Collaborative control of wave glider platforms - Local Communication and Sea State Estimation

Master Thesis


Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title
Climate change is the focus of many oceanography and marine engineering researchers, with possible links between climate change and the carbon cycle in the Southern Ocean being considered. This type of investigation requires modern and cost-effective tools to conduct surveys and collect data from the ocean. The self-propelled unmanned surface vessel, the Liquid Robotics Wave Glider, was designed primarily as a marine research tool and offers several advantages over existing research vessels and other tools employed for data acquisition in the ocean. The main advantages are its robustness at sea, i.e. its ability to withstand extreme weather conditions, its propulsion energy source, which is the wave energy, and its customisable electronics payload. The inter-platform communication strategy of the Wave Glider inspired a few engineering questions, one of which is the focal point of this research: whether Low Power Wide Area Network (LPWAN) technology can be used to set up a local communication system enabling the collaboration of two or more Wave Gliders and reduce the cost, in terms of power and communication channels, involved in the communication with the Wave Glider platforms during missions. This research considers various LPWAN technologies available on the market and proposes LoRaWAN technology for the local communication system. LoRaWAN was selected as it presented a robust radio modulation and had growing support in the industry. In this research, a LoRa-based network of two nodes was developed, implemented and tested over the surface of the ocean. It was found that the system performs well over a distance of 1 km with both antennas having one end at the mean surface level of the sea. With the intention to increase the range of the platform and achieve a reliable and robust system, the research continued with the study of the influence of the surface waves on the proposed local communication system by exploring, firstly, the impact of seawater and, secondly, the wave height on signal transmission. The first study investigated the influence that the electromagnetic properties of seawater may have on the transmission of signals from one node to the second through simulations using the computational electromagnetic package FEKO. It revealed that, at the frequency of operation, which was 868 MHz, seawater reacted as a lossy conductor and reflected the signal upward, with negligible power penetrating the surface of the ocean. The subsequent study reviewed the statistical properties of the ocean surface waves in a sea of deep waters and proposed a relationship between the wind speed (or surface wave elevation), the antenna height, the distance separation between the two nodes and the probability of the presence of a line of sight (LoS) between the two nodes. This relationship quantifies the expected result that the probability of the LoS diminishes as the wind speed or the distance between the two nodes increases, whereas it improves with an increase in the antenna height. The last part of the research focused on initial works on sea state estimation using the lossless wave equation and Kalman Filter to provide 3D sea surface elevations that would be used to change to the probability of the LoS calculated previously in the research. Indeed, using the local communication to share the point-wise sea state data can be exploited to estimate the sea state over a rectangular region delimited to include these points. Sea state estimation is expected to enhance the joint navigation and coordination of the platforms and consequently, boost the probability of the LoS through the transmission at the crest of the waves. During the development of the Kalman Filter model, it was discovered that the sample time and the sample space significantly affect the performance and the stability of the discretised models. However, a carefully selected sampling time and sample space exhibited a stable system model. The results of the Kalman filtering were a realistic sea state estimate with a minimum error at the locations in the surrounding of the measurements.