面向风电机组最大功率点跟踪的转矩曲线增益动态优化

Torque Curve Gain Dynamic Optimization for Maximum Power Point Tracking of Wind Turbines

最优转矩法是实现风电机组最大功率点跟踪的常用方法。为了提高其在湍流风速下的风能捕获性能,转矩曲线的增益系数设定需要全面考虑湍流风速特征的影响。现有研究通过构建最优参数模型实现增益系数随时变湍流风速特征的动态优化。然而,由于转矩曲线的最优增益系数同时与平均风速、湍流强度及湍流频率三个特征指标存在复杂的非线性函数关系,构建最优参数模型需要基于大量的湍流风速样本,巨大的动态仿真计算量非常耗时,很难实现批量化工程应用。为此,该文发现最优转矩增益系数与最优转矩法对应的风能捕获效率存在强相关性,进而将基于三变量的最优参数模型降维简化为单变量,并在此基础上提出了转矩增益动态优化的最大功率点跟踪控制方法。该方法在同样全面考虑湍流风速特征、维持高风能捕获效率的同时,显著降低了构建最优参数模型的计算与时间成本。基于FAST软件的仿真验证了该方法的有效性。

The optimal torque(OT)method is a common method to realize the maximum power point tracking(MPPT)of wind turbines.By modifying the torque curve gain,the wind energy capture efficiency of wind turbine operating in MPPT mode under turbulent conditions can be improved.However,there is a precondition that the torque curve gain should be reasonably designed comprehensively considering the influence of turbulence characteristics.The existing researches construct the optimal parameter model describing the relationship between optimal torque curve gain and turbulence characteristics offline,then,based on the model,realize the dynamic optimization of torque curve gain according to the time-varying turbulence characteristics online.Nevertheless,because there is a complex three-dimensional nonlinear function relation between the optimal torque curve gain and turbulence characteristic indexes including average wind speed,turbulence intensity and turbulence frequency,it is necessary to traverse search the corresponding optimal torque gain based on a large number of turbulent wind speed series with different characteristics,so as to obtain sufficient samples for constructing the optimal parameter model.The corresponding huge amount of dynamic simulation calculation is very time-consuming and difficult to be applied in batch engineering.It was found in this paper that the wind energy capture efficiency corresponding to the OT method()can reflect the comprehensive effect of multiple factors affecting the MPPT of wind turbine.And the results based on Spearman rank correlation analysis show that the correlation coefficient between and the optimal torque curve gain is greater than 0.9,which is a very strong correlation.By comparison,the Spearman correlation coefficient between optimal torque curve gain and three turbulence characteristics is about 0.5,indicates that the degree of correlation is only moderate.Therefore,it was proposed to use as a single-valued characterization index to construct a quantitative relationship with the optimal torque gain,which can reduce the dimension of the optimal parameter model based on three variables to a single variable.On this basis,an MPPT control method was proposed to optimize the torque curve gain online according to and the optimal parameter model.Since the actual wind turbine adopts an improved MPPT control method with dynamic adjustment of torque curve gain,this paper constructed a digital twin of wind turbine using the OT method in programmable logic controller(PLC),and made it run synchronously with the actual wind turbine to realize the acquisition of P^(OT)_(favg).Simulation results based on FAST software developed by National Renewable Energy Laboratory show that,the proposed method reduces the time required to construct the optimal parameter model from 112.7 days to 2.3 days,not only retains the comprehensive consideration of turbulence characteristics so as to maintain high wind energy capture efficiency,but also significantly reduces the computing capacity demand and time cost,which is convenient for batch customization and rapid deployment.At the same time,based on the Beckhoff CX5130 PLC commonly used by the current batch wind turbine products,the engineering feasibility of synchronous operation of the wind turbine digital twin and actual wind turbine was tested and verified.It should be noted that,when the wind turbine model parameters used to construct the optimal parameter model and the wind turbine digital twin do not match the actual wind turbine parameters,the application effect of this method will be decreased.Therefore,how to calibrate the wind turbine model parameters based on operating data,and update the optimal parameter model and wind turbine digital twin in a timely manner remains to be further studied.