Intelligent algorithm based modeling of renewable and green energy resources for microgrid optimization

Khamis, Alias (2025) Intelligent algorithm based modeling of renewable and green energy resources for microgrid optimization. Doctoral thesis, Universiti Teknikal Malaysia Melaka.

[img] Text (24 Pages)
Intelligent algorithm based modeling of renewable and green energy resources for microgrid optimization (24 Pages).pdf - Submitted Version

Download (593kB)
[img] Text (Full Text)
Intelligent algorithm based modeling of renewable and green energy resources for microgrid optimization.pdf - Submitted Version
Restricted to Registered users only

Download (5MB)

Abstract

The reduction of fossil fuels, rising oil prices and environmental awareness have attracted attention to the use of renewable energy (RE)-based distributed generation (DG) systems. Among the various types of renewable energy-based DG, photovoltaic (PV) and fuel cell (FC) technology have shown great potential in electricity generation due to rapid technological development, high efficiency, clean operation and slight influence by weather conditions. To ensure optimal DG output, the RE system must be coordinated using a voltage controller and optimisation techniques to determine the optimal DG output voltage and power value. To improve the AC bus arrangement, battery power is connected to a down/up converter to ensure continuous power flow between the Alternating Current (AC) bus and the battery. In order to control the voltage source inverter (VSI) of the PV/fuel cell/battery cell system, conventional methods of control voltage modes and currents with improved controllers of the artificial intelligence (AI) of both the internal current control loop and the output voltage were built. The proposed tuned Artificial Neural Network (ANN) controller has an advantage over the Adaptive Neuro-Fuzzy Inference System (ANFIS) controller while maintaining the simplicity and robustness of the Proportional Integral (PI) controller. The inverter-based DG model is applied to the microgrid system to review its effectiveness as a complete model as well as to evaluate the performance of its use in large network systems. Since the VSI model is built on a P-Q control scheme that allows separate control of active and reactive power output, DG can operate based on active and reactive power reference on the inverter. A new smart technique has been developed to manage active and reactive power reference for DG by using ANN to ensure that the DG unit operates at optimal power values while reducing the amount of power loss as well as maintaining the voltage profile within acceptable limits. The results showed that the proposed tuned ANN technique could accurately predict the active and reactive power references of DG with minimal error. A comparison was made between the ANN DG controller and the ANFIS DG controller for the power management strategy in terms of the generation by standard forecasting metrics. The comparison between the proposed AI controller and the conventional PI controller has been conducted, and the results showed that the proposed tuned artificial NN technique could accurately predict the active and reactive power references of DG with minimal error. For active power of Battery, is 0.23%, Fuel Cell is 0.23%, reactive power of Battery is 0.0175%, Fuel Cell is 0.097%, Photovoltaic PV1, 0.078% and PV2 is 0.021%. At the end of the research, the AI controller was evaluated/validated for effectiveness by comparative means also conducted to assess the performance and forecast accuracy of the tuned AI that has been chosen by forecasting metrics, which show good estimation performance in only 1.6E-14% for the coefficient of determination (R²), 5.86E-05% for root mean square error (RMSE), 9.1E-06% mean absolute error (MAE) and 0.011% for mean absolute percentage error (MAPE).

Item Type: Thesis (Doctoral)
Uncontrolled Keywords: Artificial neural network, Adaptive neuro‚ Fuzzy inference system, Distribution generation, Microgrid
Subjects: T Technology
T Technology > TK Electrical engineering. Electronics Nuclear engineering
Divisions: Faculty Of Electrical Technology And Engineering
Depositing User: Norhairol Khalid
Date Deposited: 10 Oct 2025 07:51
Last Modified: 10 Oct 2025 07:51
URI: http://eprints.utem.edu.my/id/eprint/29006
Statistic Details: View Download Statistic

Actions (login required)

View Item View Item