Software-defined control of an emulated hydrogen energy storage for energy internet ecosystems.

The increasing heterogeneity and scalability of the Internet of everything, especially, the Energy Internet (EI), is a prompt for novel engineering paradigms. The current infrastructure of hardware technologies besides the available software stack is becoming obsolete for handling the advanced requirements of the smart energy ecosystem. Accordingly, software definitions of the Internet of Things (IoT) stack functions are gaining interest to provide more flexible and scalable implementations. This paper investigates a sophisticated softwarization explicit hybrid model predictive control strategy for energy storage facility systems through the accessor design pattern. This strategy is presumed to support and facilitate the integration of higher-level analytic and control functionalities within the IoT stack at edges and nodes. Simulation is conducted with the Ptolemy II software to evaluate the execution semantics of the proposed strategy. In addition, real-time experiments on a lab-scale hydraulic tank process with both abrupt and gradual commands test scenarios showed the efficiency and reliability of the accessor-based controllers for achieving the control objectives for the benchmark storage facility with an abstract implementation. The analysis shows that this process can serve as a benchmark to mimic the charge and discharge cycles of storage elements in EI ecosystems. • A new paradigm for hydrogen energy storage interfacing within energy Internet ecosystems is proposed and investigated. • An actor-oriented approach is applied for implementing real-time control systems of hydrogen storage. • Software-defined model predictive control is implemented within a Node.js accessor host. • A benchmark process is developed as an emulator for hydrogen energy storage to evaluate the accessors design pattern. [ABSTRACT FROM AUTHOR]