Method and device for selecting high-resistance grounding fault line in distribution network based on phase space trajectory recognition

Abstract

The present invention discloses a method and device for line selection of a high-resistance grounding fault in a distribution network based on phase space trajectory identification, the method comprising: obtaining transient zero-sequence current data of each line of a distribution system when a high-resistance grounding fault occurs in the system, constructing a corresponding phase space trajectory based on the phase space reconstruction principle; judging the line where the high-resistance grounding fault occurs according to the distance change trend from a vector point to an origin on the phase space trajectory. The present invention does not need to extract a certain harmonic frequency from the collected transient zero-sequence current signal, and utilizes a lower sampling frequency and phase space trajectory fitting to automatically filter out high-frequency harmonics and noise. The present invention does not need to set a line selection setting threshold, and the line selection result is not affected by the amplitude and phase angle errors of the transient zero-sequence current sequence collected from any line of the distribution system, and has stable line selection accuracy, thereby ensuring the reliability of the line selection criterion.

Description

Technical Field

[0001] The present invention belongs to the field of distribution network fault line selection, and specifically refers to a distribution network high-resistance grounding fault line selection method and device based on phase space trajectory recognition. Background Art

[0002] The topology of distribution lines is complex and their operation is closely related to the climate, people's work and life behaviors. Ground faults are prone to occur. Once a fault occurs at a certain location in the power grid, if it is not located and repaired immediately, it will affect the safety of people's personal property and cause significant economic losses. At the same time, with the rapid development of the national economy, considering that the distribution network is an important link closely related to the power supply of users, considering the influence of operation and maintenance, equipment reasons, external factors, user influence, natural factors and design and construction, it is necessary to put forward higher requirements for power quality and effectively improve the reliability and stability of the power grid. Therefore, accelerating the construction and application of distribution network automation is a key factor in improving the reliability of distribution network power supply. In the distribution automation system, fault line selection is the core content. Its main function is: when a ground fault occurs in the distribution network, the line is accurately selected in a short time, and the fault line is quickly isolated to restore power supply to the non-fault area, thereby minimizing the scope and time of power outages caused by the fault. It can be seen that fault line selection is the premise and basis for fault isolation, troubleshooting and power supply restoration, which is of great significance to improving power supply reliability.

[0003] The existing fault line selection methods have low resistance to transition resistance, insufficient line selection reliability, and low accuracy for quantitative analysis methods using set setting values. In addition, most line selection methods focus on low-resistance grounding faults, and insufficient research on line selection for high-resistance grounding faults. When a high-resistance grounding fault occurs, the characteristic electrical quantities are often relatively weak and are easily interfered by external noise. Fault detection and processing will also be more difficult, and line selection methods based on steady-state characteristics are not accurate enough. Compared with steady-state signals, the transient characteristics of high-resistance grounding faults are more obvious. The line selection method based on the transient main resonant component needs to completely strip out the transient component, but because the resonant frequency of the transient main resonant component is close to the power frequency component at this time, it is difficult to effectively distinguish and extract it, and the implementation of the method is relatively difficult; the artificial intelligence method for high-resistance grounding fault line selection requires a large number of data samples for analysis, which takes a long time, and the sampling criteria are not unique, and it does not have a complete physical interpretation, and the accuracy of the detection results is questionable. In addition, in actual projects, when power workers connect current transformers and cable shields to the power grid, it is inevitable that operational errors will occur, resulting in reverse or wrong connections. There are also situations where different types of current transformers are used in different lines or current transformers age and cause transmission errors. These factors will affect the reliability of existing line selection technology and cause significant interference to accurate fault line selection. Summary of the invention

[0004] The present invention provides a method and device for selecting a line for a high-resistance grounding fault in a distribution network based on phase space trajectory recognition. The method utilizes the fact that the characteristics of a transient process after a fault are far more obvious than steady-state quantities, converts collected data into an intuitive phase space trajectory, amplifies the transient characteristic change process of each line, and does not require setting a setting threshold, so the reliability and sensitivity of line selection are high.

[0005] In order to achieve the above technical objectives, the present invention adopts the following technical solutions:

[0006] A line selection method for high-resistance grounding fault in distribution network based on phase space trajectory identification obtains transient zero-sequence current data of each line in the distribution system when a high-resistance grounding fault occurs in the system, and constructs the corresponding phase space trajectory based on the phase space reconstruction principle; according to the distance change trend from the vector point to the origin on the phase space trajectory, the line where the high-resistance grounding fault occurs is judged.

[0007] Furthermore, if the distance from the vector point to the origin on the phase space trajectory shows a decreasing trend, the line is judged to be a high-resistance grounding fault line, otherwise the line does not have a high-resistance grounding fault.

[0008] Furthermore, the transient zero-sequence current data obtained refers to the zero-sequence current time series of each line in the first h cycles from the moment when a high-resistance grounding fault occurs in the system.

[0009] Furthermore, when it is monitored in real time that the effective value of the zero-sequence voltage of the power distribution system is greater than the voltage setting value, it is determined that a high-resistance grounding fault occurs in the system.

[0010] Furthermore, the method for constructing the phase space trajectory based on the phase space reconstruction principle is:

[0011] The acquired transient zero-sequence current data is expressed as the zero-sequence current time series i k ={i k1 ,i k2 ,…,i km ,…,i ks}, set the delay time τ of phase space reconstruction to 1 / 4 power frequency period; where k is the line number, the value range is [1, n], n is the number of lines in the distribution system; m represents the sampling point number, m∈[1, s], s represents the total number of sampling points;

[0012] The zero-sequence current i of each sampling point in the transient zero-sequence current data is km Map it into two-dimensional space for reconstruction, and obtain the two-dimensional space vector of each sampling point with the origin as the starting point The two-dimensional space vector ends of the zero-sequence currents of all sampling points are connected in sequence to form a phase space trajectory in the two-dimensional space; are the unit vectors of the horizontal and vertical axes respectively.

[0013] Furthermore, the method for determining the high-resistance grounding fault line according to the distance change trend is:

[0014] Extract the distance from each vector point to the origin on the phase space trajectory, that is, calculate the length of each vector in the two-dimensional space, and obtain the zero-sequence current time series i of each line k The corresponding distance time series d in two-dimensional space k ={d k1 ,d k2 ,…,d km ,…,d ks};

[0015] Calculate the distance time series d k The ratio coefficient between two distances separated by one power frequency cycle m∈[1,s-4τ];

[0016] It is determined whether there is a value less than 1 in the ratio coefficient sequence of each line. If so, the line is determined to be a high-resistance grounding fault line.

[0017] A distribution network high-resistance grounding fault line selection device based on phase space trajectory recognition, comprising a phase space trajectory construction module and a fault line selection module;

[0018] The phase space trajectory construction module is used to: obtain transient zero-sequence current data of each line of the power distribution system when a high-resistance grounding fault occurs in the system, and construct a corresponding phase space trajectory based on the phase space reconstruction principle;

[0019] The fault line selection module is used to determine the line where the high-resistance grounding fault occurs according to the change trend of the distance from the vector point to the origin on the phase space trajectory.

[0020] Furthermore, the line selection method of the fault line selection module is: if the distance from the vector point to the origin on the phase space trajectory shows a decreasing trend, the line is judged to be a high-resistance grounding fault line, otherwise no high-resistance grounding fault occurs in the line.

[0021] Furthermore, the method for the phase space trajectory construction module to construct the phase space trajectory based on the phase space reconstruction principle is:

[0022] The acquired transient zero-sequence current data is expressed as the zero-sequence current time series i k ={i k1 ,i k2 ,…,i km ,…,i ks}, set the delay time τ of phase space reconstruction to 1 / 4 power frequency period; where k is the line number, the value range is [1, n], n is the number of lines in the distribution system; m represents the sampling point number, m∈[1, s], s represents the total number of sampling points;

[0023] The zero-sequence current i of each sampling point in the transient zero-sequence current data is km Map it into two-dimensional space for reconstruction, and obtain the two-dimensional space vector of each sampling point with the origin as the starting point The two-dimensional space vector ends of the zero-sequence currents at all sampling points sequentially form phase space trajectories in the two-dimensional space; are the unit vectors of the horizontal and vertical axes respectively.

[0024] Furthermore, the method of the fault line selection module for determining the high-resistance grounding fault line according to the distance change trend is:

[0025] Extract the distance from each vector point to the origin on the phase space trajectory, that is, calculate the length of each two-dimensional space vector, and obtain the zero-sequence current time series i of each line k The corresponding distance time series d in two-dimensional space k ={d k1 ,d k2 ,…,d km ,…,d ks};

[0026] Calculate the distance time series d k The ratio coefficient between two distances separated by one power frequency cycle m∈[1,s-4τ];

[0027] It is determined whether there is a value less than 1 in the ratio coefficient sequence of each line. If so, the line is determined to be a high-resistance grounding fault line.

[0028] The beneficial effects of the present invention are:

[0029] (1) By taking advantage of the fact that obvious transient characteristics are generated within a short period of time after a fault occurs, the transient zero-sequence current signal after the fault occurs is analyzed, which effectively reduces the influence of weak steady-state characteristics of faults and difficulty in detection used in traditional fault line selection methods on line selection accuracy and improves the reliability of fault line selection;

[0030] (2) There is no need to extract a certain harmonic frequency from the transient zero-sequence current signal collected by the device, thus avoiding the error generated in this process. By using a lower sampling frequency and phase space trajectory fitting, high-frequency harmonics and noise are automatically filtered out, thereby improving the anti-interference ability of the high-resistance grounding fault line selection method of the present invention and ensuring the reliability of the line selection criterion;

[0031] (3) Based on the phase space trajectory change characteristic analysis of the fault and normal line fitting, the fault line can be clearly distinguished, and there is no need to set the line selection setting threshold. It has stable line selection accuracy, high line selection reliability and sensitivity;

[0032] (4) Each line is judged independently, and there is no need for group comparison. The line selection result is not affected by the current transformer model, wiring condition, or degree of damage, that is, it is not affected by the amplitude and phase angle errors of the transient zero-sequence current sequence collected from any line in the distribution system, thus maintaining good line selection accuracy and reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is the zero-sequence current distribution when the neutral point of the distribution system is grounded through the arc suppression coil grounding system;

[0034] Figure 2 It is the second-order transient analysis equivalent circuit when the neutral point of the distribution system is grounded through the arc suppression coil grounding system;

[0035] Figure 3 It is the phase space trajectory of the non-fault line within 5 cycles after a high-resistance ground fault occurs in the distribution system;

[0036] Figure 4 It is the phase space trajectory of the fault line within 5 cycles after a high-resistance ground fault occurs in the distribution system;

[0037] Figure 5 It is a simulation model of 35kV arc suppression coil distribution system;

[0038] Figure 6 It is the dynamic waveform of zero-sequence current of each line of the distribution system in Example 1;

[0039] Figure 7 This is the trend of the distance ratio coefficient of each line in the distribution system of Example 1;

[0040] Figure 8 It is the dynamic waveform of zero-sequence current of each line of the distribution system in Example 2;

[0041] Fig. 9 This is the changing trend of the distance ratio coefficient of each line in the distribution system of Example 2. DETAILED DESCRIPTION

[0042] The following is a detailed description of an embodiment of the present invention. This embodiment is based on the technical solution of the present invention, and provides a detailed implementation method and a specific operation process to further explain the technical solution of the present invention.

[0043] The present invention provides a method for selecting a line for a high-resistance grounding fault in a distribution network based on phase space trajectory identification, which obtains transient zero-sequence current data of each line in the distribution system when a high-resistance grounding fault occurs in the system, and constructs a corresponding phase space trajectory based on the phase space reconstruction principle; according to the distance change trend from a vector point to an origin on the phase space trajectory, the line where the high-resistance grounding fault occurs is judged.

[0044] In a specific embodiment, the following steps are included:

[0045] 1) Real-time monitoring of the voltage changes at each node of the power distribution system. Once the effective value of the zero-sequence voltage in the system is greater than the voltage setting value, it indicates that a high-resistance grounding fault has been detected in the system.

[0046] 2) The real-time fault information collected by the field terminal equipment installed along the feeder is used to extract the zero-sequence current time series i of the first five cycles of each line in the entire distribution system after the high-resistance grounding fault. k ={i k1 ,i k2 ,…,i km ,…,i ks}, where k is the line number, ranging from [1, n], n is the number of lines in the distribution system; m represents the sampling point number, m∈[1, s], s represents the total number of sampling points.

[0047] 3) Based on the principle of phase space reconstruction, the delay time τ of phase space reconstruction is set to 1 / 4 of the power frequency cycle (5ms); then the zero-sequence current i of each sampling point in the transient zero-sequence current data is km Map it into two-dimensional space for reconstruction, and obtain the two-dimensional space vector of each sampling point with the origin as the starting point The two-dimensional space vector ends of the zero-sequence currents of all sampling points are connected in sequence to form a phase space trajectory in the two-dimensional space; are the unit vectors of the horizontal and vertical axes respectively.

[0048] 4) Extract the distance from each vector point to the origin on each phase space trajectory, that is, the length of each vector in two-dimensional space, expression d km 2 =x km 2 +y km 2 , d km is the distance from the vector point corresponding to the sampling point m on the phase space trajectory of line k to the origin of the two-dimensional space; km ,y km are the horizontal and vertical coordinates of the vector point corresponding to the sampling point m on the line k phase space trajectory, Thus, the zero-sequence current time series i of each line is obtained k The corresponding distance time series d in two-dimensional space k ={d k1 ,d k2 ,…,d km ,…,d ks}.

[0049] 5) Calculate the distance time series d kThe ratio coefficient between two distances separated by one power frequency cycle m∈[1,s-4τ]. Thus, we can get the sequence of all ratio coefficients, namely the distance ratio coefficient sequence μ k = {u k1 ,u k2 ,…,u km ,…,u ks}.

[0050] 6) Determine the distance ratio coefficient sequence μ of each line k Is there a distance ratio coefficient less than 1 in the circuit? If so, the line is judged to be a high-resistance grounding fault line, otherwise it is a normal line.

[0051] The principle of fault line selection according to the distance variation trend of the present invention is analyzed as follows:

[0052] 1. Analysis of transient characteristics of resonant grounding system faults

[0053] The present invention is aimed at line selection for high-resistance grounding faults in distribution systems. Since the transition resistance of the fault point of high-resistance grounding is usually hundreds to thousands of ohms, the line reactance of the distribution network is very small in comparison and can be ignored in the equivalent circuit. When a neutral point grounding system is grounded through an arc suppression coil, the zero-sequence current distribution is as follows: Figure 1 As shown. There are n lines in total, and the fault is assumed to occur on the nth line. f is the zero-sequence voltage at the fault point, u 0f is the neutral point zero-sequence voltage; the zero-sequence equivalent resistance R on the fault line in the zero-sequence equivalent network is approximately 3 times the fault point transition resistance R f ; C k (k=1,2,…,n) represents the capacitance of each line to ground, and the total capacitance of the system to ground is i k is the zero-sequence current of the kth line. Its value on the normal line k is equal to the capacitance current of the line to ground. quite; is the current flowing through the arc extinguishing coil L P The inductor current.

[0054] Figure 2 is the second-order transient analysis equivalent circuit, where U f is the zero-sequence voltage source, R is approximately expressed as 3 times the transition resistance, L P It is expressed as the equivalent inductance of the arc suppression coil, and C is the total capacitance of the system to ground.

[0055] according to Figure 2 The differential equation for establishing the equivalent circuit is:

[0056]

[0057] The characteristic root can be obtained by formula (1):

[0058]

[0059] In a resonant grounding system, an overdamped state may occur only when the capacitive current to ground is small or the fault occurs near the busbar. Therefore, in most cases, a high-resistance grounding fault in the system appears as an underdamped state. Figure 2 The middle resistor R satisfies: The transient current of line fault is a decaying sinusoidal component. Taking the underdamped state of system fault as an example, the attenuation factor δ and the resonant frequency ω f They are:

[0060]

[0061]

[0062] Where ω is the power frequency of the system, v represents the compensation detuning degree of the arc suppression coil of the distribution network. Since the compensation degree of the arc suppression coil generally does not exceed 10%,

[0063] Through arc suppression coil L P The zero sequence current is:

[0064]

[0065] in, φ represents the initial phase angle of the fault.

[0066] It can be seen from formula (3) that the attenuation factor δ is inversely proportional to the distributed capacitance C and resistance R to the ground. For a certain system, the larger the transition resistance, the slower the decay rate of the transient component of the fault current, and the transient process in the underdamped state lasts longer than that of the low-resistance grounding fault. Therefore, in the case of a high-resistance grounding fault, although the fault electrical characteristic quantity is not obvious, the fault line can be selected through its transient process.

[0067] From formula (4), we can see that for a certain system, the larger the transition resistance, the higher the resonant frequency ω f The higher the maximum frequency, the greater the power frequency. Because the transition resistance of the distribution network system is generally several thousand ohms in the case of a high-resistance grounding fault, the attenuation factor δ is small, and the arc suppression coil of the resonant grounding system compensates the system's ground capacitance current as much as possible, and the detuning degree generally does not exceed ±10%. In the case of underdamping, the resonant frequency of most faults is equivalent to the power frequency, that is, ω f ≈ω, at this time, the free oscillation component contained in the transient component is no longer concentrated in the high frequency band but is close to the power frequency, which will have a significant impact on the amplitude of the fundamental frequency component.

[0068] 2. Distribution characteristics of zero-sequence current phase space trajectory under high-resistance grounding fault

[0069] As a typical power system, the power system always generates various time-varying state quantity sequences. The phase space is represented as a space corresponding to a series of d-dimensional vector coordinates constructed by different delay times τ of the state variable time series in the power system. Each possible state of the system corresponds to a point in the d-dimensional phase space. The points corresponding to the states of the system at different times are connected in time order in the phase space to form a curve, which is the trajectory of the system operation process in the phase space.

[0070] The time series of state quantities generated by the power system can also be mapped into the corresponding phase space according to the phase space principle. However, due to the large number of state variables in the power system, the dimension of its phase space is generally very high. If the state evolution information of all characteristic quantities is collected in order to obtain all the information contained in a power system, the engineering workload is huge and the practicality is low. In actual engineering applications, usually only the time-varying sequence of one or some state quantities can be measured, such as the current and voltage sampling values ​​of certain nodes. The phase space reconstruction method believes that the evolution of any component in the system is determined by other components that interact with it. Therefore, the information of these related components is implicit in the development process of any component. The use of phase space reconstruction will have a significant amplification effect on the characteristics of the original components. The embedding theorem points out that a d-dimensional dynamic system can be mapped to a d-dimensional system that is topologically the same as the original dynamic system. E The analysis and prediction of d-dimensional systems can be carried out in this reconstructed phase space, where d E ≥2d+1, that is, all the information of the original system can be extracted from the time series of any component and an equivalent state space can be reconstructed, which theoretically proves the feasibility of the phase space reconstruction method.

[0071] The embedding dimension of phase space and the delay time τ are two key factors for the effective embedding of state variables, and appropriate values ​​must be selected according to the actual situation. It can be seen from the embedding theorem that the higher the embedding dimension is, the more comprehensive the system information it contains, but the corresponding delay time will become longer, and the amount of calculation will increase exponentially. Because the one-dimensional transient zero-sequence current time series of each line is used for phase space reconstruction, the transient circuit has a short existence time, and it is required to use these transient characteristic quantities to quickly complete the selection of the fault line in a short time, so its speed requirements must be considered during design. After comprehensive consideration, the present invention chooses to embed the transient zero-sequence current time series of each line into the 2-dimensional phase space to amplify the transient characteristic change process of each line after the fault, thereby realizing the accurate selection of the fault line and having a better line selection effect, which will also be verified in the following simulation.

[0072] Since in the case of a high-resistance grounding fault, the resonant frequency is basically equal to the power frequency as analyzed in the first point above, the optimal delay time of the standard sinusoidal signal is calculated based on the mutual information method in chaos theory, and therefore the optimal delay time τ of the present invention is 1 / 4 power frequency cycle (5ms), and the transient zero-sequence current of each line is reconstructed in 2D phase space. And according to the actual data collection situation of the current power grid substation, the sampling frequency of the transient zero-sequence current is selected as 4kHz, that is, 80 sampling points per cycle, and it can be determined that a delay of 5ms corresponds to 20 sampling points. Since the resonant frequency of the transient current in the case of a high-resistance grounding fault is close to the power frequency, rather than concentrated in the high frequency band, the sparse sampling points can also well display the transient process of the free oscillation component; moreover, the lower sampling frequency can better filter the noise, so that the constructed phase space trajectory is less affected by the system noise, and the anti-noise ability of the method and the reliability of line selection are improved.

[0073] The method of constructing a two-dimensional space vector from transient zero-sequence current is shown in equation (6).

[0074]

[0075] in, are the unit vectors of the horizontal and vertical axes respectively; m is the sampling time point number; τ represents the delay time; k represents a line in the distribution network, and its value range is [1, n].

[0076] (1) Phase space trajectory distribution characteristics of non-fault lines

[0077] In the case of a high-resistance grounding fault, since the non-fault line k only contains power frequency components and high-frequency transient ground capacitance circuit components, the transient zero-sequence current of the non-fault line k can be expressed as:

[0078]

[0079] Because when the transition resistance is large, ω f ≈ω, equation (7) can be simplified as:

[0080]

[0081] Among them, i k represents the transient zero-sequence current of the non-fault line k; C k is the ground capacitance of the non-fault line k, and the value range of k is [1, n-1]; D 1 =-L P C k ω 2 B; M 1 =L P C k (δ 2 A 1-A 1 ω 2 -2δA 2 ω); N 1 =L P C k (δ 2 A 2 -A 2 ω 2 +2δA 1 ω).

[0082] According to equation (6), the phase space trajectory is constructed, and the constant term in equation (8) is a k (t), the amplitude of the power frequency sinusoidal component is b k , we can construct the phase space trajectory, whose horizontal and vertical coordinates are equations (9) and (10) respectively, and the phase space trajectory is shown in equation (11).

[0083]

[0084]

[0085] (x k (t)) 2 +(ya k (t+τ)) 2 =b k 2 (11)

[0086] Among them, a k (t) = M 1 e -δt cos(ωt)+N 1 e -δt sin(ωt), expressed as the horizontal coordinate of the center of the phase space trajectory; a k (t+τ)=M 1 e -δ(t+τ) cos(ωt+π / 2)+N 1 e -δ(t+τ) sin(ωt+π / 2) is the ordinate of the center of the phase space trajectory. Both are periodic oscillation components that increase with time, and whose amplitude gradually decreases and tends to 0; b k =D 1 =-L P C k ω 2 B, whose absolute value is expressed as the radius of the phase space trajectory; is the system power phase voltage, C k is the capacitance of line k to ground.

[0087] From equations (9) to (11), it can be seen that when a high-resistance ground fault occurs in the distribution network, the two-dimensional phase space trajectory formed by the transient zero-sequence current of each non-fault line is the center of the circle (a k (t), a k (t+τ)) changes with time and has a radius of b k The phase space trajectory of the non-fault line within 5 cycles after the distribution network fault occurs can be drawn, such as Figure 3 As shown, this is the phase space trajectory formed by the transient zero-sequence current corresponding to the non-fault line 1. At this time, the phase space trajectory presents a spiral line developing from the inside to the outside.

[0088] (2) Phase space trajectory distribution characteristics of fault line

[0089] Due to the existence of inductive current generated by the arc suppression coil in the fault line n, its transient zero-sequence current is expressed as:

[0090]

[0091] Among them, i n Represents the transient zero-sequence current of fault line n; C n is the capacitance of fault line n to ground, D 2 =L P ω 2 (CC n )-1,M 2 =-[A 1 +L P (CC n )(δ 2 A 1 -A 1 ω 2 -2δA 2 ω)],N 2 =-[A 2 +L P (CC n )(δ 2 A 2 -A 2 ω 2 +2δA 1 ω)],N 2 、M 2 All are negative values.

[0092] Similarly to the analysis process of the non-fault line, let the constant term in equation (12) be a(t), the amplitude of the power frequency sinusoidal component be b, and construct its phase space trajectory and its horizontal and vertical coordinates as shown in equations (13), (14), and (15), respectively.

[0093] x=a(t)+b sin(ωt+φ) (13)

[0094] y=a(t+τ)+b sin(ωt+φ+π / 2) (14)

[0095] (xa(t)) 2 +(ya(t+τ)) 2 =b 2 (15)

[0096] Where a(t) = M 2 e -δt cos(ωt)+N 2 e -δt sin(ωt), expressed as the horizontal coordinate of the center of the phase space trajectory; a(t+τ)=M 2 e -δ(t+τ) cos(ωt+π / 2)+N 2 e -δ(t+τ) sin(ωt+π / 2) is expressed as the ordinate of the center of the phase space trajectory. Similar to the center coordinate analysis of the normal line, both are periodic oscillation components that increase with time, gradually decrease in amplitude and tend to 0; b(t) = D 2 =L P ω 2 (CC n )-1, and its absolute value is expressed as the phase space trajectory radius.

[0097] From equations (13) to (15), it can be seen that when a high-resistance grounding fault occurs in the distribution network, the 2D phase space trajectory formed by the transient zero-sequence current of the fault line is a spiral curve with a center (a(t), a(t+τ)) that changes with time and a radius of b. The phase space trajectory of the fault line within 5 cycles after the distribution network fault occurs can be drawn, as shown in Figure 4 As shown, this is the phase space trajectory formed by the transient zero-sequence current corresponding to the fault line 4. At this time, the phase space trajectory presents a spiral line whose radius decreases continuously over time and tends to a stable value.

[0098] 3. Implementation criteria for the phase space trajectory identification and line selection protection method of distribution network ground fault

[0099] Due to the weak fault electrical characteristics in the case of high-resistance grounding fault, the characteristic amplitude and phase of the fault line and the normal line in the steady state are not much different, and it is impossible to distinguish intuitively. From the above phase space plane diagram, it can be seen that the trajectory of each line is basically a spiral line rotating around the origin. With the change of time, the radius of the phase space trajectory curve of the non-fault line gradually increases and eventually reaches a stable maximum value, while the radius of the phase space trajectory curve of the fault line shows a spiral decrease trend and eventually reaches a stable minimum value. Therefore, using this feature, the distance from each vector point on each phase space trajectory to the origin can be extracted, and the distance from each point on the phase space trajectory of the non-fault line and the fault line to the origin over time can be calculated respectively as follows:

[0100] d k (t) 2 =x k (t) 2 +y k (t) 2 (16)

[0101] Among them, d k (t) is the distance from each point on the phase space trajectory of line k to the origin of the 2D space, k∈[1,n]; x k (t), y k (t) are the horizontal and vertical coordinate values ​​of each vector point on the phase space trajectory of each line.

[0102] Because each vector point in the time series of the distance from each line to the origin at an interval of one power frequency cycle represents the distance state at the same time in the corresponding different power frequency cycles, taking one power frequency cycle as the time scale, if the distance from the vector point to the origin at each moment in the next cycle is respectively greater than the distance value of the vector point at the same moment in the previous cycle, it can be explained that the distance from the origin of the line is increasing, otherwise it is decreasing. In order to judge the changing trend of the distance expression of each line from the origin with time, in the preferred embodiment of the present invention, each vector point in the time series of the distance from the origin is sampled at an interval of one power frequency cycle, and the ratio coefficient of the distance of each line to time is calculated:

[0103]

[0104] It can be seen from formula (17) that the capacitive free oscillation component current generated in the transient process of the non-fault line gradually attenuates over time to offset the fundamental frequency component, and the transient zero-sequence current of the line gradually increases. Therefore, the vector points on the next power frequency cycle are always greater than the vector points at the same time of the previous power frequency cycle, and their distance ratio coefficient is always greater than 1; while the zero-sequence current of the fault feeder shows a trend of attenuation in the few cycles after the fault occurs due to the inductive transient zero-sequence current generated by the arc suppression coil. There is a situation where the vector points on the next power frequency cycle are smaller than the vector points at the same time of the previous cycle. Therefore, the fault line selection criterion for the distribution network in the case of high-resistance grounding fault can be set as:

[0105] μ k ≤1 (18)

[0106] If equation (18) is satisfied, line k can be judged as a faulty line, otherwise it can be judged as a normal line, thereby achieving accurate fault line selection.

[0107] 4. Reliability analysis of a new criterion for fault line selection in distribution networks using phase space trajectory identification

[0108] In actual projects, due to the complexity of the power distribution system, power workers are likely to make operational errors during the wiring process, reversely connecting current transformers and cable shields or connecting them to the grid incorrectly; and different lines use different types of current transformers, and current transformers age and have transmission errors. This will affect the accuracy of current data collection on each line and cause significant interference to the accurate fault line selection. Overall, the appearance of these errors can be uniformly regarded as a certain proportional error in the amplitude and phase of the zero-sequence current during the acquisition, that is, there is a vector difference between the transient zero-sequence current time series collected after a line fault and the actual time series. At this time, the origin distance d of the phase space trajectory of the normal line and the fault line when the data error occurs can be obtained k The expression of ′ over time is:

[0109]

[0110] From formula (19), we can see that compared with the actual origin distance of the distribution network, the origin distance d with error k 'Data is the actual data times. Draw the ratio coefficient μ′ of the distance between the origin of the phase space trajectory of each line:

[0111]

[0112] The result obtained by formula (20) is the same as that of formula (17), that is, μ k ′=μ k The existence of such errors does not affect the line selection result of criterion (18), and criterion (18) can still be used for fault line selection, and the accuracy and reliability of fault line selection are still guaranteed.

[0113] In summary, the criterion adopted by the present invention is not affected by the current transformer model, wiring condition, degree of damage in use, and wiring condition of the cable shielding wire. That is, the present invention adopts a criterion for independent judgment of each line, does not need group comparison, is not affected by the amplitude and phase angle errors of the transient zero-sequence current sequence collected by any line in the distribution system, and still maintains good line selection accuracy and reliability.

[0114] Example 1

[0115] In order to verify the effectiveness of the distribution network ground fault line selection method based on phase space trajectory recognition described in this invention, a 35kV arc suppression coil distribution system simulation model was built using PSCAD, and a single-phase high-resistance ground fault was simulated and analyzed. Figure 5 There are 4 outgoing lines in the model, including L 1 , L 4 are two overhead lines; L 2 It is an overhead-cable hybrid line; L3 The main transformer parameters of the system are set as follows: rated voltage 110 / 35kV, rated capacity 100MVA, system ground capacitance current 84.07A, arc suppression coil detuning degree -5%, inductance value 1.263H. The main parameters of the line are shown in Table 1. 4 A ground fault point F is set on the circuit. Since the transient characteristics of the first few cycles after the system fault are most obvious, the zero-sequence current data within five cycles after the fault is collected at a sampling frequency of 4 kHz in the simulation and a phase space trajectory model is constructed for line selection analysis.

[0116] Table 1 Line parameters

[0117]

[0118] At 0.405s, a high-resistance ground fault occurs at the ground fault point F. The transition resistance at the fault point is 4000Ω, and the initial phase angle of the fault is 90°. The dynamic waveforms of the zero-sequence currents of the system lines are as follows: Figure 6 As shown, the data of five cycles after the fault occurs are analyzed. According to formula (17), the origin distance ratio coefficient μ of each line phase space trajectory is calculated respectively. k The curve of the change of the ratio coefficient of the origin distance of each line in the first five cycles under high-resistance grounding fault is as follows: Figure 7 shown.

[0119] Depend on Figure 7 As shown, line L 1 , L 2 , L 3 In the five cycles after the fault occurs, the origin distance ratio coefficient is greater than or equal to 1, and there is no value less than 1. The origin distance ratio coefficient of each line when the fault occurs in half a cycle can be extracted separately for analysis, μ 1 =2.012, μ 2 =2.012, μ 3 =2.018>1; and line L 4 The origin distance ratio coefficient of is always slightly less than 1, which is obviously different from the other three lines. When the fault occurs for half a cycle, μ 4 =0.869<1. In the five cycles of acquisition, the results always meet the requirements of criterion (18). It can be intuitively judged that the fault occurs in line L. 4 superior.

[0120] In order to verify the universality of the criterion under the high-resistance grounding fault of the distribution network, multiple groups of fault states with different initial phase angles and transition resistances are randomly set at the grounding fault point F, and the data of the first five cycles after the fault occurs are analyzed. Table 2 shows the distance ratio coefficients of each line and the line selection results at the half-cycle of the fault under different fault states. It can be seen from Table 2 that no matter how the fault state changes, the distance ratio coefficient μ of the fault line at the half-cycle after the high-resistance grounding fault occurs k The distance ratio coefficient of the normal line is always less than 1, and is clearly distinguished by the constant 1 as the dividing line. The new criterion for fault line selection in distribution network identified by phase space trajectory proposed in this paper can accurately and reliably select the fault line, and the line selection result is not affected by the time of fault occurrence and the size of the transition resistance at the fault point.

[0121] Table 2 Distribution network line selection results under different fault conditions

[0122]

[0123] Example 2

[0124] In order to verify the reliability of the distribution network ground fault line selection method based on phase space trajectory recognition described in the present invention, which is not affected by the amplitude and phase angle errors of the transient zero-sequence current sequence collected by any line of the distribution system, the 35kV distribution network simulation model line L is set 1 The polarity of the upper analog current transformer is reversed, and the collected zero-sequence current data is the opposite of the original data; Line L 2 and L 4 Simulation: Due to the long service life of the current transformer, the amplitude and phase of the measured current data have different degrees of error. Assuming that the line L 2 The current transformer detection amplitude error is 2%, and the phase error is 18°; Line L 4 The current transformer detection amplitude error is -3%, and the phase error is 9°. The distribution network model is set to have a high-resistance grounding fault with a transition resistance of 6000Ω and a fault initial phase angle of 0° at the grounding fault point F at 0.3965s. The dynamic waveforms of the zero-sequence currents of the system lines are as follows: Figure 8 As shown, it can be seen that at this time line L 1 The phase of the zero-sequence current data collected on the line is opposite to the actual data, showing significantly different characteristics from other lines, which greatly interferes with the fault line selection; while line L 2 and L 4 Due to the existence of transmission errors, the amplitude and phase of the collected current data are also different from the actual ones, which affects the line selection results.

[0125] Using the criterion proposed in the present invention, the data of five cycles after the fault occurs are analyzed, and the change curve of the ratio coefficient of the origin distance of each line is as follows: Fig. 9Although there are certain errors in the acquisition of zero-sequence current of each line after the fault occurs, the line L 1 , L 2 , L 3 Within five cycles after the fault occurs, the origin distance ratio coefficient is still greater than or equal to 1. When the fault occurs for half a cycle, μ 1 =1.549, μ 2 =1.653, μ 3 =1.550>1; and line L 4 The origin distance ratio coefficient is slightly less than 1. When the fault occurs in half a cycle, μ 4 =0.901<1, within the five cycles of acquisition, the result still meets the requirements of criterion (18), accurately judging the fault line L 4 , which verifies the reliability of the method proposed in this invention.

[0126] The above embodiments are preferred embodiments of the present application. Ordinary technicians in this field can also make various changes or improvements on this basis. Without departing from the overall concept of the present application, these changes or improvements should fall within the scope of protection required by the present application.

Claims

1. A method for selecting a high-resistance grounding fault line in a distribution network based on phase space trajectory recognition, characterized in that: Obtain the transient zero-sequence current data of each line in the distribution system when a high-resistance grounding fault occurs in the system, and construct the corresponding phase space trajectory based on the phase space reconstruction principle; determine the line where the high-resistance grounding fault occurs based on the distance change trend from the vector point to the origin on the phase space trajectory; The method for constructing phase space trajectory based on the phase space reconstruction principle is: The acquired transient zero-sequence current data is expressed as a zero-sequence current time series , set the delay time for phase space reconstruction is 1 / 4 of the power frequency period; k is the line number, and its value range is [1, n ], n is the number of lines in the power distribution system; m It is represented as the sampling point number. , s Indicates the total number of sampling points; The zero-sequence current of each sampling point in the transient zero-sequence current data Map it into two-dimensional space for reconstruction, and obtain the two-dimensional space vector of each sampling point with the origin as the starting point , the two-dimensional space vector ends of the zero-sequence current of all sampling points are connected in sequence to form the phase space trajectory in the two-dimensional space; , are the unit vectors of the horizontal and vertical axes respectively; The method for judging the high-resistance grounding fault line based on the distance change trend is: Extract the distance from each vector point to the origin on the phase space trajectory, that is, calculate the length of each vector in the two-dimensional space, and obtain the zero-sequence current time series of each line Corresponding distance time series in two-dimensional space ; Compute distance time series The ratio coefficient between two distances separated by one power frequency cycle , ; It is determined whether there is a value less than 1 in the ratio coefficient sequence of each line. If so, the line is determined to be a high-resistance grounding fault line.

2. The method according to claim 1, characterized in that If the distance from the vector point to the origin on the phase space trajectory shows a decreasing trend, the line is judged to be a high-resistance grounding fault line, otherwise the line does not have a high-resistance grounding fault.

3. The method according to claim 1, characterized in that The transient zero-sequence current data obtained refers to the current of each line from the moment when a high-resistance grounding fault occurs in the system. h The zero-sequence current time series of one cycle.

4. The method according to claim 1, characterized in that When the real-time monitoring shows that the effective value of the zero-sequence voltage of the distribution system is greater than the voltage setting value, it is judged that a high-resistance grounding fault has occurred in the system.

5. A distribution network high-resistance grounding fault line selection device based on phase space trajectory recognition, characterized in that: It includes a phase space trajectory building module and a fault line selection module; The phase space trajectory construction module is used to: obtain transient zero-sequence current data of each line of the power distribution system when a high-resistance grounding fault occurs in the system, and construct a corresponding phase space trajectory based on the phase space reconstruction principle; The fault line selection module is used to: determine the line where the high-resistance grounding fault occurs according to the distance change trend from the vector point to the origin on the phase space trajectory; The method for constructing the phase space trajectory by the phase space trajectory construction module based on the phase space reconstruction principle is: The acquired transient zero-sequence current data is expressed as a zero-sequence current time series , set the delay time for phase space reconstruction is 1 / 4 of the power frequency period; k is the line number, and its value range is [1, n ], n is the number of lines in the power distribution system; m It is represented as the sampling point number. , s Indicates the total number of sampling points; The zero-sequence current of each sampling point in the transient zero-sequence current data Map it into two-dimensional space for reconstruction, and obtain the two-dimensional space vector of each sampling point with the origin as the starting point , the two-dimensional space vector ends of the zero-sequence currents of all sampling points sequentially constitute the phase space trajectory in the two-dimensional space; , are the unit vectors of the horizontal and vertical axes respectively; The method for the fault line selection module to determine the high-resistance grounding fault line according to the distance change trend is: Extract the distance from each vector point to the origin on the phase space trajectory, that is, calculate the length of each two-dimensional space vector to obtain the zero-sequence current time series of each line Corresponding distance time series in two-dimensional space ; Compute distance time series The ratio coefficient between two distances separated by one power frequency cycle , ; It is determined whether there is a value less than 1 in the ratio coefficient sequence of each line. If so, the line is determined to be a high-resistance grounding fault line.

6. The device according to claim 5, characterized in that The line selection method of the fault line selection module is: if the distance from the vector point to the origin on the phase space trajectory shows a decreasing trend, the line is judged to be a high-resistance grounding fault line, otherwise no high-resistance grounding fault occurs in the line.

Images