A Real-Time Water Sediment Regulation Scheme in the Yellow River: An Improved Dynamic Time-Warping Method Coupled with the Fuzzy Set Algorithm

The Yellow River is a typical high sediment loads basin, and its sedimentation restricts the normal operation of reservoirs and the social and economic development of the basin. A study of the water-sediment regulation scheme (WSRS) of the Yellow River is of great significance to managing sediment transport and prolonging the life of reservoirs. However, water-sediment regulation is a complex process including comprehensive action of multiple disciplines such as hydrology, engineering, and environmental science. The existing WSRS cannot provide an effective operation scheme in a short time due to its slow calculation efficiency for water and sediment transport. To address this issue, we propose an optimal WSRS integrating the improved dynamic time warping (DTW) method and the fuzzy set. The improved operation scheme was conducted from two dimensions: time-series process and spatial characteristic attributes. The specific process is to use the fuzzy C-means with data-driven centers and fast dynamic time warping (FCMDD-FDTW) algorithm to find the water and sediment process that is most similar to the current scenario from the historical regulation scheme library; then, according to the spatial characteristic attributes such as numerical attributes, element attributes, and fuzzy attributes of the historical regulation scheme library, obtain the attribute similarity between each historical scheme and the current scenario; finally, select the scheme with the greatest similarity to the time-series process and spatial characteristic attributes of the current scenario for regulation guidance. Results show that the FCMDD-FDTW algorithm shortens the matching time by 6.78–10.39 and 1.01–2.39 s, respectively, compared with the Euclidean distance and fast dynamic time warping (FDTW) algorithms; the new method shortens the matching time efficiency by 60 and 55 s, respectively, compared with the traditional matching method, and the matching efficiency is increased by 40% and 39%, respectively. The research results can provide the technical support for water-sediment regulation in the Yellow River.