[Objective] Aiming at the issues of inaccurate position estimation and insufficient commutation current regulation in traditional sensorless brushless DC motor (BLDCM) at low-to-medium speeds and commutation regions, which lead to significant torque ripple and degraded operational stability, this paper proposes a cooperative control method that balances position estimation accuracy and commutation current regulation performance. [Methods] Within the dual vector field-oriented control framework, the position observation and current control were collaboratively designed. A sliding mode observer based on a double power piecewise smooth (DP-PS-SMO) was constructed, where different power terms were utilized to achieve rapid convergence in large error regions and high-precision estimation in small error regions. Additionally, a piecewise smooth function was employed to enhance the continuity of back electromotive force and electrical angle estimation. Finite control set model predictive current control (FCS-MPCC) was introduced, where the estimated electrical angle was used to predict the current response under the action of the inverter’s finite voltage vectors. The optimal voltage vector was selected through a cost function to suppress torque fluctuations caused by commutation discontinuity. Finally, simulation verification was conducted on the Matlab platform. [Results] The simulation results demonstrated that the proposed method effectively reduced current fluctuations in the commutation region and suppressed torque ripple caused by commutation errors. Compared with traditional sensorless control strategies, torque ripple was significantly mitigated under all four operating conditions. [Conclusion] The cooperative design of DP-PS-SMO and FCS-MPCC effectively mitigates torque ripple issues in sensorless BLDCM at low-to-medium speeds and commutation regions, enhancing system operational stability and control reliability without the need for additional position sensors.