Improved particle filter-based inclined detection ionized layer MUF short-term forecasting method

Abstract

The invention relates to the technical field of short-wave communication, in particular to an improved particle filter-based tilt detection ionized layer MUF short-term forecasting method, which comprises the following steps: initializing a system; carrying out importance sampling by using an unscented Kalman filter algorithm, extracting particles, and carrying out weight updating; judging whether resampling is needed or not, and if the resampling is needed, obtaining a resampled particle set and a corresponding weight through a random disturbance resampling algorithm; updating the state according to the obtained weight and particle swarm; and when the estimated value does not exceed the threshold value, adding one to the cycle index, returning to continuously use the unscented Kalman filter algorithm to perform importance sampling, otherwise, outputting the current particle swarm as a forecasting result; The problems of particle degradation, insufficient diversity and large calculated amount of a particle algorithm are solved, the method can be qualified for a nonlinear and non-Gaussian system state model, and when short-wave communication testing is carried out at an actual station, the frequency prediction precision is high, and the connectivity rate can reach 87% or above.

Description

technical field [0001] The invention relates to the technical field of short-wave communication, in particular to a method for short-term prediction of ionospheric MUF with tilt detection based on improved particle filter. Background technique [0002] The ionosphere is a complex system that changes with time and altitude. It has a vital influence on the propagation of short-wave sky waves, ground-space and inter-satellite information links, and the operation of space platforms. It is the intermediate medium of links and is also affected by solar radiation. , atmospheric circulation and geomagnetic environment and other factors. Accurately obtaining the maximum usable frequency (MaximumUsable Frequency, MUF) in the actual communication process is of great significance in short-wave communication and ionospheric detection. [0003] The frequency prediction of the ionosphere is also mainly divided into two types, long-term forecast and short-term forecast. Compared with long-...

AI Technical Summary

Problems solved by technology

[0005] (1) Insufficient particle degradation and diversity of the particle filter algorithm

It is due to the fact that the particles generated by the prior distribution do not necessarily reflect the real distribution, and a large number of replicated particles in resampling will lead to a decrease in particle diversity

[0006] (2) Resampling technology is inefficient

Resampling techniques such as polynomial resampling, residual resampling, layered resampling, and system resampling can solve the problem of particle weight degradation, but the diversity of particles is still missing

[0007] (3) Only applicable to linear systems

Due to the complexity and changeability of the ionosphere, the frequency change is assumed to be linear, which is an ideal situation and does not match the actual situation; some traditional methods, such as the Kalman filter short-term forecast method, will fail

[0008] (4) Low prediction accuracy and poor communication connectivity

Most ionospheric short-term forecast methods rely on empirical models, and apply the forecast results to actual short-wave communication links. The connectivity rate cannot reach the general use standard.

Images