Resonant Grounding Distribution Network Arc Fault Modeling And Fault Line Selection Based On Multi-disturbance Trend Identification

The distribution network is directly connected to users,and its power supply reliability and power supply safety have high requirements.The distribution network adopts a small current grounding method,which can continue to operate for a period of time when a single-phase ground fault occurs,ensuring the reliability of power supply.However,in this case,the fault current is small,the fault characteristics are weak,and it is susceptible to environmental factors such as noise,which causes the fault line to be misjudged,making the fault difficult to remove in time,may cause indirect phase failure or long-term single-phase grounding,leading to human and animal steps Accidents such as electric shock have caused serious hidden dangers to power supply safety.Especially in the resonance distribution network that is grounded by the arc suppression coil,due to the compensation effect of the arc suppression coil,the fault current is further reduced,and the phase of the zero sequence current of all lines is the same,which further increases the difficulty of line selection.Although there have been a lot of researches on single-phase ground fault line selection in resonant grounded distribution networks at home and abroad,the actual results are still not very satisfactory.The reason is that the complexity of distribution network faults is an important factor relative to transmission networks.The existing line selection methods have poor adaptability to complex faults,so there is an urgent need to study complex fault simulation methods and then study effective line selection methods to improve the reliability of fault line selection.Therefore,this paper mainly conducts research from increasing the complexity of the fault arc simulation model and constructing a more reliable fault line selection algorithm by strengthening the difference after multiple disturbances,mainly including the following:(1)According to the existing arc fault modeling methods,the arc is regarded as a stable law change process,which leads to the inconsistency of the simulation results and the actual situation.This paper proposes a random arc length distribution network arc fault model: according to the conditions of ignition Determine the initial arc length interval,and determine the arc length variation range according to the confidence interval;by fitting the arc length change rate,the phenomenon of free combustion of the fault arc in the air is described in detail,which is required for subsequent research on complex fault characteristics and line selection Effective digital simulation laid the foundation.(2)Considering the complexity of the fault type of the distribution network,a large amount of simulation data with different fault conditions is required to test the line selection algorithm,and a method for generating sample data with fault label simulation data is proposed,which is generated by automatically modifying the electromagnetic transient batch simulation.The fault data recorder file is named after the fault parameters to provide data support for future data-driven fault identification and line selection.(3)For the single-phase ground fault of the resonant grounding system,the fault and non-fault characteristics are not obvious,which leads to the problem of poor line selection accuracy.On the basis of the arc suppression coil and resistance line selection,according to different timing,different strength and Different directional parallel resistances are put into different disturbance directions,and two characteristic measures of interval comprehensive correlation coefficient and interval change trend are proposed to identify the zero-sequence current trend of the line,and then construct a criterion to achieve line selection.A large number of simulations show that the method can correctly identify the direction and size of the changing trend of the interval,and reliably select the faulty line.This method is simple to implement,is compatible with existing parallel resistance line selection methods,and is sequentially switched relative to the same direction.This method can produce differential disturbances in different directions of intensity,making the breakdown of fault and non-fault features more obvious;multiple times The disturbance provides a clear and clear interval for waveform trend comparison.The use of correlation,trend and other steady-state quantity features in the disturbance interval is more conducive to the identification of faulty lines with time-varying fault transition resistance.There is no need to set a threshold to improve the reliability of line selection Sex and effectiveness.The research work in this paper helps to improve the effectiveness of digital simulation of complex single-phase ground faults and the generation efficiency of massive random simulation data samples,and to provide data support for verifying the performance of line selection algorithms and researching data-driven line selection algorithms;Disturbance interval change trend detection overcomes the difficulty of selecting a fixed threshold for traditional line selection criteria and helps to enhance the applicability of complex line faults based on arc suppression coil and resistance line selection technology.