Icing Potential Index of Aircraft Icing Based on Fuzzy Logic

Aircraft icing, a cumulative hazard, is one of the major weather hazards affecting aviation. It reduces aircraft efficiency by increasing weight, reducing lift, decreasing thrust, and increasing drag. Icing also seriously impairs aircraft engine performance and causes false indication on flight instruments, loss of radio communications and failures of control panel, brakes, and landing gear. Therefore, the prediction of aircraft icing is one of the key research focuses.In order to establish an aircraft icing potential index with more reasonable threshold and easy to adopt, 372 aircraft icing cases and corresponding observation data from 2014 to 2017 provided by Beijing Weather Modification Office are analyzed based on fuzzy logical principles. The membership function of temperature and relative humidity derived from those data analysis is used to calculate the initial possibility of icing. On this basis, the membership function representing the influence of vertical velocity and cloudiness on the initial possibility of icing is determined using the national pilot reports (PIREPs) in 2016 and corresponding ERA5 reanalysis data to screen different forms of membership functions. Based on membership functions of temperature, relative humidity, vertical velocity and cloudiness, the icing potential index (I_p) can be calculated by using output from numerical weather prediction model.According to ERA5 reanalysis data, 61 icing cases and 45 non-icing cases are used to test the effectiveness of I_p. Results show that the accuracy, missing alarm rate and false alarm rate of I_p are 80.2%, 9.4% and 10.4%. Compared with the commonly used icing index (I_c), the accuracy of I_p is better, the missing alarm and false alarm reduce significantly. However, it should be noted that the difference between aircraft type and flight speed of different aircraft icing cases in this study is not discussed, and it is assumed that effects of vertical velocity and cloud cover on the initial possibility of icing are independent, which need further study.In summary, the established icing potential index (I_p) based on fuzzy logical principles is efficient and feasible, and provides information for pilots to avoid high-risk areas of icing in the air. Combining with the regional numerical weather prediction model, it can output the possibility of icing in certain areas under certain meteorological conditions and provide reference for pilots to avoid high-risk areas of icing in the air.