RAGA-PPE Model for Power Quality Comprehensive Evaluation

ABSTRACT

This paper proposes RAGA-PPE comprehensive assessment model specifically for high-dimensional index system issue in order to conduct comprehensive evaluation of the power quality of multi-index regional power grid from three dimensions of index, time and region through adjustment of time factor size. The model adopts PPE to convert power index of multidimensional data to lower-dimensional subspace and compress multiple power indexes to single power index for systematic comprehensive evaluation, which overcomes shortcomings such as influences of human and subjective factors and local optimization in conventional assessment methods. Improved RAGA based on real coding is used for optimization of target function to reduce optimization workload and search out the optimal projection direction rapidly. The optimal projection direction subject to optimization reflects the degree of importance of various power indexes for comprehensive evaluation of power quality. Examples prove that the evaluation model is simple in computation, good in applicability, and accurate and objective in evaluation result, providing a new method and thought for decision of comprehensive evaluation for power quality.

Keywords:

Power quality, comprehensive evaluation, projection pursuit, RAGA, least variance method, temporal dynamics.