Local Defects Diagnosis For Cable Based On Broadband Impedance Spectroscopy

Cables play an important role in supplying power and transmitting control signals in modern industry. In these cables. it is known that permanent faults are always induced in the actual operation. If permanent faults of cables occur, power failure or loss of control would surge up, which leads to serious economic losses and social impact. Cable permanent failure often starts from local latent defects. These local defects of cable include: local aging of cable cause by high temperature, high humidity, partial discharge, even radiation; and local damage of cable caused by external mechanical forces. The existence of local latent defects in a cable is a great threat to the security and stability of large electrical system. Therefore, the diagnosis of portions in a cable that are severely degraded before breakdown occurs is of prime importance.The common used diagnosis method for cable can’t find out local latent defects. The partial discharge (PD) measurement is a proven way to diagnose cables with weak points. However, the detection of PDs within a recorded data set is not easy, and the location of PD sources is limited by the high frequency attenuation of PD pulses as they propagate through the cable. Therefore. there is an urgent need to develop new techniques for diagnosis of cable with potential points of failure.The broadband impedance spectroscopy (BIS) is a technique being developed in recent years which may become an effective method for diagnosing cable with local degradations. It utilizes cable impedance measurements over a broad range of frequencies to obtain information on the dielectric properties. With algorithms developed, properties are extracted from these impedance measurements to characterize the condition of the cable. The representative local defects diagnosis technologies for cable based on cable BIS are mainly LIRA method and IFFT method. However, the relevant research is still in the initial stage abroad. A complete theory about cable BIS has not yet formed, and there is no reliable data analysis method for cable BIS. In the domestic, the related technology is still in the blank.Aiming at the shortcomings of the existing research, this paper presents a novel method based on broadband impedance spectroscopy (BIS) to diagnose local defects in a cable. The characteristic of cable BIS is systematically studied. The impact of local defect on the cable BIS is fully discussed. New method is proposed to locate the local defects and assess their defect conditions, and the method is verified by simulation and experiment.Firstly, the characteristics of cable BIS is studied, and the relation of the cable BIS and cable electric properties is fully discussed. Then the cable BIS measurement technique is studied, and the method and instrument for cable BIS measurement are selected. The results show that high frequency cable BIS has three features, namely, periodicity, mutability and attenuating. The three features of BIS are mainly related to the cable propagation coefficient. The BIS varies fast around the maxima of the impedance magnitude spectroscopy and the zero crossing points of impedance phase spectroscopy. This is means these points are very sensitive to the changes of cable parameters. Therefore, the cable BIS can be used to obtain information of cable condition. To meet the requirements of the BIS measurement for all kinds of cables, the upper frequency limit of the measuring instrument should be more than5.5MHz.Secondly, the impact of local defects on the cable BIS is fully studied, and the sensitivity of cable BIS to local defects is discussed. Then a model to compute the BIS of cable with local defects is put forward, which clears up the relation of cable BIS and local defects. The results show that a local defect in cable will result in changes of local propagation coefficient. Then the transmission signal will be in catadioptric at the defect, which makes the cable BIS into a function of characteristic impedance, propagation coefficient and position of the local defect. After1%of the cable total length is mild aged, the change of impedance amplitude is more than40%at part of the frequency points. Meanwhile, the change impedance phase is more that500%at part of the frequency points. Therefore, the cable impedance spectroscopy is very sensitive to local defects, and it is possible to utilize the BIS to diagnosis local defects for cable.Next, this paper presents a novel method based on the integral transform algorithm to precisely locate local defects in a cable. Then a procedure based on the particle swarm optimization (PSO) is worked out to minimize the error between the measured impedance of the cable and that calculated by a model, and then the characteristic parameter for each of the defects is figured out to assess the insulation condition. The performance of the proposed technique has been evaluated quantitatively through computer simulations. The results show local defects can be clearly located with a spatial resolution as short as0.05m even if the cable length is3.5km. The performance of the locating method can be improved by selecting an appropriate frequency range of the BIS. The impact of load at the end on the locating results can be neglected. There is no information which can lead to misjudgment in the diagnosis graph of integral transform method. Thus, the integral transform method proposed in this paper is superior to the LIRA and IFFT method. After the PSO procedure is implemented, parameters can be extracted to indicate the severity of the degradation. The effect of every degraded portion on the locating results is independent. Therefore, the proposed method exhibits a good performance when dealing with cables carrying on multiple degraded portions.At last, the method to obtain the initial characteristic parameters of actual cable is studied, in order to accumulate basic data for the diagnosis of cable local defects. Then local aging diagnostic test, local damage diagnostic test and mixed defects diagnosis test were carried out on actual cables, which verified the validity of the presented diagnosis method.