Error Registration For Temporal And Spatial Alignment On HFSWR

The problem of spatial and temporal alignment for multi-sensor system has been a great problem in the field of information fusion. To search for the registration method thus to promote the system integrated performance is of great meaning.This article is based on the High Frequency Surface Wave Radar (HFSWR). The HFSWR takes the advantage of the high-frequency electromagnetic waves with vertical polarization. The waves can propagate over the sea level without attenuation caused by long distance. Thus, HFSWR can implement over the horizon detection. Because of the special detection characteristic of HFSWR, the temporal and special alignment in this system is more complex than usual system. With consideration of the detection theory as well as the application of the spatial geometry, this article proposes new method for spatial alignment. The error after transformation is analyzed.Error can be divided into two types in HFSWR: one is called system error, which is constant or can be considered slowly changed; the other is called random error, which is considered as gauss distribution with mean zero. This article manages to found a model with consideration of the two errors simultaneously. KALMAN Filter theory is used here. As a result, the system state vector is made up of target location and speed as well as the system error. Consider the nonlinear measurement system, this article pulls in the Extended KALMAN Filter, Iterated Extended KALMAN Filter and the Particle Filter, estimating the target location and the system error. Error caused by temporal alignment can be divided into several parts. Thisarticle analyzes the error caused by different sampling periods of two sensors in the system. Least Square Algorithm and the Lagrange Interpolation Algorithm are used for temporal alignment. Simulations show that the two are both effective. However, the Lagrange Interpolation is more widely used.