Development of Intelligent Controlled Microgrid for Power Sharing and Load Shedding

An online trained intelligent voltage controller is developed in this article to improve the transient responses of the voltage and frequency of an islanded microgrid with droop control and achieve fast load shedding. Although the droop control algorithm has been extensively adopted in microgrid, some shortcoming of the droop control, such as poor disturbance rejection and slow transient response, seriously affect the power quality and stability in islanded microgrid. Moreover, when the power consumption of the demand exceeds the generated power of the distributed generators in microgrid, the decreased frequency will lead to the microgrid blackout owing to the slow load shedding characteristics of the droop control. Hence, in this article, an online trained Petri probabilistic wavelet fuzzy neural network (PPWFNN) controller is developed as the voltage controller to replace the traditional proportional--integral controller in a battery energy storage system for an islanded microgrid with droop control to achieve fast load shedding. The network structure and the online learning algorithm of the proposed PPWFNN are detailedly introduced. Finally, some experimental results are provided to validate the effectiveness of the microgrid using the proposed PPWFNN controller for the improvements of the power sharing and load shedding.

Automated summary

This article develops an online trained intelligent voltage controller to improve the transient responses and achieve fast load shedding in a microgrid with droop control. The proposed controller, a Petri probabilistic wavelet fuzzy neural network, replaces the traditional proportional-integral controller in a battery energy storage system to enhance power sharing and load shedding.