You are on page 1of 32

Hybrid Differential Evolution with Biogeography-Based Optimization for Solution of Economic Load Dispatch

Click to edit Master subtitle style

4/8/12

Contents

Introduction ELD problem formulation Biogeography based optimization (BBO) Differential Evolution (DE) DE/BBO approach Test problems and results Conclusion References

4/8/12

Introduction
In power generation our main aim is to generate the required amount of power with minimum cost. Economic load dispatch means that the generators real and reactive power are allowed to vary within certain limits so as to meet a particular load demand with minimum fuel cost This allocation of loads are based on some constraints such as voltage constraints, generator capacity constraints, etc.

Click to edit Master subtitle style

4/8/12

Economic Dispatch Formulation

System generating costs have to be represented mathematically

Costs of generators

operating

(dispatching) with

Indirect costs associated constraints on the system

Then to find the minimum system cost given

Generator costs and System constraints

4/8/12

Generator Costs

The cost to generate a MWh can vary widely

For thermal units we have well-defined equations to calculate generating costs. For other generating units (e.g., hydro and nuclear) the cost is difficult to quantify.

In all practical cases, the fuel cost of generator can be represented as a quadratic function of real power generation
4/8/12

The objective function Ft is given as

ELD problem with losses (ELDQCTL)

subject to constraints:

Real Power Balance Constraint: Generator Capacity Constraints: The transmission losses may be expressed using B-coefficients as
4/8/12

ELD with prohibited zones and ramp rate limits (ELDPOZR) Here the objective function is to be minimized subject
to additional constraints of ramp-rate limits.

As generation increases: As generation decreases:

Prohibited Operating Zones: These are the range of output power of a generator where the operation causes undue vibration of the turbine shaft. Hence mathematically the feasible operating zones of unit can be described as follows:

4/8/12

ELD with valve-point loading (ELDVPL)

A valve-point effect is the rippling effect added to the generating unit curve when each steam admission valve in a turbine starts to open. In ELD with Valve point loadings, the objective function Ft is represented by a more complex formula:

4/8/12

Biogeography-Based Optimization

Biogeography is the study of geographic distribution of biological organisms. It describes how species migrate from one island to another, how new species arise, and how species become extinct BBO is an application of biogeography to optimization problems. It is modeled after the immigration and emigration of species between the islands

4/8/12

Habitat Suitability Index (HSI): Some islands are more suitable for habitation than others.

Habitats with a high HSI have a low immigration rate but a high emigration rate. Habitats with low HSI have a high species immigration rate. This immigration of new species to low HSI habitats may raise the HSI of the habitat

Suitability Index Variables (SIVs): Habitability is related to features 4/8/12

Species model of a single habitat

S1 is like a low HSI island, while S2 is like a high HSI island. S1 has a high immigration rate and low emigration rate. S2 has a low immigration and high emigration rate

4/8/12

Basic BBO Algorithm


1.

Initialize a set of solutions to a problem Compute fitness (HSI) for each solution Compute S, , and for each solution Modify habitats (migration) based on , Mutation

2.

3.

4.

5.

4/8/12

Habitat Migration
emigrating islands (individuals) ---

immigrating island (individual)

= the probability that the immigrating individuals solution feature is replaced


4/8/12

1313

Differential Evolution (DE)

4/8/12

Differential evolution is a stochastic direct search optimization method. DE also depends on initial random population generation, which is then improved using selection, mutation, and crossover repeated through generations until the convergence criterion is met. DE adds the weighted difference between two population vectors to a third vector

Representation
High X Low

X1

X2

XD1

XD

Solutions are represented as vectors of size D with each value taken from some domain. May wish to constrain the values taken in each domain above and below

4/8/12

Maintain Population -NP

We will maintain a population of size NP

X1 X2

X1, 1 X2, 1

X1, 2 X2, 2

X1,D1 X2,D1

X1,D X2,D

XNP

XNP ,1

XNP ,2

XNP, D-1

XNP, D

4/8/12

Initialisation
High
0.78 0.42 0.83 0.22

Low Xi
Xi,1 Xi,2 Xi,D -1 Xi,D

Xi,j=Low(i)+U(0,1)(High(i)-Low(i)) Different Uniform U(0,1) variable value for each i and j.


4/8/12

Recombination

For each member i of the population G pick randomly three other different members r1, r2, Xr3 and r3.
Xr3,1 Xr3,1 Xr2 Xr1 Xr3,D1 Xr3,D Xr2,1 Xr1,1 Xi,1 Xi,1 Xr2,2 Xr1,2 Xi,2 Xi,2 Xr2,D1 Xr1,D1 Xi,D-1 Xi,D-1 Xr2,D Xr1,D Xi,D Xi,D

Temp

Zi Xi

4/8/12

Generate a random value k in 1..D

Recombination and Mutation

We progressively define the elements of the temporary candidate Z.


For I in 1..NP Generate random ki in 1..D for J in 1..D if [(Ui,j(0,1)<CR)or j=ki] Zi,j=Xr1,j + F x (xr3,j-xr2,j) else Zi,j=Xi,j
4/8/12

Recombination in pictures
Xr3 Xr2 Xr3,1 Xr2,1 i,1 Fi,1 Xr1,1 Ti,1 Xr3,1 Xr2,2 i,2 Fi,2 Xr1,2 Ti,2 Xr3,D1 Xr2,D1 i,D1 Fi,D -1 Xr1,D1 Ti,D-1 Xr3,D Xr2,D i,D Fi,D Xr1,D Ti,D

x +

= = =

i Fi Xr,1 Ti

Zi

Ti,1

Ti,2

Xi,D-1

Ti,D

Xi 4/8/12

Xi,1

Xi,2

Xi,D-1

Xi,D

Selection for Next Generation


Fitness
Zi Xi Ti,1 Xi,1 Ti,2 Xi,2 Xi,D-1 Xi,D-1 Ti,D Xi,D

is the ith individual of next generation.

4/8/12

DE/BBO Algorithm

Generate the initial population P Evaluate the fitness for each individual in P While the termination criterion is not satisfied For each individual calculate species count probability For each individual calculate Immigration rate and emigration 4/8/12 rate

Hybrid migration operator

4/8/12

Contd

for i=1 to NP Evaluate the offspring if is better than

4/8/12

Test Problems and Results

4/8/12

Test Case 1 (ELDPOZR)

A 3 generators system with ramp rate limit and prohibited operating zone is considered. Pd=300 MW Output(MW) DE/BBO BBO APSO GA 2PHASE N.N
P1 P2 P3 207.637 87.2833 15.0000 207.9926 86.0125 16.0723 310.0774 10.0774 3260.1748 3620.1799 .017 200.528 78.2776 33.9918 312.797 12.8364 3634.3127 3634.3127 194.26 50 79.62 323.89 24.011 3737.20 165 113.4 34 312.45 12.45 3652.6000 -

Total Power 309.920

Ploss

9.9204

Total Cost($/hr)3619.7565 Av. Cost($/hr) 3619.7568 Time(sec.) 0.015

4/8/12

Test Case 2 (ELDVPL)

A system with 40 generators with valve point loading is used here


Output (MW) P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 P13 P14 P15 P16 P17 P18 P19 P20 DE/ BBO ICA_ PSO Output (MW) P21 P22 P23 P24 P25 P26 P27 P28 P29 P30 P31 P32 P33 P34 P35 P36 P37 P38 P39 P40 DE/ BBO ICA_ PSO BBO BBO 110.7998 110.7998 97.3999 179.7331 87.9576 140.00 259.5997 284.5997 284.5977 130.00 168.7998 94.00 214.7598 394.2794 394.2794 304.5196 489.2794 489.2794 511.2794 511.2794 111.0465 111.5915 97.60077 179.7095 88.30605 139.9992 259.6313 284.7366 284.7801 130.2484 168.8461 168.8239 214.7038 304.5894 394.2761 394.2409 489.2919 489.2919 511.2997 511.3073 Total cost ($/h) 110.80 110.80 97.41 179.74 88.52 140.0 259.60 284.60 284.60 130.00 168.80 94.00 214.76 394.28 394.28 304.52 489.28 489.28 511.28 511.28 523.2794 523.2794 523.2794 523.2794 523.2794 523.2794 10.00 10.00 10.00 97.00 190.00 190.00 190.00 164.7998 200.00 200.00 110.00 110.00 110.00 511.2794 121420.89 523.417 523.2795 523.3793 523.3225 523.3661 523.4362 10.05316 10.01135 10.00302 88.47754 189.9983 189.9881 189.9663 164.8054 165.1267 165.7695 109.9059 109.9971 109.9695 511.2794 121426.95 523.28 523.28 523.28 523.28 523.28 523.28 10.00 10.00 10.00 96.39 190.00 190.00 190.00 164.82 200.00 200.00 110.00 110.00 110.00 511.28 121413.2

4/8/12

Test Case 3 (ELDQCTL)

A system with 38 generators is taken here. Fuel cost characteristics are quadratic

4/8/12

4/8/12

Conclusion

Economic dispatch determines the best way to minimize the current generator operating costs to meet the demand. DE has good exploration ability in finding the region of global minimum. BBO has good exploitation ability in global optimization problem as solutions get fine tuned gradually as the process goes on. 4/8/12

REFERENCES
1. Aniruddha Bhattacharya and Pranab Kumar Chattopadhyay, Hybrid Differential Evolution With Biogeography-Based Optimization for Solution of Economic Load 4/8/12

4/8/12

You might also like