RANK PROCESSING OF VIBRATION SENSOR SIGNALS FOR SIGNALING THE LANDING OF A AMPHIBIOUS AIRCRAFT UNDER CONDITIONS OF A PRIORI UNCERTAINTY

Abstract

The purpose of the work is to use a rank model of signal processing for signaling the landing of an amphibious aircraft. Rank processing refers to nonparametric methods of detecting a signal against a background of interference. Nonparametric methods are used if the functional type of the input data distribution is unknown and only the most general differences between the presence and absence of a signal are indicated. Almost all nonparametric detectors contain devices as a component element that perform some invariant transformation of the S array of sample values of X. As a result of this transformation, a new array Z = SX is formed, the distribution of elements of which is precisely known in the absence of a signal. The transformation S, which is chosen heuristically, allows us to reduce the problem of detecting a signal against a background of interference with an unknown distribution to the problem of testing a simple hypothesis regarding the distribution of the array Z. Research objectives: 1) preliminary digital filtering of amphibious aircraft flight records for the use of rank processing; 2) conducting an experiment to obtain the characteristics of a rank detector used to signal the landing of an amphibious aircraft; 3) analysis of the results obtained. A model of signal processing of a vibration sensor for signaling the landing of an amphibious aircraft is proposed. The model consists of a bandpass filter (PF), a COEX calculator, a divider, a rank detector and a reference sample generator. The rank detector allows you to reduce the task of detecting the vibration sensor signal against the background of interference to the task of testing a simple hypothesis regarding the distribution of ranks. The nonparametric Watson agreement criterion is used to make a decision on the presence of a drive signal. The proposed processing model provides the following parameters of the flood alarm system: 1) insensitivity to changing characteristics of signals and interference, 2) the decision-making algorithm guarantees high quality of detection in conditions of significant a priori uncertainty. The results of the conducted studies show: 1) the distribution of ranks in the absence of a signal is always approximated by a uniform distribution law. In situations where a signal is present in the mixture, the uniform distribution is destroyed, and the presence of a signal is determined by a set threshold, 2) the sensor based on rank criteria provides high quality detection of the amphibious aircraft landing signal. The proposed approach to solving the detection problem can find a place in many applied problems where there is a priori uncertainty. For example, in radar, sonar, communications, medicine and other fields of science and technology.