ENHANCEMENT OF SHORT-TERM LOAD FORECASTING BASED ON PARALLEL HYBRID WAVELET NEURAL NETWORK

• Main Author: SOVANN , NARIN
• Format: Thesis
• Language: English
• Published: 2016
• Subjects: TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://utpedia.utp.edu.my/21856/1/2016%20-%20%20ELECTRICAL%20-%20ENHANCEMENT%20OF%20SHORT-TERM%20LOAD%20FORECASTING%20BASED%20ON%20PARALLEL%20HYBRID%20WAVELET%20NEURAL%20NETWORK-SOVANN%20NARIN-MASTER%20OF%20SCIENCE%20ELECTRICAL%20AND.pdf; http://utpedia.utp.edu.my/21856/

Short-term load forecasting (STLF) is the prediction of load demands from one hour to one week which crucially is used for operation and planning of the electric power system. Load demands are nonstationary processes and sensitive to the weather conditions. Due to these challenges, STLF requires a new model that can achieve accuracy and robustness of load forecasting. This work proposes a hybrid model to -improve the accuiacy and certainty tor one-day ahead (from 1 hour to 24 hours) load forecasting. This proposed method is Parallel Hybrid Wavelet Neural Network (PWNN) which comprises of Wavelet Transform (WT), hybrid particle swarm optimization and Levenberg-Marquardt algorithm (PSO-LM) and neural network (NN) based on parallel prediction method.