Hybrid Metaheuristic Algorithm for Short Term Load Forecasting
Electric load forecasting is undeniably a demanding business due to its complexity and high nonlinearity features. It is regarded as vital in electricity industry and critical for the party of interest as it provides useful support in power system management. Despite the aforementioned situation, a...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
United Kingdom Simulation Society
2016
|
| Subjects: | |
| Online Access: | http://umpir.ump.edu.my/id/eprint/16362/ http://umpir.ump.edu.my/id/eprint/16362/ http://umpir.ump.edu.my/id/eprint/16362/ http://umpir.ump.edu.my/id/eprint/16362/1/peoco2016_IJSSST_0035.pdf |
| Summary: | Electric load forecasting is undeniably a demanding business due to its complexity and high nonlinearity features. It is regarded as vital in electricity industry and critical for the party of interest as it provides useful support in power system management. Despite the aforementioned situation, a reliable forecasting accuracy is essential for efficient future planning and maximize the profits of stakeholders as well. With respect to that matter, this study presents a hybrid Least Squares Support Vector Machines (LSSVM) with a rather new Swarm Intelligence (SI) algorithm namely Grey Wolf Optimizer (GWO). Act as an optimization tool for LSSVM hyper parameters, the inducing of GWO assists the LSSVM in achieving optimality, hence good generalization in forecasting can be achieved. Later, the efficiency of GWO-LSSVM is compared against three comparable hybrid algorithms namely LSSVM optimized by Artificial Bee Colony (ABC), Differential Evolution (DE) and Firefly Algorithms (FA). Findings of the study revealed that, by producing lower Root Mean Square Percentage Error (RMSPE), the GWO-LSSVM is able to outperform the identified algorithms for the data set of interest. |
|---|