Multichannel based Integrated Navigation for Scalable Maritime Applications

Kurzfassung

The capability to provide automated, onboard Position, Navigation and Timing (PNT) data compliant with the accuracy, integrity, continuity and availability requirements in accordance to the different phases of vessel navigation is considered as one core element of the International Maritime Organization (IMO) e-Navigation strategy. This capability, known as resilience of the system, is fundamental for any system supporting safety of live critical maritime applications. In order to achieve the resilience, satellite based, shipside and landside components should be fused by an “integrated PNT system”. However, the heterogeneity of the ship sensors together with the long life-cycles of vessels constitutes a practical challenge for the design of an onboard PNT module. Therefore the German Aerospace Center (DLR) promotes an open and scalable architecture for the shipborne “PNT Module”, which serves as a front-end between an integrated PNT system and shipside applications like Integrated Navigation Systems (INS), the Automatic Identification System (AIS) and the Electronic Chart Display and Information System (ECDIS). As the core component of a PNT Module, the PNT Unit processes and integrates the data of radio navigation systems and services as well as shipborne sensors (like speed log, gyro compass). The goal of the PNT Module is the provision of PNT information and associated integrity information in accordance with changing performance requirements during berth to berth navigation, taking into account the different grades of ships navigational sensors. The application of data fusion techniques within the PNT Unit improves the resilience of PNT information and enables the accuracy estimation by integrity monitoring functions. In order to fulfill these objectives, the redundancy of PNT parameters should be provided by taking redundant sensors or by applying different measuring techniques. In a first development phase of the PNT Unit, following sensors and GNSS services are taken into consideration, including three GNSS receivers, non-dedicated GNSS compass composed of distributed antennas, code- and phase based differential GNSS (DGNSS) services, Inertial Measurement Unit (IMU) and other ship sensors. To perform a robust and efficient data fusion, two fundamental architectures can be considered. The first consists in the fusion of all the available data using a centralized filter. This method yields the optimality in the mathematic aspect. However, a modeling failure of a single sensor might fail the entire system. The other approach is a multi-channel architecture, where each channel is also a filter and fuses the data of a subset of sensors. Different channels are running independently, so that a sensor failure might only affect its own channel rather than the entire system. Considering that the sensors might have individual outages or synchronization problems, the multi-channel approach provides not only a higher resilience but also the simplicity in the algorithm design. Furthermore this architecture has the advantage being adaptable for different grades of ships navigational sensors, to support the scalability of carriage requirements, and to enable a stepwise rollout into the maritime traffic system. This paper focuses on the multi-channel system architecture of the PNT Unit. The minimal task of each channel is to provide position and velocity information regarding the consistent common reference point (CCRP). The attitude information is needed in order to transfer the position from its measuring point onto CCRP. Similar to that, the angular rate information is needed for the transformation of velocity information. Based on these requirements, following channels are implemented at the first stage: Channel type 1: Single GNSS + GNSS compass. This is the basic configuration. Although a single GNSS (main GNSS antenna) allows the position and velocity estimation, multiple distributed GNSS antennas will construct the backup system in case that the main GNSS antenna does not work properly. Channel type 2: Single GNSS +IMU+GNSS compass, where the integration of single GNSS and IMU is made with a tightly-coupled architecture and the attitude information from GNSS compass is integrated in a loosely-coupled architecture. This channel can work in almost any maritime operation area. The use of IMU will, at one hand, smooth the results from single GNSS, and on the other hand, enable the error detection in the GNSS measurement domain. Considering the long antenna baseline on the ship, the attitude estimation based on GNSS compass and IMU will provide more accurate attitude estimation than the conventional GNSS/IMU integration. Channel type 3: IALA Beacon DGNSS+ IMU+ GNSS compass. This configuration can be applied if the correction data from an IALA Beacon DGNSS station are available. The DGNSS results are integrated with IMU output in a loosely-coupled architecture together with the attitude information from GNSS compass. Channel type 4: Phase based (RTK) DGNSS+ IMU+ GNSS compass. This configuration yields the highest accuracy in the sub dm regime, but of cause requires data from a RTK reference station. The RTK results are integrated with the IMU output in a loosely-coupled architecture together with the attitude information from GNSS compass. This paper is structured as follows. Firstly, an overview of the multichannel architecture is given. Secondly each of the applied channels will be described in detail and investigated regarding its advantages and disadvantages. For this purposes test data are used, collected during a measuring trip, which was carried with the vessel “Baltic Taucher II” in the Baltic Sea. Based on the data analysis from this measurement campaign, we will present the results from different channels and show how the entire system can benefit from the multichannel approach.