Affiliations: [a] College of Engineering, Honghe University, Mengzi, Yunnan, China
| [b] Center of Comprehensive Testing, Quality and Technical Supervision of Honghe Prefecture, Mengzi, Yunnan, China
| [c] Power Plant, YEIG Malong New Energy Generation CO., LTD, Yunnan Energy Investment, Malong, Yunnan, China
Correspondence:
[*]
Corresponding author. Zhimei Duan, College of Engineering, Honghe University, Mengzi, Yunnan, China. E-mail: [email protected].
Abstract: Harmonic pollution and reactive power consumption are the two most important factors affecting power quality. Power electronics, as the largest harmonic source, will produce a large number of harmonics while running, and will also lose reactive power, reduce the power factor of the system and affect power quality. Firstly, the current situation of reactive power and harmonic suppression based on fuzzy control is briefly introduced. Then, the harmonic component and power factor of the power grid are taken as two objective functions to model the power grid. The improved particle swarm optimization (PSO) genetic algorithm is used to establish the harmonic suppression and reactive power compensation model. To improve the design efficiency, aiming at the complexity of the algorithm and the field measurement data, the reactive power compensation device is quickly solved on the Hadoop cloud computing platform to achieve the global optimal compensation effect, to improve the convergence speed and global search ability of the particle swarm optimization algorithm. Thus, based on ensuring the power factor, the harmonic component is restrained and the safety and reliability of power grid operation are improved.
Keywords: Harmonic suppression, reactive power compensation, cloud computing, parallelization
