Satellite signal cycle slip detection method based on reinforcement learning

Abstract

The invention discloses a satellite signal cycle slip detection method based on reinforcement learning, and the method comprises the steps: 1, carrying out the high-order differential processing of the carrier phase of a satellite signal, amplifying the cycle slip information, and eliminating the influence of integer ambiguity and other errors; 2, converting the carrier phase time sequence into a vector set through phase-space reconstruction, and constructing a cycle-slip-free training sample set by using a cycle-slip-free carrier phase value; 3, defining a prediction variance, a cycle slip detection statistic and a reward function of reinforcement learning; 4, constructing and training a network model of a reinforcement learning method until the number of iterations reaches the number of samples; and 5, judging whether the satellite signal has cycle slip or not. According to the invention, active learning can be realized, feedback can be obtained from the environment, and effective detection of cycle slip is realized.

Description

technical field [0001] The invention belongs to the field of satellite navigation and positioning, in particular to a satellite signal cycle slip detection method based on reinforcement learning. Background technique [0002] Satellite positioning is based on the principle of ranging, and its positioning methods can be divided into: pseudo-range positioning, carrier phase positioning and differential positioning. Due to the high precision of carrier phase measurement, high-precision satellite navigation applications (high-precision positioning, direction finding, attitude, etc.) all use carrier phase measurement method, but the whole cycle jump of carrier phase (referred to as cycle jump) will affect the accuracy of measurement results. Therefore, the detection and repair of cycle slips is a key problem that must be solved in this field. [0003] Cycle slips refer to the jump or interruption of the entire cycle count caused by the loss of lock of the satellite signal in the...

AI Technical Summary

Problems solved by technology

[0013] The high-order difference method and polynomial fitting method are suitable for single-frequency receivers, but they can only detect large cycle slips of more than 5 cycles, and cannot detect small cycle slips; the ability of the pseudo-range phase combination method to detect cycle slips depends on the pseudo-range measurement. accuracy, so it is not suitable for single-frequency receivers; the ionospheric residual method requires dual-frequency carrier phase values, and is not suitable for single-frequency receivers; High; the wavelet method requires two or more stations to obtain double-difference observations, and the complexity is relatively high

Images