Prediction of Market Power Using SVM as Regressor Under Deregulated Electricity Market

This paper proposes a methodology to utilize support vector machines (SVM) as a regressor tool for predicting market power. Both the companies, i.e., Generation (Gencos) and the Distribution (Discos), can utilize this tool to forecast market power on thei

  • PDF / 303,301 Bytes
  • 13 Pages / 439.37 x 666.142 pts Page_size
  • 54 Downloads / 202 Views

DOWNLOAD

REPORT










Keywords Support vector machine Market power Regressor Kernels Nodal must run share

1 Introduction Market power Must Run Generation (MRG): To analyze the load variation and transmission line constraints, market power indices like MRS and nodal must run share (NMRS) are considered [1]. Taking account on generation and transmission constraints, MRG is

K. Shafeeque Ahmed  Fini Fathima  R. Ananthavijayan  P.K. Shanmugam (&) S.K. Sahoo  R.C. Naidu School of Electrical Engineering, VIT University, Vellore 632014, Tamil Nadu, India e-mail: [email protected] S.K. Sahoo e-mail: [email protected] R.C. Naidu e-mail: [email protected] © Springer Science+Business Media Singapore 2016 M. Pant et al. (eds.), Proceedings of Fifth International Conference on Soft Computing for Problem Solving, Advances in Intelligent Systems and Computing 437, DOI 10.1007/978-981-10-0451-3_54

605

606

K. Shafeeque Ahmed et al.

the contribution of individual generator to the total load. MRG can be obtained by solving the following linear optimization equations:

where e Pg Pd Plmax F

Minimize Pgk

ð1Þ

Subject to eT ¼ ðPg  PdÞ ¼ 0

ð2Þ

0  Pg  Pgmax

ð3Þ

Plmax  FðPg  PdÞ  Plmax

ð4Þ

Vector with all one; Power dispatch vector; Demand vector; Line limit vector; and Distribution factors matrix [1]

Equation (2) denotes power balance equation; Eqs. (3) and (4) imply output limits of generator and limits of transmission lines, respectively. Must Run Share (MRS): MRS indicates the variations in the market power with respect to dynamic load. MRSk of generator k in a power market is given by =Pd MRSk ¼ Pgmust k

ð5Þ

where Pd is the total demand in the power market. The generator is said to have market power only if the MRS is greater than zero. Nodal Must Run Share (NMRS): NMRSk,i is the minimum contribution of the must run generator k to feed the load at node i and to analyze the geographical difference of market power NMRSk;i ¼ Pgmust k;i =Pdi i ¼ 1; 2 . . . N

ð6Þ

The NMRS is determined from the following expression: NMRSk;i ¼

Pgmust ½M 1 ik Pgmust k;i k ¼P 1 Pdi j2N ½M ij Pgj

ð7Þ

where Pgmust k;i is the contribution of the must run generator k to Pdi, N is the number of buses, Pdi is the load at bus i, j is the bus which is directly connected to bus i through transmission lines, and ½M  is the distribution matrix. This matrix is used to show how the power supplied at a node is contributed from all the generators in system and the complete discussion is available in [1]. A complete survey on various indices, used to measure the market power, is available in [2].

Prediction of Market Power Using SVM as Regressor …

607

2 Support Vector Machines SVM has been extensively used as a classifier and regressor in various power system domains like fault diagnosis, load forecasting, stability assessment, security assessment, market clearing price estimation/prediction, and so on. In [3], Correa-Tapasco et al. have combined SVM as a regressor and Chu-Beasley genetic algorithm (CBGA) for locating faults in power