Fast experimental energy characterization of a controlled rotary axis

The goal of the paper is to present a methodology and the related instrumentation for fast energy characterization of industrial machines. Standard energy measurements prescribe acquiring power flowing into the system and into every significant subsystem of the machine using a considerable number of power meters (or repeated measurements) that determine high installation times, costs and efforts. Often, such measurements are missing additional information about machine activity, needed to understanding energy dissipations. Fast experimental characterization method proposed in the paper assumes to measure solely the total power absorbed by the machine and to estimate the consumptions of machine subsystems by acquiring additional signals, e.g. states and variables, via a newly developed wireless, non-invasive measurement system. Knowledge of these states allows to create an intermediate-precision energy models of the system. A case study is performed to identify and analyse such model of axis in an industrial robot. Apart from power, two additional signals are acquired: axis acceleration via accelerometer and current of one of the motor phases, using a current transducer. Techniques of sensor data fusion are employed to obtain more reliable results. Axis energy models are identified and compared to find the best match for the considered application.