Research On Fault Location Of Multi-terminal And Multi-section Hybrid Lines Based On Traveling Wave Metho

As an important component of power system equipments,the line fault will have a serious impact on power transmission.It is very important to locate the fault location and determine the fault phase quickly for restoring the power supply and eliminating the hidden trouble of the transmission line.The problem of fault location and fault phase selection for overhead-cable hybrid transmission line,especially for multi-terminal hybrid transmission line,has not been solved well.In view of this,this paper studies the fault location and fault phase selection of multi-terminal multi-segment overhead-cable hybrid transmission line.(1)The traveling wave transmission characteristics of overhead-cable hybrid transmission line are studied,which are different from those of single uniform transmission line.The existing signal decomposition methods have the problem of mode aliasing in the detection of fault traveling wave front,combining the Empirical Fourier Decomposition(EFD)algorithm which can improve the problem with the Teager Energy Operator(TEO)which reflects the instantaneous energy change,a fault traveling wave front method based on EFD-TEO is proposed,and its effectiveness and accuracy are verified.(2)Aiming at the problem of fault location for three-terminal multi-segment hybrid transmission line,a method for determining the fault branch is proposed by analyzing the time relationship of the initial fault traveling wave arriving at each measuring point and T-node.Aiming at the problem of fault location of multi-terminal and multi-segment hybrid transmission line,according to the characteristics of the multi-terminal hybrid transmission line,a fault branch separation iterative judgment method of the multi-terminal hybrid transmission line is proposed,two topological structure separation conditions are given,the multi-terminal hybrid transmission line structure is subjected to topological separation according to the time when the fault initial traveling wave reaches each end measuring point,and the separated transmission line are subjected to iterative calculation by using a corresponding method,so as to realize that accurate judgment of the fault branch of the multi-terminal hybrid transmission line.After the fault branch is judged,the time difference of the fault initial traveling wave reaching the endpoint of the branch and the T-node is further utilized,and the time difference is compared with the time difference of the fault traveling wave propagating from the connecting point of each overhead line and cable of the branch to the end measuring point and the T-node,so that the overhead line segment or the cable segment where the fault is located is determined.Finally,the fault distance is calculated by using the two-terminal traveling wave method.The electromagnetic transient simulation software PSCAD/EMTDC is used to build a three-terminal and multi-terminal multi-segment overhead-cable hybrid transmission line model,and the proposed fault branch determination and fault location method are verified.(3)Accurate identification of fault phase is also an important link in the process of line fault detection.Deep learning is applied to the problem of fault phase selection of transmission line,and a fault phase selection method of multi-terminal hybrid transmission line based on convolutional neural network is presented.Based on the multi-terminal hybrid transmission line fault branch judgment result,the fault current and voltage data of one cycle after the fault of the fault branch are preprocessed and then used as feature input.The convolutional neural network model is built using Tensor Flow deep learning framework,and the fault phase selection model is trained and tested.The accuracy of the fault phase selection method is verified by fault simulation data.