Optimal PMU configuration method for distribution networks based on information entropy gain

The high penetration of distributed renewable energy in distribution networks, coupled with complex load characteristics, imposes higher demands on automated dispatching systems. Currently, the low redundancy in distribution network measurement data necessitates optimizing the configuration of measurement devices to enhance online perception capabilities. This study addresses the challenge of multi-source heterogeneous measurement data fusion in existing distribution network measurement systems by proposing a PMU optimal configuration method based on information entropy theory, aiming to improve state estimation accuracy and system observability.First, the differences in measurement characteristics among various devices are analyzed, and a multi-source measurement data fusion strategy is proposed to resolve the inconsistent sampling periods. Second, information entropy theory is introduced to quantify the time-series fluctuations and information gain of measurement data. A time-series measurement information entropy model is established, and an optimization configuration model is constructed with the objective of maximizing PMU information entropy gain. Finally, simulations are conducted using the IEEE 33-node system and a 51-node system in a Shanghai region. Results demonstrate that the proposed method significantly reduces state estimation errors through dynamic weight adjustment and multi-source data fusion, while the PMU configuration scheme balances economic efficiency with information gain. • Developed an entropy-based model to quantify temporal dynamics of AMI/FTU/PMU measurements, overcoming static-data limitations. • Proposed dynamic weight fusion resolving time asynchrony via PMU-synchronized multi-source data integration. • Maximized entropy gain in PMU placement for information yield beyond traditional topological observability. • Validated on 51-node Shanghai grid enabling real-time state estimation for renewable-dense distribution networks.