Research On Optimal Control Of Wind Storage System Based On Cloud Theory

As the world environment deteriorates,more and more countries are committed to energy transition.As a pollution-free renewable energy,wind power has become one of the important research objectives of energy transformation at present.However,due to its own uncertainties,wind power will affect the voltage and stability of the grid in the grid-connection process,and even cause serious safety accidents and large-scale power outages.In order to ensure the safe and stable operation of wind farms and power grids,wind power needs to be accurately predicted.However,the current technology cannot fully predict wind power.In order to ensure the actual grid-connected power can be produced according to the pre-planned output,a certain capacity of energy storage is usually configured.In addition,the current construction cost of energy storage is very expensive.So as to increase the service life of energy storage and the utilization rate of wind power,it is necessary to design a reasonable energy storage configuration and output control model.This thesis takes the historical wind power data of actual wind farm as the research basis,combines the improved interval prediction method to configure the energy storage capacity,and designs a controller to control the energy storage output.Specific research contents are as follows:（1）Wind power interval prediction based on hybrid semi-cloud model.Because of the uncertainty of wind power,the deterministic distribution function can’t accurately describe the distribution of error in interval prediction.In order to improve the fitting accuracy of the error data,a wind power interval prediction method based on hybrid semi-cloud model and nonparametric kernel density estimation was proposed.Firstly,the left and right half regions of the data are divided by the peak probability density of the prediction error.Secondly,the cloud model is used to describe the characteristics of the left and right side of the data respectively,and the distribution characteristics of the error are described by the parameters of the cloud,and a hybrid half cloud model conforming to the concept is established.On this basis,the semi-cloud model is used to generate enough conceptual cloud droplets,and the cloud droplets are used for nonparametric kernel density estimation.Then,with the minimum width of the prediction interval as the optimization objective,the prediction results of the interval under different confidence levels were obtained.Finally,the model is verified by simulation according to the actual historical data of wind farms in the United States.The fitting accuracy of different methods is compared by relative entropy,and the range prediction width（ud_i）,range coverage(M_PICP)and comprehensive evaluation F value are selected to evaluate the range prediction results under different methods.The simulation results show that the proposed method can effectively improve the accuracy of interval prediction,and the comprehensive evaluation of prediction interval is better than the traditional method,which lays a foundation for the subsequent research of energy storage configuration and output control.（2）Energy storage capacity configuration based on wind power prediction.Based on the interval prediction of wind power,this chapter tracks the fluctuation of wind power energy in the fluctuation interval and compensates the prediction error in the interval.When the prediction error exceeds the upper limit of the prediction interval,the upper limit of the interval is taken as the compensation output of the energy storage.Similarly,when the prediction error is lower than the lower limit of the interval,the lower limit of the interval is taken as the compensation output of the energy storage,and the storage capacity is configured according to the SOC requirements of the energy storage.Finally,the energy storage capacity configuration results under different confidence levels are obtained.（3）Energy storage output control considering the energy storage SOC based on cloud controller.Energy storage equipment plays a very important role in the process of wind power grid connection.In order to reduce the use cost of energy storage and increase the service life of energy storage,this paper designs a cloud controller to control the real-time output of energy storage to ensure that the SOC of energy storage is in the specified state.Firstly,according to the control requirements,the SOC at the previous time and the energy storage output at the current time（i.e.,the compensation power）were selected as the prior input of the controller,and the correction coefficient K₁ was selected as the posterior output of the controller.Then,the logic control rules of the cloud controller were established according to the working state requirements of the energy storage,that is,when the SOC of the energy storage was close to the critical value,the controller output K₁ to correct the energy storage output.Finally,based on the configuration of energy storage capacity,the actual historical data of a wind farm is selected to verify the control effect of the controller,and the control effect of the cloud controller designed in this paper is compared with that of the fuzzy controller by the proportion of the working dead time of the energy storage.The simulation results show that the cloud controller designed in this paper can make the proportion of energy storage dead time less than 0.35%,and the effect is better than the control effect of fuzzy controller.（4）Energy storage output control considering wind power grid connection requirements based on double-layer cloud controller and PSO.When the wind farm is connected to the grid,the volatility of wind power will cause safety problems in the power system.Therefore,pre-smoothing of wind power is often considered to reduce the impact of wind power uncertainty on the power system.Based on this,this paper proposes to use cloud controller to control the real-time output of energy storage.Firstly,the real-time fluctuation of wind power is selected as the input of the prior component of the controller,and the correction coefficientK₂ is selected as the output of the posterior component.Then,the logic control rules of the cloud controller were established according to the grid-connected fluctuation requirements of wind power,that is,when the fluctuation of wind power exceeded the specified requirements,the controller corrected the energy storage output to ensure that the fluctuation of wind power was within the specified range.In addition,in order to ensure the service life of energy storage,the SOC controller and wind power flattening controller are jointly applied to the output control of energy storage,and the PSO algorithm is used to solve the entropy value of the flattening controller,so as to ensure that the impact on the SOC controller is minimized under the premise of meeting the requirements of wind power grid connection.Finally,the actual historical data of wind farms in the United States were selected to verify the effect of the double-layer cloud controller.For different entropy values,the effectiveness of the entropy value selection was evaluated by the self-defined evaluation index and the wind power fluctuation over-limit proportion ratio,and the control effect before and after wind power leveling was displayed.The simulation results show that the cloud controller optimized by PSO algorithm can make the fluctuation of wind power meet the specified requirements under the premise of ensuring the working state of the energy storage.