Abstract:

The bias estimation for sensors on moving base is one of difficult sensor registration problems. Using the true position of cooperation targets, the gesture angle biases of moving base are converted to the measurement biases, and a decoupling model for computing all kinds of biases is proposed in polar coordinates. Based on the decoupling model, the generalized least square method is used to estimate the gesture angle biases and measurement biases in real time. A Monte Carlo simulation is used to analyze the performance of the method. The simulation results show the effectiveness and the strong convergence rate of the proposed algorithm.