A Drone Localization System Based On Multiple Time Delays Fusion

In recent years,there are many vicious incidents about drones,which not only seriously endangers people's lives and property,but also has a negative i mpact on the national level.Therefore,research on countermeasures of drones i s a hot work direction;drone countermeasures can be carried out from three a spects of audio,video and radio frequency.Audio technology has become the f ocus of this article because of its low cost and easy implementation.In this paper,a drone monitoring system based on the audio signals of th e drone rotor vibration is proposed.The main task is to use the audio signals collected by the system to detect and locate the drone.Therefore,the innovatio nal work of this paper can be divided into two parts:The first part is the calculation optimization of arrival time delay(TDOA).This paper describes a series of noval improvements about Generally Cross Co rrelation.It regards the problem of computing delay as a problem of tracking t arget.Thus,it uses Kalman filter to connect multi frames data and correct the predictions of delay by feedback of candidate delay,so the precision can be i mporved.Based on experiments,this paper also verifies the effectiveness of ab ove method.The second part is the localization of drones.This paper describes the tra ditional time delay fitting algorithm,and describes the limitations of this metho d which is vulnerable to noise interference based on the actual data processing results.In order to solve this limitation,a new set of regular term based posi tioning algorithm is proposed in this section,which effectively reduces the inte rference of noise to positioning.Finally,the positioning results and error analys is of multiple groups of experimental data confirm the proposed method The al gorithm is superior to the traditional method.