Intelligent power electronics for smart Industry 4.0 applications

Use of power converters

intelligente Leistungselektronik Anwendungen
© Fraunhofer IISB

Power converters form the basis for modern electrical power supplies in the home (e.g., mains adapters for IT and communications), in industrial plants (e.g., in voltage converters to supply DC networks), and in mobility (e.g., in drive inverters and charging devices for electric vehicles). They are responsible for converting electrical energy (e.g., alternating voltage into direct voltage), switching loads, and controlling electrical drives, as well as numerous other applications. In the process, they continuously record parameters such as current and voltage and how these vary over time – and therefore have the potential to directly determine the status of the system they are connected to, which may be a power grid, a production plant, a battery, or an electric motor.

Cognitive converters create smart systems

Our goal is to harness the relationship between electrical parameters and the complete system by exploring and applying “cognitive power electronics”. In these applications, a cognitive converter uses additional embedded electronics to analyze data regarding electronic power components and makes intelligent decisions based on data derived from the connected system. This requires the recording of relevant electrical parameters within the converter, the identification and validation of correlations between electrical and system parameters, and finally the translation of the relationships identified into intelligent algorithms. Depending on the degree of decision-making autonomy afforded to electronic power components, this results in everything from smart systems to cyber-physical systems (CPS).

Interplay between data and analytics expertise

The »Intelligent power electronics« application involves an interplay between several levels of expertise at the ADA Lovelace Center: the »Data« area of expertise researches solutions for data processing and analysis within the power converter. Ultimately, the actual intelligence of the cognitive converter is achieved in the »Analytics« area of expertise through the adaptation and implementation of suitable machine learning procedures and algorithms.