Research On Problems About Current Transformer Saturation Detection And Compensation

As a system fault occurs, network parameters of the power system change suddenly.As a result, the amplitude of the fault current up to a lot of times of the load current, andexponentially decaying DC offset appares in the fault current. In this condition, the magneticcore is promptly to saturation, which causes the secondary current waveform distorts. Theprotective CT saturation would cause protection relay malfunction or even tripping, while themeasuring CT would affects the operation of other secondary measurement equipment. Thispaper does comprehensive research on the problems aobut CT saturation, the main contentsare described as follows.1. Brief introduce about CT and its mathematical model. The combination of theory andsimulation analysis about the effects caused by some parameters on CT saturationcondition is performed. These parameters including primary current, exponentiallydecaying DC offset, load impedance, residual of the iron core, core area, ratio and turns ofthe CT winding and so on.2. For the first time, the paper proposes a new CT saturation detection method based onmathematical morphology (MM). By analyzing the secondary current waveform, whichhas abrupt changes where saturation begins and ends, the paper proposes a detectionmethod that combines morphological gradient wavelet (MGW) and morphologicalopening-closing and top-hat operators to accurately detect the onsets and ends ofsaturations. The performance of the proposed method has been evaluated in a variety ofCT saturation cases. The simulation studies have been indicated that the method is able todetect the saturation intervals with great accuracy, meanwhile inheriting the advantages ofanti-noise and resist harmonic. In addition, the method has high stability and wideeffective scope, which involving low computation load and short time delay.3. For the first time, the paper proposes a new method to identify CT saturation intervalsusing the improved morphological gradient and morphological decomposition. Themethod features the waveform properties of the distorted secondary current, which hasabrupt changes where saturation begins and ends. Two algorithms, the improved morphological gradient and the morphological decomposition, are designed to extract thesaturation intervals, respectively, and the detection results are combined using AND logicto give the final detection result. The simulation results show that the proposed methodcan accuratedly detect onsets and ends of saturation under a variety of operation condtions.Moreover, the method has the advantages of low computation load, short time delay, highprecision, stability and wide effective scope.4. The paper proposes s a new compensation method to reconstruct healthy secondarycurrent waveform based on linear regression. After detecting the onsets of the saturationbased on the above detection method, the proposed compensation method is triggered toreconstruct the healthy secondary current by estimating the parameters of fundamentalcomponent and exponentially decaying DC offset component based on the knownamplitude of each sampling point in the unsaturated intervals, respectively, until the end ofthe saturaton is detected. Comparing with the other compensation methods, the proposedmethod has the advantages of low computation burden and high compensation precision.Moreover, the method doesnot introduces truncation error and has a great potential to beimplemented in real time. The simulation results have been indicated that the proposedcompensation method is capable of reconstructing the healthy secondary current withgreat accuracy under different conditions.5. Developes a system on chip (SoC) whose core is FPGA, implements the saturationdetection method based on the morphological gradient and morphological decomposition.Firstly, developing a development board, built a hardware platform; Secondly, developinga software platform: design the FPGA circuit and build the μC/OS-II soft kernel; At last,writing the detection algorithm and creating the test platform. Complete the developmentof the SoC. Experimental results show that, in the SoC, the detection method is effective,and the hardware and software platform has a strong reliability.