Wind Power Forecasting using an Artificial Neural Network for ASPCS

doi:10.4236/epe.2013.54B080

Paper Menu >>

Journal Menu >>

Energy and Power Engineering, 2013, 5, 414-417

doi:10.4236/epe.2013.54B080 Published Online July 2013 (http://www.scirp.org/journal/epe)

Wind Power Forecasting using an Artificial Neural

Network for ASPCS

Kazuma Hanada1, Takataro Hamajima1, Makoto Tsuda2, Daisuke Miyagi2, Takakazu Shintomi3,

Tomoaki Takao4, Yasuhiro Makida5, M as at ak a Kajiwara6

1Department of Electrical and Electronics Systems, Hachinohe Institute of Technology, Aomori, Japan

2Department of Electrical Engineer ing, Tohoku University, Miyagi, Japan

3Advanced Research Institute for the Sciences and Humanities, Nihon University, Tokyo, Japan

4Department of Engineering and Applied Sciences, Sophia University, Tokyo, Japan

5High Energy Accelerator Research Organization, Ibaraki, Japan

6Iwatani Corporation, Tokyo, Japan

Email: [email protected]

Received April, 2013

ABSTRACT

In order to use effectively renewable energy sources, we propose a new system, called Advanced Superconducting

Power Conditioning System (ASPCS) that is composed of Superconducting Magnetic Energy Storage (SMES), Fuel

Cell-Electrolyzer (FC-EL), hydrogen storage and DC/DC and DC/AC converters in connection with a liquid hydrogen

station for fuel cell vehicles. The ASPCS compensates the fluctuating electric power of renewable energy sources such

as wind and photovoltaic power generations by means of the SMES having characteristics of quick response and large

Input-Output power, and hydrogen energy with FC-EL having characteristics of moderate response and large storage

capacity. The moderate fluctuated power of the renewable energy is compensated by a trend forecasting method with

the Artificial Neural Network. In case of excess of the power generation by the renewable energy to demand it is con-

verted to hydrogen with EL. In contr ast, sh or tage of the electr ic pow er is made up with FC. The f aster fluctu ation power

that cannot be compensated by the forecasting method is effectively compensated by SMES. In the ASPCS, the SMES

coil with an MgB2 conductor is operated at 20 K by using liquid hydrogen supp lied from a liquid hydrogen tank of the

fuel cell vehicle station. The necessary storage capacity of SMES is estimated as 50 MJ to 100 MJ depending on the

forecasting time for compensating fluctuation power of the rated wind power generation of 5.0 MW. As a safety case, a

thermo-siphon cooling system is used to cool indirectly the MgB2 SMES coil by thermal conduction. In this paper, a

trend forecasting result of output power of a wind power generation and the estimated storage capacity of SMES are

reported.

Keywords: Renewable Energy; Artificial Neural Network; Forecasting; SMES; Liquid Hydrogen; Fuel Cell and

Electrolyzer

1. Introduction

The applications of renewable energy are needed in the

context of climate change and energy resource. But,

when renewable energy generations, such as wind and

photovoltaic power generations, affected by weather

condition are introduced in large quantities, an existing

power system may become unstable. Therefore, we pro-

pose a new system shown in Figure 1, called Advanced

Superconducting Power Conditioning System (ASPCS)

that is composed of Superconducting Magnetic Energy

Storage (SMES), Fuel Cell-Electrolyzer (FC-EL), hy-

drogen storage and DC/DC and DC/AC converters in

connection with a liquid hydrogen station for fuel cell

vehicles [1-3].

AC/DC

DC/DC

DC/AC Utility

Grid

DC/AC ACLoad

DCLoad

DC/DC

カードル

Comp.

貯蔵

ディスペンサ

熱交換器

Comp.

Ref.

MgB2

SMES

Vacuum

Chamber

LH2間接冷却

LH2 ローリー

Controller

（含未来予

測技術）

5MWクラス自然エネルギー源

1MWクラスハイブリッド貯蔵システム

商用電力系統

液体水素ステーション

BUS

28 kl

水素系信号系電気系

CompressedH

Dispenser

(Prediction

Technology)

Heat

Exchanger

Line ElectricSignal

Ta nk

Ta nk er

StationforVehicl e

1MW‐classHybridStorageSystem

5MWRenewableEnergyResou rces

UtilityGrid

Indirectcooling

Figure 1. Advanced Superconducting Power Conditioning

System.

K. HANADA ET AL. 415

The ASPCS compensates the fast fluctuating electric

power of the renewable energy sources by means of the

SMES having characteristics of quick response and large

Input-Output power, and hydrogen energy with FC-EL

having characteristics of moderate response and large

storage capacity. The moderate fluctuated power of the

renewable energy is compensated by a trend forecasting

with the Artificial Neural Network (ANN) [4, 5 ]. In case

of excess of the power generation by the renewable en-

ergy to demand it is converted to hydrogen with EL. In

contrast, in case of shortage the electric power is made

up with FC. The faster fluctuation that cannot be com-

pensated by the forecasting method is effectively com-

pensated by SMES. In the ASPCS, the SMES coil with

an MgB2 conductor is operated at 20 K by using liquid

hydrogen supplied from a liquid hydrogen tank of the

fuel cell vehicle station. The necessary storage capacity

of SMES is estimated as 50 MJ to 100 MJ depending on

the forecasting time for compensating fluctuation of the

rated wind power generation of 5.0 MW. As a safety, a

thermo-siphon cooling system is used to cool indirectly

the MgB2 SMES coil by thermal conduction.

In this paper, a forecasting result of output power of a

wind power generation and the necessary storage capac-

ity of SMES are reported.

2. Artificial Neural Network

ANN is a model of biological neural network, and is

composed of a lot of artificial neurons that collect output

signals through a transfer function of Equation (1) when

they are received input signals.

( )tanh( )

fxx ee





 (1)

The ANN shown in Figure 2 is a layered ANN. The

layered ANN is formed three layers; an input layer that

receives input signal, an output layer that produces out-

puts signals and a hidden layer that makes processes of

signals.

Weights are set in connection between neurons. The

outputs of each neuron inputs into other neurons after

Inputlayer HiddenlayerOutputlayer

Inputdata

Outputdata

Figure 2. Layered ANN.

they are multiplied by the weights. Sin ce the weights are

usually initialized with random values, they are adjusted

to get desirable outputs. This is a learning process. A

back propagation method that is one of the learning me-

thods of ANN is used in this paper. The ANN can find

relations between input data and the output data easily by

the learning process.

3. K-means Method

A k-means method is one of clustering methods to

perform data classification [6, 7]. And the method classi-

fies given data into k clusters. The step of k-means me-

thod is as follows:

(i) Determine number k of the clusters.

(ii) Assign clusters for the data at random.

(iii) Calculate the center of each cluster with the

assigned da t a .

(iv) Calculate distances between the data and the

cluster-centered, and assign the data to the nearest cluster

again.

(v) When the allotment of all data into th e clusters

does not change by the process mentioned above, the

calculation is over. Otherwise, the above process is re-

peated after the cluster centers are recalculated from

newly assigned clusters.

4. Simulation

4.1. Output Power Forecasting

In this study, a wind power generation of rated output 5

MW having time series data of the output power shown

in Figure 3 is used. The future trend output is forecasted

for 50 seconds needed for control of FC-EL by the ANN

model shown in F igure 4.

Let the output power time series of the wind power be

{x1, x2, …, xn}. The ANN learns relation between xt,

xt-10, …, xt -60 that are the output power at present time t to

60 seconds before and xt+50 that is the output power at 50

100

020000 40000 60000 80000

Output power[%]

time[sec]

Figure 3. Output power of a wind pow er generation.

K. HANADA ET AL.

416

seconds from the present time, and forecasts output pow-

er 50 seconds in future. The forecasting result by the

ANN is shown in Figures 5, 6 and Ta ble 1. Table 1 in-

cludes the result of moving average that is calculated by

xt = (xt-10+ xt-20+ xt-30+ xt-40+ xt-50)/6 to compare with the

result by the ANN. The ANN can forecast the output

power of the wind power generation at 50 seconds in

future, and it is shown that the resu lt of the ANN is b etter

than that of the moving av erage.

4.2. The Necessary Storage Capacity of SMES

When FC-EL is operated based on the forecast men-

tioned above, charging or discharging power by SMES

compensates the difference between actual output power

and the forecasted output power as shown in Figure 7.

The required storage capacity of SMES shown in Table

2 is determined by alternately charging or discharging

power. In addition, distribution of the charging or dis-

charging power by SMES is shown in Figures 8 and 9.

From the above result, the storage capacity of SMES

using the ANN is smaller than that using moving aver-

age.

5. Conclusions

In this paper, the output power of the wind power gen-

eration is forecasted for ASPCS that is composed wind

t-10

t-60

t+50

Figure 4. The ANN forecasting output power of a wind

power generation.

100

010000 20000 30000 40000 50000 60000 70000 80000

Outputpower[%]

time[sec]

Actualou tput powerl Forecastedoutputpower

Figure 5. Forecasting result of output power of a wind

power generation by ANN.

10000 10100 10200 10300 10400 10500 10600 10700 10800 10900 11000

Outputpower[%]

time[sec]

Actualoutputpower

Forecastedoutputpower(ANN)

Forecastedoutputpower(movingaverage)

Figure 6. Forecasting result of output power of a wind

power generation by ANN (between 10000 and 11000 sec-

onds).

Table 1. Forecasting result of output pow er of a wind power

generation.

Forecasting method ANN Moving average

Average error [MW] 0.028 –0.001

Distribution [(MW)2] 0.025 0.028

Maximum absolute error [MW] 0.685 0.725

‐0.8

‐0.6

‐0.4

‐0.2

0.2

0.4

0.6

0.8

Charge/dischargepowerofSMES[MW]

Figure 7. Charge/discharge power of SMES.

Table 2. Storage capacity of SMES.

Forecasting method ANN Moving average

Maximum charge capacity[MJ] 72.0 73.5

Maximum discharge capacity[MJ] –61.8 –83.4

Average[MJ] 1.45

–0.05

Absolute average[MJ] 6.47 7.83

K. HANADA ET AL.

417

power generation, SMES and FC-EL and supplies

smoothed power from this system. As a result, the ANN

has better accuracy than the moving average method and

is able to forecast the output power, and the required sto-

rage capacity of SMES is smaller. In the future study, we

optimize structure of th e ANN and improve the selection

method of learnin g dat a .

‐100 ‐50050 100

frequency[%]

charge/dischargepowerofSMES[MJ]

6. Acknowledgements

This work is supported by the Advanced Low Carbon

Reduction Technology R&D of Japan Science and

Technol o gy Agency.

REFERENCES

[1] T. Hamajima, et al., “Application of SMES and Fuel Cell

System Combined with Liquid Hydrogen Vehicle Station

to Renewable Energy Control,” IEEE Transactions on

Applied Superconductivity, Vol. 22, No. 3, 2011, p.

5701704. doi:10.1109/TASC.2011.2175687

Figure 8. Distribution of charging/discharging energy of

SMES (ANN).

[2] T. Shintomi, et al., “Design Study of SMES System

Cooled by Thermo-siphon with Liquid Hydrogen for Ef-

fective Use of Renewable Energy,” IEEE Transaction

Supercond., Vol. 22, No. 3, 2011, p. 5701604.

doi:10.1109/TASC.2011.2178575

‐100 ‐50050 100

frequency[%]

charge/dischargepowerofSMES[MJ]

[3] T. Nakayama, et al., “Optimization of SMES Compensa-

tion Capacity for Stochastic Power Using a Kalman Fil-

ter,” TEION KOGAKU, Vol. 45, No. 3, 2010, pp. 99-106.

doi:10.2221/jcsj.45.99

[4] D. E. Rumel hart, et al., “Parallel Distributed Processing,”

MIT Press, 1986.

[5] T. J. Sejnowski, et al., “Parallel Networks that Learn to

Pronounce English Text,” Complex System, Vol. 1,1987,

pp. 145-168.

[6] J. B. MacQueen, “Some Methods of Classification and

Analysis of Multivariate Observations,” Proceedings of

5th Berkeley Symposium on Math, Stat. and Prob., 1967,

pp. 281-297.

[7] E. W. Forgy, “Cluster Analysis of Multivariate data: effi-

ciency vs. Interpretability of Classifications,” Biometrics,

Vol. 21, 1965, pp. 768-769.

Figure 9. Distribution of charging/discharging energy of

SMES (moving average).