Research On Application Of Intelligent Joint Control Power Generation Optimization Dispatching Of Cascade Hydropower Stations

South China is rich in hydro-power resources.The storage capacity of hydro-power stations s low in small and medium-sized basins is small,and the intelligent level of cascade hydro-power stations in basins.The cascade power stations fail to form a coordinated and constrained joint generation dispatching mode.When the short-term runoff and the joint control power generation dispatching are not coordinated,the water will be abandoned,and the contradiction between the storage capacity,water head and the optimal efficiency of the generator units will be occur,which will affect the comprehensive benefits of the cascade power stations in the whole basin.With the development of modern intelligent algorithms,the generation scheduling of cascade hydro-power stations is gradually studied in depth.This paper takes the intelligent joint control power generation optimization of cascade hydro-power stations in small and medium-sized basins as the direction,combines the relationship between rainfall,runoff,water level and the operating power generation efficiency of power units in each power station reservoir area in the basin.Research on model optimization methods such as modern genetic algorithm and particle swarm optimization algorithm.In the optimization process of hydro-logical runoff forecast model and power station dispatching model.The dispatching platform based on the control architecture of model and algorithm is designed and established to improve the comprehensive economic benefits of cascade hydro-power stations in the basin have been improved.The main research work is as follows:(1)Calculation and optimization of hydro-logical runoff prediction model in the basin.The intelligent joint control power generation of cascade hydro-power stations in the basin.Mostly it takes the distribution and transformation of water energy as the dispatching direction,the short-term rainfall variable as the main input parameter of runoff prediction model directly affects the accuracy of runoff prediction.Xin’anjiang River Model is established to predict runoff.The calculation of the model is optimized through genetic algorithm,which improves the accuracy of the model under the constraints of rainfall changes and other conditions during the calculation process.IT provides accurate prediction parameters for the intelligent joint control power generation of hydro-power stations in the subsequent basin.(2)Focusing on the multi-objective generation scheduling problem of cascade hydro-power stations in the basin.The particle swarm optimization algorithm is introduced into the study of joint control power generation.The joint control power generation model was based on the particle swarm optimization algorithm.The model calculation dimension constraint was added,which realizes the rapid calculation and rapid convergence of the model.Finally,the hydro-power load of cascade stations in the basin was scheduled and distributed through the model optimization calculation.Under the condition that the runoff forecast value is taken as the scheduling constraint.The research shows that the PSO algorithm is suitable for the optimization of the joint control power generation dispatching model of cascade hydro-power stations in the basin,which not only improves the reasonable distribution of hydro-power in the basin,but also improves the comprehensive power generation efficiency of the basin。(3)Based on the actual demand of intelligent joint control power generation of cascade hydro-power stations in the basin,the design scheme of the model system is determined.The system architecture and joint control power generation process are designed using the key technologies of database technology.Realizing real-time collection of runoff data,real-time transmission of load distribution,maximum reuse of data resources,and effective combination of algorithms and models with the scheduling platform software.Realize the downloading of dispatching and distribution load data and the actual data feedback.Realized the model modification to obtain the optimal solution.