A complementary filter consisting of a GPS-receiver and an accelerometer

Abstract

Complementary filters have become widely used to measure the angular position of moving objects in space due to their simplicity and high accuracy. If at first they were used mainly in navigation systems, now they are beginning to be widely used in navigation systems. In such complementary filters, the complementary filter most often includes a GPS receiver and an accelerometer. The problem of reducing the influence of high-frequency interference of the GPS receiver on the accuracy of the system is considered. Direct filtering of the output signal of the GPS receiver has a significant disadvantage, which is the delay of the filter of the measured signal. This is often the main obstacle of using filters. Complementary filters due to the combination of two meters make it possible to filter the signal without affecting the passage of the useful signal. This is the main advantage of complementary filters. It should also be noted that they are very simple to implement, which has led to the widespread use of complementary filters in systems for measuring the motion of objects. The paper proposes a second-order filter structure in which there is no integration of the accelerometer signal that significantly simplifies the implementation of the filter and eliminates the problem of the influence of the non-zero accelerometer signal. The presence of zero offset accelerometer and integration procedures is one of the main problems in solving navigation problems.
 In the absence of an integrator, the question arises of taking into account the initial value of the output signal of the GPS receiver. This issue in the work is solved by the circuit method. In complementary filters built on the compensation scheme, this problem is solved quite simply. The article considers a simpler complementary filter, built on the filtration scheme. An algorithm for taking into account the initial conditions is proposed for such a filter structure. The influence of the structure of the transfer functions of the filter on the efficiency of its use is analyzed. It is shown that the complementary filter can significantly reduce the influence of high-frequency noise of the GPS receiver. The block diagram of the filter in the Matlab-Simulink environment and the simulation results that show the effectiveness of the proposed complementary filter structure are presented.
 Such systems can be used in small moving objects like UAVs, in robotics, in the educational process.