Research On Key Technologies Of Airport Pavement Foreign Object Detection Rada

Foreign Object Debris(FOD)refers to foreign objects on airport runways that may pose a risk to aircraft.In order to detect FOD in time,domestic and foreign research teams have developed several detection devices,but these devices have been suffering from the problem of poor detection performance for small size FOD.To address this difficult issue,this paper proposes Refined Feature Classification-Clutter Map(RFC-CM)detection technology and distributed FOD detection processing technology from the analysis of the clutter model,and completes the algorithm verification by AIRCAS-FOD detection system.The experimental results showed that the above technical proposal effectively improved the detection capability of the system for weak FOD.The main contributions of this paper are as follows:1.The analysis of airport road surface clutter modeling is completed based on measured data.The statistical model of clutter on the road surface and the clutter simulation method are introduced,the spatial and temporal correlation characteristics of the clutter on the road surface are analyzed,and the verification of the statistical model and the spatial and temporal correlation characteristics is completed by the measured data,which provides a theoretical basis for the subsequent Constant False Alarm Rate Detector(CFAR)technology and the distributed FOD detection and processing technology.2.To address the high false alarm and high miss detection problems of existing detection techniques for weak FOD detection,a Bi Parametric Clutter Map technique is studied based on the clutter model of the runway surface.The simulation results show that the technique has excellent detection performance for weak FOD in the runway,but will produce missed detection and serious false alarms in the clutter edge and grassland.In order to adapt to the practical application requirements,the RFC-CM-CFAR technology is proposed on this bases.The technique uses the amplitude gradient characteristics of the clutter on the ground for terrain classification,and then uses different detection methods to achieve the detection of weak FOD according to the clutter model of different terrains..In order to verify the effectiveness of the technique,an outfield experiment was designed and conducted.The experimental results show that the RFC-CM-CFAR technique can effectively detect targets with a particle size of 1.5 cm.In addition,compared with other techniques,the KAPPA coefficient of this technique is 0.95,and the detection performance is significantly better than that of the conventional CFAR technique.3.To further detect targets with particle size less than 1.5 cm,a distributed FOD detection and processing technique is proposed.The core of this technique is to use the different spatial correlation characteristics of the target and the clutter to improve the signal to clutter ratio(SCR)of the weak FOD through the accumulation of multi-station echoes and achieve the effective detection of the weak FOD.Firstly,the clutter space decorrelation constraint is proposed by combining the clutter correlation features;Secondly,the "ping-pong" correlation parameter model is proposed in consideration of the deterioration of the performance caused by the non-ideal orthogonal signal;Thirdly,in view of the poor estimation accuracy of the correlation parameters in the low SCR condition by the traditional peak method,the pre-estimation method is proposed.Finally,semi-physical simulation experiment is completed based on the measured clutter data.The results show that the distributed FOD detection processing technique improves the SCR of weak targets by 6 dB.