Abstract: For the problem of cost limitation and weak performance, the MEMS vehicle gyroscope has become the bottleneck of electronic stability program. In this paper, the data of multiple gyroscopes were fused based on Kalman filter algorithm, combined with the measured error value of this type of gyroscope in the real scene and the difference method to eliminate the influence of vehicle hybrid noise on its real measured value. The results show that compared with single gyroscope, the RMS error of virtual gyroscope was reduced by 49.74%. Based on the virtual gyro technology, a vehicle centroid sideslip angle integral estimator was designed. Compared with a single gyroscope, the estimation error of vehicle sideslip angle was reduced by 82.5%. Furthermore, the driving stability of the vehicle was improved under the limit condition and a good balance between cost and performance was achieved.