A fast longitudinal differential protection method for flexible DC distribution lines

Abstract

The present invention relates to a fast longitudinal differential protection method suitable for flexible direct current distribution lines. By constructing a protection start criterion based on the SW normality detection method, and targeting the problem that the transient characteristics of short-circuit currents under different fault transition resistances are greatly different and the differential current is easily affected by factors such as noise, abnormal data, CT transmission error, signal synchronization error, and abnormal data, the current sum mutation amount, current difference mutation amount, and their cumulative amount are constructed. Based on this, a fault identification principle with strong anti-interference ability is proposed, which is combined with the SW start criterion to realize fast line protection. The present invention can reliably identify line short-circuit faults and is less affected by interference factors; and the protection judgment result is not affected by changes in the distribution network operation mode. It is applicable to various distribution network architectures such as radial, double-end power supply, and ring, and can effectively improve the protection speed of flexible direct current distribution lines and the sensitivity of high-resistance fault identification.

Description

Technical Field

[0001] The present invention belongs to the technical field of distribution network line protection, and in particular to a fast longitudinal differential protection method suitable for flexible direct current distribution lines. Background Art

[0002] Flexible DC distribution networks have broad development prospects as a highly efficient carrier for diverse "source-load-storage" power systems due to their technical and economic advantages, including facilitating the integration of renewable energy, energy conservation and consumption reduction, future urban grid capacity expansion, and efficient and flexible conversion and utilization of electricity. However, several key issues remain to be addressed. Among them, relay protection, as the first line of defense to ensure the safe and stable operation of distribution networks, has become a research hotspot. The difficulty faced in protecting flexible DC distribution lines lies in the fact that after a line short circuit, the fault current increases rapidly and has a high peak value. However, power electronic devices have a weak ability to withstand transient overcurrents. A single line short-circuit current can affect all lines in the DC field. Protection must use minimal effective information within a few milliseconds to quickly and accurately identify faults.

[0003] DC line protection is mainly divided into single-ended protection and longitudinal protection. In single-ended protection, the existing technology applies traveling wave protection to multi-terminal flexible DC systems, and uses voltage, current traveling waves and voltage differentials to form a comprehensive criterion to improve protection selectivity and reliability. In the existing technology, the curvature algorithm is used to extract the sudden difference in the current transient waveform during faults inside and outside the zone, and the protection criterion is formed in combination with the current sudden change. There is also an RL model-based algorithm in the existing technology, which uses the differential equation of the fault to calculate the fault distance and construct distance-based protection. Based on the establishment of a complex frequency domain equivalent model of the distribution network, the existing technology proposes a protection strategy based on the voltage waveform of the current-limiting reactor. However, the single-ended protection principle still has some shortcomings. For example, the distribution network has many short lines, complex topology and low voltage, which makes it difficult to identify and capture the traveling wave head; the transient current and voltage changes during high-resistance faults are not obvious, and the selectivity and reliability of transient characteristic change rate protection and distance-based protection are often poor when taking into account factors such as noise and measurement errors; high-frequency boundary protection relies on the primary system to provide additional conditions such as current-limiting reactors, which increases the design requirements and complexity of the primary system;

[0004] Longitudinal protection uses electrical quantities at both ends of the line to form protection criteria, has clear action boundaries, is easier to set and has higher selectivity than single-end protection. Based on the analysis of influencing factors such as synchronization error, load type and measurement error, the existing technology proposes differential quick-break protection suitable for DC distribution network, but it is only for bipolar short-circuit faults and does not consider the influence of transition resistance. The existing technology proposes high-voltage DC line differential protection with the ability to resist distributed capacitance current of the line, and its applicability in the distribution system remains to be verified. The existing technology proposes longitudinal protection based on the short-time energy of line differential current. The current energy function expands the difference in transient characteristics between the fault line and the non-fault line, but it is also susceptible to factors such as noise and abnormal data. In terms of longitudinal directional protection, the existing technology proposes protection based on the direction of current mutation, but this solution relies on the cooperation of upstream and downstream line measurement devices to achieve protection. Its applicability in complex distribution networks with multiple branch lines remains to be verified. The existing technology proposes a longitudinal protection based on the cosine similarity of the current mutation amount. The existing technology uses the t-test to analyze the correlation of the full fault current waveform and establishes a protection strategy based on this. The current correlation protection is easy to set and has a certain ability to tolerate high resistance. However, when a high-resistance fault is superimposed on strong noise, there is still a problem of reduced reliability.

[0005] In addition, related research has proposed protection principles based on intelligent algorithms such as artificial neural networks and deep learning. However, the sample training process for intelligent protection is complex, and retraining is often required when the system operating mode changes.

[0006] In summary, most protection schemes find it difficult to meet the "four properties" requirements of flexible DC distribution line protection while taking into account influencing factors such as transition resistance, noise, abnormal data, distributed capacitance and communication delay. Summary of the Invention

[0007] The purpose of the present invention is to overcome the shortcomings of the existing technology and propose a fast longitudinal differential protection method suitable for flexible DC distribution lines, which can reliably identify line short-circuit faults and is less affected by interference factors; and the protection judgment result is not affected by changes in the operation mode of the distribution network. It is suitable for various distribution network architectures such as radial, double-end power supply and ring, and can effectively improve the protection speed of flexible DC distribution lines and the sensitivity of high-resistance fault identification.

[0008] The present invention solves the technical problem by adopting the following technical solutions:

[0009] A fast longitudinal differential protection method for flexible DC distribution lines, comprising the following steps:

[0010] Step 1: Detect the normality of the differential current based on the SW method and calculate the positive differential current SW test value W p Sum current negative differential current SW test value W n ;

[0011] Step 2: Determine the positive differential current SW test value W p Or the current negative differential current SW test value W n Is it less than the critical value P? If so, proceed to step 3, otherwise return to step 1;

[0012] Step 3: Calculate the cumulative amount of positive line current and value mutation ∑Δi (M+N)p , the cumulative amount of the positive line current difference mutation ∑Δi (M-N)p , the cumulative amount of negative line current and value mutation ∑Δi (M+N)n , the cumulative amount of the negative line current difference mutation ∑Δi (M-N)n ;

[0013] Step 4: Get the sum of the cumulative amount of the positive line and the difference And the sum of the accumulated amount of the negative line and the difference

[0014] Step 5: Determine the sum of the accumulated amount of the positive line and the difference Are both greater than 0? If so, proceed to step 6; otherwise, proceed to step 7.

[0015] Step 6: Determine the sum of the accumulated amount of the negative line and the difference Are they all greater than 0? If they are all greater than zero, the interval level is short-circuited and the end occurs; otherwise, the positive electrode in the area is short-circuited and the end occurs;

[0016] Step 7: Determine the sum of the accumulated amount of the negative line and the difference Are they all greater than 0? If they are all greater than zero, the negative electrode in the area is short-circuited and the process ends; otherwise, return to step 1.

[0017] Wherein, M is the M side number of line MN, N is the N side number of line MN, i is the current, p is the positive pole, and n is the negative pole.

[0018] Moreover, in step 1, the positive direction of the current is from the busbar to the line, and the negative direction of the current is from the line to the busbar.

[0019] Moreover, the positive differential current SW test value W in step 1 p Sum current negative differential current SW test value W n The calculation method is: Use SW test statistic W to calculate:

[0020]

[0021] Where, L is the total number of differential current samples, i diffj is the differential current sampling value, i' diff i is the differential current sampling value diffj In non-descending order, is the mean value of the differential current, a j is the expected value of the standard normal distribution order statistic calculated by linear regression.

[0022] Moreover, the critical value P in step 2 set Find i' through the test statistic W value table diff The critical value for 90% similarity to the standard normal distribution.

[0023] Moreover, the specific calculation method of the cumulative amount of line current sum mutation and the cumulative amount of line current difference mutation in step 3 is:

[0024]

[0025] Δi M+N (j)=|i M (j)+i N (j)|-|μ M+N |

[0026] Δi M-N (j)=|i M (j)-i N (j)|-|μ M-N |

[0027]

[0028] Among them, i M (j), i N (j) are the currents measured at the jth sampling point on both sides of the line, |μ M+N |、|μ M-N | are the absolute values of the mean current sum and the mean current difference in the last data window before the line differential current abnormality is detected, N is the length of the fault type identification data window, Δi M+N (j) is the line current and value mutation amount, Δi M-N (j) is the sudden change of line current difference.

[0029] Moreover, the criteria for steps 5, 6 and 7 are:

[0030] like and If both are greater than 0, there is a positive grounding short circuit fault in the area:

[0031]

[0032] like and greater than 0, or If it is less than 0, there is a negative grounding short circuit fault in the area:

[0033]

[0034] like or Less than 0, and If it is greater than 0, it is considered that there is an inter-pole short circuit fault in the zone:

[0035]

[0036] If the above situation does not exist, there is no short circuit in the area, where

[0037]

[0038] Among them, k p is the positive electrode proportional coefficient, k n is the negative electrode proportional coefficient, ∑Δi (M+N)p is the cumulative amount of positive line current and value mutation, ∑Δi (M-N)p is the cumulative amount of the positive line current difference mutation, ∑Δi (M+N)n is the cumulative amount of negative line current and value mutation, ∑Δi (M-N)n It is the cumulative amount of the sudden change in the negative line current difference.

[0039] The advantages and positive effects of the present invention are:

[0040] The present invention constructs a protection start-up criterion based on the SW normality detection method. The start-up criterion is not affected by noise and can be reliably started under both inter-pole short-circuit faults and single-pole high-resistance faults. At the same time, in order to address the problem that the transient characteristics of short-circuit currents under different fault transition resistances are greatly different and the differential current is easily affected by factors such as noise, abnormal data, CT transmission errors, signal synchronization errors, and abnormal data, the current sum mutation amount, current difference mutation amount and their cumulative amount are constructed, and based on this, a fault identification principle with strong anti-interference ability is proposed, which is combined with the SW start-up criterion to achieve rapid line protection. The present invention can reliably identify line short-circuit faults and is less affected by interference factors; and the protection judgment result is not affected by changes in the operation mode of the distribution network. It is applicable to various distribution network architectures such as radial, double-end power supply and ring, and can effectively improve the protection speed of flexible direct current distribution lines and the sensitivity of high-resistance fault identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is a flow chart of the present invention;

[0042] Figure 2 It is a schematic diagram of the sliding window of the present invention;

[0043] Figure 3 This is a schematic diagram of the relationship between variables in each data window in the fault type identification link of the present invention;

[0044] Figure 4 A structural diagram of the distribution network constructed for an embodiment of the present invention. DETAILED DESCRIPTION

[0045] The present invention is further described below in conjunction with the accompanying drawings.

[0046] A fast longitudinal differential protection method for flexible DC distribution lines, such as Figure 1 As shown, the following steps are included:

[0047] Step 1: Detect the normality of the differential current based on the SW method and calculate the positive differential current SW test value W p Sum current negative differential current SW test value W n .

[0048] The positive direction of line current is defined as the flow from busbar to line, and the reverse direction is the flow from line to busbar. diff =i M +i N , are the currents on both sides of the line respectively. The differential current sampling value i diffj (j=1,2,…,L) are sorted in non-descending order, denoted as i' diff :

[0049] i′ diff1 <i′ diff2 <… <i′ diffL

[0050] Use SW test statistic W to test the positive differential current SW value W p Sum current negative differential current SW test value W n Perform the calculation:

[0051]

[0052] Where, L is the total number of differential current samples, i diffj is the differential current sampling value, i' diff i is the differential current sampling value diffj In non-descending order, is the mean value of the differential current, a j is the expected value of the standard normal distribution order statistic calculated by linear regression.

[0053] Step 2: Determine the positive differential current SW test value Wp Or the current negative differential current SW test value W n Is it less than the critical value P? If so, proceed to step 3, otherwise return to step 1. The critical value P is found by looking up the test statistic W value table i' diff The critical value for 90% similarity to the standard normal distribution.

[0054] Step 3: Calculate the cumulative amount of positive line current and value mutation ∑Δi (M+N)p , the cumulative amount of the positive line current difference mutation ∑Δi (M-N)p , the cumulative amount of negative line current and value mutation ∑Δi (M+N)n , the cumulative amount of the negative line current difference mutation ∑Δi (M-N)n .

[0055] At the same time, using Figure 2 and Figure 3 The sliding window with an interval of k shown in FIG is used as a data window, that is, a fault type identification data window, to observe the data, such as Figure 2 As shown, W x That is, the fault type identification data window, W x-1 That is, the last data window before the line differential current abnormality is detected, k is the sliding window interval, and W x-1 Internal calculation |μ M+N | and |μ M-N |, W x Internal calculation|i M (j)+i N (j)| and|i M (i)-i N (j)|, and the line current and value mutation Δi are obtained M+N (j) and the current difference mutation Δi M-N (j):

[0056] Δi M+N (j)=|i M (j)+i N (j)|-|μ M+N |

[0057] Δi M-N (j)=|i M (j)-i N (j)|-|μ M-N |

[0058]

[0059] Among them, i M (j), i N (j) are the currents measured at the jth sampling point on both sides of the line, |μ M+N |、|μM-N | are the absolute values of the mean current sum and the mean current difference in the last data window before the line differential current anomaly is detected.

[0060] In order to improve the protection of the discriminant data window Δi M+N and Δi M-N The difference between the sum of the line current and the value of the sudden change is calculated by using the cumulative amount of the line current and the cumulative amount of the line current difference sudden change to calculate the cumulative amount of the positive line current and the value of the sudden change ∑Δi (M+N)p , the cumulative amount of the positive line current difference mutation ∑Δi (M-N)p , the cumulative amount of negative line current and value mutation ∑Δi (M+N)n , the cumulative amount of the negative line current difference mutation ∑Δi (M-N)n Perform the calculation:

[0061]

[0062] Step 4: Get the sum of the cumulative amount of the positive line and the difference And the sum of the accumulated amount of the negative line and the difference

[0063] Considering that when the fault type is a high-resistance ground short circuit, the transient current is small, it is possible that ∑Δi (M+N)n ,∑Δi (M-N)n is also small, so the positive electrode proportional coefficient and the negative electrode proportional coefficient k are introduced p and k n This will further expand the difference between short circuits and other faults or abnormal conditions within the zone and improve protection reliability. The basis for determining steps 5, 6, and 7 is as follows:

[0064] like and If both are greater than 0, there is a positive grounding short circuit fault in the area:

[0065]

[0066] like and greater than 0, or If it is less than 0, there is a negative grounding short circuit fault in the area:

[0067]

[0068] like or Less than 0, and If it is greater than 0, it is considered that there is an inter-pole short circuit fault in the zone:

[0069]

[0070] If the above situation does not exist, there is no short circuit in the area, where

[0071]

[0072] Among them, k p is the positive electrode proportional coefficient, k n is the negative electrode proportional coefficient.

[0073] Step 5: Determine the sum of the accumulated amount of the positive line and the difference Are both greater than 0? If so, proceed to step 6; otherwise, proceed to step 7.

[0074] Step 6: Determine the sum of the accumulated amount of the negative line and the difference Are they all greater than 0? If they are all greater than zero, the interval levels are short-circuited and the end occurs; otherwise, the positive poles in the area are short-circuited and the end occurs.

[0075] Step 7: Determine the sum of the accumulated amount of the negative line and the difference Are they all greater than 0? If they are all greater than zero, the negative electrode in the area is short-circuited and the process ends; otherwise, return to step 1.

[0076] According to the above-mentioned fast longitudinal differential protection method applicable to flexible DC distribution lines, the following is constructed on the PSCAD / EMTDC platform: Figure 4 The six-terminal ring-shaped flexible DC distribution network shown is calculated to verify the effect of the present invention.

[0077] The constructed six-terminal ring flexible DC distribution network has a sampling frequency of 10kHz and a fault identification data window length of 1ms. Taking line L1 protection R11 and R21 as an example, simulations were performed to verify different faults.

[0078] The voltage level of the ring flexible DC distribution network is ±10kV, and the voltage level of the AC main grid connected to the ring network is 110kV. Figure 4As shown, T1 and T2 are MMC converters connected to the AC mains. T1 uses constant voltage control to maintain system voltage, while T2 uses constant power control. T3, T4, and T5 are VSC converters, using constant power control, connected to medium- and low-voltage AC loads. T6 and T7 are dual-active bridge-type DC transformers, which maintain the low-voltage side voltage at ±750V and ±400V respectively through single-phase shift control. The photovoltaic power source uses maximum power point tracking control and is connected to the DC distribution network via DC transformer T6. Distribution network grounding methods include grounding the neutral point of the converter transformer valve side via a high-resistance grounding and directly grounding the neutral point of the VSC split capacitor. F1 to F8 are system fault points. The specific fault types and transition resistance data are shown in Table 1.

[0079] Table 1 Fault type and transition resistance

[0080]

[0081] The process of the present invention is used to determine the short-circuit faults in the zone in Table 1, and the results are shown in Table 2:

[0082] Table 2 Short-circuit fault identification results in different zones

[0083]

[0084] The data in Table 2 show that the protection of intra-zone inter-pole short circuit and single-pole ground short circuit can reliably identify the correct fault type and is not affected by transition resistance and noise.

[0085] Since various interference factors may occur in the power distribution network, the reliability of the present invention is tested under different influences.

[0086] (1) Detection of short-circuit faults taking into account interference factors.

[0087] Optical fiber communication has a data transmission speed of approximately 200 km / ms, and the communication delay for distribution lines less than 40 km is approximately 0.2 ms. Taking the case of a positive-pole ground short circuit at F1 through a 50Ω transition resistor and a metallic short circuit between poles at F2 as an example, the reliability of the scheme was verified by considering a signal synchronization error of 0.3 ms. The scheme's reliability was also verified by considering the case of a positive-pole ground short circuit at F1 through a 50Ω transition resistor and a metallic short circuit between poles at F2, with mutual inductor errors of +10% and -10% on both sides of L1. The simulation results, shown in Table 3, show that the protection system can reliably distinguish between internal and external faults even in the presence of interference.

[0088] Table 3 Short-circuit fault identification results considering interference factors

[0089]

[0090]

[0091] (2) Detection of abnormal interference with normal operation.

[0092] Taking the occurrence of abnormal data I, abnormal data II, and CT disconnection on the positive electrode R11 side of line L1 at 0.5s as an example, the reliability of the proposed protection is verified. The corresponding results are shown in Table 3. When there is abnormal interference during normal operation, the protection does not operate.

[0093] Table 4 Identification results of abnormal interference in normal operation

[0094]

[0095] (3) Detection of asymmetric ground fault on the AC side.

[0096] When an asymmetric ground fault occurs on the converter transformer valve side, the DC voltage on the MMC fluctuates periodically, and the protection of the adjacent line detects the current change. After a single-phase ground fault occurs on F7 at 0.5 seconds, the protection on line L1 fails to operate, as shown in Table 5. The scheme is unaffected by the AC fault.

[0097] Table 5 Protection identification results under AC fault

[0098]

[0099] (4) Testing the reliability of protection for switching heavy loads.

[0100] Switching heavy loads at adjacent nodes of a line can cause a sudden change in line current. The reliability of the scheme is verified by switching a 4MW heavy load ACLOAD1 at 0.5s. The results are shown in Table 6. The protection operates in both cases, so switching heavy loads does not affect protection reliability.

[0101] Table 6 Protection identification results under heavy load switching

[0102]

[0103] (5) Protection reliability detection of changes in distribution network operation mode.

[0104] Taking the case where line L2 is disconnected and the distribution network switches from closed-loop operation to open-loop operation, and line L1 experiences an inter-electrode metallic short circuit, a positive electrode ground short circuit through a 50Ω transition resistor, and a negative electrode metallic short circuit at 0.5s as an example, the protection reliability is analyzed. The simulation results are shown in Table 7. The protection operates reliably and determines the corresponding fault type, and is not affected by changes in the system operation mode.

[0105] Table 7 Short circuit identification results within the zone during open-loop operation

[0106]

[0107]

[0108] According to the above detection, the present invention aims to solve the problem that the transient characteristics of short-circuit currents under different fault transition resistances vary greatly and the differential current is easily affected by factors such as noise, abnormal data, CT transmission errors, signal synchronization errors, and abnormal data. It constructs the current sum mutation amount, the current difference mutation amount, and their cumulative amount, and accordingly proposes a fault identification principle with strong anti-interference ability, which is combined with the SW start-up criterion to achieve rapid line protection. The present invention can reliably identify line short-circuit faults and is less affected by interference factors; and the protection judgment results are not affected by changes in the distribution network operation mode. It is applicable to various distribution network architectures such as radial, double-end power supply, and ring, and can effectively improve the protection speed of flexible direct current distribution lines and the sensitivity of high-resistance fault identification.

[0109] It should be emphasized that the embodiments described in the present invention are illustrative rather than restrictive. Therefore, the present invention includes but is not limited to the embodiments described in the specific embodiments. Any other embodiments derived by those skilled in the art based on the technical solutions of the present invention also fall within the scope of protection of the present invention.

Claims

1. A fast longitudinal differential protection method for flexible DC distribution lines, characterized by: The following steps are involved: Step 1: Detect the normality of the differential current based on the SW method and calculate the positive differential current SW test value W p Sum current negative differential current SW test value W n ; Step 2: Determine the positive differential current SW test value W p Or the current negative differential current SW test value W n Is it less than the critical value P set If it is less than, proceed to step 3, otherwise return to step 1; Step 3: Calculate the cumulative amount of positive line current and value mutation ∑Δi (M+N)p , the cumulative amount of the positive line current difference mutation ∑Δi (M-N)p , the cumulative amount of negative line current and value mutation ∑Δi (M+N)n , the cumulative amount of the negative line current difference mutation ∑Δi (M-N)n ; Step 4: Get the sum of the cumulative amount of the positive line and the difference And the sum of the accumulated amount of the negative line and the difference Step 5: Determine the sum of the accumulated amount of the positive line and the difference Are both greater than 0? If so, proceed to step 6; otherwise, proceed to step 7. Step 6: Determine the sum of the accumulated amount of the negative line and the difference Are they both greater than 0? If they are both greater than zero, the inter-electrode short circuit in the zone is terminated; otherwise, the positive electrode in the zone is short-circuited and terminated. Step 7: Determine the sum of the accumulated negative circuit quantities and the difference Are they all greater than 0? If they are all greater than zero, the negative electrode in the area is short-circuited and the process ends; otherwise, return to step 1. Wherein, M is the M side number of line MN, N is the N side number of line MN, i is the current, p is the positive pole, and n is the negative pole.

2. A fast longitudinal differential protection method for flexible DC distribution lines according to claim 1, characterized in that: In step 1, the positive direction of the current is from the busbar to the line, and the negative direction of the current is from the line to the busbar.

3. The fast longitudinal differential protection method for flexible DC distribution lines according to claim 1, characterized in that: The positive differential current SW test value W in step 1 p Sum current negative differential current SW test value W n The calculation method is: Use the SW test statistic W to calculate: Where, L is the total number of differential current samples, i diffj is the differential current sampling value, i' diff i is the differential current sampling value diffj In non-descending order, is the mean value of the differential current, a j is the expected value of the standard normal distribution order statistic calculated by linear regression.

4. The fast longitudinal differential protection method for flexible DC distribution lines according to claim 1, characterized in that: The critical value P in step 2 set Find i' through the test statistic W value table diff The critical value for 90% similarity to the standard normal distribution.

5. The fast longitudinal differential protection method for flexible DC distribution lines according to claim 1, characterized in that: The specific calculation method of the cumulative amount of line current sum mutation and the cumulative amount of line current difference mutation in step 3 is: Δi M+N (j)=|i M (j)+i N (j)|−|µ M+N | Δi M-N (j)=|i M (j)–i N (j)|−|µ M-N | Among them, i M (j), i N (j) are the currents measured at the jth sampling point on both sides of the line, |μ M+N |、|μ M-N | are the absolute values of the mean current sum and the mean current difference in the last data window before the line differential current abnormality is detected, L is the length of the fault type identification data window, Δi M+N (j) is the line current and value mutation amount, Δi M-N (j) is the sudden change of line current difference.

6. The fast longitudinal differential protection method for flexible DC distribution lines according to claim 1, characterized in that: The criteria for steps 5, 6 and 7 are: like and If both are greater than 0, there is an inter-pole short circuit fault in the zone: like and greater than 0, or If it is less than 0, there is a positive short circuit fault in the area: like and Less than 0, or If it is greater than 0, it is considered that there is a negative short circuit fault in the area: If the above situation does not exist, there is no short circuit in the area, where Among them, k p is the positive electrode proportional coefficient, k n is the negative electrode proportional coefficient, ∑Δi (M+N)p is the cumulative amount of positive line current and value mutation, ∑Δi (M-N)p is the cumulative amount of the positive line current difference mutation, ∑Δi (M+N)n is the cumulative amount of negative line current and value mutation, ∑Δi (M-N)n It is the cumulative amount of the sudden change in the negative line current difference.

Images