Abstract: Aiming at the problem that it is difficult to equalize the measurement noise and delay in the traditional speed measurement method under the application of low-precision position sensors, a simplified Kalman filter speed measurement algorithm is designed for the application of dual-inertia servo system. The observed model of the core Kalman filtering part is simplified to a first-order model by state pre-estimation, and the way of constructing the first-order model as well as the way of designing the additional required speed prediction and shaft torque pre-estimation are given, the computational cost and debugging complexity are greatly reduced compared to the higher-order model-based Kalman filter speed measurement while maintaining the measurement error and noise balance. Finally, the proposed speed measurement algorithm is compared with other speed measurement algorithms, and the simulation and experimental results show that compared with the traditional algorithm, the error is smaller under the same measurement delay; and compared with the Kalman filter speed measurement with the same computational cost, it also has better error suppression ability.