Performance improvements for calculations of third party risk around airports

摘要

During the past two decades, NLR has been performing research for third party risk around airports, and a calculation model and methodology for different types of airports are developed. The NLR third party risk model is used to evaluate the risk for people living and working around an airport. The design of new or changed air routes and runway infrastructure at airports require that third party risk studies are conducted to determine the impact for the airport surroundings. Due to the increase of traffic, the availability of improved individual flight track data, and the need for detailed calculations on a denser grid, the time needed for a risk calculation has increased significantly. It becomes challenging when more scenarios are involved and hence more risk calculations are needed in the decision making with respect to airport development and land use planning. In this paper, we present the research of using parallel programming hardware to improve the performance of the calculation model, and in particular the use of graphics cards (GPUs) to carry out massive parallel operations. Because of different phases in the calculations, and the relation between adjacent grid cells, the translation of this model into a parallel implementation is not straightforward. Results show that a calculation for Schiphol airport, based on the traffic for a calendar year and a dense grid that took a week to complete in the original implementation, will finish well within one hour in the improved parallel implementation, with identical results. Other findings of the research show that the dedicated focus on improved performance has also helped finding and solving performance issues in the original model implementation.