Optimizing Ship Pilotage with Intelligent Information Services: Integrating GIS-Based Big Data Positioning and Neural Network Approaches

Abstract

The development of the shipping industry has put forward new requirements and challenges for its pilotage and services. To promote navigation development and ensure safe and efficient ship navigation, experiments have been conducted to combine geographical information systems (GIS) big data with neural systems, resulting in a proposed intelligent ship pilotage service method. This paper presents a novel approach to building a ship pilot system using GIS technology. The system is enhanced with the introduction of the Faster-RCNN model to improve its positioning function, and Gaussian distribution is employed to optimize the loss function. Finally, the system’s ship pilot service set time parameters are solved based on ship entry and exit scheduling to achieve intelligent navigation and services for ship piloting. The data showed that on the MarineT dataset, the research method (GIS big data positioning-neural network) achieved its maximum fitness value at 36 iterations of the system, with a value of 99.78. At the same time, when the system ran 66 times, the average absolute percentage error obtained by the research method infinitely approached 0. In addition, based on the AIS dataset, when the recall rate of the four algorithms was 0.800, the accuracy of the research method was 0.873, with the highest numerical value. Practical applications have shown that when the system iterated 51 times, the total waiting time for ship piloting in and out of the port quickly decreased to 177.92 hours, which is significantly better than manual scheduling time. The aforementioned findings suggest that the implemented system has the capability to considerably decrease piloting time, deliver cutting-edge technical aid for the current expansion of the shipping sector, and establish a stable basis for future intelligent shipping technology.

Keywords

• gis positioning
• big data
• neural network
• ships
• pilot
• intelligent services
• information service