarb-Shanno minimization algorithm to calculate the minimum value of the target function, wherein a state variable corresponding to the minimum value is a long wave estimate of observation data;
extract space multi-scale information from the actually-measured data by subtracting the long wave estimate from the actually-measured marine environment data to serve as new actually-measured marine environment data, adjust the filtering parameter to decrease it, replace the actually-measured marine environment data with the new actually-measured marine environment data, and repeat the steps of calculating a target function value, calculate a gradient value of the target function and calculate a minimum value of the target function to sequentially extract the secondary long wave and short wave information from the actually-measured marine environment data; and
update the background field data to form a final data assimilation analysis field.

2. The system of claim 1 , wherein the processor is configured to:
calculate a target function value wherein the target function is defined as

J
⁡

(
w
)

=

1
2

⁢

w
T

⁢
Bw

+

1
2

⁢

(

HBw
-
d

)

T

⁢

O

-
1

⁡

(

HBw
-
d

)

where J represents the target function, w=B −1 (x−x b ), B represents a background field error covariance matrix, x b represents a background field vector, x represents a state variable, O represents an observation field error covariance matrix, H represents an interpolation operator from a mode grid to an observation point, d=y−Hx b , y represents a measurement vector, and the background field is obtained by data assimilation step forward integration by an ocean numerical model;

give an initial estimated value of w by setting the initial estimated value of the w as 0, and indirectly assuming the initial estimated value of the state variable x as the value x b of the background field;
calculate Bw by parallel recursive filtering,
wherein a transfer function S (z) of the parallel recursive filter is defined as:
S ( z )= S 1 ( z )+ S 2 ( z )+ S 3 ( z )+ S 4 ( z )

wherein the parallel recursive filter is a low-order parallel form converted from a four-order van Vliet recursive Gauss filter, wherein, the function S i (z), i=1, 2, 3, 4 is calculated according to the following formula:

S
1

⁡

(
z
)

=

e
1

⁢
z

+

f
1

(

z
-

d
1

)

⁢

(

z
-

d
2

)

S
2

⁡

(
z
)

=

e
2

⁢
z

+

f
2

(

z
-

d
3

)

⁢

(

z
-

d
4

)

S
3

⁡

Assignees

• Zhao Yuxin
• Univ Harbin Eng

Inventors

• Zhao Yuxin
• Liu Chang
• Zhang Xuefeng
• Liu Liqiang
• Li Gang
• Gao Feng
• Li Ning
• Shen Zhifeng
• Zhang Zhenxing
• Qi Zhao

Classifications

• G06F30/00

  Computer-aided design [CAD] · CPC title

• H03H2017/021

  Wave digital filters · CPC title

• G16Z99/00

  Subject matter not provided for in other main groups of this subclass · CPC title

• H03H17/0202

  Two or more dimensional filters; Filters for complex signals (multidimensional convolutions G06F17/153) · CPC title

• H03H17/0282 Primary

  Sinc or gaussian filters (H03H17/0671 takes precedence) · CPC title