Method for predicting a characteristic resulting from the swell on a floating system for at least two future time steps

Patent metadata

Field	Value
Publication number	US-10697422-B2
Application number	US-201815965068-A
Country	US
Kind code	B2
Filing date	May 17, 2018
Priority date	Oct 27, 2015
Publication date	Jun 30, 2020
Grant date	Jun 30, 2020

Abstract

The present invention relates to a method allowing short-term prediction of the wave motion (force, elevation, . . . ), from a time series of past wave measurements. The prediction method according to the invention is based on the estimation of the variable coefficients of an autoregressive model while allowing multi-step minimization, that is over a horizon of several time steps in the future, of the prediction error.

First claim

The invention claimed is:

1. A method of predicting an effect of future wave motion on a floating system used to control generation of power by the floating system comprising:
a) measuring at least one resultant characteristic of an effect of wave motion on the floating system for at least one time step;
b) predicting future resultant characteristics of wave motion on the floating system for at least two future time steps by i) constructing at least one autoregressive wave model which relates at least one resultant characteristic of a future time step to the measured at least one characteristic by using time-variable coefficients, ii) determining the time-variable coefficients from use of a random walk model, and iii) determining the future resultant characteristic of wave motion for the at least two future time steps by using of the at least one autoregressive wave model, the determined time-variable coefficients and the measured at least one the characteristic of wave motion from at least one time step; and
c) using the predicted future resultant characteristics of wave motion to control the generation of power which is one of electrical, pneumatic or hydraulic power by the floating system.

2. A method as claimed in claim 1 , wherein the at least one resultant characteristic of wave motion is at least one of force exerted by the waves on the floating system and elevation of waves related to the floating system.

3. A method as claimed in claim 1 , wherein the time-variable coefficients are determined by using at least one of an extended Kalman filter and a linear Kalman filter bank.

4. A method as claimed in claim 2 , wherein the time-variable coefficients are determined by using at least one of an extended Kalman filter and of a linear Kalman filter bank.

5. A method as claimed in claim 1 , wherein the random walk model is expressed by a relationship:
a j (k+1)=a j (k)+η j (k), for allowing calculation of evolution at time step k+1 of each variable coefficient a i of the autoregressive wave model, starting from a value a j (k) at time step k and corresponding stochastic uncertainty η j (k) at time step k.

6. A method as claimed in claim 2 , wherein the random walk model is expressed by a relationship:
a j (k+1)=a j (k)+η j (k), for allowing calculation of evolution at time step k+1 of each variable coefficient a i of the autoregressive wave model, starting from a value a j (k) at time step k and corresponding stochastic uncertainty η j (k) at time step k.

7. A method as claimed in claim 3 , wherein the random walk model is expressed by a relationship:
a j (k+1)=a j (k)+η j (k), for allowing calculation of evolution at time step k+1 of each variable coefficient a i of the autoregressive wave model, starting from a value a j (k) at time step k and corresponding stochastic uncertainty η j (k) at time step k.

8. A method as claimed in claim 4 , wherein the random walk model is expressed by a relationship:
a j (k+1)=a j (k)+η j (k), for allowing calculation of evolution at time step k+1 of each variable coefficient a i of the autoregressive wave model, starting from a value a j (k) at time step k and corresponding stochastic uncertainty η j (k) at time step k.

9. A method as claimed in claim 1 , wherein the autoregressive wave model is expressed with a formula:
ŷ(k|k−1)=x AR (k−1) T a(k), with ŷ(k|k−1) being a predicted characteristic at time step k, x AR (k−1) being a vector of the at least one measured resultant characteristic prior to time step k and a(k) being a vector of the time-variable coefficients of the at least one autoregressive wave model at time step k.

10. A method as claimed in claim 2 , wherein the autoregressive wave model is expressed with a formula:
ŷ(k|k−1)=x AR (k−1) T a(k), with ŷ(k|k−1) being the predicted characteristic at time step k, x AR (k−1) being a vector of the at least one measured resultant characteristic prior to time step k and a(k) being the vector of the time-variable coefficients of the at least one autoregressive wave model at time step k.

11. A method as claimed in claim 7 , wherein the time-variable coefficients a(k) are determined by using an extended Kalman filter and the future resultant characteristic is determined for N future time steps later than time step k, by:
(1) considering time step p=k;
(2) constructing vector x AR (p−1) T of the measured at least one resultant characteristic prior to time p;
(3) determining a characteristic ŷ(p+1|k) for time step p+1 by using a vector x AR (p−1) T and vector a(k) of the time-variable coefficients; and
(4) repeating (2) and (3) for N future time steps.

12. A method as claimed in claim 9 , wherein the time-variable coefficients a(k) are determined by using an extended Kalman filter and the predicted future resultant characteristics are determined for N future time steps later than time step k, by:
(1) considering time step p=k;
(2) constructing vector x AR (p−1) T of the measured at least one resultant characteristic prior to time p;
(3) determining a characteristic ŷ(p+1|k) for time step p+1 by using a vector x AR (p−1) T and vector a(k) of the time-variable coefficients; and
(4) repeating (2) and (3) for N future time steps.

13. A method as claimed in claim 1 comprising for future time steps p:
(1) constructing the autoregressive wave model;
(2) determining the time-variable coefficients of the at least one autoregressive model of the time step p by use of an adaptive Kalman filter bank; and
(3) determining the at least one resultant characteristic by using the at least one autoregressive model of time step p and variable coefficients of the at least one autoregressive model of time step p.

14. A method as claimed in claim 13 , wherein the at least one autoregressive wave model is expressed by a formula:

y
^

⁡

(

k
+
h

|
k

)

=

∑

j
=
1

p

⁢

a

j
,
h

⁡

(
k
)

⁢

y
⁡

(

k
-
j
+
1

)

with ŷ(k+h|k) being a characteristic predicted at time step k+h;
y(k−j+1) being a characteristic measured at time step k−j+1; and
a j,h (k) being time-variable coefficients of the at least one autoregressive model.

Assignees

• Ifp Energies Now

Inventors

• Nguyen Hoai-Nam
• Tona Paolino

Classifications

• G01C13/004

  vertical movement · CPC title

• F03B13/16 Primary

  using the relative movement between a wave-operated member, {i.e. a "wom"} and another member, {i.e. a reaction member or "rem"} · CPC title

• Y02E10/30

  Energy from the sea, e.g. using wave energy or salinity gradient · CPC title

• Y02E10/20

  Hydro energy · CPC title

• F05B2260/8211

  of the weather · CPC title