Optimal Placement of Dynamic Voltage Restorer for Voltage Sag Mitigation using Artificial Neural Network
| dc.contributor.author | Ghimire, Archana | |
| dc.date.accessioned | 2022-02-01T11:10:33Z | |
| dc.date.available | 2022-02-01T11:10:33Z | |
| dc.date.issued | 2020-07 | |
| dc.description | Voltage sag is one of the most critical and frequently occurring events in electric power systems and more prominent in radial distribution systems. | en_US |
| dc.description.abstract | Voltage sag is one of the most critical and frequently occurring events in electric power systems and more prominent in radial distribution systems. When the uninterrupted operation of a load is desired, there is a necessity of a mechanism to dynamically compensate the voltage in the grid. Custom power electronics devices are used to achieve such dynamic compensation of which Dynamic voltage restorer (DVR) is extensively used because of its fast operation, the competence of compensating active and reactive power, fewer harmonics injection and its ability to operate for both voltage sags/swells apart from being compact in size and cheaper in cost. Based on the nature of the network and equipment served, the unit of DVRs to be placed in a system also varies. It becomes not only economically infeasible to provide DVR to every load in a system, but also, such placement of a large number of DVRs violate the operation of the system. Thus, the optimal placement of this device must be determined so that it maximizes the system performance. This study focuses on the Artificial neural network method in order to optimally locate DVR in a radial distribution network. The working model of a DVR developed in MATLAB/SIMULINK simulation environment has been implemented to restore the node voltages back to the pre-fault conditions dynamically. In order to analyze the influence of various parameters affecting the operation of DVR in radial distribution networks, sensitivity analysis has been implemented. The developed DVR model has been optimally located in the network using the ANN-based approach, which uses the Levenberg Marquardt backpropagation algorithm with a target of minimizing the voltage deviation of the buses from their rated operating values. Optimization results obtained by using ANN have then been verified by finding the optimal location using a different approach, which involves the minimization of System Average RMS Frequency Index (SARFI). Sensitivity analysis shows that the developed DVR model can bring the bus voltages back to a safe operating range above 0.9p.u. for more than 90% of the simulated cases. Optimal placement of DVR is found to be at line 2-3 for IEEE 15 bus system and line 2-6 for Thimi-Sallaghari radial distribution systems. Simulation results with DVR placed in the optimal location for both the systems show a consequential enhancement in voltage profile of systems with DVR being able to restore the voltage sags at neighboring buses to more than 90% of the nominal rated voltage at each bus. | en_US |
| dc.identifier.citation | MASTER OF SCIENCE IN ENERGY SYSTEMS PLANNING AND MANAGEMENT | en_US |
| dc.identifier.uri | https://hdl.handle.net/20.500.14540/8022 | |
| dc.language.iso | en | en_US |
| dc.publisher | Pulchowk Campus | en_US |
| dc.subject | Voltage Sag | en_US |
| dc.subject | Dynamic Voltage Restorer (DVR) | en_US |
| dc.title | Optimal Placement of Dynamic Voltage Restorer for Voltage Sag Mitigation using Artificial Neural Network | en_US |
| dc.type | Thesis | en_US |
| local.academic.level | Masters | en_US |
| local.affiliatedinstitute.title | Pulchowk Campus | en_US |
| local.institute.title | Institute of Engineering | en_US |