Detection and elimination method of "outlier value" in flight data based on aircraft motion equation

Abstract

The invention relates to a method for detecting and eliminating "outlier values" of flight data based on aircraft motion equations, belonging to the technical field of flight mechanics and information processing, and used for checking whether there are "outlier values" in flight data that are greatly different from other measured values. In order to overcome the misjudgment of "outlier value" caused by the prior art method only relying on its own measurement data, the present invention proposes a method for judging and eliminating the threshold value of "outlier value" based on the aircraft motion equation. The speed of change is to predict the flight state and parameters step by step according to the effectively simplified output response, and judge and eliminate the "outlier value" in the measurement according to the set threshold value at each step; due to the possibility of "outlier value" in the fast-changing state or parameters in the actual measurement The characteristic is larger than the slow variable, and the output response has engineering precision. The method proposed by the invention is more effective for eliminating flight data "outliers".

Description

Detection and elimination method of "outlier value" in flight data based on aircraft motion equation technical field The present invention relates to a kind of flight data " outlier " detection and elimination method, particularly relate to based on aircraft motion equation Flight data "outlier" detection and elimination method. Background technique In the actual aircraft airborne measurement and control system, the test quantity is often due to factors such as vibration, coding, and the error of the measuring equipment itself. Due to the inaccuracy of the measurement data, especially in the measurement of acceleration and angular velocity, there are often one or more jump points, these Jump points differ significantly from other measurements and are often referred to as "outliers" in the measurement. Although the number of these "outliers" is small, due to A large amplitude will bring large errors to flight state estimation, navigation, aircraft performance analysi...

AI Technical Summary

Problems solved by technology

The first class method is owing to not utilizing the information contained in the observed value, and credibility is not high, easily causes misjudgment; The 2nd

Although the second type of method has strong adaptability, the algorithm is complicated, which causes an additional burden on the system and is not convenient for online work; the third type

Most of the methods are based on the detection of outliers when filtering is optimal, and the ability to detect outliers in non-optimal filtering is insufficient. Moreover,

No simplification of the equations of motion of the aircraft gives predictive relationships for flight parameters

Images