energies

Article
Hybrid GA-PSO Optimization of Artificial Neural
Network for Forecasting Electricity Demand
Atul Anand * ID
and L Suganthi
Department of Management Studies, College of Engineering, Guindy, Anna University, Chennai,
Tamil Nadu 600025, India; [email protected]
* Correspondence: [email protected]; Tel.: +91-944-439-5222




Received: 18 January 2018; Accepted: 15 March 2018; Published: 23 March 2018


Abstract: In the present study Artificial Neural Network (ANN) has been optimized using a
hybrid algorithm of Genetic Algorithm (GA) and Particle Swarm Optimization (PSO). The hybrid
GA-PSO algorithm has been used to improve the estimation of electricity demand of the state of
Tamil Nadu in India. The ANN-GA-PSO model uses gross domestic product (GSDP); electricity
consumption per capita; income growth rate and consumer price index (CPI) as predictors that affect
the electricity demand. Using the historical demand data of 25 years from 1991 till 2015 it is found that
ANN-GA-PSO models have higher accuracy and performance reliability than single optimization
models such as ANN-PSO or ANN-GA. In addition, the paper also forecasts the electricity demand
of the state based on “as-it-is” scenario and the scenario based on milestones set by the “Vision-2023”
document of the state.

Keywords: electricity demand; ANN; PSO; GA; hybrid optimization; forecasting

1. Introduction
Electricity reforms have liberalized the electricity sector in many countries. The salient features
have been unbundling of generation, transmission and distributions entities; a competitive market
with in countries and creation of an independent regulator for access to transmission infrastructure.
In the prevailing deregulated markets, forecasting of electricity demand has emerged as a
key research field [1–3]. Many research tools and algorithms have been developed for electricity
demand forecasting. Most of modeling techniques fall under parametric or non-parametric categories.
Parametric techniques [4–8] are incapable of adapting to any type of environmental or societal
changes. Many parametric techniques such as Auto regressive Integrated Moving Average (ARIMA),
Exponential technique and Multiple Linear Regression when used for electricity demand forecasting
do not yield the desired accuracy [9]. In order to overcome the respective drawbacks of the parametric
techniques and to provide the ability of global search non-parametric (artificial intelligence) techniques
are preferred by researchers [10–12].
Artificial Neural Network (ANN) is very popular amongst researchers due to its adaptability
over wide range of problems involving decision making in uncertain situations. This has led to
the rapid developments of hybrid models [13]. Many variants of ANN involving hybridization by
learning techniques such as Backward Propagation (BP), Genetic Algorithm (GA) and Particle Swarm
Optimization (PSO) have been proposed by several researchers. The use of ANN with different
optimization methods is also useful to forecast the electricity demand. Amjadi, N. and Keynia
Farshid [14] presented a stochastic search technique based on hybridization of ANN for load forecasting
problem. According to them the hybridized ANN algorithm allows effective search of the solution
space without falling in local minima. Abdul, H. et al. [15] also came up with similar conclusion about
ANN model that was trained for short term load forecasting. Cincotti, S. et al. [16] has highlighted the

Energies 2018, 11, 728; doi:10.3390/en11040728 www.mdpi.com/journal/energies

Energies 2018, 11, 728 2 of 15

usability of computational intelligence for forecasting electricity prices. According to them trained
ANN model leads to improvement in mean average percentage error (MAPE). Hybrid ANN-BP model
has been considered by Fuliang Yin et al. [17] using historical load data for training the neural network.
It is observed by them, that ANN with back propagation algorithm improves the training time and
convergence towards solution.
Hybrid ANN-GA optimization forecasting models have wide range of applications. In GA search
follows the principles of evolution and natural genetics. According to Goldberg [18] GA produces
near optimal solutions by following robust search processes. G Aenables optimizing of weights of
demand equations. Canyurt et al. [19] studied the dependence of total energy demand of Turkey as a
function of economic indicators in linear, quadratic and exponential forms. Ceylon and Ozturk [20],
Haldenbilen and Ceylon [21], Assarch et al. [22] analyzed the total energy demand of Iran based on
GA. Hybrid ANN optimized with PSO has been successfully applied for load forecasting.
Hybrid ANN optimized with PSO model has been successfully applied for demand forecasting.
Bi T et al. [23] and Lu N et al. [24] have used radial basis function neural network for forecasting (RBF).
Banda E et al. [25] have presented that time series models lead to large forecasting errors due to their
sluggishness to adapt to changing load characteristics. According to their findings ANN-PSO model
gives improved results as compared to ANN-BP. Yang S D et al. [26] have incorporated PSO algorithm
with an adaptive weight factor to improve the performance.
The remainder of the paper is organized as follows: Section 2 introduces the Electricity sector
in Tamil Nadu; Section 3 presents methodology used for research; Section 4 shows the features of
ANN-GA-PSO models; Section 5 brings out the results and discussion; Section 6: Conclusions.

2. The Tamil Nadu Electricity Sector

For more than a decade, the energy sector in Tamil Nadu has witnessed a high growth of industrial
activity coupled with exploding domestic electricity demand in the consumer segment. Both of these
factors have led to a large deficit in power availability over the last few years. Table 1 brings out the
fact that generation capacity has not kept pace with electricity consumption. The immediate solution
to the predicament has been the buying of power through short term contract. According to the report
of Central Electricity Authority of India (CEA) the electricity deficit of the state in the year 2013 was
around 17.5% as compared to 2.8% in the year 2008. Hence there is a dire necessity to forecast the
electricity demand by the year 2023 to facilitate the investments in the sector.

Table 1. Key electricity demand determinants.

Electricity (kWr) Income Growth GSDP Price Demand

Year
Consumption Rate per Capita (%) (Billion Rs) Index (in mWh)
1991 295 10.97 4.81 48 17,173
1992 303 11.9 5.27 55 19,130
1993 334 12.9 5.74 65 20,289
1994 350 13.9 6.2 79 23,193
1995 421 14.8 6.6 82 24,610
1996 435 15.7 7.1 85 25,805
1997 449 16.8 7.5 89 26,943
1998 459 17.9 8 92 27,862
1999 496 18.8 8.5 94 30,434
2000 510 14.7 10.9 101 33,418
2001 539 15 10.88 103 36,578
2002 708 15.2 15.01 107 38,529
2003 740 15.3 17.56 109 46,130
2004 780 15.5 18.66 110 49,712
2005 860 17.23 17.73 115 51,282
2006 960 19.99 20.44 117 49,485
2007 1000 12.58 12.98 124 56,493
Energies 2018, 11, 728 3 of 15

Table 1. Cont.

Electricity (kWr) Income Growth GSDP Price Demand

Year
Consumption Rate per Capita (%) (Billion Rs) Index (in mWh)
2008 1000 13.73 14.4 136 53,506
2009 1080 18.83 19.53 151 57,212
2010 1040 17.27 18.07 166 60,518
2011 1074 18.06 16.7 163.02 61,897
2012 1118 18.29 17.66 159.01 66,391
2013 1161 16.3 19.98 157.39 72,987
2014 2130 17.89 42.27 143.52 74,990
2015 2007 12.94 38.45 138.77 77,218

Factors Affecting Electricity Demand

Electricity consumption of a state is a function of man affecting factors such as gross state domestic
product (GSDP), consumer prices index, energy per capita and income parameters. The following
factors reflect their major impacts on electricity demand:

(1) GSDP: Even though the linkage between GSDP growth and electricity demand growth are not as
strong as it was in the past, it is worth considering the impact on the society of high GDP growth
itself since they are linked to each other. A high GSDP growth rate year after year means higher
manufacture of products and provision of services at an unprecedented pace leading to higher
electricity demand. The electricity demand continues to grow in the state because of high level to
continue in a business as usual scenario.
(2) Electricity consumption per capita (E.Con) has increased from 510 kWh in year 2000–2001 to
1065 kWh in 2011–2012, that is more than 100% increase. Hence per capita consumption has been
taken as an independent factor.
(3) Income growth rate (per capita): The vision 2023 document of the state of Tamil Nadu aims at
doubling the per capita income by 2023. It is also seen that any increase in family income leads to
spurt in consumption.
(4) Consumer Price Index (CPI): Prices have an indirect impact on the electricity demand by
affecting the purchase of luxury goods such as air conditioners, washing machines etc.

3. Methodology
In this section, ANN that is optimized by hybrid GA-PSO algorithm in the linear and quadratic
forms models the electricity demand. The results of ANN-GA-PSO and A-G-P-Q are compared with
ANN with single optimization with GA and PSO algorithms.

3.1. Artificial Neural Network

ANN resembles human brain in its origin. It consists of a large number of neurons interconnected
to form a complex and non-linearly connected array of parallel network. The most common form
of ANN is multilayered perceptron (MLP) which has an input layer with one or more hidden layer
connected to one output layer. In present research we have considered a multilayer perceptron (MLP)
that has three neurons layers. The first one is the input layer which is in the direct contact with the
input data. The middle one is called the hidden layer and it has no contact with outside system.
It connects data from the input layer and sends them to the next layer. The last one is the output
layer that sends out results. Table 2 gives the network information of ANN about the input layer that
is made up of four factors namely, electricity consumption (E.Con), income growth rate, GSDP and
Consumer price index. The hidden layer has been used as the activation function. The output layer
comprises of one unit representing electricity demand as the dependent variables. The in-sample data
is split into two subsets, namely, the training set and the validation set. The training set is then used to
Energies 2018, 11, 728 4 of 15

train ANN-GA-PSO models until the training error ratio criterion of 0.001 is achieved. The Table 3
shows the sum of squares error, relative error, stopping rule and the training time of the ANN.

Table 2. ANN-GA-PSO Network Information.

1 E.Con (electricity consumption)

2 Income growth rate
Input Layer Factors
3 GSDP
4 CPI
Number of Units 59
Number of Hidden Layers 1
Hidden Layer(s) Number of Units in Hidden Layer 1 6
Activation Function Hyperbolic tangent
Dependent Variables 1 Electricity Demand
Number of Units 1
Output Layer Rescaling Method for Scale Dependents Standardized
Activation Function Identity
Error Function Sum of Squares

Table 3. Model Summary: ANN-GA-PSO.

Sum of Squares Error 0.004

Relative Error 0.001
Training
Stopping Rule Used Training error ratio criterion (0.001) achieved
Training Time 0:00:00.23

3.2. Particle Swarm Optimization (PSO)

Particle swarm optimization algorithm was developed by Kennedy and Eberhart in 1995 [27].
Over the years PSO has become a popular population based derivative free algorithm. A variant of
PSO was developed by Shi and Eberhart [28,29] by inserting a time dependent variable that improves
the convergence of the search process. In the previous research paper by the authors [30] following
equations have been used for the particle position and the velocity of the particles:

vt+1 = vt + R1 ∗ C1 ∗ ( g − xt ) + R2 ∗ C2 ∗ ( p − xt ) (1)

x t +1 = x t + v t +1 (2)

where C1 and C2 are knowledge factors, R1 and R2 are random numbers, g is the location of the leader,
p the personal best location, vt is the velocity at iteration “t” and xt is theposition at iteration “t”.
This equation reveals the particle leader location to each particle.
Decreasing the variable enables the slowing down of the speed of the particles around the leader
location and provides a balance between exploration and exploitation. PSO finds an optimal point from
the random set of points with the help of a fitness function, so that the random points are initialized
between the ranges of values of the past two years, which might find the point that matches the straight
line formed by the data. This new point is the predicted value for the next year.
GA-PSO hybrid algorithm was first proposed by Bates and Granger [31]. According to them
linear combination of two forecasting models have a distinct the advantages over individual models.
For their application in electricity domain, Nazari et al. [32] proposed a model using two metaheuristic
algorithms, namely GA and PSO for forecasting energy demands. They found that the exponential
model derived from the PSO model is the best model. Unler [33] proposed PSO based demand
forecasting model for Turkey using gross domestic product, population as predictors of energy
demand. Younes M et al. [34] provided a solution to the economic dispatch problem using a hybrid
method genetic algorithm-particle swarm optimization (GA-PSO). They found that GA-PSO provides
flexibility fast convergence, less computational time for non-linear characteristics of power systems.
Energies 2018, 11, 728 5 of 15

Araby EE El et al. [35] proposed that a two layered hybrid PSO-SLP (Successive Linear Programming)
approach that is suitable for nondifferentiable and discontinuous objective functions. Jarrndal
and Hamdan [36] have described a combined approach of artificial neural networks (ANN) with
particle-swarm-optimization (PSO) and genetic algorithm optimization (GA) for short and mid-term
load forecasting. The model identifies the relationship among load, temperature and humidity using
a case study of Sharjah City in United Arab Emirates. They have found that ANN is one of the
powerful artificial intelligence techniques for load forecasting which is independent of the human
experience [37,38]. In the hybrid algorithm PSO is used as a main frame while GA is used as local
search that enables PSO to jump out of the local optima. In this way GA-PSO-NN gives a superior
generalization capability, low prediction error and optimum network.
When ANN is optimized by a single optimization method such as GA or PSO then it suffers
from well-known drawbacks. In the present study, we propose a hybrid algorithm called GA-PSO,
which lead to better optimization results. GA-PSO combined optimization algorithm can fully combine
merits of single optimization models without their disadvantages. In order to test the accuracy of the
models, we have compared the forecast results of ANN-GA-PSO models with other models using
single optimization of ANN by GA, single optimization of ANN with PSO, ANN with backward
propagation, ARIMA, HOLTS and linear regression. Mean absolute percentage error has been used
as an indicator of quality of prediction. It is worth mentioning that, for the sake of comparison
among different techniques electricity demand is derived using the same for all the modeling methods.
Results point out that ANN optimized by both GA-PSO in quadratic form (A-G-P-Q) gives the best
performance followed by ANN-G-P model. Consequently A-G-P-Q model is used to forecast the
electricity demand until 2025 based on “as-it-is” scenario and scenario as per the “Vision document”
of the state.

4. ANN-GA-PSO Models
In order to successfully predict Tamil Nadu’s electricity demand efficiently and precisely a hybrid
GA-PSO based ANN model is proposed here in two-form estimation method.

4.1. Two Form Estimation Method

The authors have used the following equations for the GA-PSO optimization:

N
DGA−pso−linear = ∑ (Yi ∗ Xi + Wo ) (3)
i =1

N  
∑ (Yi ∗ Xi + Wo ) + ∑ iN=1 Kij ∗ Xi ∗ X j + ∑ iM 2


DGA−Pso−Quadratic = =1 Ui ∗ Xi (4)
i =1

where D is the electricity demand; Xi , X j are the factors affecting ith and jth factors affecting electricity
energy demand; Wo , Yi , k ij and Ui are the coefficients and N is the number of demand-affecting factors.
PSO searches for the best fitted members that minimize the error. PSO optimizes the weights of
socio economic indicators by using both linear and quadratic regression models. Based on these two
variations of PSO, models have been named ANN-PSO (Linear) and ANN-PSO (Quadratic) respectively.
In PSO-Quadratic, the coefficients of the input variables are calculated as per the Equation (4). For the
Quadratic PSO model the quadratic terms are introduced in the following evolution equations:

vt+1 = vt + R1 ∗ C1 ∗ sign ( g − xt ) ∗ ( g − xt )2 + R2 ∗ C2 ∗ sign ( p − xt ) ∗ ( p − xt )2 (5)

x t +1 = x t + v t +1 (6)
Energies 2018, 11, 728 6 of 15

Quadratic PSO algorithm improves the diversity of the swarm leading to higher performance in
global optimization. Quadratic PSO projects the input variables for the years 2001 to 2015 based on the
data from 1991 to 2000 as input.
In GA, N represents the number of the particles in the population; f i as the fitness value for the
individual i. The population size particles are reproduced on the position of the particles using the
following equation.
fi
pi =
( f s − f max )
where f max is thelargest fitness value in the generation and pi represents the probability for the
selection of the individual i. The crossover and the mutation operations are implemented with pi and
pm according to following equations:

X tA+1 =∝ ∗ XBt + (1− ∝) ∗ X tA

X tA+1 =∝ ∗ X tA + (1− ∝) ∗ XBt

where X tA and XBt are cross over chromosomes. ∝ is a parameter that is constant.

4.2. GA-PSO Hybrid Optimization Algorithm

In most of the research papers on the subject, either GA or PSO has been used as single
optimization algorithm [39,40]. But our research puts forward hybrid GA-PSO algorithm, where GA
and PSO are applied serially for providing the best optimizing solution for ANN. PSO optimization
is applied to a population of 100 particles and the position and velocity of particle that give the
best objective function is arrived at and is designated as ‘pbest’. This current best fitness position
is compared with the global best. The best global position obtained after PSO optimization is taken
as selection value for the GA optimization. In our research paper, the GA further optimizes the best
solution thrown up by the PSO. It has been found in our research that hybrid optimization of GA-PSO
therefore gives a better solution as compared with single optimization by GA or PSO.As shown in
Figure 1 the iterative approach of GA-PSO followed in the study is as follows:

Step 1: First, we initialize a population size of 100 and assign positions and velocities of particles.
The number of weights and biases are used to calculate the fitness function for all the particles.
Step 2: The best position value achieved by particle p is set as pbest. The pbest with best value is set as
gbest and this value is stored.
Step 3: The desired optimization fitness function f (x) is evaluated for each particle.
Step 4: The evaluated fitness value fp of each particle is compared with its pbest value. If fp < pbest
then pbest = fp and bestxp = xp, where xp represents the current coordinates of particle p and
bestxp represents the coordinates corresponding to particle p’s best fitness so far.
Step 5: After objective function value is calculated for new positions of each particle the overall best
fitness value of the swarm becomes the gbest value of the swarm.
Step 6: Next, the velocity and location of the particle is updated according to Equations (1) and (2).
The best position is fed into the General Algorithm as selection.
Step 7: The calculation is stopped when the maximum number of iteration reaches 200 or if the
convergence occurs before it otherwise Loop to step 3 until convergence. In the present study,
the convergence occurs around 50 iterations as shown in Figure 2.
Step 8: The pop size of M particles obtained by GA and M particles are combined to form new pop
size particles.
Step 9: Let gen = gen + 1, then step 3 is carried out.
Step 10: The best fitness values and solutions, namely, the position are outputted.
Energies 2018, 11, 728 7 of 15

Figure 1. Flow chart of ANN-GA-PSO.

Figure 2. Convergence Speed of ANN-GA-PSO algorithm.

Energies 2018, 11, 728 8 of 15

4.3. Computational Environment and Data Management

All the GA and PSO techniques have been developed in open source SCILAB environment.
For application of ARIMA, HOLTS and Linear models standard econometric toolboxes of IBM SS
Software (Version 2, developed by IBM, Armonk, NY, USA) has been used for ANN simulation.
It is designed to provide the necessary tools as a part of standard ANN algorithms and relevant
analysis. In this research study, the GSDP data is measured in rupees and per capita energy intensity
in KWh. The coefficients of predictors as obtained from GA-PSO optimization are depicted in Table 4.
The factors are normalized according to the following equation for optimum functioning of ANN.

n(x) = (x − xmin)/xmin (7)

Table 4 shows the relative values of the independent variables GA-PSO optimization that are used
for ANN simulation where E.Con, Income, GSDP, CPI are the input variables. Table 5 indicates the
coefficients of Equation (3) obtained by using GA-PSO optimization.

Table 4. Normalized Values of GA-PSO-Quadratic.

Year E.Con Income GSDP CPI Sq-E.Con Sq-Income Sq-GSDP Sq-CPI

2001 0 0.153846 0 0 0 0.023669 0 0
2002 0.313544 0.153846 0.363636 0.038835 0.09831 0.023669 0.132231 0.001508
2003 0.372913 0.153846 0.636364 0.058252 0.139064 0.023669 0.404959 0.003393
2004 0.447124 0.230769 0.727273 0.067961 0.19992 0.053254 0.528926 0.004619
2005 0.595547 0.307692 0.636364 0.116505 0.354677 0.094675 0.404959 0.013573
2006 0.781076 0.538462 0.818182 0.135922 0.61008 0.289941 0.669421 0.018475
2007 0.855288 0 0.181818 0.203883 0.731517 0 0.033058 0.041568
2008 0.855288 0.076923 0.272727 0.320388 0.731517 0.005917 0.07438 0.102649
2009 1.003711 0.461538 0.818182 0.466019 1.007435 0.213018 0.669421 0.217174
2010 0.929499 0.307692 0.636364 0.61165 0.863969 0.094675 0.404959 0.374116
2011 0.992579 0.384615 0.545455 0.582524 0.985213 0.147929 0.297521 0.339335
2012 1.074212 0.384615 0.636364 0.543689 1.15393 0.147929 0.404959 0.295598
2013 1.153989 0.230769 0.818182 0.524272 1.33169 0.053254 0.669421 0.274861
2014 2.953618 0.384615 2.818182 0.398058 8.723858 0.147929 7.942149 0.15845
2015 2.723562 0 2.454545 0.349515 7.417791 0 6.024793 0.12216
2016 3.022263 0.153846 2.909091 0.38835 9.134076 0.023669 8.46281 0.150815
Year X12 X13 X14 X23 X24 X32 X34 Demand
2001 0 0 0 0 0 0 0 0
2002 0.048237 0.055944 0.012176 0.055944 0.005975 0.055944 0.014122 0.053338
2003 0.057371 0.097902 0.021723 0.097902 0.008962 0.097902 0.03707 0.261141
2004 0.103183 0.167832 0.030387 0.167832 0.015683 0.167832 0.049426 0.359068
2005 0.183245 0.195804 0.069384 0.195804 0.035848 0.195804 0.074139 0.40199
2006 0.420579 0.440559 0.106166 0.440559 0.073189 0.440559 0.111209 0.352862
2007 0 0 0.174379 0 0 0 0.03707 0.544453
2008 0.065791 0.020979 0.274024 0.020979 0.024645 0.020979 0.087379 0.462792
2009 0.463251 0.377622 0.467749 0.377622 0.215086 0.377622 0.381289 0.56411
2010 0.286 0.195804 0.568529 0.195804 0.1882 0.195804 0.389232 0.654492
2011 0.381761 0.20979 0.578201 0.20979 0.224048 0.20979 0.317741 0.692192
2012 0.413158 0.244755 0.584037 0.244755 0.209111 0.244755 0.345984 0.815053
2013 0.266305 0.188811 0.605004 0.188811 0.120986 0.188811 0.42895 0.99538
2014 1.136007 1.083916 1.175712 1.083916 0.153099 1.083916 1.121801 1.050139
2015 0 0 0.951925 0 0 0 0.857899 1.11105
2016 0.464964 0.447552 1.173695 0.447552 0.059746 0.447552 1.129744 1.211196
Energies 2018, 11, 728 9 of 15

Table 5. Coefficients of GA-PSO Linear.

Year E.Con Income GSDP CPI x1 x2 x3 x4 x5

2016 2167 14.88 43.06 142.9 −1.93 0.91 −1.005 −1.63 −1.14
2017 2341 17.11 48.23 147.22 −1.94 0.54 −1.56 −0.41 −1.25
2018 22 19.68 54 151.64 −1.98 −0.1 −0.179 −1.64 −1.12
2019 2730 22.63 60.5 156.19 −2 0.39 −1.05 −0.86 0.77
2020 2949 26.03 67.76 160.87 −2 −1.4 −0.68 −0.76 0.99
2021 3185 29.93 75.89 165.7 −1.99 0.19 −1.58 −1.23 −0.63
2022 3439 34.4 85 170.7 −2 −1 −1.95 −1.2 −0.31
2023 3715 39.58 95.2 175.8 −1.99 0.28 0.28 −1.71 0.62
2024 4012 45.52 106.6 181 −2 0.41 −0.9 −0.21 −0.49
2025 4333 52.35 119.42 186.5 −2 −1.26 0.127 −0.76 0.03

4.4. Evaluation of the Forecast Performance

Root Mean Square Error (RMSE) or Mean Absolute Error (MAE) is commonly used as a measure
of forecasting performance. However, Fatai and Armstrong [41] have negated RMSE or MAE, as both
are scale dependent and RMSE is affected by outliers that are common in electricity forecasting.
Weron R [42] has asserted that MAPE is the most popular evaluation index that works well in load
forecasting. Therefore, in order to compare predictive accuracy of the ANN-GA-PSO models, we have
used mean absolute percentage error (MAPE) as the evaluation index. The MAPE and forecasting
accuracy (τ) have been defined as follows:

100 n
n t∑
MAPE = |( At − Ft )/At |
=1

τ = 1 − |( At − Ft )/At | if |( At − Ft )/At | < 1

τ = 0 if |( At − Ft )/At | ≥ 1

where At is the actual value and Ft is the forecast value.

The total electricity demand of Tamil Nadu from year 2001 to 2015 has been used as a benchmark
to test the effectiveness and superiority of the proposed ANN-GA-PSO models. First, ARIMA
(1,0,1), HOLTS and linear models have been employed to calculate electricity demand. Secondly
the simple optimization of ANN is performed by GA and PSO separately and the results are
tabulated under ANN-GA and ANN-PSO respectively. The optimum weight coefficients of GA-PSO
optimization are obtained from Equations (3) and (4) for ANN-GA-PSO in linear and quadratic forms
respectively. For the sake of verifying the validity and superiority of the proposed ANN-GA-PSO
models, the comparison is also made with ANN-BP model.

5. Results
Table 6 and Figure 3 shows results of ANN-G-P and A-G-P-Q models in both linear and quadratic
forms along with simple optimization models, ANN-PSO and ANN-GA. Figure 4 compares the errors
of linear, Time series models (Holts and ARIMA), ANN-GA, ANN-PSO, ANN-G-P and A-G-P-Q
models. Table 7 and Figure 5 compare the MAPE values of different models. It can be seen that
MAPE of A-G-P-Q (0.2%) and ANN-G-P (0.3%) are far better than MAPE of single optimized models
of ANN-GA (0.42%) and ANN-PSO (0.4%). Table 8 depicts the forecasting accuracy (τ) of different
models. It is clear that τ of A-G-P-Q model at 0.78 followed by ANN-G-P at 0.7 are far superior to single
optimization models. Figure 6 compares the result of the ANN-G-P (Linear) and A-G-P-Q (Quadratic)
model against the actual values of the electricity demand from the year 2001 to 2015. ANN-G-P and
A-G-P-Q are in close agreement with the actual values. The forecasts of A-G-P-Q are compared with
actual demand on a logarithmic scale in Figure 7. It is seen that the relationship between the two is
Energies 2018, 11, 728 10 of 15

linear and the slope is 0.99. Thus A-G-P-Q model is best suited for forecasting the electricity demand
for the year 2016 to 2025.

Figure 3. Performance of models.

Figure 4. Error comparison of various models.

Figure 5. MAPE (in %).

Energies 2018, 11, 728 11 of 15

Figure 6. Comparison of ANN-G-P & A-G-P-Q.

Figure 7. Forecasting by A-G-P-Q model.

Table 6. Performance of different models.

Year Act Total ANN-Pso Linear Holts ANN-BP ANN-G-P ARIMA ANN-GA A-G-P-Q
2001 36,578 36,206 39,441 37,643 36,434 36,705 36,018 36,582
2002 38,529 38,302 43,532 40,247 38,987 38,854 39,876 38,618 38,827
2003 46,130 46,180 44,614 42,787 46,337 46,109 43,671 46,192 46,238
2004 49,712 50,054 45,595 45,829 49,786 49,484 49,214 49,323 49,731
2005 51,282 51,007 48,299 48,925 51,254 51,540 51,458 51,179 51,611
2006 49,485 49,394 50,630 51,870 49,643 49,949 52,069 49,707 49,640
2007 56,493 56,927 53,094 54,343 56,282 56,546 53,244 56,795 56,586
2008 53,506 53,257 56,060 57,267 53,719 53,792 56,676 53,201 53,404
2009 57,212 57,172 61,235 59,603 57,404 57,720 58,214 57,303 57,383
2010 60,518 60,737 64,208 62,076 60,205 60,465 62,391 60,302 60,522
2011 61,897 62,353 64,090 64,631 62,098 61,757 64,313 62,024 62,011
2012 66,391 66,593 63,920 67,069 66,515 66,282 65,730 66,713 66,378
2013 72,987 73,023 64,302 69,712 72,635 73,126 69,779 73,164 73,160
2014 74,990 74,890 78,675 72,748 74,464 75,084 72,866 74,898 75,109
2015 77,218 77,285 75,235 75,681 77,818 76,870 78,189 76,930 77,242
Energies 2018, 11, 728 12 of 15

Table 7. MAPE VALUES (percentage).

Linear Holts ARIMA ANN-BP ANN-GA ANN-P ANN-G-P A-G-P-Q

6.07 0.85 3.02 0.44 0.42 0.4 0.3 0.22

Table 8. Forecasting Accuracy (τ) .

Linear Holts ARIMA ANN-BP ANN-GA ANN-P ANN-G-P A-G-P-Q

0 0.15 0 0.56 0.58 0.6 0.7 0.78

5.1. Future Estimation

The future estimation of the electricity demand of Tamil Nadu has been evaluated under two
scenarios. Scenario 1 (as it is) assumes the energy consumption to grow at the rate of 5%, income at
the rate of 12%, GSDP at 11% and CPI at 2%. Scenario 2 considers the VISION Document 2023 [43]
goals of the state as expected growth rate of energy consumption as 8%, income growth as 15%, GSDP
as 12% and CPI at 3%. Table 9 shows the tabulated results of the forecasted electricity demand for
scenario 1 and scenario 2 using A-G-P-Q model. Figure 8 shows the forecasted electricity demand as
per scenario 1 and scenario 2. The projected electricity demand as per scenario 2 are on the higher
side throughout except for the year 2020. The electricity requirement for the year 2025 is 84 GWh as
compared to 87.8 GWh as per scenario 1 and scenario 2 respectively. The state of Tamil Nadu will have
to find resources for fulfilling the demand of 87.8 GWh if it wants to achieve the goals set up by the
Vision 2023 document.

Table 9. Demand Forecast.

Year Scenario 1 Scenario 2

2016 80,881 80,537
2017 81,213 83,324
2018 81,142 82,726
2019 82,137 84,301
2020 83,044 81,074
2021 82,752 83,469
2022 83,029 85,331
2023 83,826 87,581
2024 83,401 85,636
2025 84,263 87,825

Figure 8. Forecasts as per scenario 1 and scenario 2 using A-G-P-Q model.

Energies 2018, 11, 728 13 of 15

5.2. Relationship between GSDP and Electricity Demand

According to Kostyannikova D [44], the causality and co-integration relationship between GSDP
and electricity demand are not uniform across countries due to difference in policies and energy
structure. Our present study shows that electricity demand and GSDP are co-integrated in the case
of Tamil Nadu. As shown in Figure 9 one percent increase in total energy consumption leads to an
increase of 0.86 in GDP while one percent increase in GSDP will raise total energy consumption by
0.79 percent.
Our research shows that in case of Tamil Nadu, causality exists between GSDP and electricity
demand. Hence it will be possible to increase the GSDP by investing in bridging the electricity
demand gap.

Figure 9. Relationship between Electricity demand and GSDP.

6. Conclusions
This study has proposed a novel algorithm based on PSO and GA for optimizing ANNs in
linear and quadratic forms for forecasting of electricity demand. ANN has been optimized by the
hybrid optimizing algorithm of PSO and GA in linear and quadratic forms. Single optimized ANN
(ANN-GA, ANN-PSO) have been compared with hybrid optimized ANN’s (ANN-GA-PSO, A-G-P-Q).
ANN-GA-PSO models in linear and quadratic forms have demonstrated 28% and 48% improvement
over ANN-GA model and 25% and 43% improvement over ANN-PSO model. ANN-GA-PSO models
can solve the problem of over fitting and falling in local minimum in data set ANN-GA-PSO model
shave been used to explore the relationship between electricity demand and GSDP of Tamil Nadu
state which is seen as co-integrated. ANN-GA-PSO models can be used for resource planning and for
bridging the energy gap in the state to achieve the goals set out in the Vision document of the state.

Acknowledgments: The author will like to thank the Tamil Nadu Electricity Board, Electronics Corporation of
Tamil Nadu and the Centre for Development of Advanced Computing for their continuous support. Venu, Eshani
and Amay for their assistance in trouble shooting in programming.
Author Contributions: L Suganthi proposed the idea and gave inspiration for the development. Atul Anand
collected the data and established the forecasting model.
Conflicts of Interest: The authors declare that there is no conflict of interest.
Energies 2018, 11, 728 14 of 15

References
1. Schweizer, V.J.; Morgan, M.G. Bounding US electricity demand in 2050. Technol. Forecast. Soc. Chang. 2016,
105, 215–223. [CrossRef]
2. Perez-Garcia, J.; Moral-Carcedo, J. Analysis and Long term forecasting of electricity demand trough a
decomposition model. A case study for Spain. Energy 2016, 97, 127–143. [CrossRef]
3. Granger, C.W.J. Combining forecast-twenty years later. J. Forecast. 1989, 8, 167–173. [CrossRef]
4. Parikh, J.; Purohit, P.; Maitra, P. Demand projections of petroleum products and natural gas in India. Energy
2007, 32, 1825–1837. [CrossRef]
5. Zhang, M.; Mu, H.; Li, G.; Ning, Y. Forecasting the transport energy demand based on PLSR method in
China. Energy 2009, 34, 1396–1400. [CrossRef]
6. Limanond, T.; Jomnonkwao, S.; Srikaew, A. Projection of future transport energy demand of Thailand.
Energy Policy 2011, 39, 2754–2763. [CrossRef]
7. Jenkins, G.M.; Reinsel, G.C. Time Series Analysis: Forecasting and Control; Holden Day: San Francisco, CA,
USA, 1976.
8. Ediger, V.S.; Akbar, S. ARIMA forecasting of primary energy demand by fuel in Turkey. Energy Policy 2007,
35, 1701–1708. [CrossRef]
9. Tepedino, C.; Guarnaccia, C.; Iliev, S.; Popova, S.; Quartieri, J. A Forecasting Model Based on Time Series
Analysis Applied to Electrical Energy Consumption. Int. J. Math. Models Methods Appl. Sci. 2015, 9, 432–445.
10. O’Connell, N.; Pinson, P.; Madsen, H. Benefits and Challenges of electrical demand response: A critical
Review. Renew. Sustain. Rev. 2014, 39, 686–699. [CrossRef]
11. Trotter, I.M.; Blkesjo, T.F.; Feres, J.G.; Hollanda, L. Climate Change and Electricity Demand in Brazil:
A stochastic Approach. Energy 2016, 102, 596–604. [CrossRef]
12. Fan, G. A method to estimating the parameters of logistic model and application. Math. Econ. 2010, 27,
105–110.
13. Kandananond, K. Forecasting Electricity Demand in Thailand with an Artificial Neural Network Approach.
Energies 2011, 4, 1246–1257. [CrossRef]
14. Amjady, N.; Keynia, F. Day ahead forecasting of electricity markets by a mixed data model and hybrid
forecast method. Int. J. Electr. Power Energy Syst. 2008, 30, 533–546. [CrossRef]
15. Abdul Hamid, M.B.; Abdul Rahman, T.K. Short Term Load Forecasting Using an Artificial Neutral Network
Trained by Artificial Immune System Learning Algorithm. In Proceedings of the 12th International
Conference on Computer Modelling and Simulation, Cambridge, UK, 24–26 March 2010.
16. Cincotti, S.; Gallo, G.; Ponta, L.; Raberto, M. Modeling and forecasting of electricity spot prices:
Computational intelligence vs. Classical econometrics. AI Commun. 2014, 27, 301–314.
17. Yin, F.; Wang, J.; Guo, C. Advances in Neural Networks ISNN 2004. In Proceedings of the International
Symposium on Neural Networks, Dalian, China, 19–21 August 2004; Springer: Berlin/Heidelberg,
Germany, 2014.
18. Goldberg, D.E. Genetic Algorithms in Search, Optimization, and Machine Learning; Addison-Wesley Publishing
Corporation, Inc.: Boston, MA, USA, 1989.
19. Canyurt, O.E.; Ozturk, H.K. Application of genetic algorithm (GA) technique on demand estimation of fossil
fuels in Turkey. Energy Policy 2008, 36, 2562–2569. [CrossRef]
20. Ceylan, H.; Ozturk, H.K. Estimating energy demand of Turkey based on economic indicators using genetic
algorithm approach. Energy Convers. Manag. 2004, 45, 2525–2537. [CrossRef]
21. Haldenbilen, S.; Ceylan, H. Genetic Algorithm approach to estimate transport energy demand in Turkey.
Energy Policy 2005, 33, 89–98. [CrossRef]
22. Assareh, E.; Behrang, M.A.; Assari, M.R.; Ghanbarzadeh, A. Application of PSO and GA techniques on
demand estimation of oil in Iran. Energy 2010, 35, 5223–5229. [CrossRef]
23. Bi, T.; Yan, Z.; Wen, F.; Ni, Y.; Shen, C.M.; Wu, F.F.; Yang, Q. On-line fault section estimation in Power Systems
with radial basis function neural network. Int. J. Electr. Power Energy Syst. 2002, 24, 321–328. [CrossRef]
24. Lu, N.; Zhou, J. Particle Swarm Optimization-Based RBF Neural Network Load Forecasting Model.
In Proceedings of the 2009 Asia-Pacific Power and Energy Engineering Conference, Wuhan, China,
27–31 March 2009.
Energies 2018, 11, 728 15 of 15

25. Banda, E.; Folly, K.A. Short Term Load Forecasting Based on Hybrid ANN and PSO. In Advances in Swarm
and Computational Intelligence, Proceedings of the International Conference in Swarm Intelligence, Beijing, China,
25–28 June 2015; Springer: Cham, Switzerland, 2015; pp. 98–106.
26. Yang, S.D.; Li, X. A new ANN optimized by improved PSO algorithm combined with chaos and its
application in short term load forecasting. In Proceedings of the 2006 International Conference on
Computational Intelligence and Security, Guangzhou, China, 3–6 November 2006; Volume 2.
27. Kennedy, J.; Eberhart, R.C. Particle swarm optimization. In Proceedings of the IEEE International Conference
on Neural Networks, Perth, Australia, 27 November–1 December 1995; pp. 1942–1947.
28. Shi, Y.; Eberhart, R.C. Parameter selection in Particle Swarm Optimization. In Evolutionary Programming;
Springer: Berlin/Heidelberg, Germany, 1998; Volume 7, pp. 611–616.
29. Shi, Y.; Eberhart, R.C. Fuzzy adaptive particle swarm optimization. In Proceedings of the 2001 Congress on
Evolutionary Computation, Seoul, Korea, 27–30 May 2001.
30. Anand, A.; Suganthi, L. Forecasting of Electricity Demand by Hybrid ANN-PSO Models. Int. J. Energy
Optim. Eng. 2017, 6, 66–83. [CrossRef]
31. Bates, J.M.; Granger, C.W. The combination of forecasts. Oper. Res. Soc. 1969, 20, 451–468. [CrossRef]
32. Nazari1, H.; Kazemi1, A.; Hashemi, M.-H. Selecting the appropriate scenario for forecasting energy demands
of residential and commercial sectors in Iran using two metaheuristic algorithms. Iran. J. Manag. Stud. 2016,
9, 101–123.
33. Unler, A. Improvement of energy demand forecasts using swarm intelligence: The case of Turkey with
projections to 2015. Energy Policy 2008, 36, 1937–1944. [CrossRef]
34. Younes, M.; Farid, B. Genetic Algorithm-Particle Swarm Optimisation (GA-PSO) for Economic Load Dispatch.
Prz. Elektrotech. 2011, 87, 369–372.
35. EL-Araby, E.E.; Yorino, N. A hybrid PSO technique for procuring VAR ancillary service in the deregulated
electricity markets. Int. J. Electr. Power Energy Syst. 2010, 32, 664–670. [CrossRef]
36. Jarndal, A.; Hamdan, S. Forecasting of peak electricity demand using ANN GA and ANN-PSO approaches.
In Proceedings of the 7th International Conference on Modeling, Simulation, and Applied Optimization
(ICMSAO), Sharjah, UAE, 4–6 April 2017.
37. Ekonomou, L. Greek long-term energy consumption prediction using artificial neural networks. Energy 2010,
35, 512–517. [CrossRef]
38. Bunn, D.W. Forecasting with more than one model. J. Forecast. 1989, 8, 161–166. [CrossRef]
39. Donaldson, R.G.; Kamstra, M. Forecast combining with neural networks. J. Forecast. 1996, 15, 46–91.
[CrossRef]
40. Chen, H.Y. Research on combination forecasting model based on effective measure of forecasting methods.
Forecasting 2001, 20, 72–73.
41. Fatai, K.; Oxley, L.; Scrimgeour, F.G. Modeling and Forecasting the demand for electricity in New Zealand:
A comparison of alternative approaches. Energy J. 2003, 24, 75–102. [CrossRef]
42. Weron, R. Electricity price forecasting: A review of the state-of -art with a look in to the future. Int. J. Forecast.
2014, 30, 1030–1081. [CrossRef]
43. Vision Tamil Nadu 2023. Available online: www.spc.tn.gov.in (accessed on 18 January 2018).
44. Kostyannikova, D. Economic Growth and Energy Consumption in OECD Countries: A Causality Analysis.
Master’s Thesis, City College of New York, New York, NY, USA, 2012.

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).