The Research On The Characteristics Of The Two-phase Mixture Discharges Under The DC Voltage

With the continuous development of the power network in china, the direct-voltage (DC) power transmission possesses such advantages as high transmission capacity, long working distance and saving the occupied line land and space. More and more DC transmission lines will go into the construction and operation in the future. But most of the built and constructing lines will go by the area with special and harsh circumstances and climate. The severe environment can greatly impact the insulation of the DC transmission lines. The research on the two-phase mixture discharge (TPMD) under the DC voltage is particularly important to improve the safe operation of the DC transmission network.This paper has investigated the discharge regulations on the two-phase mixture (TPM) under the DC voltage based on the discharge research under the lightning impulse voltage. Some kinds of TPMs are investigated by the discharge experiments and theoretical analysis under DC voltage. The breakdown voltage and the percentages of the discharge path in TPMs are compared with those in air. The results reveal the laws on the TPMD under DC voltage; A physical model on the selection of DC discharge path has been put forward and the neural network control model has been applied to simulate the optimal path for the discharge development.By the experimental results from the effect of the DC polarity, the macroparticle volume fraction, the dielectric constant and the diameters on the selection path, this work reveals that the breakdown voltage and the percentages of the discharge path in TPMs also exist the effects of macroparticle sizes:When D<0.01mm, under positive or negative DC voltage, the percentage of the selection of TPMs is always smaller than50%and the discharge path selects the air but TPM, the breakdown voltage is higher than that in air; When D>0.1mm, under positive or negative DC voltage, the percentage of the selection of TPMs is higher than50%and the discharge path selects the TPM but air, the breakdown voltage is lower than that in air; When0.01mm<D<0.1mm, under negative DC voltage the percentage the selection of TPMs is higher than50%and the discharge path selects the TPM but air, the breakdown voltage is lower than that in air; but under positive DC voltage the percentage the selection of TPMs is smaller than50%and the discharge path selects the air but TPM, the breakdown voltage is higher than that in air.The mechanism on the effects of macroparticle sizes has been done a systematic analysis. The research includes:1) the interactions between the macroparticles and the electrical field, the electrons and ions or the photoionization;2) the surface traps of the macroparticles capture the electrons and ions and the macroparticles occupy the discharge space and reduce the channel length which obstructs the initiation and propagation of the avalanche;3) these macroparticles in TPMs distort the electrostatic field, interact with ions, electrons or photons and are charged by diffusion, so produce corresponding enhancements or decreases in ionization and excitation as the avalanche front encounters them which affect the TPMD under DC voltage.On the basis of the streamer theory and probability and statistics theory, the paper uses the strength of an electric field from the Poisson’s equation as the criterion of the streamer development and supposes that the breakdown time of the streamer development meets Weibull distribution and the distorted field value decides the direction for the streamer development in the TPMD space. Based on the above study the physical model on the selection of positive discharge path has been set up. The development of the DC discharge path can be affected by the local electric field. The objective function (optimal path) corresponds to the energy function of the Hopfield neural network. The states of the neurons of this network will correspond to the sequence of nodes which is determined by the values of the local electric field. According to the stability theory of continuous Hopfield neural network, when the energy function tends to the minimum value and the states of the neurons also tend to an equilibrium point, the sequence of nodes is the optimal path for the discharge development. Comparing with measuring results of the experiment, this model could simulate well the selective probability of the discharge path under DC voltage.