A performance prediction framework for protection and control applications in substation automation

Protection and control is critical in substation automation in order to safely operate and manage the expensive primary equipment. Protection and control can be developed as a composition of function blocks which are executed periodically on processing devices. To ease the architectural design of substation automation by evaluating the feasibilty of the mapping of protection and control functions on processing devices, this paper presents a performance prediction framework. Instead of a classical time consuming worst case execution time analysis, the proposed framework relies on a pragmatic approach by evaluating the probability of the CPU utilization prediction. The framework takes advantages of the composability property of the protection and control functions by analyzing the performance of each function block individually and then combining them into various configurations. The paper's contribution is the definition of the performance prediction framework as well as its empirical validation against a real-world case study.