Embedded Applications of MS-PSO-BP on Wind/Storage Power Forecasting

TELKOMNIKA, Vol.15, No.4, December 2017, pp. 1610~1624
ISSN: 1693-6930, accredited A by DIKTI, Decree No: 58/DIKTI/Kep/2013
DOI: 10.12928/TELKOMNIKA.v15i4.6720  1610
Received July 3, 2017; Revised August 30, 2017; Accepted November 2, 2017
Embedded Applications of MS-PSO-BP on
Wind/Storage Power Forecasting
Jianhong Zhu*
1
, Wen-xia Pan
2
, Zhi-ping Zhang
3
1,2
College of Energy and Electrical Engineering, HoHai University, JiangSu NanJing，P.R.China
1
Institute of Electrical Engineering, Nantong University, Nantong, JiangSu Province, China
3
Jiangsu Longyuan Wind Power Co., Ltd, Jiangsu Nantong, R.China
*Corresponding author, e-mail: jh.zhu@ntu.edu.cn
Abstract
Higher proportion wind power penetration has great impact on grid operation and dispatching,
intelligent hybrid algorithm is proposed to cope with inaccurate schedule forecast. Firstly, hybrid algorithm
of MS-PSO-BP (Mathematical Statistics, Particle Swarm Optimization, Back Propagation neural network)
is proposed to improve the wind power system prediction accuracy. MS is used to optimize artificial neural
network training sample, PSO-BP (particle swarm combined with back propagation neural network) is
employed on prediction error dynamic revision. From the angle of root mean square error (RMSE), the
mean absolute error (MAE) and convergence rate, analysis and comparison of several intelligent
algorithms (BP, RBP, PSO-BP, MS-BP, MS-RBP, MS-PSO-BP) are done to verify the availability of the
proposed prediction method. Further, due to the physical function of energy storage in improving accuracy
of schedule pre-fabrication, a mathematical statistical method is proposed to determine the optimal
capacity of the storage batteries in power forecasting based on the historical statistical data of wind farm.
Algorithm feasibility is validated by application of experiment simulation and comparative analysis.
Keywords: wind power schedule forecast, MS-PSO-BP, storage, optimal capacity
Copyright © 2017 Universitas Ahmad Dahlan. All rights reserved.
1. Introduction
Due to instability generation and inaccurate scheduling forecast of wind power, high
proportion grid affects the stability of power system operation, it should maintain an optimal
balance between the power and the load from time to time [1]. Hence, wind power prediction
systems are developed based on the numerical weather prediction, incorporated with statistical
model or some other advanced research methods, such as the artificial neural network, support
vector machines, Kalman filter, grey relational analysis, fuzzy logic methods, wavelet
transformation, as well as physical methods [2-9]. Rasit Ata affirmed artificial neural network on
wind power prediction and schedule forecast by comparing with other intelligent algorithms
within 30 years [10]. However, due to the randomness of wind power fluctuations, the existing
prediction algorithms are difficult to reflect the characteristics of the power fluctuations
perfectly [11].
In addition, European countries, such as Germany, Denmark and Spain, punishment
mechanism is drawn up to deal with the overrun error of schedule forecast. Northern Europe, for
example, may be subject to an average tariff of 12% penalty for poor schedule forecast. Indian
government also issued a decree in 2013 that wind power projects of installed capacity higher
than 10MW must provide ultra-short-term wind power prediction. If the difference between
actual output and schedule forecast is more than 30%, the wind farm is required to pay the fine.
Hence, some researchers are still working on the hybrid optimization algorithm to improve the
accuracy of wind power prediction.
In practical application, the accuracy improvement of schedule forecast can also be
achieved by external means, such as releasing positive or absorbing negative estimated energy
errors, which can be realized by an energy buffer device [12]. Batteries have flexible charge-
discharge characteristic and relative mature energy storage technology, which can be used to
absorb the redundant power, correct unforeseen owed power supply, reduce wind power
schedule forecast errors and improve the accuracy of wind power schedule forecast. At the

TELKOMNIKA ISSN: 1693-6930 
Embedded Applications of MS-PSO-BP on Wind/Storage Power Forecasting (Jianhong Zhu)
1611
same time, it may achieve a fast power regulation to improve the stability of the wind power
system and the reliability of the power supply by a small storage capacity [13-18].
In this paper, a wind power schedule forecast error correction method is proposed by
means of a hybrid algorithm integrated with a physical technique. Firstly, back propagation (BP)
neural network in power pre-fabrication is proposed not only containing Mathematical statistics
(MS), but also considering artificial neural network integrated with particle swarm algorithm
(PSO-BP). Secondly, analysis and comparison of several intelligent algorithms (BP, RBP, PSO-
BP, MS-BP, MS-RBP, MS-PSO-BP) are done to verify the availability of the proposed prediction
method. Finally, an improved wind power schedule forecast correction system based on storage
batteries is used to improve schedule forecast accuracy, optimal capacity of the storage battery
is studied, practical operating data are used in MATLAB simulation.
2. Power Prediction Technique Based on MS-PSO-BP
2.1. Mathematical statistics (MS) sample data pretreatment
In theory, a single wind turbine output power can be obtained by formula (1) [19], while
in actual wind farm, because of the influence of the external environment and different wind
turbines characteristics, error exists on the actual power curve distribution, as shown in
Figure 1.
32
2
1
wpm VRC•P  (1)
Figure 1. Normal wind speed-power curve
Due to numerical weather prediction error, wind farm actual power output does not quite
coincide with the prediction. In neural network training and prediction process, numerical
weather prediction data are essential, so it is necessary to study the relationship between
numerical weather forecast and the actual power output.
In this paper, the sampling period used for numerical weather forecast is 10 min, mainly
including wind speed, wind direction and temperature data. According to the formula (1), the
wind speed is the most important factor that influences the turbine power output, so the
research is mainly on the relationship between wind speed and wind power. Figure 2 is shown
the relationship between wind farm power output and the forecast wind speed.
It is seen from which that relationship is not accord completely with the wind power
curve in Figure 1. To the same forecast wind speed，the actual power output may be different.
The training sample of neural network as such historical statistical data will influence the training
effect, lower convergence rate. For large amount data, mathematical statistics is an effective
analysis method. To analyze and predict the relationship more precisely between the actual
output power and wind speed of wind farm, probability statistics method is used. Wind speed
partition is done to realize convenient analysis, partition statistics is referred to IEC61400-12
standard. The wind speed data collected should cover range from -1m/s (cut into the wind

 ISSN: 1693-6930
TELKOMNIKA Vol. 15, No. 4, December 2017 : 1610 – 1624
1612
speed) to 1.5 m/s multiplied by 85% of the rated wind speed. According to Bin methods [20],
wind speed range collected is adopted 2 m/s to 20 m/s, divided by 1 m/s，the center value of
each bin is the integer times of 1 m/s, each bin will contain a lot of wind speed scatters.
Figure 2. 100MW wind farm power output and wind speed forecast relations
By this means, the wind speed data from the numerical weather prediction are divided
into several partitions. In each BIN interval, wind farm actual power output are more dispersed.
To get a detailed analysis of each BIN range of power distribution, taking the 9.5 m/s ~ 10.5 m/s
wind speed range as an example, kernel density estimation algorithm is used to compute
density distribution characteristics of the power output [21], the calculation formula is shown in
formula (2), where f is probability density correspond to iP , P is power output point of BIN
range(9.5 m/s ~ 10.5 m/s). According to the formula (2), wind farm output power distribution can
be calculated among the wind speed range, as shown in Figure 3.
 , i
f P KSDE P    (2)
Figure 3. 9.5m / s ~ 10.5m / s wind speed- power output distribution of wind farm
Similarly, statistics for actual power output correspond to each BIN interval is done, it can
be seen from the Figure 3 and Figure 4, power distribution presents certain regularity in each
BIN interval.

1613
Figure 4. Power distribution of different BIN ranges confidence low limit
  1Ndown Nup
P P P P     (3)
At first, the power output probability distribution of each BIN interval is concentrated on
the peak probability density, and decreased symmetrically on both sides. Next, along with the
increased wind speed BIN ranges, probability distribution curve correspond to the peak also
increases gradually, moving to left. To revise small probability power output, bin estimation
theory is used to construct an estimated interval ,Ndown NupP P   , making estimation range cover P
within the probability 1  . The calculation method is as shown in formula (3). NdownP and NupP is the
confidence low limit.
Upper limit of parameters P in the Nth interval, no longer labeled N in the following, downP
and upP represents same meaning as NdownP and NupP . Some BINs probability distribution in the
interval [0，100000] are not symmetrical completely in Figure 4, divided mainly into three
conditions in Figure 5, peak to the left (a), peak center (b), peak to the right (c).
Figure 5. BINs probability distribution (left/mid/right)
To guarantee larger probability of power points appeared in the estimates range,
considering the probability symmetry distribution on both sides of the peak, the formula (4) (5)
(6) are adopted to estimate intervals of above three conditions respectively.

 ISSN: 1693-6930
1614
 max
0 0
1
( ) ( ) 1 , ( ) 1
2
upP P
F x f P dP f P dP       (4)
     max max
max 0
1 1 1
( ) ( ) ( ) 1 , 1 ( ) 1 1
2 2 2
up
down
P P P
P P
F x f P dP f P dP f P dP             (5)
 max100000
0
1
( ) ( ) 1 , ( ) 1 1
2down
P
P
F x f P dP f P dP        (6)
When probability integral of left half of peak is less than  1
1 1
2
  , the probability density
curve is as shown in Figure 5 (a), it can be seen 0downP , upP can be obtained by the formula (4),
distribution range is as upP,0 .When probability integral is less than  1
1 1
2
  , but greater than
 1
1
2
 , the probability density curve is as shown in Figure 5 (b). According to the formula (5), downP
and upP can be obtained, the distribution range is  updown PP , . When probability integral is greater
than  1
1 1
2
  , the probability density curve is as shown in Figure 5 (c), 100000upP , downP can be
obtained by the formula (6). The distribution range is 100000,downP .Once the confidence level is
selected, the confidence lower limit and the confidence upper limit of each BIN can be obtained,
using these values as midpoints of each BIN interval, confidence lower limit curve downcurveP and
confidence upper limit curve upcurveP can be fitted by the interpolation algorithm, as shown in
Figure 6.
According to the above curves, neural network training data are revised, smaller
probability data points beyond confidence curve are amended as formula (7).Where
*
P is the
actual output power of wind farm, P is the corrected power output, which will be served as the
training samples of neural network.
*
* *
*
dowmcurve downcurve
downcurve upcurve
upcurve upcurve
P P P
P P P P P
P P P
 

  


(7)
Figure 6. Confidence upper and lower limit schematic
2.2. Revised BP Neural Network model
At present, the representative models of the neural network are BP (back
propagation) neural network and RBF (radial basis function) neural network [22,23]. They
have some advantages, such as strong robustness and fault tolerance, self-learning, self-
organization, adaptability, and can approach to arbitrary complex nonlinear relationship.
The paper used them in wind power prediction, and prediction effects are compared and
analyzed with each other.

1615
2.2.1. BP neural network model
The typical BP neural network shows three layers structure, including the input layer,
middle layer and output layer. Middle layer can be designed as single hidden layer or multi-
hidden layers structure. The core of the algorithm is forward information dissemination and error
back propagation, the process is done again and again, and the weights of each layer and
threshold is adjusted continuously, finally the error is reduced to an acceptable level. Assume
number of input layer nodes is n, the middle layer number p, output layer m, so : n m
f R R is
completed. The input and output topology structure is as shown in Figure 7.
……
…………
n mq
……
WV
j
1X
iX
nX
1z
kz
pz
1y
jy
my
Figure 7. Three-layer neural network topology of BP
The output of the node j of middle layer is as formula (8).Where i along to [1 ,n], j along
to [1 ,p], the output of the node k is as formula (9).
n
u f w x i n j p
j i j i j
i
),( 1 ,1
1
1
      

(8)
2
1
,1
p
k j k j k
j
y f v Z K m

 
      
 
 (9)
Among formula (8)(9), the k [1 ,m], 1
f is the transfer function of hidden layer, 2
f is the
transfer function of output layer, i
x represents each neuron input of the input layer, i j
w is the
weight of the input layer to the middle layer, j
 is the middle layer node threshold, j k
v is the
connection weight from the middle layer to output layer, k
 is the output layer thresholds, the
initializations of weights and thresholds are produced by random, and the random initial value
tends to reduce the convergence speed, easy to make the training results fall into local
minimum value.
2.2.2. RBF neural network model
The Radial Basis Function Neural Network (Radial Basis Function Neural Network, the
RBFNN) is a kind of feed forward Neural Networks. Compared with the BP neural network, RBF
neural network not only has a physiological basis, but a simple structure, concise training and
fast convergent speed. RBF neural network also has the three layers structure. The weight
between input layer and hidden layer is fixed to 1, only the weight between hidden layer and
output layer is adjustable. Number of input layer nodes is n, middle layer p, output layer m, so
: n m
f R R .The input and output model is as shown in Figure 8.The output of the j hidden node
is as formula (10).
   j j jh X c (10)
1
m,1
h
j i j i j
i
y w X c j

     (11)

 ISSN: 1693-6930
1616
…
Input layer Output layerHidden layer
…
 1 1,x c 
 2 2,x c 
 ,k kx c 

1x
2x
px
1iw
2iw
kiw
 if x
Figure 8. RBF neural network structure
In the formula (11)， 1 2, nX x x x    is the input vector, 1 2,j j j j nc c c c    is the first j
hidden node of RBF data center.  j
  is the activation function of the hidden nodes, generally,
the Gaussian function is taken as  
2
2
u
u e 

 .It can be seen that the key to establish a RBF
network model is determining the number of hidden layer of RBF network h and data center j
c ,
the width of the radial basis functionσ,the connection weight from output neurons to the hidden
layer neurons is i j
w .Like BP neural network, the RBF neural network, for key parameters,
initialization is generated randomly, these random initialization values affect the neural network
training in a certain extent.
2.2.3. Improved Neural Network under adaptive mutation Particle Swarm Optimization
(PSO)
RBF neural network and BP neural network are used widely, also having some
deficiencies, such as long training time, easy falling into local minimum value, and so on.
Although RBF neural network is better due to its convergence speed and local minimum value
problem, but parameter initialization value is generated randomly, so the neural network training
is affected in a certain extent. In order to be able to solve these problems, the PSO algorithm is
introduced to improve neural network algorithm.
PSO basic idea is inspired by the birds swarm behavior regularity, a simplified model of
swarm intelligence is then established. It is an optimization algorithm based on iterative process.
The first is to initialize a group of particles, each particle has two characteristics, position and
velocity. The position of each particle is representative for a possible solution of optimization
problem, and the velocity of the particle is expressed on the direction and distance of flight.
Then optimal particle in the solution space is searched through iteration. For each iteration, the
particle individual position is updated by tracking individual extremum best
P and group extremum
best
G , until the optimal particle is found. Setting a group consisted of m particles and fly at a
certain speed in D dimensional search space, then the particle swarm can be expressed on
1 2, , DX X X X   L , i
X represents the ith particle's position, also represents a possible solution of
problem, it can be expressed with matrices 1 2
, ,
T
i i i i D
X X X X   L .The fitness value of the each
position of particle can be calculated by the substitution of i
X into the objective function. best
P
represents the ith particle speed, it can be expressed on 1 2
, ,
T
i i i i D
V V V V   L . Particles update
speed and position according to formula (12) (13).
             1 1 2 2
1i j i j j i j j i j
V t V t C R P t X t C R G t X t          (12)
     1 1i j i j i jX t X t V t    (13)
Among formula (12) (13), j = 1, 2,..., d, represents particle dimension, t is the times of
iteration, 1
C and 2
C are learning factors, 1
R and 2
R represent random number between 0~1,  as

1617
inertia weight,  i j
V t represents the ith particle current speed in the tth generation,  i jX t
represents the ith particle current location in the tth generation. In the general PSO algorithm,
inertia weight represents the impact of the historical rate on the current speed. Where the larger
 is, the stronger global search ability particles have. While the lower  is, the stronger portion
search ability particles have. Once 0  , it means that the particles lose ‘memory’. To give
attention to both global and local search, take  as linear gradient, in formula (14), i t er is
current iteration number, max
i t er is the maximum number of iterations, max
 is inertia weight initial
value, mi n
 represents inertia weight ultimate value. To prevent particles from blind search, the
particle's position and speed are limited respectively within a certain range[ ]max max
X ，X and
[ ]max max
V ，V .
 max max mi n
max
i t er
i t er
      (14)
In order to avoid the ‘precocity’ and low iterative efficiency of PSO algorithm, the
mutation is introduced to PSO algorithm, the principle is that the population is initialized at
certain probability after each updating, so expanding search space of the dwindling population
during the process of iteration, ensure the optimal location be searched before jumping out, thus
improving the global convergence of the algorithm.
2.2.4. BP Neural Network combined with PSO on prediction model
The improved PSO algorithm is used to optimize parameters of neural network
prediction model [24]. The steps are as following.
Step 1 Building particle swarm, a neural network topology structure is established
according to the input and output sample. The parameters be to optimize are coded to the
individual particles of real vector population.
Step 2 The initialization of particle swarm parameters, mainly including the size of the
population, learning factor, particle position and velocity interval, number of iterations, etc.
Step 3 Calculating the particle fitness value, according to the input and output sample,
the fitness function value of each particle is calculated, the current position is set to itself optimal
location, the position of optimal particle in initialization population is set to the global optimal
position.
Step 4 Loop iteration, PSO algorithm formula (12) (13) are used to update particle
velocity and position.
3. Wind/Storage Dynamic Correction of Schedule Forecast
It can be seen from literatures that prediction error exists inevitably in theoretical
prediction algorithms due to various complex factors. The storage system can be used to
absorb the redundant energy, correct unforeseen owed power supply, lower prediction errors
and improve the accuracy of wind power forecast due to flexible charge-discharge
characteristic. In Figure 9, the wind/storage system is mainly composed of wind turbines and
vanadium battery package. The power relationship is as shown in formula (15). Where, a
P is the
actual power output of the wind farm, b
P is charging and discharging power of battery energy
storage system， d
P is the whole Wind/Storage system output. The key is how to determine b
P
dynamically. From the foregoing description, MS-PSO-BP neural networks has been used to
give modified predicted power '
)(P
P t firstly, once the actual power output is greater than the
predicted wind farm output ( ( )a p
P P t ), then the energy storage system is charged to guarantee
the accuracy of the forecast wind power ( ( )d p
P P t ), and b
P is set to a negative value. On the
contrary, when ( )a p
P P t , b
P is set to a positive value, the energy storage system is discharged.
When ( )a p
P P t and '
( )d P
P P t , b
P is set zero, the battery energy storage system is kept in
holding state. e
p is the prediction error tolerance. ( )error
p t is the error between modified
prediction power and real-time power. Algorithm flowchart is as shown in Figure 10. If the

 ISSN: 1693-6930
1618
modified prediction power value from MS-PSO-BP neural network lies in the allowed range
error, the predictive value is regarded as optimal forecasting power of the moment. On the
contrary, if the predictive value is out of the range, then the energy storage system is triggered
to amend the predicted value.
Gearbox
Gearbox
...
AC/DC
..
Pa
Pb
Pd
Wind Turbines
Energy
storage
system
Grid
Figure 9. Configuration of wind farm combined with storage battery
d b a
P P P  (15)
Figure 10. Wind/storage amending algorithm flowchart

1619
4. Simulation and Analysis
4.1. BP/RBP/PSO-BP prediction without training sample pretreatment
In order to verify the effectiveness of the proposed method, a real wind farm data
samples are used to test the validity of the algorithm, the wind farm is located in the coastal
areas of Jiangsu, and installed capacity is 100 MW. Full-year data of 2014 are used in
mathematical statistics, the first two months date of 2015 are used as training sample, the data
of March are used in test, data sampling period is 10 min, mainly including wind speed, wind
direction, temperature and wind power data, all mathematical operation are performed by
normalized process. To distinguish the pros and cons of the different algorithms, root mean
square error (RMSE)and mean absolute error(MAE) are used to measure wind power prediction
error, and the calculation formula is as (16) (17).
 
2
1
n
Mi Pi
i
P P
RMSE
C n




g
(16)
1
n
Mi Pi
i
P P
MAE
C n




g
(17)
MiP is the real power of the moment i, PiP is the prediction power of the moment i, C is
power capacity of the field wind farm, n is as the number of sample. Taking wind power on
March 1, 2015 as prediction object, using BP neural network, PSO-BP neural network, RBF
neural network respectively to predict 96 wind power points from 0:00 to 24:00. Prediction
results are shown in Figure 11. The RMSE and MAE results are as shown in Table 1. It can be
seen that prediction results of three neural network algorithms are improved compared with
existing wind farm forecast accuracy.
Figure 11. Wind power prediction results of BP/RBP/PSO-BP/actual field
Table 1. Error Statistics
Algorithm RMSE (%） MAE（%）
Existing wind farm prediction 15.09 12.12
RBF 14.37 12.18
BP 14.57 12.21
PSO-BP 13.52 11.77
BP neural network prediction results are similar to RBF neural network, but the
prediction accuracy of BP neural network algorithm is improved obviously through amendment
of PSO algorithm. RMSE is reduced by 10.40%, and MAE is reduced by 2.89% compared with
existing prediction. Considering uncertainty of the neural network algorithm, the above methods
are tested repeatedly, prediction results are analyzed. The wind farm power output are predicted

 ISSN: 1693-6930
1620
from March 1 to March 20, and the RMSE and MAE results are as shown in Appendix Table 1
and Appendix Table 2. It can be seen that RBF neural network, BP neural network and PSO-BP
neural network prediction results are uncertain for single prediction, so the average prediction
error are analyzed overall, average prediction error of the 20 days ago on the March are as
shown in Figure 12. Prediction error of RBF neural network prediction system is similar to
existing wind power prediction. BP neural network overall training effect is better than RBF
neural. Its average RMSE is reduced by 8.59% compared with RBF neural network and reduced
by 7.13% compared with existing prediction system of wind farm. Furthermore, PSO-BP neural
network is 2.7% lower than the average RMSE, the same for the MAE, reduced by 8.44%
compared with the existing wind power prediction system, reduced by 2.83% and 2.83%
respectively compared with the BP and RBF prediction algorithms.
Figure 12. Comparison chart of the average prediction error on wind power
4.2. BP/RBP/PSO-BP prediction with training sample pretreatment
To verify effectiveness of training data pretreatment for prediction precision and
convergence speed for the neural network algorithm, the proposed mathematical statistics
method is used to modify neural network training samples. Selecting confidence level
95.01   , according to formula (4) (5) (6), the corresponding confidence lower limit curve
downcurveP and confidence upper limit curve upcurveP are obtained. Using the formula (7) to correct
training data, the diagram is shown in Fig 13. After correction, data points of small probability
have been processed, preventing the "mixed" from affecting on neural network training. The
results of average RMSE and MAE are shown in Table 2.
Figur 13. Data revised schematic diagram

1621
Table 2. Prediction Error Comparison Before and After the Training Data Pretreatment
Algorithm RMSE（%）
RMSE（%)
Date revised
MAE（%）
MAE（%）
Date revised
Existing wind farm
prediction
16.40 12.79
RBF 16.65 16.17 12.91 12.85
BP 15.23 14.80 12.05 11.70
PSO-BP 14.83 13.89 11.71 11.54
Data show that prediction errors of three algorithms after sample pretreatment are
decline in various degrees, RMSE of RBF, BP, PSO-BP reduced by 2.90%, 2.82% and 6.34%
respectively, and the MAE just reduced by 0.4%, 2.90% and 0.4% respectively, it can be seen
that the accuracy after pretreatment is improved effectively. Another advantage is the training
speed, some "special points" are prevented from slow convergence or even not convergence.
Figure 14 and Figure 15 show training effects of the BP neural network and RBF neural network
under pretreatment or no. It can be seen from Figure 14, under the condition of the same
training target, training with data pretreatment has better training effect on training speed and
precision compared with that untreated. RMSE reaches steady state (0.019) only after 28 times.
Figure 14. BP neural network training with and
without data pretreatment
Figure 15. RBF neural network training with
and without data pretreatment
While without data pretreatment, it needs 46 times training and RMSE reaches the steady
state (0.024). It can also be seen from the Figure 15, under the condition of the same training
target, RBF neural network training after data pretreatment need only 23 times to reach the
target value of 0.01. Without pretreatment, no training goals can be arrived even after a
maximum of 200 times. Comparing Figure 14 with Figure 15, it can also be seen that algorithms
convergence speed and convergence results of BP are all inferior to that of RBF, showing
feasibility and effectiveness of the experimental results.
4.3. Wind/storage system physical amending for schedule forecast
The simulation experiment is done in a wind farm of 100000 kilowatts in eastern coastal
of China. The historical statistical data are concluded within 4 consecutive days in June, such as
wind speed, the wind field actual output power. The data during previous two days are adopted
as training samples for prediction model. The measured data in the next two days are regarded
as the calibration data, which are used to verify the accuracy and yet the validity of the
algorithm.
The results of simulation are shown in Figure 16. Prediction errors simulation are as the
basis to determine the battery capacity needed. In order to prevent the prediction process from
the occasional error, experiments are conducted 1000 times, each maximum capacity max
C of
battery charge or discharge is stated, and a high level of confidence max
C is selected for
referenced battery capacity. When certain capacity is supplied in the wind power system, the
5 10 15 20 25 30 35 40 45 50 55 60
0
0.02
0.04
0.06
0.08
0.1
0.12
Epochs
MeanSquaredError(mse)
BP
Processed BP
Goal
0 20 40 60 80 100 120 140 160 180 200
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
0.05
0.055
Epochs
MeanSquaaredError(mse)
RBF
Processed RBF
Goal

 ISSN: 1693-6930
1622
results of reducing the prediction errors are as shown in Figure 17 and Figure 18. As can be
seen from Figure 17, the power output Pd of the wind farm is smooth more with battery energy
storage system, and is close to the prediction curve of wind power. Figure 18 shows that the
battery capacity can also meet the needs of short-term wind power correction. Meanwhile, it is
concluded from the experiments that RMSE of wind power forecast has been reduced to
9.3535e+003W, 10 minutes predictive error integration is reduced to 2.1941e+004Ah, the
battery energy storage system can minify wind power prediction error effectively.
Figure 16. Modified Power prediction error integration
Figure 17. Power output forecast correction
with battery energy storage
Figure 18. Storage battery SOC change during
correction
5. Conclusion
This paper combines the mathematical statistics and BP neural network on wind power
prediction, PSO algorithm is used to improve prediction precision. Based on which,
Wind/Storage system is used to amend wind farm power forecast. Simulation results show that
the proposed pretreatment (mathematical statistics method) can improve the neural network
training speed and precision. In addition, the PSO algorithm can also improve the prediction
precision of the BP neural network effectively. Compared with the current wind farm forecasting
strategy, RMSE of PSO-BP can be reduced by 6.34%, and the MAE reduced by 0.4%.
Moreover, the schedule forecast accuracy can be improved effectively by physical Wind/Storage
dynamical correction, and experiments show that RMSE of wind power forecast has been
reduced to 9.3535e+003W, the energy storage system can minify wind power prediction error
effectively. However, limited by the technical conditions, the battery capacity is still an important
bottleneck in application all along. The battery capacity can be insufficient for big share on the
access of large wind farms, but it can be used as a power fine-tuning in large wind power.
0 20 40 60 80 100 120 140 160 180 200
-150
-100
-50
0
50
100
150
200
Time(T=15min)
Capacity(MAh)
0 50 100 150
0
20
40
60
80
100
Time(T=15min)
P(MW)
Forecast corrected with battery
Forecast corrected without battery
0 50 100 150 200
0
0.2
0.4
0.6
0.8
1
Time(T=15min)
SOC

1623
Acknowledgements
This work is a part of the Natural Science Foundation of China (No.
51377047,No.51407097). Experimental data are taken from Jiangsu Longyuan Wind Power
Co., Ltd..The authors would like to thank for the supports from both the Ministry of Science and
Technology of China and National Natural Science Foundation of China and Jiangsu Longyuan
Wind Power Co., Ltd. of China.
References
[1] KM Nor. Renewable Distributed Generation Models in Three-Phase Load Flow Analysis for
Smart Grid. Telkomnika. 2013; 11(4): 661-668.
[2] Tastu J, Pinson P, Trombe PJ, Madsen H. Probabilistic forecasts of wind power generation
accounting for geographically dispersed information. IEEE Transactions on Smart Grid.
2014; 1: 480-489.
[3] Quan H, Srinivasan D, Khosravi A. Short-term load and wind power forecasting using
Neural Network-based prediction intervals. IEEE Transactions on Neural Networks and
Learning Systems. 2014; 2: 303-315.
[4] Venayagamoorthy GK, Rohrig K, Erlich I. One step ahead: short-term wind power
forecasting and intelligent predicti-ve control based on data analytics. IEEE Power and
Energy Magazine. 2012; 5: 70-78.
[5] Can W, Zhao X, Pinson P. Direct interval forecasting of wind power. IEEE Transactions on
Power Systems. 2013; 4: 4877-4878.
[6] Ming Y, Shu F, Wei JL. Probabilistic short ter-m wind power forecast using componential
sparse bayesian learning. IEEE Transactions on Industry Applications. 2013; 6: 2783-
2792.
[7] Potter CW, Negnevitsky M. Very short-term wind forecasting for Tasmanian power
generation. IEEE Transactions on Power Systems. 2006; 2: 965-972.
[8] Yongqian L, Jie S, Yongping Y, Wei JL. Short-term wind power prediction based on
wavelet transform Support Vector Machine and Statistic-characteristics analysis. IEEE
Transactions on Industry Applications. 2012, 4: 1136-1141.
[9] Abdollah KF, Abbas K, Saeid N. A New Fuzzy-Based Combined Prediction Interval for Wind
Power Forecasting. IEEE Transactions on Power Systems. 2016; 1: 18-26.
[10] Rasit A. Artiﬁcial neural networks applications in wind energy systems: a review.
Renewable and Sustainable Energy Reviews. 2015; 49: 534–562.
[11] Tewari S, Geyer CJ, Mohan N. A Statistical model for wind power forecast error and its
application to the estimation of penalties in liberalized markets. IEEE Transactions on
Power Systems. 2011; 4: 2031-2039.
[12] Hamideh B, Saifur R, Manisa P. Sizing Energy Storage to Mitigate Wind Power Forecast
Error Impacts by Signal Processing Techniques. IEEE Transactions on Sustainable Energy.
2015; 6(4):1457-1465.
[13] Li Q, Choi SS, Yuan Y, Yao DL. On the determination of battery energy storage capacity
and short-term power Dispatch of a wind farm. IEEE Transactions on Sustainable Energy.
2011; 2: 148 -158.
[14] Zheng X, Ding J, Li M, et al. Using STATCOM with Energy Storage to Enhance ACDC
System Stability. Telkomnika. 2016; 14(2): 423-430.
[15] Ogimi K, Yoza A, Yona A, et al. A study on optimum capacity of battery energy storage
system for wind farm operation with wind power forecast data. Harmonics and Quality of
Power (ICHQP), 2012 IEEE 15th International Conference on. IEEE. 2012: 118-123.
[16] Mohod SW, Aware MV. Micro wind power generator with battery energy storage for critical
load. IEEE Systems Journal. 2012; 1: 118-125.
[17] Teleke S, Baran M E, Bhattacharya S, et al. Optimal control of battery energy storage for
wind farm dispatching. IEEE Transactions on Energy Conversion. 2010; 25(3): 787-794.
[18] D Kottick, Blau M, Edelstein D. Battery energy storage for frequency regulation in an
island power system. IEEE Transactions on Energy Conversion. 1993; 3: 455-459.
[19] Zhou S, Lu ZX. Wind power and electric power system. BeiJing: Chinese Electric Power
Publishing house. 2011: 228.

 ISSN: 1693-6930
1624
[20] Kalavade A, Lee EA. The extended partitioning problem: hardware/software mapping and
implementation-bin selection. IEEE Sixth International Workshop on Rapid System
Prototyping. 1995: 12-18.
[21] Li CH, Sun ZH, et al. Kernel Density Estimation and its application to clustering algorithm
construction. Journal of Computer Research and Development. 2004; 10.
[22] Cong S. Toolbox for MATLAB neural network theory and application. BeiJing: China
Science and Technology University Press. 1998.
[23] P Campolucci, A Uncini, Piazza F, Rao BD. On-line learning algorithms for locally recurrent
neural networks. IEEE Transactions on Neural Networks. 1999; 10(2): 253-271.
[24] Pingzhou T, Zhaocai X. The research on BP Neural Network model based on guaranteed
convergence Particle Swarm Optimization. Second International Symposium on Intelligent
Information Technology Application. 2008; 2: 13-16.
Appendix
Appendix Table 1 . RMSE Contrast of Algorithm Proposed on Paper with on-Site Original
Forecast
Date
Wind Forecasting
System
Training data uncorrected Training data corrected
RBF BP PSO-BP RBF BP PSO-BP
3.1 15.09 14.37 14.56 13.51 15.46 14.41 14.31
3.2 11.54 11.73 12.37 9.99 12.44 12.65 9.95
3.3 24.69 25.32 21.63 20.14 23.18 20.27 20.96
3.4 24.15 20.36 24.39 23.25 21.60 23.71 21.18
3.5 10.51 15.30 13.26 18.82 12.36 12.30 12.37
3.6 8.21 12.44 10.00 11.32 6.22 11.61 9.69
3.7 7.74 10.88 8.74 10.76 6.96 8.11 9.03
3.8 33.46 33.20 33.24 32.77 33.73 33.76 30.99
3.9 21.99 27.73 14.90 12.82 21.86 13.38 13.72
3.1 36.35 28.47 22.44 19.33 24.02 20.04 22.01
3.11 4.63 6.93 5.20 3.74 5.79 6.48 5.12
3.12 14.5 14.44 12.14 12.56 11.70 1045 10.05
3.13 12.17 12.06 13.27 12.88 25.63 13.44 10.54
3.14 12.16 12.03 14.35 9.39 12.24 13.11 11.20
3.15 11.7 14.89 12.67 13.40 13.35 11.51 12.63
3.16 10.22 10.30 10.31 9.03 10.04 9.04 9.02
3.17 13.72 14.72 12.60 15.73 11.44 13.33 12.42
3.18 33.56 28.01 25.24 28.21 29.42 27.59 21.87
3.19 13.3 12.23 12.91 10.74 11.93 11.70 12.60
3.20 8.21 7.64 10.25 8.13 13.94 9.08 8.13
Mean 16.395 16.65 15.23 14.83 16.17 14.80 13.89
Appendix Table 2. MAE Contrast of Algorithm Proposed on Paper with on-Site Original Forecast
Date
Wind Forecasting
System
Training data uncorrected Training data corrected
RBF BP PSO-BP RBF BP PSO-BP
3.1 12.12 12.18 11.59 11.46 12.18 12.14 12.42
3.2 8.94 9.24 9.37 10.05 9.06 8.67 10.25
3.3 18.56 22.61 16.39 16.67 22.72 15.83 17.47
3.4 16.81 16.28 22.35 20.71 16.19 18.75 19.47
3.5 8.22 12.74 10.84 11.22 12.24 11.20 8.52
3.6 7.07 9.98 7.94 8.81 9.49 8.79 8.98
3.7 5.41 6.11 7.59 6.06 5.89 6.12 2.35
3.8 26.20 26.68 26.53 27.36 26.88 24.81 26.78
3.9 18.80 17.93 11.22 9.82 18.53 11.11 10.55
3.1 25.01 19.92 9.85 14.44 20.07 15.73 15.47
3.11 3.37 5.71 5.20 3.71 4.66 3.96 4.84
3.12 11.44 13.04 9.01 10.74 12.85 7.63 8.35
3.13 9.91 8.15 14.25 9.12 9.00 10.06 9.92
3.14 9.71 8.35 9.95 9.37 8.32 11.09 11.16
3.15 10.05 12.16 12.84 9.95 12.00 11.36 9.54
3.16 8.23 7.83 7.32 6.85 7.76 6.62 7.17
3.17 11.67 12.45 9.80 9.44 12.41 9.69 10.83
3.18 27.87 23.01 23.18 21.52 22.99 23.79 19.09
3.19 10.66 8.80 9.30 10.45 8.77 10.05 8.90
3.2 5.68 5.04 6.51 6.47 5.08 6.66 8.82
Mean 12.79 12.91 12.05 11.71 12.85 11.70 11.54

Embedded Applications of MS-PSO-BP on Wind/Storage Power Forecasting

More Related Content

What's hot (17)

Similar to Embedded Applications of MS-PSO-BP on Wind/Storage Power Forecasting (20)

More from TELKOMNIKA JOURNAL (20)

Recently uploaded (20)

Embedded Applications of MS-PSO-BP on Wind/Storage Power Forecasting