Control and estimation problems based on inertial measurements in industrial mechatronic systems

The price of inertial platforms (based on MEMS sensors) is dropping, permitting the use of these sensors in various field, like, for instance, the industrial one. The inertial measurements can be used raw or can be elaborated in order to obtain high level information like the orientation of an object. The thesis approach the problem of using the inertial platforms firstly as orientation sensor, showing some algorithms for the estimation, and then as a way to increase the performance of a mechatronic system. The first application shows an innovative anti sway system for bridge crane, which uses the inertial platform as angle sensor; the estimated oscillation angle is then used inside a control loop to move the crane without sway, increasing the safety and the productivity. To be noted that the control is invisible to the operator, leading to the so called Human In the Loop (HIL) structure. Then, a condition assessment algorithm, based on the analysis of the spectrogram of the acceleration using the Dynamic Time Warping, is explained for a circuit breaker. The last chapter shows an impact detection algorithm for an automatic access gate, based on a sensor fusion between acceleration and the motor torque in order to estimate the friction coefficient and, on this parameter, evaluate the presence of the impact. All the solutions proposed aim to show how these inertial sensors can be used in an innovative way to attain better performance from an industrial mechatronic system.