Prediction and control of Hopf bifurcation in a large‐scale PV grid‐connected system based on an optimised support vector machine. Issue 14 (19th January 2018)

• Record Type: Journal Article
• Authors: Li, Sheng; Wei, Zhinong; Ma, Yanan; Cheng, Jielei
• Is Part Of: Journal of engineering. Volume 2017:Issue 14(2017)
• Page Start: 2666
• Page End: 2671
• Publication Date: 2018-01-19
• Subjects: photovoltaic power systems -- load shedding -- bifurcation -- power grids -- support vector machines -- power engineering computing -- particle swarm optimisation -- genetic algorithms -- power system dynamic stability -- power generation control; large‐scale PV power station -- control method -- online prediction -- emergency control capability -- load fluctuation -- load bus voltage -- oscillatory collapse -- emergency load‐shedding measure -- equilibrium point -- current operating point -- HBI prediction value -- particle swarm optimisation -- genetic algorithm optimisation -- SVM algorithm -- HBI index -- system voltage stability evaluation process -- UHB point position -- large‐scale photovoltaic power station -- saddle‐node bifurcation -- voltage collapse -- power system -- unstable Hopf bifurcation -- optimised support vector machine -- large‐scale PV grid‐connected system; Engineering -- Periodicals; Engineering; Electronic journals; Periodicals; 620.005
• DOI: 10.1049/joe.2017.0836
• Languages: English
• ISSNs: 2051-3305

Abstract

Abstract : The unstable Hopf bifurcation (UHB) of a power system generally precedes the saddle‐node bifurcation, so the UHB point can be the system critical point of voltage collapse. However, after the large‐scale photovoltaic (PV) power station is connected to the system, the fluctuation characteristic of PV output can lead to the change of UHB point position and the evaluation process of system voltage stability is also affected. Therefore, according to the feature that the UHB point has the minimum value of hopf bifurcation index (HBI) index, a method for predicting the equilibrium HBI value is proposed based on an optimised support vector machine (SVM) algorithm (including particle swarm optimisation SVM and genetic algorithm optimisation SVM). Once the HBI prediction value of the current operating point (the equilibrium point) is less than or equal to the preset HBI alert value, it is possible to take an emergency load‐shedding measure to remove a small amount of load, and the position of UHB point under maximum PV output is used to set the amount of load‐shedding, so the oscillatory collapse of the load bus voltage caused by a sudden increase of PV output or load fluctuation can be avoided. Taking a large‐scale PV power station access to the classic three‐node system as an example, the prediction and control method is verified. Simulation result shows that the method is simple and practical, as well as suitable for on‐line prediction and control of UHB point in a PV grid‐connected system, which is able to improve the prevention and emergency control capability of the system voltage stability.