CN-115524767-B - Civil aviation weather short-time forecasting method based on weather radar echo extrapolation

Abstract

The invention discloses a civil aviation weather short-time prediction method based on weather radar echo extrapolation, which constructs a motion vector field of a radar echo map for radar waves by an optical flow method combined with a pyramid algorithm, extrapolates the motion vector field to obtain a predicted radar echo map of future time, and fuses an ADS-B terminal system for display and early warning. The method realizes 2-hour extrapolation prediction of weather radar echo, is beneficial to providing references for weather, control and flight personnel to grasp weather, realizes a diversified platform organically integrating weather, control and flight, is beneficial to establishing consistent situational awareness for weather, control and flight personnel, and ensures flight safety. Therefore, control and flight personnel can effectively and timely acquire the latest weather information and dangerous weather early warning information, and the control and flight personnel are combined with the ADS-B technology to guide the aircraft to avoid dangerous weather according to the aircraft flight path and the weather early warning information, so that the utilization rate of an airspace is improved.

Inventors

• LONG YANYAN
• LI FEI
• Kang Xianbiao

Assignees

• 中国民用航空飞行学院

Dates

Publication Date

20260508

Application Date

20221027

Claims (1)

1. 1. A civil aviation weather short time forecasting method based on weather radar echo extrapolation is characterized by comprising the following steps: S1, acquiring radar echo data of multiple radars and ADS-B terminal system data, preprocessing the radar echo data by using a bilinear interpolation algorithm and a Lanbert projection method, and splicing the preprocessed multiple radar echo data into a radar echo diagram; s2, correcting pixel points in the reduced radar data echo diagram by using a pyramid algorithm to obtain a reduced optical flow field, performing displacement estimation on the reduced optical flow field by using an optical flow method, and iteratively calculating an accurate optical flow field to obtain a radar echo diagram motion vector field, wherein the method specifically comprises the following steps of: s21, scaling down the spliced radar echo data graph obtained in the step S1; s22, correcting pixel points in the reduced radar data echo diagram by using a pyramid algorithm to obtain a reduced optical flow field, wherein the method comprises the following steps of: s221, setting initial pixel point displacement Calculating the estimated position of the initial pixel point in the next frame of image: Wherein, the The estimated position of the initial pixel point in the next frame image is taken as an X, and the position of the initial pixel point in the current frame image is taken as an X; s222, calculating an intermediate variable of the optical flow field according to the radar return wave field displacement, wherein the calculation mode is as follows: Wherein Deltab (X) and A (X) are optical flow field intermediate variables, A 1 (X) is a symmetrical matrix coefficient corresponding to a first frame radar echo image pixel point, For the symmetric matrix coefficient corresponding to the radar echo image pixel point of the next frame after displacement estimation, B 1 (X) is the vector coefficient corresponding to the radar echo diagram of the first frame after displacement estimation; S223, iterating by using the obtained intermediate variable of the optical flow field to obtain a radar echo diagram motion vector field; s23, performing displacement estimation on the reduced optical flow field by using an optical flow method, and iteratively calculating an accurate optical flow field to obtain a radar echo diagram motion vector field; S3, based on the motion vector field, extrapolating a radar echo moving track of a future time by adopting a semi-Lagrangian method, and superposing the radar echo moving track and the existing time echo map to generate a predicted radar echo map, wherein the method specifically comprises the following steps of: S31, calculating a radar echo moving track of future time by using a semi-Lagrangian method; the specific calculation method for calculating the radar echo moving track of the future time by using the semi-Lagrangian method in S31 is as follows: S311, dividing the forecast period τ into a plurality of intervals, expressed as: τ=N×Δt wherein N is the number of intervals, and Deltat is the time length of each interval; s312, overlapping the pixel point displacement in each interval to obtain an extrapolated image; s313, iterating the displacement of the pixel points in each section of interval to obtain a radar echo moving track of future time; The specific calculation method for iterating the displacement of the pixel point in each section of interval in S313 is as follows: Wherein a n+1 is the displacement of a single period obtained after iteration; is the pixel point in the optical flow field The speed at the position, n is the iteration number; s32, superposing the future time radar echo moving track obtained by the calculation in the S31 and the existing time-identified radar echo map to obtain a predicted radar echo map; s4, fusing the predicted radar echo diagram with system data acquired by the ADS-B terminal system, setting an alarm threshold and displaying the alarm threshold in a graphical interface mode, so that control and flight personnel can effectively and timely acquire the latest meteorological information and dangerous weather pre-warning information, and the latest meteorological information and dangerous weather pre-warning information are combined with an ADS-B technology to guide an airplane to avoid dangerous weather.

Description

Civil aviation weather short-time forecasting method based on weather radar echo extrapolation Technical Field The invention relates to the field of weather prediction, in particular to a civil aviation weather short-time prediction method based on weather radar echo extrapolation. Background At present, control and weather in an empty pipe system are lack of organic fusion, so that in severe weather, a controller can only consult weather conditions of the weather person in a telephone way, and the telephone dictation way enables the controller to lack of visual understanding of the severe weather, and a pilot cannot timely and effectively master the weather with various changes. Therefore, it is significant to develop a system that fuses weather radar information with ADS-B information. The system is oriented to control and flight personnel, and provides a meteorological information access platform organically combined with weather, so that the control and flight personnel can effectively and timely acquire latest meteorological information and dangerous weather early warning information, and the system is combined with an ADS-B technology, and guides an airplane to avoid dangerous weather according to airplane tracks and the meteorological early warning information, so that the utilization rate of an airspace is improved. Disclosure of Invention Aiming at the defects in the prior art, the invention provides a civil aviation weather short time forecasting method based on weather radar echo extrapolation. In order to achieve the aim of the invention, the invention adopts the following technical scheme: a civil aviation weather short time forecasting method based on weather radar echo extrapolation comprises the following steps: s1, acquiring radar echo data of multiple radars and ADS-B terminal system data, and preprocessing the radar echo data to obtain a radar echo diagram; S2, calculating a radar echo diagram motion vector field obtained in the step S1 by using an optical flow method combined with a pyramid algorithm; s3, extrapolating a predicted radar echo diagram of future time based on the radar echo diagram motion vector field obtained in the step S2; S4, fusing the predicted radar echo diagram obtained in the step S3 with ADS-B terminal system data, setting an alarm threshold and displaying the alarm threshold in a graphical interface mode. Further, the method for preprocessing the radar echo data to obtain the radar echo map in the S1 is as follows: preprocessing radar echo data by using a bilinear interpolation algorithm and a Lanbert projection method, splicing radar echo data graphs of a plurality of radars obtained after preprocessing, and acquiring echo pictures of thunderstorm activities according to an empirical threshold. Further, the step S2 specifically includes the following steps: s21, scaling down the spliced radar echo data graph obtained in the step S1; S22, correcting pixel points in the reduced radar data echo diagram by using a pyramid algorithm to obtain a reduced optical flow field; s23, performing displacement estimation on the optical flow field subjected to absolute shrinkage by using an optical flow method, and performing iterative calculation on an accurate optical flow field to obtain a radar echo diagram motion vector field. Further, the step S22 specifically includes the following steps: s221, setting initial pixel point displacement Calculating the estimated position of the initial pixel point in the next frame of image: Wherein, the The estimated position of the initial pixel point in the next frame image is taken as an X, and the position of the initial pixel point in the current frame image is taken as an X; the step S22 specifically includes the following steps: s221, setting initial pixel point displacement Calculating the estimated position of the initial pixel point in the next frame of image: Wherein, the The estimated position of the initial pixel point in the next frame image is taken as an X, and the position of the initial pixel point in the current frame image is taken as an X; s222, calculating an intermediate variable of the optical flow field according to the radar return wave field displacement, wherein the calculation mode is as follows: Wherein Deltab (X) and A (X) are optical flow field intermediate variables, A 1 (X) is a symmetrical matrix coefficient corresponding to a first frame radar echo image pixel point, For the symmetric matrix coefficient corresponding to the radar echo image pixel point of the next frame after displacement estimation,B 1 (X) is the vector coefficient corresponding to the radar echo diagram of the first frame after the displacement estimation. S223, iterating by using the obtained intermediate variable of the optical flow field to obtain a radar echo diagram motion vector field. Further, the step S3 specifically includes the following steps: S31, calculating a radar echo moving track of future time by using a semi-Lagra