Optimization of Revenue Generated by Hydro Power Plant by Bat Algorithm (BA)

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 5

Optimization of Revenue Generated by Hydro

Power Plant by Bat Algorithm(BA)


Shrey Gupta Kapil Sharma
Department of Computer Engineering Department of Computer Engineering
Delhi technological University Delhi technological University
Delhi,INDIA Delhi,INDIA
Email Id: shreygupta111@gmail.com Email Id: kapil@ieee.org

Abstract ±Electricity is prime driver of growth and has varied optimization by using particle swarm optimization (PSO) to
range of uses spanning transportation,production of goods, maximize the benefit cost ratio.In this model ,Author had made
industries , domestic uses and many. In countries , like India , the certain modifications to cost data and design parameters and
demand for electricity has always been more than the supply and used PSO,an optimization tool,which is a metaheuristic
hence the country face power shortage. Electricity is generated by
algorithm that provides a population based search.Different
non-renewable resources like coal , oil , fossil fuels , nuclear power
and natural gas and from renewable resources like wind optimisation algorithms have different performances in
energy,hydroelectric power plants,biomass ,solar different type of problems [2].
energy,reservoirs.Hydroelectric power plants are major source of Bat Algorithm was formulated by Author [3] for continuous
electricity in india with presently installed capacity of 41,997.42 constrained optimization problems.In comparison to other
MW as on July 31, 2015 which is 15.22% of total electricity metaheuristic algorithms like particle swarm optimization ,
generation in India. Hydroelectricity means production of genetic algorithms and Harmony Search , Bat Algorithm is
electrical power through the use of the gravitational force of proven to be potentially more powerful and it delivers most
falling or flowing water. It is significant to optimize the revenue optimized solutions.Thus,making use of Bat Algorithm to
generated by the hydroelectric power plant to meet the
optimize the revenue generation of hydro power plant is the
implementation cost and for expansion of hydroelectric power
plants. A lot of research work has been done in optimization of most optimized technique to deliver great results in short
hydroelectric flow. In this paper, a new calibrated revenue time.Bat Algorithm is an evolutionary strategy based on the
optimization model is developed with Multi-Objective Bat echolocation behaviour of the bats. In this method,a number of
Algorithm (MOBA) for hydro power plants . This model can be bats are initialised with random values which then find the best
used to increase the revenue generated by the hydroelectric flow among them and on the principal of the survival of the
from the dam wih respect to given constraints in less time. Bat fittest,the most fit bat is selected for evolution and hence
Algorithm is the latest metaheuristic algorithm which is based on converges towards the better solution in each evolution
the behaviour of bats . Bat Algorithm has received increased (iteration) and hence optimising the result.Randomly allocation
attention in many research fields recently.Bat algorithm based
of values to the parameters of the bats give the advantage of
optimization model can predict the flow of water through turbine
in hydroelectric power plant in order to maximize the revenue. bats' travelling through every possible optimized solution,thus
always giving the most optimized solution.
Keywords ± Application of bat-algorithm ; Bat algorithm; In this model,we have optimized the revenue generated by
electricity production ; hydroelectric flow ; Optimization ; revenue hydro power plant by the flow of water from the dam to run the
maximization. turbines which generates hydroelectricity.Overflow water is
carried through spill flow lines.The prices of electricity varies
I. INTRODUCTION with time,thus generating more electricity at the time of higher
The Economically optimal design of hydro power plant by prices will lead to higher revenue . We have implemented this
maximizing revenue and minimizing cost and power loss has model which tells at what time how much water should go
been known for a long time. However, these models are not through turbines and how much should go through spillway.
fully optimized due to pronounced complexity of hydro power
plant and its optimization model. In the recent years several II. LITERATURE REVIEW
optimization techniques have emerged including evolutionary Authors [4] has studied artificial bee colony (ABC) algorithm
methods that have proved great significance in optimization and presented that ABC is well suited for cluster analysis,
problems where it was difficult to apply mathematical analysis constrained problem optimization, structural optimization, and
due to the entanglement of the problem.Hydro power plant expert system. ABC has played important role in medical, MRI
revenue optimization is an optimization problem. Authors [1] image classification, face recognition, bioinformatics etc. ABC
proposed an optimization model for hydro power plant is also an metaheuristic algorithm based on the behavior of

978-9-3805-4421-2/16/$31.00 2016
c IEEE 3576
honey bees in their colony and has evolutionary working BA iteratively finds the optimal solution of the target problem
strategy,hence useful for optimization problems.Optimization through a guided search (i.e., the movements of a number of
techniques are highly required in almost every problem domain bats) over the solution space, by using fitness function and
to achieve best results in least time. Authors [5] has proposed evolving bats. K bats are applied to find the optimal solution in
Bacterial Foraging Optimization Algorithm (BFOA) to a solution space.Multi-Objective Bat algorithm is used to keep
estimate the parameters of cost estimation. models and to in consideration various factors,functions and constraints.Bat
determine which cost estimation model gives best results based Algorithm is a highly efficient algorithm with remarkable
on the Least Square Estimation fitting technique [6]. These results.Recent researches have proved the superiority of Bat-
models can be used further to solve various optimisation Algorithm over other metaheuristic algorithms like Simulated
problem. Author [7] and author [8] had discussed the Annealing (SA),Particle Swarm Optimization(PSO) , Genetic
importance of hydroelectricty as a renewable alternative for Algorithm (GA)[13]. Bat Algorithm can be hybridised with
electricity generation and various constraints on the model other algorithms like COCOMO-BAT Algorithm,which can
including the variations in water flow,regulation of turbines and reduce the errors significantly [14].Most real-world engineering
various researches so far.They have also discussed the control optimization problems are nonlinear with complex
model for hydroelectric energy-value optimization that constraints.BA has been tested against eight non-linear
focusses on the the load on the turbines and hourly sequence of optimisation tasks which served as the benchmark and BA
flow of water. showed excellent results,better than other metaheuristic
The process of hydroelectricity generation has evolved over algorithms [15].
years resulting in large amount of hydroelectricity being
x Initialize the parameters (pulse frequency (f) ,pulse rate(r)
produced at very low costs.It has evolved from running the and loudness(A)).
turbine by water to produce electric spark in the air between x Initialize the positions of K Bats(Randmly)
two conductors producing a light to a dynamo connected to a x For the construction-step index n = 1 : N,
turbine in a flour mill providing street lighting to hydro power ʹ For the Bat index k = 1 : K,
plants to produce high amounts of power that can be stored and ʹ Consecutively update the position of K-th bat.
ʹ Select a solution among the best solutions.
transmitted to longer distances. Bat-Algorithm can be used to ʹ Create new solutions(evolution) by moving bats randomly
optimize the results of turbine based energy production model or around the best selected solution.
[9]. x Different weights are assigned to different functions and
Bat Algorithm is easy to implement and superior in terms of then above process is carried out.Make the final solution
accuracy and efficiency compared to other algorithms.It has decision according to the selected bats' positions.
good convergence property and results shows that it is better in
quality of solution than PSO and other metaheuristic
algorithms.This algorithm is capable of finding the feasible Fig. 1. Steps in the Bat-optimization model
solution near global optimal solution with less computation
cost [10], [11]. IV. FORMULATION OF THE HYDROELECTRIC FLOW
OPTIMIZATION PROBLEM
III. BAT ALGORITHM
The objective of hydroelectric flow optimization for electric
Bat-inspired algorithm(BA) is a meta power generation is to schedule the committed generating unit
heuristic optimization algorithm developed by Xin-She Yang in outputs so as to generate the maximum revenue while
2010.Bat algorithm is based on the echolocation behaviour satisfying all units and operational constraints of the power
of micro bats with varying pulse rates of emission and loudness system. The hydroelectric flow problem is a constrained
[12].Following are the idealised rules of the bat algorithm- optimization problem and it can be mathematically expressed
1. All bats are able to identify their preys or objects by as follows:
echolocation.This echolocation helps them to sense 1. The empirical equation for the turbine, and relates the
distance between the object and the bat. electricity produced to the turbine flow and reservoir level
2. %DWV À\ UDQGRPO\ ZLWK YHORFLW\ Y L  DW SRVLWLRQ [ L 
turbineFlow( t  1) *
ª k1 * ( Storage(t )  Storage(t  1))  k º
ZLWKD¿[HGIUHTXHQF\IPLQDQGYDULHVZDYHOHQJWKȜ Electricty (t )
¬« ¼»
2
2
and loudness A to search for prey. They can
automatically adjust the wavelength (or frequency) of 2. The reservoir storage level as a function of the flow into and
their emitted pulses and adjust the rate of pulse out of the reservoir.
emission r ‫[ א‬0,1], depending on the proximity of their Storage ( t ) Storage ( t  1)  't * >inFlow(t  1)  spillFlow (t  1)  turbineFlow( t  1) @
target; 3.Revenue generation.
3. We can generally keep the loudness(A) and rate of
pulse emission constant,after tuning to attain best
revenue ¦ turbineFlow(i ) * price(i )
solution.
Where t : time,

2016 International Conference on Computing for Sustainable Global Development (INDIACom) 3577
k1 , k 2 :constants VI. RESULTS AND COMPARISONS
Water can leave the reservoir either through the spillway or via We ran our system on various sizes of test cases and it was
the turbine. Water that leaves through the turbine is used to observed that Bat-Algorithm based optimized solution is faster
generate electricity. We can sell this electricity at prices that are than other solution,giving the similar profit optimization. In
determined by various market factors. We control the rates at optimization problems,we need find an alternative with the
which water flows through the spillway and the turbine. Our most cost effective or highest achievable performance under
goal is to find values for these rates that will maximize revenue the given constraints, by maximizing desired factors and
over the long term. minimizing undesired ones. In comparison, maximization
Constraints: Certain constraints need to be taken into means trying to attain the highest or maximum result or
consideration. Both the reservoir level and downstream flow outcome without regard to cost or expense.In this paper we
rates must remain within specified ranges. The maximum level have made use of bat-algorithm in order to solve hydroelectric
of flow that the turbine can handle is set, but the spillway level flow optimization problem by increasing the revenue and
is unrestricted. Additionally, the reservoir level at the end of minimising the the computation time.It is interesting to note
the simulation must be the same as the level at the start of the that the bat-algorithm doesn't give the same solution on every
simulation. The power plant will continue to operate beyond run due to its characteristics of randomly searching bats with
the period for which we are optimizing, so we need to prevent random velocity in every iteration ,which makes it relatively
the solver from finding a solution that empties the reservoir. better method to solve such problems.Due to randomisation,it
is able to cover all local minima/maxima,hence finding the best
solutions in the solution space.The following table shows the
V. SYSTEM MODEL comparisons :
We have used Multi-Objective Bat Algorithm(MOBA)in
MATLAB.Our algorithm contains following steps: Decision Non-linear method Bat-algorithm
x We have taken two objective function,1.Total waterinlet variables (fmincon)
and outlet;2.Total revenue generated.These function are
Time Revenue Time Revenue
assigned weights sequentially and then the BA is called for
(sec) (Rs) (sec) (Rs)
each set of weights.
x BA parameters(population size,number of 50 71.85 2349.7 27.73 2352
generations,loudness,pulse rate) are set in accordance to 100 268.26 4311 121.3 4155
maximise the result.
x The results are maximised when there is nill spill flow and
complete turbine flow to generate hydroelectricity.Cost
function computes for each bat and the best solution's bat
is selected to evolve for next generation.
x Under the constraints ,like Max change in Turbine and
Spill Flow < 500 CFS, each dimension of the bat is
qualified by the previous dimension,hence one dimension
is computed at a time.
x Finally the best bat of last generation is choosen for the
solution.
FLOW CHART

Fig. 2. Solution by non-linear method


a) Shows the prices of water with respect to time.
b) Shows the solution,that is,the amount of water that should run through
turbine with respect to time.

3578 2016 International Conference on Computing for Sustainable Global Development (INDIACom)
The above plots of the Bat-algorithm shows that there can be
many possible solutions to an optimization problem,and bat
algorithm is capable enough to find them.Though both the
above solution are generating the optimal revenue
approximately,but the computation time is significantly
different.The time taken by the non-linear model is constant,but
the time taken by bat-algorithm can be changed and optimized
by tuning its parameter.In above illustrations,it is clear that bat
algorithm is way faster than non-linear model,it can become
more fast by tuning the parameters further.The main objective
of this paper is to show that bat-algorithm can be employed
easily in varied ranges of problems ,like optimization problems,
car side problem,mathematical problem,engineering design
problems etc., and it provides significant solutions that are
Fig. 3. Solution by Bat-Algorithm method widely acceptable.It takes relatively very less time to span
a) Shows the prices of water with respect to time.. through the solution space and generate best solution.It
b) Shows the solution,that is,the amount of water that should run through provides better scope to accomodate wide number of
turbine with respect to time. constraints and different objectives.This model,hence,is reliable
and useful.
VII. CONCLUSION
Optimization problems in engineering often concerns multiple
objective,hence the problem is called multi-objective
optimization problem,for which multi-objective bat
algorithm(MOBA) is used.Different objectives often conflict
each other,and hence true optimal solution doesn't exist.In such
situations some tradeoffs and approximations are often
needed.Multi objective bat algorithm has the capability to
consider all the objectives,span through all the best possible
solutions in little time and generate the optimal solution.As no
solution is a true solution in this case,hence bat algorithm tends
to provide different solution,giving the same optimal result.The
bat algorithm is a pwerful tool for the problem solver as it
Fig. 4. Alternate solution by Bat-Algorithm method provide the power to solver to accomodate all the constraints
a) Shows the prices of water with respect to time.. and conditions .Proper parmeter tuning can optimize the
b) Shows the solution,that is,the amount of water that should run through solution further.Bat algorithm has the future scope to develop
turbine with respect to time. hybrid algorithm and optimize present algorithms,which can be
highly useful in data centric problem to solve them in less
time.There are also several models that are used for a particular
problem and give the best and most optimized solution.But
those solutions are difficult to be expanded and reused for other
problems.Unlike those,bat algorithm can be employed in vast
ranges of problems. If we can properly control the parameter in
accordance to the problem,if we can make the search faster and
move the bats towards the required solution at a faster rate,then
number of iteration would reduce manifolds and hence
optimising the solution further.Therefore bat algorithm is a
promising technique to solve optimization problem with high
accuracy and less time.

REFERENCES
[1] O.P. Rahi, A.K. Chandel, and M.G. Sharma, "Optimization of Hydro
Fig. 5. Results of large-scale optimization problem solved with quadprog. Power Plant Design by Particle Swarm," in International Conference on
Top left: variations in the price of electricity over the time period; bottom Communication Technology and System Design 2011, Coimbatore, 2012,
left: optimal flow rates for turbine flow and spillway flow; top right: total pp. 418-425.
revenue from electricity sold over the time period; bottom right: total [2] Sankalap Arora and Satvir Singh, "A Conceptual Comparison of Firefly
volume of water stored in the reservoir over the time period Algorithm,Bat Algorithm and Cuckoo Search," in International

2016 International Conference on Computing for Sustainable Global Development (INDIACom) 3579
Conference on Control, Computing, Communication and Materials,
Punjab, 2013.
[3] Xin-She Yang, "A New Metaheuristic Bat-Inspired Algorithm," in Nature
Inspired Cooperative Strategies for Optimization (NISCO 2010) (Eds. J.
R. Gonzalez et al.), Berlin, 2010, pp. 65-74.
[4] Kuldeep Singh Kaswan, Sunita Choudhary, and Kapil Sharma,
"Applications of Artificial Bee Colony Optimization Technique: Survey,"
in 2nd International Conference on Computing for Sustainable Global
Development, Delhi, 2015, pp. 8.106-8.110.
[5] Neha Mittal and Kapil Sharma, "Ranking of Software Cost Estimation
Model Using Bacterial Foraging Optimization Algorithm," in 2nd
International Conference on Computing for Sustainable Global
Development, Delhi, 2015, pp. 8.99-8.105.
[6] Bushra Khalid and Kapil sharma, "Ranking of Software Reliability
Growth Models Using Bacterial Foraging Optimization Algorithm," in
2nd International Conference on Computing for Sustainable Global
Development, delhi, 2015, pp. 8.82-8.87.
[7] Nand Kishor, R.P. Saini, and S.P. Singh, "A review on hydropower plant
models and control," in Renewable and Sustainable Energy Reviews,
Roorkee, 2006, pp. 776-796.
[8] Aris P. Georgakakos, Huaming Yao, and Yongqing Yu, "Control Model
for Hydroelectric Energy-Value Optimization," Journal of Water
Resources Planning and Management, vol. 123, no. 1, pp. 30±38, January
1997.
[9] Tamiru Alemu Lemma and Fakhruldin Bin Mohd Hashim, "Use of Fuzzy
Systems and Bat Algorithm for Exergy Modeling in a Gas Turbine
Generator," in IEEE Colloquium on Humanities, Science and Engineering
Research, Penang, 2011, pp. 305-310.
[10] S. Biswal, A. K. Barisal, A. Behera, and T. Prakash, "Optimal Power
Dispatch Using BAT Algorithm," in Energy Efficient Technologies for
Sustainability (ICEETS), 2013 International Conference, Nagercoil, 2013,
pp. 1018-1023.
[11] Koffka Khan and Ashok Sahai, "A Comparison of BA, GA, PSO, BP and
LM for Training Feed forward Neural Networks in e-Learning Context,"
I.J. Intelligent Systems and Applications, vol. 4, no. 7, pp. 23-29, june
2012.
[12] Xin-She Yang, "Bat Algorithm: Literature Review and Applications," Int.
J. Bio-Inspired Computation, vol. 5, no. 3, pp. 141-149, August 2013.
[13] Ahlem Yamina GHERBI, Hamid BOUZEBOUDJA, and Fatiha LAKDJA,
"Economic Dispatch Problem using Bat Algorithm," Leonardo Journal of
Sciences, no. 24, pp. 75-84, June 2014.
[14] Neha Gupta and Kapil Sharma, "Optimizing Intermediate COCOMO
Model Using BAT Algorithm," in 2nd International Conference on
Computing for Sustainable Global Development, Delhi, 2015, pp. 8.88-
8.92.
[15] Xin-She Yang and Hossein Amir Gandomi, "Bat Algorithm: A Novel
Approach for Global Engineering Optimization," Engineering
Computations, vol. 29, no. 5, pp. 464-483, November 2012.

3580 2016 International Conference on Computing for Sustainable Global Development (INDIACom)

You might also like