Name and designation of thesis supervisor: 

Prof. Dr. Md. Kamal Hosain 

Head 

Electronics & Telecommunication Engineering (E.T.E.) 

Rajshahi University of Engineering & Technology (RUET) 

Rajshahi-6204, Bangladesh. Cell: +8801716130151 

Email: kamaleteruet@gmail.com 

Thesis Keywords: Bidirectional Converter, Multilevel Inverter, Grid Ties PV inverter, Induction Motor Drive

This thesis proposes a grid-interactive bidirectional solar water pumping system with seven level Neutral Point Clamped (NPC) Multilevel Inverter (MLI) fed induction motor (IM). The main goal of this research is to establish an efficient water pumping system with a reduction of harmonics in the grid input current and to establish bidirectional power flow between the grid and the PV array to minimize power loss and uninterrupted operation. DTC-SVM control strategy has been adopted to control the induction motor to accomplish water pumping. A two-layer artificial Recurrent Neural Network (RNN) based back emf Model reference adaptive system (MRAS) estimates the motor flux and speed. Torque, stator current, and rotor flux ripples can all be dampened using the control strategy in this research, leading to better overall dynamic performance. Sensorless DTC-SVM control has been implemented to make the system cost-effective and reliable. A third-order integrator regulates the DC link voltage to maintain power flow between the grid and the water pump. Implementing a seven-level NPC to the DTC-SVM scheme led to an impressive motor drive performance and reduced harmonic content in the grid current when the utility grid feeds the IMD.SVPWM modulation scheme has been used for generating_the_gate_signals_to_the_three_phase_NPC_MLI._A_bidirectional_voltage_sourceconverter (VSC) is implemented for ensuring uninterrupted water pumping. The complete system is modeled in MATLAB/Simulink using the Sim-power system toolbox, and its performance and effectiveness under various adverse conditions are assessed.

• To use an efficient control for the induction motor drive (IDM) so that the motor's speed doesn't fluctuate despite the system's change in the load torque.

• To use efficient converters to reduce switching loss and obtain low input current THD

• To design and implement a seven-level neutral point clamped MLI for power conversion from the solar PV array and to supply the IDM for water pumping.

• To ensure efficient bidirectional power flow in and out of the grid with an improved control technique.

• A higher-order multilevel inverter can be used in order to obtain a better sinusoidal output waveform and to mitigate the THD level.

• An improved neural network can be used in order to increase the efficiency of flux estimation and motor speed estimation.

•  PV power short-term forecasts can be done with the help of Artificial Intelligence and improved neural networks.

• With the use of artificial intelligence, it is possible to build a connection between the many weather-related input data, such as irradiance, temperature, and spectrum, and the predicted output of the system, which includes motor speed and DC Link voltage.

• The flux estimation technique can be improved by implementing a higher-order integrator.

• The bidirectional power flow and the power-sharing process can be improved