Research On Landing Risk Assessment Of Carrier- Based Aircraft Based On Multi-Attribute Decision Making

Carrier-based aircraft is an important weapon in ocean war. Carrier-based aircraft safely landing is one of the most key in the chain link of operations by 80% of accidents being from landing. Due to the loss and the harmfulness of carrier-based aircraft, carrier-based aircraft landing technique is extremely important to the researchers. Not only Automatic Carrier Landing System (ACLS) and deck are improved continuously, also landing risk assessment of carrier-based aircraft is researched gradually. For increasing accuracy rate, two methods of landing risk assessment are improved by the guide of pilot and ACLS, including dynamic multi-attribute decision making (DMADM) based on Mean-Variance and landing risk decision making based on multi-attribute preference. Two methods assess landing risk of the multi-flights, achieving the comprehensive assessment of flight performance.Firstly, this paper introduces aircraft relevant coordinates. On the basic of analyzing carrier-based aircraft force situation and motion state, carrier-based aircraft landing system is built. Due to the limitations and the lack of traditional landing risk assessment and multi-attribute decision making, dynamic multi-attribute decision making of carrier-based aircraft landing risk based on Mean-Variance is improved. The improved method combines mean with variance, ananlyzing and judging, getting order plementing the relevance of different moments and reflecting the importance of landing stability.Some reasons and the lack of optimizing model are analysed by multi-attribute group decision applying in landing risk assessment. Multi-attribute group decision making with preference is presented. The optimized model is built by two-dimension triangular fuzzy number expectation value, getting attribute weights.The coherence of system data and specialistic information is implemented. Unknown attribute weights are solved.In the MATLAB environment, carrier-based aircraft system model is simulated, acquiring landing information and distributions of touchdown points on deck. According to comparing improved method and traditional method, the same or similar scheme orders is better solved by the improved dynamic decision making method. And acquired attribute weights can approximate standard attribute weights by the improved preference group decision making, getting a more accurate order result. The feasibility and availability of the improved methods are proved by simulation results.In basis of vast experiments for two flights, statistical result shows that the improved dynamic multi-attribute decision making is better applied in the condition of a pilot flighting multiple routes. And the improved methods is universality for the similar appications.