Navigation Path Estimation using UFIR Filter over G28U7FTTL GPS Receiver Data

A previously plotted path is an important requirement in real navigation of unmanned vehicles to avoid diversion or collisions. An important challenge is to obtain specific paths in small farmlands with the minimal economical investment. With this aim, the Global Position System (GPS) technology can be used that allows an object position determination on the entire Earth with highest precision. The cost of the GPS receiver is proportional to accuracy, which is susceptible to disturbances and noise. The GPS sampling time in commercial devices is one pulse-per-second that changes the measurement size per lapse according to the vehicle velocity. In this work we propose using low cost GPS receivers to obtain strategic points around a small farmland and build a high-resolution path. Using these points, a reference path is obtained using a cubic interpolation of GPS lapses and matching the length of samples. The iterative unbiased finite impulse response (UFIR) filter is applied to smooth the reference path. The results are compared to the Savitzky-Golay (S-G) filter-based estimates. It is shown that the two- and three-state UFIR filters produce an acceptable accuracy and outperform the S-G filter.