Study On Dynamic Decision-making Model Of Reservoir Forecast And Operation Process And System Implementation

Focusing on the changes and uncertainties in the operation process,it is currently a hot topic to study the forecasting operation of reservoirs that adapt to dynamic changes.This paper is problem-oriented,focusing on the integration of forecasting,scheduling,and decision-making.The focus is on integrating evaluation decisions that adapt to dynamic changes in the scheduling process.According to the regulation cycle of the reservoir or the period of the scheduling plan,the scheduling process is distributed in sections in time,and the rolling update and nested advancement mechanism are used to realize the process dynamic scheduling.Taking reservoir as the key of regional water resource allocation decision,this paper analyzes the spatial distribution and business relationship of reservoir in the region,and realizes decomposition and aggregation on the platform;the scheduling scheme and real-time implementation are realized the implementation schemes feed each other to realize the organic combination of planning and scheduling.Using modern information technology,the operation mechanism is integrated into the operation process,so that the reservoir business can "move" to adapt to various changes.The main research contents are as follows:(1)A problem-based process decision-making model for reservoir forecasting and scheduling that adapts to changes is proposed.Based on the analysis of the problems existing in the actual operation of the reservoir,focusing on the uncertain factors and dynamic changes.Based on the entire process of the reservoir dispatching cycle,closely related to the forecasting and decision-making,relying on a series of mechanisms to establish innovalive decision-making model.Following the approximate approximate correctness,the intervalized adaptation mechanism is designed.The rolling correction mechanism for the forecast scheduling process is constructed.The multi-time scale scheduling scheme and implementation scheduling scheme are nested mutual feed-back mechanism.On an integrated platform,these mechanisms are integrated with forecasting,operation,and decision-making to form a reservoir forecasting and dispatching model that adapts to changes.(2)On the basis of a large number of reservoir prediction research results,combined with the scheduling period and process,according to the prediction period,the quality of the prediction results,the risk of the prediction results,etc.Based on the approximate correct theory,a combined comprehensive scheme of runoff forecast interval is proposed.Using component technology to technology to componentize the existing forecasting model,a forecasting model library is established.From the forecasting model method library,selecting multiple model methods to obtain multiple forecasting results and form forecast intervals,and formula expressions are tried for the forecast intervals.(3)Using component technology,a reservoir scheduling method library is established,combined with the forecasting model method library,a rolling correction mechanism based on the scheduling process is realized.In the reservoir operation cycle,the forecast dispatch is integrated,and the process is divided into stages.The dispatch plan is continuously revised and adjusted for the incoming water forecast in the face of the time period.The dispatch execution plan changes with the change of the incoming water forecast.Adapt to changes in time on the operational flow chart of forecasting and scheduling.Taking reservoir multi-obj ective scheduling as an example,according to the characteristics of the MOEA/D-AWA algorithm,the multi-obj ective scheduling of the reservoir based on the MOEA/D-AWA algorithm is implemented;the MOEA/D-PFR algorithm is used to solve the irregular Pareto front-shaped multi-obj ective Problem;Through the verification analysis of the two algorithms,it further illustrates the role of the method library in adaptive scheduling.(4)A nested mutual feed mechanism of the multi time scale adjustment plan scheme and the implementation scheme is proposed.In the scheduling process of multiple time scales,according to the mutual feedback relationship between the two sets of schemes and the rolling correction mechanism of forecast scheduling,the mutual feedback process of the planned scheme and the implementation scheme and the guiding scheduling operation mechanism were designed to improve the execution of the operation scheme under each time scale,and at the same time,the rolling nesting mutual feed integration of reservoir predictive operation in the operation process is realized(5)Based on the Comprehensive integrated platform and forecast operation method,the development,implementation and integrated application of the dynamic decision-making system for reservoir forecast operation are carried out for the rolling nested mutual feed mechanism of reservoir forecast operation and the integrated integrated implementation in the operation process.The Operation related to reservoir forecasting and scheduling begin with the description of process flow charts and knowledge graphs.At the same time,it depends on the mechanism in the flow chart to carry out the dynamic decision-making process.From the spatial scale decomposition of Beijing-Tianjin-Hebei water resources transfer,to the dynamic forecasting and scheduling of Miyun Reservoir.The feasibility,effectiveness,rationality and flexibility of reservoir dynamic and adaptive operation are verified by the integration of complex forecast operation and operation mechanism.