Search

Underwater navigation is generally accomplished through the data fusion of INS (Inertial Navigation System) and auxiliary sensors such as DVL (Doppler Velocity Logger) sensor. However, because of the possibility of DVL outage, alternative low-cost solutions are attractive. Among these, one is using vehicle kinetic model information extracted by the Newton-Euler equation to improve INS performance, which is called model-aided navigation. In this paper, only the vehicle translational motion dynamics are used to replace DVL in underwater navigation in DVL outage. The vehicle 3D translational dynamics has been obtained by using general Newton-Euler equations. Integrating these dynamics leads to the calculation of velocities in the body reference frame. Similar to DVL measurements, the calculated velocities (called “Pseudo-DVL” in this paper) are fused with INS. In this paper, a Kalman filter is used for INS/Pseudo-DVL data fusion. Field test data collected by an equipped research AUV called SUT-III is used to evaluate the proposed approach experimentally. Also, a comparison between the integrated navigation of INS-DVL and INS-Pseudo-DVL is performed and the results are provided. The results reveal that despite having low-cost MEMS IMUs, INS/Pseudo-DVL has effectively corrected standalone INS estimation. Finally it is concluded that despite slight degradation of navigation performance, an expensive DVL sensor can be replaced by using the proposed method in DVL outage.

INS-DVL integration is a common method for underwater navigation. However, inherent errors of sensors, especially in MEMS IMUs, lead to inaccuracies in estimating the position and attitude. In this paper, dynamic motion model of AUV is used to improve MEMS INS-DVL navigation. In this method, which is called model-aided or model-based navigation, the information of the kinetic model of the vehicle (obtained from Newton–Euler equations) is used to improve the navigation performance. Previous model-aided navigation studies about AUVs have been focused on the translational dynamic model of vehicles. As the best of our knowledge, this paper is the first one which suggests using a rotational motion model of AUVs to improve the navigation performance and at the same time, reports successful implementation of method through field-testing. The utilized dynamic model, describes the rotational motion of AUV in a 3DOF form. Information extracted from the rotational dynamic model and the outputs of gyroscopes, will be fused in a secondary Kalman filter. So, before the fusion with auxiliary sensors in the main Kalman filter, the gyroscopes outputs have been improved by the rotational motion model. Therefore, the overall output will be better than the outputs of gyroscopes by itself. The main Kalman filter is a perturbation-based Kalman filter which is employed for data fusion. This algorithm is implemented on real data collected from field-test. The results show that this inexpensive method which requires no additional equipment was truly effective in improving navigation performance. Using this model, the accuracy of navigation was improved 5 to 10 times.

MEMS-IMUs have an extensive application in multifarious studies, as well as industrial and commercial areas. It is crucial to diminish their intrinsic errors in a suitable calibration procedure. In this paper, a novel calibration procedure was proposed for Inertial Measurement Units (IMUs) on a turntable. A general nonlinear model of the IMU output including the effects of bias, scale factor, misalignment, and lever arm was derived. Transformed Unscented Kalman Filter (TUKF) was utilized to perform the estimation of error parameters for gyroscopes and accelerometers. The calibration maneuvers were applied using a tri-axis turntable to create input signals. In addition, assuming the sensors not placing in the center of the table makes angular acceleration become another variable which affects the estimation of the error parameters. Therefore, a suitable angular acceleration estimator was designed utilizing the angular velocity output. According to experimental results, applying the proposed method caused 66% and 63% increase in the accuracy of gyroscopes and accelerometers outputs, respectively, compared with the calibrated signals based on the least square method.

This study presents a fault tolerant control (FTC) scheme based on fault estimation (FE) for a system subject to input saturation, uncertainty, L 2 -bounded disturbance and additive faults. In this study, the saturation is represented in polytopic form, and the disturbance is partitioned into matched and unmatched parts. For the FE scheme, a partially-decoupled unknown input observer is introduced to completely compensate for the matched part and only the unmatched part affects the FE performance. The FE and FTC schemes are integrated to ensure that the resulting closed-loop system is stable with L 2 -gain performance. Furthermore, all sufficient conditions for robust performance are derived and cast as linear matrix inequalities (LMIs). Usage of the Young relation removes equality constraints, and thus all LMIs are solvable in a single step. Numerical simulations are provided to verify the effectiveness of the proposed scheme.

Kinematic constrained navigation is a subset of model-aided navigation systems that is attractive because of its independence from extra hardware equipment. In this study, the error-based Kalman Filter is used for the data fusion process through feedback strategy, and kinematic constraints are utilized to improve Inertial Navigation System-Doppler Velocity Log (INS-DVL) navigation performance. Here, the pseudo-measurement method is used for applying kinematic constraints. Real offline data obtained from field tests of an instrumented vessel in shallow water demonstrate that applying kinematic constraints improves navigation performance and robustness against initial condition uncertainty. Repeatability of results is also investigated for three different maneuvers.

This paper presents a Fault Tolerant Control (FTC) scheme for systems with L 2 -bounded disturbances and generalized sector nonlinearity at the input. An adaptive observer-based fault estimation (FE) scheme is used to estimate the fault. The nonlinearity causes the control input (the measurable controller output, which is injected into the FE scheme) to differ from the actuator output (which enters the plant); this mismatch (together with the disturbances) causes interactions between the FE and FTC schemes, violating the separation principle and hence they cannot be designed separately. We introduce a method that integrates both FE and FTC schemes, and design them simultaneously (thus overcoming the violation of the separation principle), guaranteeing bounded stability. To obtain the sufficient conditions for stability and minimum L 2 -gain performance for disturbance rejection, we use the Sector Condition (SC) and successfully re-casted the design as Linear Matrix Inequalities (LMIs) which can be performed in a single step and yield tractable solutions. Through a numerical example, we demonstrate the effectiveness of the proposed method.

By development of ocean exploration, autonomous vehicles are employed to perform on-water and underwater tasks. Using an extended Kalman filter, Inertial Navigation System/Doppler Velocity Log (INS/DVL) integrated systems are trying to navigate in oceans and underwater environments when Global Positioning System (GPS) signals are not accessible. The dependency of DVL signals on acoustic environments may cause any DVL malfunction due to sea creatures or strong wave-absorbing material. In this paper, an improved version of evolutionary TS-fuzzy (eTS) is proposed in order to predict DVL sensor outputs at DVL malfunction moment, by utilizing an artificial intelligent (AI) aided integrated system. According to lack of input selection and shrinking, while the classic eTS suffers from soaring prediction errors and may result in instability, by adding these properties to eTS, the performance increases in long-term DVL outage. The proposed eTS-aided system makes ocean navigation purposes possible during long-term and simultaneous outage of GPS and DVL. These evolutionary fuzzy systems change their structure depending on the path which makes the trained fuzzy system more flexible with non-stationary and varying environments. The real sensor data is collected online with a test setup on a lake and then the algorithms are applied. The powerful capacity of the proposed data fusion method is demonstrated in analysis results.