Shandong Huanneng Design Institute (China)

A family of structure-dependent methods is proposed based on discrete control theory. Although the displacement and velocity expression of this family method are similar to those of the previously published method developed by Mohammad Rezaiee-Pajand, the structure-dependent parameters of this family are different from the previously published method. The family of structure-dependent methods is named the MUSE algorithm method. Based on discrete control theory, a new family of integration algorithms is proposed by using the poles of the Newmark-[Formula: see text] method. Theoretical analysis indicated that the MUSE algorithm method possesses properties of zero amplitude decay and is self-starting. Also, its Period Elongation can be reduced by parameter ‘[Formula: see text]’. Numerical examples show that parameter ‘[Formula: see text]’ introduced in this paper can improve control Period Elongation and improve the accuracy of this method.

With the deepening of information construction and the rapid development of power business, residents’ electricity consumption behaviors show different characteristics. Power grid enterprises have accumulated rich and valuable data resources, and the analysis of users’ electricity consumption experience and corresponding users’ behaviors has been paid more and more attention. That is, it is necessary to draw a picture of the electricity consumption from the electricity consumption characteristics in combination with the electricity consumption information of users. The realization of process is mostly realized by clustering. Aiming at the problem that traditional K-means algorithm is sensitive to the initial clustering center, this paper proposes an improved clustering analysis method of power consumption load based on ISODATA. When the cluster center is unknown, the clustering effect of this method is less affected by the initial center, and the clustering measurement parameters can be dynamically calculated to achieve more effective data clustering. Finally, based on the actual data, the user’s electricity consumption portrait is realized to verify the effectiveness of the method.