Title: Real-time intelligent control and cascading failure prevention in microgrid systems based on neural network algorithm: an experimental approach
Authors: Rabie Belkacemi; Sina Zarrabian
Addresses: Department of Electrical and Computer Engineering, Tennessee Technological University, Cookeville, TN, 38505, USA ' Department of Electrical and Computer Engineering, Tennessee Technological University, Cookeville, TN, 38505, USA
Abstract: This paper presents an intelligent control method based on artificial neural network (ANN) to prevent cascading failures and blackout in microgrid systems after N1 contingency condition. Microgrids have low inertia as compared to the utility power grids which makes their control very challenging. The main contribution of this work is to utilise the machine learning structure of ANN to prevent blackout and make microgrids more reliable and resilient. This method is able to relieve the congestion on lines by adaptive power re-dispatch to prevent consecutive line outages. The proposed ANN control approach is tested on an experimental test system. Experimental results show that the ANN approach provided accurate and robust control and management of the microgrid system by preventing a total system collapse. The technique is compared to a heuristic multi-agent system (MAS) approach based on communication interchanges. The ANN showed a faster and better response than the MAS.
Keywords: ANNs; artificial neural networks; failure prevention; microgrid; adaptive control; blackout; cascading failures; N–1 contingency; real-time systems; intelligent control; reliability; resilience; multi-agent systems; MAS; agent-based systems.
International Journal of Power and Energy Conversion, 2016 Vol.7 No.3, pp.292 - 311
Available online: 23 Aug 2016 *Full-text access for editors Access for subscribers Purchase this article Comment on this article