Position and orientation system (POS), which typically integrates strapdown inertial navigation system (SINS) and global navigation satellite system (GNSS), serves as a key sensor in airborne remote sensing, mobile mapping, and vehicle localization. POS can provide reliable, high-frequency and high-precision motion parameters using nonlinear Kalman Filter models based on fusion methods, such as extended Kalman filter, unscented Kalman filter, central difference Kalman filter (CDKF), and square root CDKF (SR-CDKF). Although the nonlinear parametric models are of high efficiency, there are also limitation on their capabilities of prediction and estimation, as it is often impossible to model all aspects of POS. In this paper, Gaussian processes (GP)-based is presented to enhance the capabilities of prediction and estimation for parametric CDKF. On one hand, it can estimate the state vector of POS with the nonlinear parametric CDKF on condition that trained data is limited; on the other hand, GP can take both the noise and the uncertainty in the nonlinear parametric CDKF into consideration. Consequently, the incorporation of GP into CDKF can result in further performance improvement. The proposed approach is verified in the real experiment, and shows that large performance benefits are achieved through applying the enhanced GP-CDKF(EGP-CDKF) into the SINS/GNSS integrated system.

Location awareness and navigation promote varieties of emerging applications of mobile collaborative multiple uncrewed aerial vehicles (UAVs). Cooperative UAVs fuse the global position system (GPS), inertial navigation systems (INS), peer to peer ranging radios derived from relative navigation of ultra-wideband (UWB) under complicated environments. Those information sources can be incorporated into spatiotemporal cooperation posited by the intra-user measurement of INS and GPS, and the inter-user measurement of the relative navigation of swarm UAVs. This paper considers the localization and navigation of multiple collaborative UAVs in networks with GPS/INS/UWB jammers in the case that the measurements are missed or randomly delayed by a sampling period. In a navigation situation with a partially denied navigation signals (e.g.GPS Jammers for some UAVs, UWB jammers for others, etc.), we propose an improved method of cooperation location for the swarm, allowing measurement jammers concerning the normal sigma point belief propagation (SPBP). This algorithm integrates message passing based on the Bayesian framework, a sigma point belief propagation of random packet loss (SPBP-RPL) to exploit spatiotemporal cooperation and measurement knowledge. Compared with existing general sigma point belief propagation, the advantages of the novel method are validated through a simulation of swarm UAVs with GPS/INS/UWB. Results show that the algorithm of combining spatiotemporal cooperation with measurement knowledge reduces the location uncertainty of swarm UAVs agents and improves location accuracy remarkably.

Multi-task remote sensing sensors have become attractive development directions of aerial remote sensing system, which rely on distributed position and orientation system (DPOS) to provide multi-node motion parameters to achieve superior performance. DPOS depends on transfer alignment from its master system to slave inertial measurement units to obtain multi-node motion information. However, for DPOS, there are many factors like carrier maneuver mode and external disturbance which will result in time-varying measurement noise and further degrade transfer alignment performance of DPOS obviously. In this work, a transfer alignment method based on variational Bayesian cubature RTS smoothing is developed to improve the accuracy of DPOS, which is implemented by combining the cubature RTS smoothing algorithm and variational Bayesian estimation method to deal with the time-varying measurement noise. A semi-physical simulation based on real flight experiment has been conducted, the results show that the motion parameter accuracy has achieved noticeable enhancement than the existing cubature RTS smoothing algorithm.

With the development of the novel remote sensing system such as array antenna interference synthetic aperture radar (InSAR), it is in urgent demand to design and develop an airborne distributed position and orientation system (D- POS) to provide multi-node time-spatial reference information accurately. To solve the problem, this paper describes a novel design and fabrication of DPOS, which integrates the main strapdown inertial navigation system (SINS), global position satellite (GPS), multiple slave inertial measurement units (IMUs) and distributed navigation processing computer system (DNPCS). In the proposed system, multiple digital signal processors(multi- DSPs) are used for multiple inertial navigation solutions and Kalman filter(KF) computations, and a field programmable gate array (FPGA) is used for creating an efficient interface of the GPS and multiple IMUs with the multi-DSPs. This also helps reduce the count of the total chips, resulting in a compact system. The system is designed to provide synchronous multi-node real-time high-precision navigation solutions with an update of 200Hz. In order to verify the performance of DPOS, the real flight test, which combined InSAR and DPOS, was conducted and showed the proposed system can help compensate the motion error of the high-resolution interference mapping more efficiently than the single POS.

A Position and Orientation System (POS) integrating an Inertial Navigation Systems (INS) and the Global Positioning System (GPS) is a key component of remote sensing motion compensation. It can provide reliable and high-frequency high-precision motion information using a Kalman Filter (KF) during GPS availability. However, the performance of a POS significantly degrades during GPS outages. To maintain reliable POS outputs, this paper proposes a new hybrid predictor based on modelling the nonlinear time-series data-driven INS-errors using Noisy Input Gaussian Process Regression (NIGPR), which takes the input noise into account. The proposed approach is used to learn the nonlinear INS-errors model when GPS signals are available. When GPS outages occur, it starts to predict the observation measurement, and then feeds it to a KF as a virtual update to estimate all the INS errors. The proposed approach is verified in a real airplane, which combines a POS and Synthetic Aperture Radar (SAR). Experimental results show that the proposed approach significantly improves the performance of the POS, with improvements more than 90% better than a KF and 10% better than a Gaussian Process Regression (GPR/KF) combination during various GPS outages.

Cooperative perception in the field of connected autonomous vehicles (CAVs) aims to overcome the inherent limitations of single-vehicle perception systems, including long-range occlusion, low resolution, and susceptibility to weather interference. In this regard, we propose a high-precision 3D object detection V2V cooperative perception algorithm. The algorithm utilizes a voxel grid-based statistical filter to effectively denoise point cloud data to obtain clean and reliable data. In addition, we design a feature extraction network based on the fusion of voxels and PointPillars and encode it to generate BEV features, which solves the spatial feature interaction problem lacking in the PointPillars approach and enhances the semantic information of the extracted features. A maximum pooling technique is used to reduce the dimensionality and generate pseudoimages, thereby skipping complex 3D convolutional computation. To facilitate effective feature fusion, we design a feature level-based crossvehicle feature fusion module. Experimental validation is conducted using the OPV2V dataset to assess vehicle coperception performance and compare it with existing mainstream coperception algorithms. Ablation experiments are also carried out to confirm the contributions of this approach. Experimental results show that our architecture achieves lightweighting with a higher average precision (AP) than other existing models.