Research On Realization Method Of Practical Optical Current Transducer

A novel current measurement scheme is presented to satisfy the requirement of the power system. The main contributions of this paper include data fusion arithmetic, chaos-principle-based temperature compensation method and calibration method of Optical Current Transducer (OCT). By means of theoretical studies and experiments, the Adaptive Optical Current Transducer (AOCT) was developed, calibrated and run into operation at an 110kV substation. The power system requires current transducer to have excellent performances as follows: (1) the accuracy satisfies metering purpose; (2) good frequency response and large dynamic range needed by protection, control and real-time analysis; (3) terrific stability supporting operation for a long term; (4) simple and reliable insulation. That is well known that the CT with core cannot satisfy these requirements. The frequency characteristic, accuracy and response abilities of Rogowski in both steady-state operations and faulty conditions are studied. The conclusion is achieved that Rogowski-type current transformer can not reach the accuracy of class 0.2 for metering application, only satisfy fundamental component protection purpose, but unable to suit for future dynamic applications. On the other hand, Faraday-type OCT has excellent dynamic performances. Up to now OCT still suffers from temperature affects and performance variations risk. Birefringence resulted from temperature change reduces the accuracy of OCT. Variations of performance with time define the stability of the system. Simplifying optical system can ensure good stability of OCT. The mechanisms of temperature effect are studied in this paper, indicating that the effects are complicated, nonlinear and can't be described exactly. In order to improve the accuracy of OCT, data fusion model is constructed to compensate temperature effects. Data fusion arithmetic for steady-state operation and memory relevance approach for fault condition are presented. A novel OCT, called AOCT is designed. The precision CT with optical fiber transmission provides accuracy fundamental component measurement data for the data fusion model, so that by means of the data fusion arithmetic the AOCT ensure all measurement results over the whole frequency ranges to be accurate. When faults occur, the data fusion is prohibited to avoid the influence of CT saturation, and the compensation parameters are kept according memory relevance approach. Because temperature changes so slowly than fault time that the compensation parameters before fault is still effective for measurement of faulty current. The AOCT combines high accuracy with excellent dynamic response ability. The accuracy satisfy requirements of class 0.2 of IEC60044-8 standard, and the maximum peak instantaneous error is no more than ±1%. Although the AOCT is consisted of two measurement units, the precision CT with optical readout (OECT) and the Faraday-type OCT, but less requirements of AOCT to each unit are needed than the situation of each unit working independently, which is of benefit to large-scale manufacture and application of the AOCT. Data fusion approach is presented to decrease effect of temperature to OCT, but signal-to-noise ratio (SNR) is not considered. In fact the measure accuracy is affected by interference and noise too. Noise effects have to be concerned at lower current levels. The conventional parameter estimation methods don't work well at lower SNR, so that data fusion approach can't work as expected. A novel chaos-principle-based temperature compensation method is presented to overcome the problem mentioned above. The new method is based on two key technologies put forward in this paper. One is the novel identifying method of chaos characteristic, which distinguishes the motion state of Duffing oscillator based on the envelope of the output X. It is real-time and easy to realize. The other is the novel amplitude estimation method of weak sine signal, called Duffing-ML method, which follows analysis of SNR improvement characteristic and frequency spectrum. The Duffing-ML method applies ML (maximum likelihood) estimation to the output X of Duffing system to achieve good noise immunity ability. The two key technologies are applied to data fusion arithmetic to form the chaos-principle-based temperature compensation method, which extends the measurement range of OCT, and ensure data fusion to be effective in sever noise condition. Standardizing the calibration of OCT is important for manufacture and application of OCT. According to IEC60044-8 standard, interfacing approaches are suggested; New test setups are constructed respectively for OCT with analogue and digital output. The virtual instrument technique is applied to meet the needs for calibrating small analogue signal and digital signal. Suggestions are provided for reference channel design in basis of error analysis theory. Performances of AOCT are examined by test system fulfilled in this paper. The AOCT developed in this paper had been operating in a substation in BaoDing city HeBei province for 25 months, shown stable performance. Chinese Society for Electrical Engineering (CSEE) upheld the technical evaluation conference of the AOCTin December, 2004, and the conclusion was drawn that the AOCT reaches international leading level.