Font Size: a A A

Study On Unit Commitment Of Hydropower Plant Based On Multi-Core Parallel Computing

Posted on:2018-10-22Degree:MasterType:Thesis
Country:ChinaCandidate:S H ZhuFull Text:PDF
GTID:2322330536461362Subject:Water conservancy project
Abstract/Summary:PDF Full Text Request
With the construction of huge hydropower plant projects in the southwest of China in recent years,the plant is characterized by large capacity and large number of units.It is a great challenge to the calculation of the economic operation of hydropower plant.In fact,the unit commitment of hydropower plant is a high-dimensional,discrete,non-convex and nonlinear optimization problem.The calculation is confronted with many difficulties,such as low efficiency,poor calculating accuracy and complex decision support,especially for large scale of hydropower plant.In the paper,we take Wujiangdu station as the research background,and make a deep study of the unit commitment of hydropower plant from the aspects of parallel computing,inflow forecasting,decision support construction.The main contents of this paper include the following aspects:If we still use the conventional dynamic programming method to solve the problem of the unit commitment of the large scale of hydropower station,we will be faced with the "dimension disaster" dilemma and the computational timeliness test.Therefore,we can take advantage of multi-core parallel technology to avoid the shortcomings of traditional methods.Now,we propose a new method,which is dynamic programming method based on the Fork/join parallel framework,good results are obtained.The inflow of a reservoir is a significant input for the interior of the unit commitment calculation.It is essential to predict the reservoir inflow volume because of the accuracy of inflow forecasting directly determines the selection of water head and water consumption and the end water level during the scheduling period.In this thesis,time series analysis is adopted as an inflow prediction method.The correlation between historical inflow volume and time series is suggested.Results of this experiment demonstrated the inflow forecasting accuracy rate is satisfied.We designed the unit commitment of hydropower plant optimized control system on the J2 EE technology platform based on the Web technique and the B/S software design framework.Focusing on system requirements,system architecture,principles of development and system behavioral structure design.The system have the advantages of small size,good compatibility,user-friendly and easy maintenance and software update.Through the analysis of the research on these issue,I have a deep-going comprehension of the question about power-plant optimal operation.Additionally,I have made the summary elaboration for the whole text and prospects for the future.
Keywords/Search Tags:Unit Commitment of hydropower plant, Multi-Core Parallel Computing, Fork/join Framework, Time Series Analysis, Reservoir Inflow Prediction, Software Design
PDF Full Text Request
Related items