Adaptive gyroscope drift compensation based on temporal noise modelling

Gyroscope is widely used in applications like Indoor navigation, path tracking, motion sensor fusion, Video stabilization etc. This sensor is known for its susceptibility towards bias drift and dynamic errors which compromise the performance and user experience. The static bias error, random white noise and temperature interference makes it difficult to manage the drift in real time applications for hand held devices. This paper categorizes various error sources and their impact on actual gyroscope measurements. Based on the category and experimental observations, inferences are drawn and an Adaptive Bias Correction method proposed to compensate these random and static bias errors. The paper concludes performance and stability of the proposed novel method over various gyroscope chips from different chip manufacturers. Comparison of key performance indicators derived with and without algorithms on various use cases confirms the accuracy of algorithm.