Abstract:The optimal operation of short-term cascade hydropower systems, with dozens of hydropower stations, multistage calculation periods and complex hydraulic constraints, faces a serious "curse of dimensionality" problem, and it is difficult to solve this problem directly or obtain a precise optimal hydropower dispatching plan in an acceptable time. This study presents an efficient solution framework based on the load decomposition strategy to alleviate the dimensionality problem. First, the load decomposition strategy is exploited by reducing the number of optimization stages to reduce the dimensionality of the original complicated optimization problem. By continuously adjusting the number of time periods and load of each stage, the load decomposition strategy can divide the original optimization problem into two subproblems with different optimization periods, namely, the segmented load process and burr load process. Second, a hydropower station classification method is proposed, in which all stations are divided into balanced power stations for the burr load and main dispatching stations for the base load, significantly reducing the participation of most hydropower stations in the frequent dispatch process. Finally, according to the transformation of the water balance relationship, a mathematical expression that accurately describes the complex hydraulic connection problem between the two subproblems is constructed to obtain more refined solution results. Practical project cases involving a large-scale hydropower system with 7 stations on the Wu River of China are used to test the sensitivity and efficiency of the proposed method. The sensitivity analysis indicates that the different optimization stages in this method can quickly obtain a globally optimal result and improved computational efficiency. Moreover, compared with the 24-point and 96-point traditional optimization scheduling methods, the water discharge of the presented method is reduced by 6.6% and 7.8% in the dry season and 4.0% and 4.9% in the flood season in a limited time, respectively, which suggests that the solving difficulties caused by the "curse of dimensionality" can be effectively alleviated and the solution efficiency can be greatly improved. The simulation calculations for different seasons and scales also indicate that the proposed method is an effective tool for the optimal operation of large-scale hydropower systems.

Method for LSTM-Based Cascade Hydropower Plant Scheduling

Daily water supply prediction method based on integrated learning and deep learning

A hybrid Deep Learning model based on feature capture of water level influencing factors and prediction error correction for water level prediction of cascade hydropower stations under multiple time scales

Long-, Medium-, and Short-Term Nested Optimized-Scheduling Model for Cascade Hydropower Plants: Development and Practical Application

Resilience-Based Coordinated Scheduling of Cascaded Hydro Power with Sequential Heavy Precipitation

A long-term intelligent operation and management model of cascade hydropower stations based on chance constrained programming under multi-market coupling

Daily Runoff Forecasting Using a Cascade Long Short-Term Memory Model that Considers Different Variables

Optimizing hydropower scheduling through accurate power load prediction: A practical case study

Multistep Prediction of Multiple Small Hydropower Stations’ Total Power in a Watershed within a Day Considering the Distributed Discharge TCN-LSTM Model

Exploring the application of machine‐learning techniques in the next generation of long‐term hydropower‐thermal scheduling

Medium-term peak shaving operation of cascade hydropower plants considering water delay time

Research on the Hydropower Generation Optimal Scheduling of Larger-Scale Cascade Hydropower Stations on Jinsha River and Lancang River

Intelligent Low-Consumption Optimization Strategies: Economic Operation of Hydropower Stations Based on Improved LSTM and Random Forest Machine Learning Algorithm

An optimization-based algorithm for scheduling hydrothermal power systems with cascaded reservoirs and discrete hydro constraints

Scheduling Hydrothermal Power Systems with Cascaded and Head-Dependent Reservoirs

Research on Medium- and Long-Term Hydropower Generation Forecasting Method Based on LSTM and Transformer

An Integrated Transfer Learning Method for Power Generation Prediction of Run-off Small Hydropower in Data-scarce Areas

Short-Term Load Forecasting Based on Deep Learning Bidirectional LSTM Neural Network

Daily Generation Scheduling of Cascade Hydro Plants Considering Peak Shaving Constraints

Solution framework for short-term cascade hydropower system optimization operations based on the load decomposition strategy

Research on short-term multi-level and multi-objective optimal scheduling of cascade hydropower stations under complex water resources and power constraints