Research On Active Power Hierarchical Coordinated Control Strategy Of Wind Power Cluster Based On Model Predictive Control

With the implementation of the "carbon peak,carbon neutral" strategy,wind power and other renewable energy generation is constantly connected to the grid,and the construction of a new power system with new energy as the main body is accelerated,which promotes the clean and lowcarbon transformation of the energy and power industry.However,due to the differences in spatial distribution of wind power resources and electricity consumption load,insufficient local capacity,inter-regional transmission capacity,and the inherent randomness and volatility of wind power,there is inevitably wind power abandonment.In this dissertation,it is necessary to study the active hierarchical coordination control strategy of wind power clusters under power limitation to ensure the safe and stable operation of wind farms,minimize the impact of wind abandonment,and realize the coordinated and economic operation of each wind farm.In this regard,this dissertation takes the active power hierarchical coordination control strategy of wind power clusters as the research object,firstly,the control strategy of cluster layer and subfield layer is studied according to the general structure,and then the active power coordination control strategy between cluster layer and subfield layer is studied and explored based on the model predictive control theory,and the main completed research contents are as follows:1)At the cluster layer,to fully tap the advantage of the coordination active control of clustered wind farms under limited output scenarios,a fast coordinated distribution method for curtailed wind power among clustered wind farms with limited output was proposed.Firstly,based on the thinking of hiving the cluster off,the clustered wind farms were divided into two sorts,i.e.,power rising farms and power deduction farms,and the constraint condition of each sort of wind farms was simplified respectively,then according to its own characteristics a coordinative allocation model of the curtailed wind power of wind power cluster,in which the minimum quadratic sum of the adjustment amount of wind farm output was taken as objective,was constructed.Secondly,by use of strong duality theory a fast analytical solution method for the curtailed wind power in the mentioned model.Finally,taking a certain actual wind power cluster in Henan province for example,the effectiveness of the proposed model and method were verified.Through the analysis of arithmetic examples,it is verified that the proposed method of fast coordinated allocation of wind power abandoned by wind clusters can effectively reduce the size of the problem,improve the speed of model solution,and reduce the root-mean-square value of the resulting wind power output adjustment by about 46.69% compared with the traditional method,which is conducive to improving the level of refined operation control of wind farms.2)At the wind farm layer,a wind farm active power allocation method is proposed based on a Markovian state transfer process that considers the operating states of wind turbines.Firstly,the four operating states of normal,shallow-limited,deep-limited and out-of-service are classified according to the operating performance of the wind turbines,and the corresponding classification criteria are given;secondly,based on Markov theory,a wind turbine state transfer model and the corresponding state transfer matrix are constructed;finally,with the objective of minimizing the risk of deviating from scheduling instructions,this dissertation proposes an active power allocation method for a wind farm considering the operating states of wind turbines.The analysis shows that the proposed model and method can fully consider the operating status of the units and allocate as many units as possible to those with good output performance,reducing the risk of deviation from the dispatching instructions by more than 7.78% compared to the traditional method.3)In order to bring into play the coordination effect of wind power active control at the cluster level and the wind farm level,a hierarchical coordinated control method for wind power active based on model predictive control(MPC)theory is proposed.The method divides the wind power active dispatch into two levels: the cluster allocation level and the wind farm control level for coordinated control.In the cluster layer,the wind farms are dynamically clustered according to the relationship between the predicted active power output at the next moment and the actual active power output at the current moment,and the power output of each wind farm in the cluster is optimally allocated by combining the power prediction and the system active dispatch value,and sent to the wind farm layer.The units in the wind farm layer are optimized on a rolling basis according to the dispatch indicators issued by the cluster layer and the operational status of the turbines,so that units with good output performance generate more power.In addition,in the feedback module,the prediction module is error corrected according to the real-time active power feedback data of wind turbines to improve the accuracy of wind power active power prediction.The analysis of the algorithm shows that the proposed control strategy reduces the power adjustment of wind power clusters by more than 27.84%,while reducing the risk of wind farms deviating from the dispatching instructions,improving the output schedule of units with good output performance and promoting the efficient consumption of wind power.