Extension of robust three-stage Kalman filter for state estimation during Mars entry

In the context of unknown inputs (unknown dynamic biases) and unknown measurement systematic errors, this article studies the extension of robust three-stage Kalman filter (ERThSKF) that satisfies high precision entry navigation filter requirements for Mars pinpoint landing mission. The imprecise dynamic model with uncertain parameters could produce unknown inputs. Using radiometric beacons and/or inertial measurement unit (IMU) as output observation during the Mars atmospheric entry phase, the measurement data from the radio range and/or IMU have unknown measurement systematic errors. To solve these problems, we made an ERThSKF to obtain the accurate state estimation of the uncertain non-linear Mars system with unknown inputs and unknown measurement systematic errors which can be effectively estimated and compensated. Computer simulations show that the proposed navigation filter algorithm in this study can achieve high convergence speed and minor errors, which fulfils the need of future pinpoint Mars landing missions.