Three-phase Balanced Loss Reduction Strategy For Distribution Power Network

With the development of economy and technology in China,more and more kinds of electrical appliances are used in residential areas.However,the three-phase unbalance of district distribution network will be caused by different user’s electricity usage habits.The three-phase unbalance will increase additional distribution network losses and endanger the safe use and stable operation of electrical equipment.How to effectively control the three-phase unbalance of distribution network side is one of the hot issues in power enterprise research.Therefore,this paper mainly studies the treatment of three-phase unbalarnce in distribution network side.First of all,the research background,adverse effects and research status of three-phase unbalance are introduced and two ideas of clustering analysis of user’s electricity consumption law are presented.To begin with,the load characteristic model is constructed by introducing Hilbert-Huang transform and the clustering analysis of power consumption law is earried out on the basis of the characteristic model.Then,the load curve is clustered directly based on the shape of load curve.Further more,the idea of balanced allocation strategy based on cluster analysis is introduced.In addition,the basic concept of load curve is introduced and the concept of classical load characteristic index is expounded.The feasibility and method of using Hilbert-Huang transform to construct load characteristic model are also illustrated.Subsequently,the K-means clustering algorithm for clustering analysis of users’ electricity consumption rules is introduced.Experiments were carried out with 600 users’ electricity data in a certain area of Shaanxi Province as data source to verify the validity of clustering analysis based on this feature model.Besides.because K-means clustering algorithm can not self-adaptively determine the optimal number of clusters,and using Hilbert-Huang transform to extract load features has a large amount of computation.Therefore,we start from the morphological characteristic.s of the load curve itself and directly use the improved adaptive clustering algorithm Ad_SODA to carry out the adaptive clustering analysis of the load curve.The Frechet distance and cosine similarity are used to describe the similarity between the two curves.Compared with the SODA clustering algorithm before the improvement,the algorithm is more suitable for clustering analysis directly according to the shape similarity of the user load curve.The experimental results show that the effect of Ad SODA algorithm is better than that of SODA algorithm.What’s more,an optimal commutation mathematical model is established,which takes the lowest unbalance of three phases in distribution network side and the least adjustment times of power supply phase for users as objective functions.The hill climbing algorithm is used to enhance the local optimization accuracy of MFO algorithm and this improved algorithm is used to solve the balanced allocation problem.The superiority of this method is verified by simulation experiments.Finlly,the general enineering structure of the balanced loss reduction system and the.general structure of the com mutation switch are introduced,which are mainly divided into five modules:voltage acquisition module,current acquisition,module,switching module,zero-crossing detection module and wireless communication module.The safety of phase switching of the cormmutation switch is verified by experiments.