Humans have always had the necessity of estimating their location in space for various reasons, e.g. hunting, traveling, sailing, battling, etc. Today, many other areas also demand that information, such as aviation, agriculture, multiple smartphone applications, law enforcement, and even film industry, to mention but a few. Estimating position and orientation is known as navigation, and the means to achieve it are called navigation systems. Each approach has its pros and cons, but sometimes it is possible to combine them into an improved architecture. For instance, inertial sensors (i.e. accelerometers and gyroscopes) can be integrated with magnetometers, producing an Attitude and Heading Reference System (AHRS); this process is referred to as sensor fusion. However, before sensors can be used to produce the navigation solution, calibration is often necessary, especially for low-cost devices. In this study,we perform the calibration of a triaxial consumer-grade magnetometer via an extended two-step methodology, correct small mistakes present in the original paper, and evaluate the technique in a restricted motion scenario. This technique can be implemented in-field, simply by rotating the sensors to multiple orientations; the only external information necessary is the local Earth's magnetic field density, easily estimated through reliable models. The error parameters, i.e. biases, scale factors, and misalignments, are indirectly estimated via a least squares algorithm. The calibration is first performed through software simulation, followed by hardware implementation to validate the results.
Improved hardware implementation of a TSR based MPPT algorithm for a low cost connected wind turbine emulator under unbalanced wind speeds