Grid Power Quality with FACTS Devices and Renewable Energy Sources Using Deep Learning Algorithms

Modern customers use many sensitive devices comprised of power electronics that are quite sensitive to power quality (PQ) disturbances in the supply network. From worldwide customer surveys, complaints on PQ related disturbances, such as harmonics, voltage dips, flicker, etc., are increasing every year. Poor PQ has large financial consequences for the affected customers. In extreme cases, poor PQ of the electric supply can cause financial losses to the network operators and the equipment manufacturers. All these factors have motivated interest in grid power quality. In this dissertation, the power quality in different areas including wind forecasting, FACTS devices and their state estimation, and harmonic and filters are discussed. Because the energy available from wind varies widely because wind energy is highly dependent on continually fluctuating weather related parameters such as wind speed and wind direction, wind farms have difficulties with system scheduling and energy dispatching. Therefore, the growth of wind power penetration in the emerging power system has made wind forecasting indispensable for system operators to include wind power generation in unit commitment and economic scheduling. Three novel methods for wind speed and direction forecasting are proposed in this dissertation. In each method, the design is simple with high learning capability and prediction accuracy, and does not require extensive training, parameter tuning, or complex optimization. For state estimation of FACTS devices, we propose a new approach called spanning tree maximum exponential absolute value (ST-MEAV). The novel state estimator is developed based on the combination of the maximum exponential absolute value (MEAV) and a low stretch spanning tree preconditioner. An overall algorithm is presented to show the process. A modified ST-MEAV called ST0-MEAV is also proposed to improve the computational efficiency. Furthermore, the state estimation of the two FACTS devices called interphase power controllers (IPC) and unified interphase power controllers (UIPC) is addressed. The new formulations minimize the number of additional variables needed for the SE to reduce the computational load and to simplify implementation compared to previous methods presented in the literature for similar FACTS devices to UPFC or IPFC. For harmonic mitigations in power networks with PV energy sources and offshore wind generation, two novel filters called LCL-PST (phase shifting transformer) and RLC-PST have been introduced where R, L, and C stand for a resistor, inductor, and capacitor respectively. It is shown that the proposed methods reduced the total harmonic distortion (THD) that results in higher power factor and power quality correspondingly.

Level thesis:degree_level

Doctorate Degree

Year dc:date.issued

2019

Author dc:creator

• Chitsazan, Mohammad A.

Advisor dc:contributor.advisor

• Trzynadlowski, Andrzej M.

Committee members dc:contributor.committeemember

• Fadali, Mohammed
• Ben-Idris, Mohammed
• Quint, Thomas
• Abbasi, Behrooz

• Deep Learning
• FACTS Devices
• Forecasting
• Harmonics
• Renewable Energy Integration
• State Estimation

Handle dc:identifier.uri

http://hdl.handle.net/11714/5725

OAI identifier oai:identifier

oai:scholarwolf.unr.edu:11714/5725