Control Strategy Design Of Electrified Railway Energy Storage System Based On Dual Application

With the continuous development of electrified railways,while bringing many conveniences to people,it also brings a series of problems.The peak power of the traction load will not only cause power quality problems such as negative sequence and voltage fluctuations,but also increase the capacity requirements of the system on the traction transformer.In addition,if the braking energy generated by the locomotive in the braking condition is not recovered in time,it will not only cause a waste of resources,but will also cause an impact on the grid when injected into the power grid.Solving the problem of recovery and utilization of regenerative braking energy and excessive peak power also has an important impact on the economic benefits of railways.Focusing on these two issues,this paper introduces the dual application of super capacitor energy storage to the traction load in the electrified railway system to reduce the maximum demand of the traction load and increase the capacity utilization rate of the electrified railway traction transformer,so as to improve the economic benefits of the railway sector.the goal of.Based on a large number of related literatures,this paper conducts an in-depth study on the control strategy of energy storage devices,and proposes a dualapplication control strategy for electrified railway energy storage systems,which achieves load peak elimination and regenerative braking energy recovery.The main research contents and innovations of this article are as follows:In order to improve the accuracy of the control results,the proposed control strategy in this paper incorporates a load prediction link.Therefore,the traction load prediction technique is first investigated.To overcome the shortcomings of existing traction load prediction methods,this paper first analyzes the traction load data characteristics.To reduce the difficulty of prediction,the DWT model is first used to decompose the traction load data.The selection of the mother wavelet function and the number of decomposition layers in the DWT model are demonstrated from both principle analysis and experimental validation,and are selected on the basis of merit.Then,according to the frequency difference of different sequences,the TCN model is chosen to predict medium and low frequency sequences;the SVR model is chosen to predict high frequency sequences,and the key parameters in the model are optimized using PSO to make the SVR model have higher prediction accuracy.Finally,the effectiveness of the proposed prediction model is verified in the Keras framework of python.Based on the improved traction load prediction method,this paper analyzes the current energy storage control strategy under dual application and points out two problems that still exist in it.The first problem is the possible "dead time" of energy storage during system operation,and the second problem is the influence of prediction error on control results.To address problem 1,we use a real-time SOC correction method.In order to get a better correction effect,we introduce the idea of rolling optimization and predictive control in the correction process.This reduces the computation of the correction process and avoids the problem of new load peaks caused by the correction.For problem two,after improving the prediction model,we can also add a rolling prediction process to each rolling optimization process.By reducing the prediction time scale to reduce the error of the prediction results,thus reducing the impact of the prediction error on the control results.The experimental data show that the proposed control strategy can effectively solve the above two problems,ensure the effect of dual application and improve the economy of system operation.