Inductive perturbation tolerance improvement of hybrid networks using an integrated expert system

摘要

The design and integration of an expert system is presented to solve the instability and unreliability problems caused by exogenous perturbations in hybrid networks. The proposed expert system performs a directional induction technique on the state vectors of the storage units and linearizes the non-linear dynamics on input and link voltages and currents based on the frequency of dispatch, state of charge, depth of discharge, and utilization factors of storage units. The expert system then identifies optimal, real-time control action on the power variables of the storage units that satisfies various performance and reliability targets. Results show that the expert system based inductive perturbation tolerance approach could provide improvements leading to an optimal performance improvement and enables the hybrid network to satisfy system reliability criteria by compensating for dynamic and transient disturbances as compared to other static and dynamic approaches such as heuristic-based, fitness-based or prediction-based techniques.