This paper takes advantage of the complementary characteristics of Global Positioning System (GPS) and Light Detection and Ranging (LiDAR) to provide periodic corrections to Inertial Navigation System (INS) alternatively in different environmental conditions. In open sky, where GPS signals are available and LiDAR measurements are sparse, GPS is integrated with INS. Meanwhile, in confined outdoor environments and indoors, where GPS is unreliable or unavailable and LiDAR measurements are rich, LiDAR replaces GPS to integrate with INS. This paper also proposes an innovative hybrid scan matching algorithm that combines the feature-based scan matching method and Iterative Closest Point (ICP) based scan matching method. The algorithm can work and transit between two modes depending on the number of matched line features over two scans, thus achieving efficiency and robustness concurrently. Two integration schemes of INS and LiDAR with hybrid scan matching algorithm are implemented and compared. Real experiments are performed on an Unmanned Ground Vehicle (UGV) for both outdoor and indoor environments. Experimental results show that the multi-sensor integrated system can remain sub-meter navigation accuracy during the whole trajectory.

This paper presents a low-cost real-time lane-determination system that fuses micro-electromechanical systems inertial sensors (accelerometers and gyroscopes), global navigation satellite system (GNSS), and commercially available road network maps. The system can be used for intelligent transportation systems, telematics applications, and autonomous driving. The system does not depend on visual markings or highly precise GNSS technology, such as DGPS or RTK, and it does not need explicit lane-level resolution maps. High-resolution estimation of the vehicle's position, velocity, and orientation is implemented by fusing inertial sensors with GNSS in a loosely coupled mode using extended Kalman filter. A curve-to-curve road-level map-matching is implemented using a hidden Markov model followed by a least-square regression step that estimates the vehicle's lane. The system includes a lane-change detector based on inertial sensors and the filtered vehicle's state. The system has been realized in real time and tested extensively on real-road data. Experiments showed robust map-matching in challenging road intersections and a 97.14% lane-determination success rate.

In the realm of navigation systems, Inertial Measurement Unit (IMU) sensors play a pivotal role. The advent of Micro-Electro-Mechanical System (MEMS) sensors has introduced a lightweight and cost-effective alternative for IMUs. However, MEMS IMUs come with the challenge of larger stochastic errors that accumulate over time, resulting in navigation drifts. To address this issue, the conventional approach involves fusing IMU with the Global Navigation Satellite System (GNSS) for reliable navigation. Nevertheless, this fusion setup fails in providing ubiquitous navigation during GNSS outage scenarios due to persistent IMU errors. In this paper, an efficient methodology is developed to mitigate navigation drifts by eliminating IMU errors using Light Gradient Boosting Machine (LightGBM) and Categorical Boosting (CatBoost) Machine Learning (ML) algorithms. In contrast to existing methodologies that employ high-end and expensive IMUs for training models to denoise low-cost MEMS IMUs, this paper proposes utilizing Inverse Kinematics (IK). This approach helps to derive clean IMU training data from the Position, Velocity, Attitude (PVA) values estimated through the Extended Kalman Filter (EKF) when GNSS is available and reliable. The distinctive advantage of the IK approach lies in its capacity to obtain real-time pseudo error-free IMU data without the necessity for high-end IMUs to train ML models. The proposed method undergoes testing in both Loosely coupled and Tightly coupled EKF scenarios using simulation and real dataset under varying GNSS outage durations. Comparisons are made between the denoised IMU signals and signal processing techniques such as Moving Average (MA) and Savitzky Golay (SG). Additionally, we present a comparative analysis of the proposed algorithms against Convolutional Neural Networks (CNN). Results demonstrate a noteworthy enhancement in position, velocity, and orientation estimation. Furthermore, the computation time required for model training and prediction across various algorithms is analyzed. The outcomes prove the superiority of the proposed tree-based algorithms over the conventional filtering methods and CNN in denoising IMU and improving the navigation results.

The Extended Kalman Filter (EKF) is currently a dominant sensor fusion method for mobile devices, robotics, and autonomous vehicles. Its performance heavily depends on the selection of EKF parameters. Therefore, the optimal selection of parameters is a critical factor in EKF design and use. In this paper, a methodical and efficient method of EKF parameter tuning is presented. The tuning framework uses nominal parameters generated by Gauss Markov (GM) and Allan Variance (AV) methods that are tuned by Genetic Algorithms (GA) accelerated by Design of Experiments (DoE). This framework has been implemented in MATLAB and tested using simulations and real data under a tightly coupled EKF that fuses IMU and GNSS measurements of a self‐driving car provided by the Blackberry QNX company. The results demonstrate that GA‐tuned parameters increase accuracy substantially over nominally tuned parameters, and that the DoE technique consistently improves the convergence behavior of the GA.

Tightly-coupled(TC) fusion of Inertial Measurement Units (IMUs) with Global Navigation Satellite Systems (GNSSs) is a common technique that provides high-rate positioning even under GNSS interruptions. In order to provide accurate positioning, errors of IMU and GNSS must be modelled and estimated by filtering techniques such as Extended Kalman Filter (EKF). Due to nonlinearity and stochastic characteristics of IMU and GNSS system and measurement models, robust filter design has been a challenge. Conventional design techniques use missionspecific fixed models and trial-and-error noise parameter tuning to design IMU/GNSS filters. These conventional techniques are inflexible and do not always lead to accurate designs as there are no ways to verify the filter ability to estimate sensors errors accurately. To address this challenge, this paper presents a flexible design framework and a systematic procedure for TC IMU/GNSS fusion. The framework utilizes symbolic engines to represent and linearize system and measurement models. Symbolic engines are flexible in new models and fusion algorithms development. In order to evaluate the estimation of sensors errors, an Inverse-Kinematics module is developed to generate error-free sensors measurements which can be contaminated by known errors. The filter parameters are tuned using Genetic Algorithms and the performance is evaluated based on the accuracy of estimating all states including the added known errors. The framework has been used to develop a quaternion-based EKF design and verified on real raw IMU/GNSS data. The results showed that the developed framework greatly reduces efforts to design robust and accurate fusion systems for TC IMU/GNSS integration.