Hall sensors are commonly used to detect rotor position information in permanent magnet synchronous motors (PMSM). However, due to the low resolution of Hall sensors, the speed signal directly collected by the motor during rotation contains significant random errors and noise. Using this signal directly may increase the error and jitter in the rotor position estimation, thereby affecting the control performance of the system. This paper proposes a novel position estimation method that combines the Kalman filter and phase-locked loop (PLL) to precisely obtain the rotor position. In the proposed method, to suppress the noise and errors of the speed signal, the state and observation equations are established using the Kalman filter algorithm. Additionally, to enhance the precision of the rotor position estimation, the position obtained by integrating the speed from the Kalman filter is processed using the PLL algorithm, and the PLL algorithm parameters are dynamically corrected. To verify the feasibility and accuracy of the proposed speed and position estimation method, simulations and experiments are performed, respectively. By adopting the proposed method, the speed error is reduced by 30% to 50%, and the current harmonic component is reduced by about 48.7%, which effectively improves the accuracy of the rotor position estimation.