This thesis focuses on the use of model predictive control (MPC) for the purpose of controlling a neutral-point clamped (NPC) inverter. Results are first obtained using the conventional control method of space-vector pulse-width modulation (SVPWM). Next, MPC is used to control the NPC inverter by use of a cost equation to find the optimal switching control sequence given the need to provide low total harmonic distortion, low switching frequency and low neutral-point voltage deviation. This cost equation is optimized using an iterative parametric sweep of the weights assigned to the cost equation and ranking the results based on a fitness function. Then a long short-term memory (LSTM) recurrent neural network is used to reduce the prediction error associated with estimating the voltages and current for the next state in the MPC algorithm. Last of all, an LSTM recurrent neural network is used to model the MPC algorithm.