Pumped storage unit adjusting system identification method based on deep learning

Abstract

The invention discloses a pumped storage unit adjusting system identification method based on deep learning, and the method comprises the following steps: (1) carrying out mechanism modeling on a pumped storage unit adjusting system, generating simulation data through a mechanism model, dividing the simulation data into a training set and a test set, and determining an input variable of a deep learning model; (2) a deep learning GRU model is constructed, and deep features of operation data of the pumping and storage unit adjusting system are mined; (3) improving a Harris eagle optimization algorithm, and introducing a nonlinear energy index decline strategy into escape energy; (4) dynamic adjustment of hyperparameters of the GRU model is achieved through the improved Harris eagle optimization algorithm; and (5) obtaining a predicted value by using the trained GRU model and the test set data. According to the method, the structure and parameters of the optimal model are obtained by improving the global optimization capacity of the Harris eagle optimization algorithm, so that the generalization of the GRU model is improved, accurate identification of the pumped storage adjusting system is realized, the modeling precision of the unit is effectively improved, and the control quality of the unit is further guaranteed.

Description

technical field [0001] The invention belongs to the technical field of identification and modeling of pumped-storage units, and more particularly, relates to a deep-learning-based identification method for adjustment systems of pumped-storage units. Background technique [0002] The governor of the pumped storage unit is an important part of the control system of the hydropower station, and the accuracy of its mathematical model has an important influence on the automatic control of the hydropower station and the power quality of the power grid. As an important means of modeling the control system of the pumped-storage unit, the identification of the regulation system of the pumped-storage unit is of great significance to the dynamic analysis of the pumped-storage unit and the entire power grid. Classical identification methods (fuzzy model and traditional neural network model) still have problems of over-learning and over-fitting, the identification results are not satisfac...

AI Technical Summary

Problems solved by technology

The classic identification methods (fuzzy model and traditional neural network model) still have the problems of over-learning and over-fitting, the identification results are unsatisfactory, and the modeling performance and generalization extrapolation ability need to be improved

Images