| In the context of large-scale grid connection of wind power,the actual electricity demand presents characteristics of diversification,variability,and complexity,which puts forward higher requirements for the flexibility,optimization,and accuracy of the wind farm operation.On the premise of completing the power generation task of wind farms,improving the utilization rate of wind energy and the operational efficiency of wind farms as much as possible,extending the service life of wind turbines,and ensuring that wind turbines follow control instructions have become the focus of wind farm control theory research.Designing appropriate wind farm control strategies and completing theoretical verification are of great significance for solving practical needs.However,research on wind farm control systems cannot only stay in the theoretical design stage.Ultimately,solving practical problems requires practical implementation and even industrial application verification of control strategies.This thesis focuses on the theoretical design of control strategies for the implementation of wind farm control systems-software simulation verification of theoretical effectiveness-hardware implementation-hardware simulation testing.The main research content and achievements are:(1)This thesis designs the overall implementation architecture of double layer control system at the field level and the single wind turbine level for the wind farm and the general idea for the implementation and testing of the controller at the field level and the single wind turbine level for the wind farm.In terms of wind farm control strategies,unlike previous studies that only consider field level or single wind turbine control,this thesis adopts a holistic construction of wind farm field level optimization control and single wind turbine command control,achieving an organic connection between the control logic of the two.The single wind turbine command control system takes over the power command of field-level optimization for wind turbine control and then feeds back the performance of the wind turbine after being controlled by single wind turbine command to the scheduling system.Based on the actual wind speed,a new round of field-level power optimization is carried out.In terms of the implementation and testing of controllers at the field level and the single wind turbine level for the wind farm,with the development of hardware platforms,this thesis has chosen more advanced hardware for implementation.The main frequency of the ARM board is up to 2GHz,which greatly improves its computational efficiency compared to previous MHz and KHz products,and the hardware running speed is extremely fast.The controller hardware platform built based on it can serve as the hardware foundation for laboratory research.It can be used for full hardware simulation testing with simulators simulating wind turbines,providing a research platform for laboratory-level research and laying the foundation for industrial applications.(2)This thesis adopts a distributed model predictive control strategy for wind farm field-level optimization and designs a hardware implementation architecture of the field-level controller with a central unit and multiple parallel distributed units based on the control strategy.Subsequently,specific design,development,and verification of the field-level controller are carried out.In terms of field-level optimization control,this thesis introduces distributed model predictive control into wind farm field-level optimization control,and uses a fast parallel gradient algorithm for distributed parallel solution,simplifying the dimensions of field-level optimization problems and reducing computational difficulty.In terms of field-level controller design and development,there is a lack of research on the hardware implementation and specific control performance of field-level distributed control theory.On the theoretical basis,this thesis implements a hardware architecture of a central unit and parallel distributed unit for wind farm field-level controllers,reducing hardware computing difficulty,optimizing computing resource allocation,and improving hardware computing speed.After hardware simulation,the effectiveness,correctness,and speed of the field-level controller were verified.(3)This thesis adopts a variable speed constant frequency variable pitch control strategy for the control system of the single wind turbine in the entire operating area,and based on the concept of class in C++,the hardware implementation of the single wind turbine controller is specifically designed,developed,and verified.In terms of single wind turbine command control,a variable speed constant frequency variable pitch control strategy was adopted for the entire operating area,with different control objectives for different operating areas.For the control of pitch angle,active power,constant voltage control of grid side converter,and torque control of rotor side converter,comprehensive and detailed integrated implementation has been carried out.In terms of implementation of the single wind turbine controller,unlike in the past,a relatively simple linearized wind turbine model is adopted,which only implements hardware control for a certain part of the single wind turbine(only pitch angle control,or only rotor side converter control).The wind turbine model in this thesis is a high-order refined model based on mechanical electromagnetic joint modeling,which is more suitable for real scenarios.The program of the standalone controller is written in C++and implemented in class-based form.The programming language is simple,and its control logic is easy to modify,increasing the flexibility and diversity of the control system.At the same time,the high-frequency hardware platform enables the single controller to run faster.Finally,hardware simulation verified the effectiveness and correctness of the single controller. |