Font Size: a A A

Dual-platform Co-simulation Research On Large Offshore Wind Power Generation

Posted on:2022-11-15Degree:MasterType:Thesis
Country:ChinaCandidate:L DongFull Text:PDF
GTID:2480306770993719Subject:Electric Power Industry
Abstract/Summary:PDF Full Text Request
As an efficient and low-carbon new energy power generation method,offshore wind power has developed rapidly in recent years due to its abundant resources and no land resource restrictions.At the same time,as large-scale offshore wind power is connected to the grid,modern power systems not only become increasingly large in scale,but also exponentially in complexity.The interaction between large wind farms and the source network not only brings complex stability problems to the power system,but also brings great challenges to real-time simulation analysis.Traditional real-time simulation technology is difficult to take into account the large-scale AC and DC grid and the complexity of large offshore wind farms.In order to realize the accurate realtime simulation of these two parts,this thesis takes the Jiangsu Rudong offshore wind power project as the background,and conducts in-depth research on the construction of the simulation model and the application of the system segmentation algorithm.The main research contents of the thesis are as follows:First,the general flow of electromagnetic transient computing and the interface timing of parallel computing are analyzed in detail.The widely used interface algorithms are studied,including the transmission line decoupling method,the ideal transformer model method,the state space node method,and the inductance decoupling method.Aiming at the numerical oscillation problem generated by the inductive interface,a hybrid integration method is proposed to model the interface,and the oscillation phenomenon is suppressed by changing the proportion of the trapezoidal integration method before and after the transient process,thereby improving the traditional inductive decoupling method.Second,a co-simulation platform is built based on HYPERSIM and RT-LAB,and on this basis,the feasibility analysis and simulation verification of the dual-platform are carried out.In the communication aspect of the platform,the communication content between the simulation platform and the external controller is studied,According to the mutual data interaction content and communication format requirements,the simulation platform and the hardware interface are configured at the bottom layer to realize the communication between the simulation platform and the external controller.Third,in order to ensure that the RT-LAB side wind farm still has high simulation accuracy under limited resources,the wind turbine and wind farm models are compared and equivalently modeled.The dynamic link library model of the wind turbine is used to check the hardware-in-the-loop model connected with the actual controller to ensure the consistency of the two models at the simulation level.The single-machine equivalence of small-scale wind power clusters is carried out by the method of power multiplication and cluster equivalence.On this basis,the multi-machine equivalence modeling of large-scale wind farms is completed.For the modular multi-level converter station part on the HYPERSIM side,the CPU+FPGA architecture is used to realize its refined simulation.Fourth,in view of the limitations of the single interface segmentation model,a variety of interface algorithms are proposed to segment the model of large offshore wind power systems.Based on the research on different interface algorithms,this thesis selects the appropriate model split point,and uses the appropriate interface algorithm to decouple the system,The experimental results show that the divided simulation model can be solved within each simulation step,and the simulation accuracy can be guaranteed.Finally,on this basis,the dual-platform real-time simulation of large-scale offshore wind power generation is completed.
Keywords/Search Tags:offshore wind power, interface algorithm, HYPERSIM, RT-LAB, cosimulation, model segmentation
PDF Full Text Request
Related items