Earthquake early warning method based on support vector regression

Patent metadata

Field	Value
Publication number	US-9368017-B2
Application number	US-201314011753-A
Country	US
Kind code	B2
Filing date	Aug 28, 2013
Priority date	Mar 4, 2013
Publication date	Jun 14, 2016
Grant date	Jun 14, 2016

Abstract

An earthquake early warning method for an earthquake detecting system includes utilizing support vector regression (SVR) method to build an earthquake detecting model according to the a plurality of vectors, wherein each of the vectors is corresponding to an earthquake information and a ground motion intensity; detecting a new earthquake information of a new earthquake and generating a specific vector according to the new earthquake information when the new earthquake occurs; and predicting a new ground motion intensity of the new earthquake according to the specific vector and the earthquake detecting model.

First claim

What is claimed is:

1. An earthquake early warning method for an earthquake detecting system comprising:
utilizing a support vector regression (SVR) method to build an earthquake detecting model according to a plurality of vectors, wherein each of the vectors is corresponding to an earthquake information and a ground motion intensity;
detecting new earthquake information of a new earthquake and generating a specific vector according to the new earthquake information when the new earthquake occurs;
predicting a new ground motion intensity of the new earthquake according to the specific vector and the earthquake detecting model; and
disseminating an earthquake warning according to the new ground intensity;
wherein the earthquake information comprises at least one of a peak acceleration, a peak velocity, a peak displacement, a predominate period, a cumulative absolute velocity and integral squared velocity of an earthquake initial wave of each of a plurality of earthquake records;
wherein the new earthquake information comprises at least one characteristic value of a new earthquake initial wave of the new earthquake corresponding to the earthquake information;
wherein the step of predicting the new ground motion intensity of the new earthquake according to the specific vector and the earthquake detecting model is performed according to the following equation:

y
new

=

f
⁡

(

x
new

)

=

[

∑

i
=
1

m

⁢

(

α
i

-

β
i

)

⁢

k
⁡

(

x
new

,

x
i

)

]

+

b
*

;

wherein y new is the new ground motion intensity, x 1 -x m are the plurality of vectors, x new is the specific vector, α 1 -α m , β 1 -β m and b* are coefficients derived according to the support vector regression, the plurality of vectors and the ground motion intensity corresponding to the plurality of vectors, and function k is a kernel function corresponding to a high-dimensional characteristic space;
wherein the kernel function is k(x i ,x j )=exp(−∥x i −x j ∥ 2 /2σ 2 ), where σ is a constant;
wherein the coefficients α 1 -α m , β 1 -β m are acquired via solving a quadratic programming and the quadratic programming is

max

α
,
β

⁢

[

-

1
2

⁢

∑

i
=
1

m

⁢

∑

j
=
1

m

⁢

(

α
i

-

β
i

)

⁢

(

α
j

-

β
j

)

⁢
k
⁢

(

x
i

,

x
j

)

-

ɛ
⁢

∑

i
=
1

m

⁢

(

Assignees

• Nat Applied Res Laboratories

Inventors

• Hsu Ting-Yu
• Huang Shieh-Kung

Classifications

• G08B21/10 Primary

  responsive to calamitous events, e.g. tornados or earthquakes · CPC title

• G01V1/008

  Physics · mapped topic

• G01V1/01

  Measuring or predicting earthquakes · CPC title