This paper proposes the use of Model Reference Adaptive Controllers (MRAC) to mitigate low-speed friction issues in reaction wheels, which are crucial for achieving precise pointing and stabilization in CubeSats. Given the constraints of limited volume, mass, and power in small satellite platforms, effective attitude control is paramount. This study introduces a novel approach by applying MRAC to enhance the performance of reaction wheels in space applications. The research involves a comparative analysis of three control strategies: two adaptive MRAC approaches and a classical PID controller, all utilizing velocity feedback. The adaptive MRAC strategies include a PID-parameter adaptive MRAC and a two-parameter adaptive MRAC. These controllers are designed to manage the nonlinearities and uncertainties associated with reaction wheel operation, particularly at low speeds where friction effects are prominent. Validation was conducted using a Nidec BLDC motor with a flywheel as a test device, focusing on accumulative error measurements in tracking reference speeds. The results demonstrate that the adaptive PID-parameter MRAC outperforms both the classical PID controller and the two-parameter adaptive MRAC by providing superior friction compensation and achieving the lowest cumulative error. This adaptive approach also significantly improves speed tracking performance, highlighting the overall effectiveness of MRAC in addressing low-speed friction issues and managing nonlinearities in reaction wheels.