A Real-Time State Estimation Method with Dual Filter Framework Based on MEMS-SINS/GNSS system

High-precision navigation information of a remote-controlled ship (RC-ship) is vital for ensuring the safe and effective operation by shore-based operators. However, the complexity of the waterborne environment contributes to increased uncertainty in the system model, leading to inaccuracies in state filter estimations. Additionally, Due to the delay of communication between ship and shore, the accuracy of real time state estimation will be further reduced. In this paper, A real-time state estimation method considering both navigation system error and time delay error is proposed. First of all, an adaptive interactive multi-model (IMM) algorithm based on Cubature Kalman Filter (CKF) framework is proposed to address the issue of system noise uncertainty by matching the appropriate model within the model library. Then, the uncertainty of transmission delays is modeled using a probability density function (PDF), enabling this method to handle uncertain time delays. Finally, an augmented state method considering the uncertainty of communication delay is proposed to compensate the time delay error. To validate the effectiveness of the proposed method, a series of real ship-borne experiments including different scenarios of communication delay were conducted. The results show that this method can estimate the real-time state of the ship under different sailing scene with uncertain time delay.