Research And Application Of Hydropower-Photovoltaic Multi-Scale Joint Scheduling Considering The Decomposition Of Medium And Long-Term Contracted Electricity

Under the background of "dual carbon",new power systems have developed rapidly,and photovoltaic power generation has become one of the main energy sources.Water light Plementation is an important measure to solve the new energy consumption,and also an important part of the national multi energy complementation policy.The water light joint scheduling problem has always been a research hotspot.With the deepening of power market reform,the roles of hydropower and optoelectronics have changed,and the water light joint dispatching also faces new problems:① how to effectively decompose the monthly contracted electricity in the medium and long-term market;② How to build the prediction method adapting to changes;③How to participate in the monthly assessment of water light joint short-term daily scheduling.Water optical joint scheduling under the market-oriented environment is a complex system problem with multi-scale,multi object and multi problem intersection,which requires a set of systematic management mode for unified coordination.Based on this,this paper takes the idea of "mode key problem integrated application" to carry out the research on water optical multi-scale joint dispatching considering the decomposition of medium and long-term contracted electricity.The main work is as follows:(1)Aiming at the key issues such as the decomposition of medium and long term contract electricity,adaptive prediction considering multiple factors and multiple time scales,and water optical short-term scheduling considering monthly contract electricity deviation assessment,based on the complexity of water optical joint scheduling with multiple time and space scales,a water optical joint multi-scale scheduling model under the decomposition of medium and long term contract electricity is proposed,and a hierarchical constraint mechanism from large time scale to small time scale is established The feedback mechanism and rolling correction mechanism from small time scale to large time scale form a multi-scale prediction module,a multi-scale scheduling module and a power decomposition module,establish the cross integration relationship between the modules,and provide ideas for the water optical joint scheduling problem in the new situation.(2)Focusing on the dynamic nature of the actual prediction work and the need to improve the prediction accuracy,an adaptive prediction method based on decomposition integration mode in multiple time scales is established.The mechanism of prediction error is analyzed from the perspective of prediction process.In view of the demand for multi time scale prediction and the improvement of prediction accuracy,a decomposition and integration prediction model oriented to prediction process is proposed,and an extensible database of data processing methods and prediction models with learning ability is established;According to the dynamic characteristics of the actual prediction work,an adaptive rolling feedback prediction mechanism is designed to feed back and correct the output results according to the prediction error.The mechanism is used to adapt the prediction method to ensure the stability of the prediction results and reliability.On this basis,taking the medium and long term water inflow forecast and short-term photovoltaic power generation forecast of the reservoir as examples,the application of adaptive forecasting method based on decomposition integration mode is realized.The results show that the established forecasting method can effectively solve the problem of forecasting demand and improving forecasting accuracy under multiple time scales.(3)A multi-objective monthly contracted energy decomposition model for hydropower stations is established.Considering the daily inflow runoff forecast,the model takes the minimum monthly water consumption and the water level at the end of the month closest to the reservoir operation plan level as the objective,uses the adaptive variable scale immune clonal algorithm of the hybrid EAD to solve the problem,and uses the optimal compromise solution to obtain the decomposed daily energy.This model improves the disadvantages of the traditional average distribution method,and fully considers the regulation and storage capacity of the reservoir,Realize effective decomposition of contracted electricity from monthly scale to daily scale.At the same time,a decomposition model of contracted energy of photovoltaic power station considering forecast is established,and a correction strategy based on Gini coefficient theory is proposed for the deviation between actual daily energy generation and daily decomposed energy,which can meet the requirement of correcting the remaining contracted energy after the deviation occurs.Taking specific hydropower stations and photovoltaic power stations as application objects,the proposed model is verified,and the results show that the model can reasonably and effectively decompose the monthly contracted energy,and solve the problem of effective connection between medium and long-term and short-term of water light joint dispatching.(4)Aiming at the deviation assessment of monthly contracted electricity and the effective implementation of water light joint short-term scheduling on electricity,a hydropower station level water light joint short-term optimal scheduling model considering the constraint of daily decomposed electricity is constructed.With the minimum water consumption of the reservoir,the minimum deviation of daily contracted electricity and the maximum photovoltaic consumption as the goals,NSGA-Ⅲ algorithm is used to solve the problem.At the same time,the economic dispatch model of unit in the plant is established with the minimum generation cost.The application in actual power stations shows that the proposed model can quickly provide a reasonable output plan and improve the utilization rate of water resources and photovoltaic consumption rate while minimizing the daily deviation of electricity.(5)Taking the actual hydropower station and photovoltaic power station as the object,the application verification of the water optical multi-scale joint dispatching mode proposed in this paper is carried out.The calculation and analysis from different scenarios show that the dispatching mode proposed in this paper can effectively solve the key problems in the water optical joint dispatching and provide decision-making basis for the regulation.At the same time,in response to the multi-scale stepwise constraint and feedback and rolling correction mechanism proposed in this paper,a platform+component system rapid integration technology is proposed to adapt to the changing characteristics of the system,and a water optical joint scheduling application system is built based on components,which provides a practical technology for the docking of water optical joint scheduling mode with the actual,and realizes the application.