WO2019233437A1

WO2019233437A1 - Method for evaluating security of closed-loop operation of medium-voltage distribution network

Info

Publication number: WO2019233437A1
Application number: PCT/CN2019/090120
Authority: WO
Inventors: 周自强; 张焰; 冯楠; 郭强; 连鸿波; 陈旸; 余颖辉
Original assignee: 国网上海市电力公司; 上海交通大学
Priority date: 2018-06-05
Filing date: 2019-06-05
Publication date: 2019-12-12
Also published as: CN108711851B; JP2020527320A; CN108711851A; JP6839769B2

Abstract

Disclosed is a method for evaluating the security of a closed-loop operation of a medium-voltage distribution network. The method includes: on the basis of acquired topological structures and device parameters of a closed-loop network, and acquired real-time running data before loop closing of the closed-loop network, estimating a state of a high-voltage distribution network to obtain voltage amplitude values and phase angles of 10kV buses at head ends of closed-loop feeders; selecting a ratio of an active load of each load point on each of the closed-loop feeders to the active power at the head end of the feeder and a ratio of a reactive load of each load point on the closed-loop feeder to the reactive power at the head end of the feeder as input variables, and calculating, on the basis of historical load data of each load point, a multi-order semi-invariant of each of the input variables; calculating a certainty power flow and closed-loop currents; calculating multi-order semi-invariants of the closed-loop currents; obtaining the cumulative probability distribution of the closed-loop currents; and calculating the limit violation probability of each of the closed-loop currents, and determining that a closed-loop operation has security insofar as the multiple limit violation probabilities are all less than a pre-set threshold value.

Description

Method for assessing the safety of closing operation of medium voltage distribution network

This application claims priority from a Chinese patent application filed with the Chinese Patent Office on June 5, 2018, with application number 201810568104.0, the entire contents of which are incorporated herein by reference.

Technical field

The present application relates to the field of safe and stable operation of a power system, for example, to a method for evaluating the operation safety of a closed loop of a medium voltage distribution network.

Background technique

The medium voltage distribution network adopts a closed-loop design and an open-loop operation power supply mode. Under normal circumstances, the contact switch is turned on, and the power distribution network operates as a radiating structure. During equipment maintenance or accident handling, the load can be transferred without interruption through the loop operation of the contact switch. The closing operation of the medium-voltage distribution network may generate a large closing current in the closing network, causing line current protection actions or overloading certain electrical equipment, leading to a wider range of power outages. Therefore, the operator needs to evaluate the safety of the loop operation before the loop operation.

At present, there is no systematic method for evaluating the safety of loop-closure operation in medium-voltage distribution networks. In actual production, workers usually judge whether the loop-closure operation is safe based on experience, lacking corresponding theoretical support, and low accuracy.

Summary of the Invention

The present application provides a method for evaluating the operation safety of the closed-loop operation of a medium-voltage distribution network, and solves the problem that the evaluation accuracy of the closed-loop operation safety of the medium-voltage distribution network is not high.

In one embodiment, the present application discloses a method for evaluating the operation safety of the close-loop operation of a medium-voltage distribution network. The method considers the load point load on the close-loop feeder as a random variable, and obtains the close-loop based on the probability flow theory. Probability of out-of-limits of steady-state current and transient inrush current and evaluation of the safety of loop closing operation, including:

Based on the obtained topology and equipment parameters of the loop-to-loop network, and real-time operating data of the loop-to-loop network before looping, the state of the high-voltage distribution network is estimated, and the voltage amplitudes of the 10kV busbars at the ends of the two loop-feeding feeders are obtained. With phase angle

Selecting as input variables the ratio of the active load at each load point of each combined loop feeder to the active power of the feeder end and the ratio of the reactive load at each load point of the combined loop feeder to the reactive power at the feeder end. Calculate the multi-order semi-invariants of each input variable based on the historical load data of each load point;

Based on the voltage amplitudes and phase angles of the 10kV busbars at the ends of the two loop feeders, perform deterministic power flow calculation and loop current calculation at the reference operating point to obtain the loop network system state variable reference value and loop current reference value. Obtaining a conversion matrix based on the reference value of the state loop network system variable and the reference value of the loop current;

Calculating a multi-order semi-invariant of a loop current according to the multi-order semi-invariant of each input variable and the conversion matrix;

Obtaining a cumulative probability distribution of the closing current according to a multi-order semi-invariant of the closing current;

The maximum allowable current carrying capacity of the feeder and the setting value of the current protection are used as the limit values, and the limit crossing probability of each loop current is calculated based on the cumulative probability distribution. Closed loop operation is safe.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic diagram of a loop closing operation of a medium voltage distribution network of the present application;

FIG. 2 is a schematic flowchart of a method for assessing a loop-closure operation safety of a medium-voltage distribution network; FIG.

FIG. 3 is a technical roadmap of the method for assessing the operation safety of a closed loop of a medium voltage distribution network.

Detailed ways

The present application discloses a method for evaluating the operation safety of a closed loop of a medium-voltage distribution network. The following further describes the present application with reference to the accompanying drawings and specific embodiments.

In actual production, the staff usually thinks that the loop closing operation is safe based on experience:

(1) The 10kV bus voltage amplitude difference of the feeders where the breaks on both sides of the contact switch before closing the loop are less than 10%;

(2) The load difference between the feeders on both sides of the contact switch before closing the loop is small, and the total load is not greater than the upper limit of the transmission capacity of any feeder.

Therefore, in order to ensure safety, the close-loop operation of the medium-voltage distribution network is mostly performed at night when the load is light.

However, this evaluation method based on production experience lacks corresponding theoretical support. Theoretical analysis shows that the magnitude of the closing loop current is greatly affected by the voltage amplitude difference and phase angle difference on both sides of the contact switch fracture. Therefore, it is lack of scientific basis to evaluate the safety of closing operation only based on the voltage amplitude difference between the 10kV bus bars on both sides of the closing point. In addition, comparing the total load of the feeder loop with the upper limit of the transmission capacity of the feeder loop will result in a conservative assessment of the safety of the feeder loop, which has limited reference value for the final decision of the joint loop.

Relevant research proposes to evaluate the safety of the closed loop by calculating the steady-state current and transient inrush current of the closed loop. However, due to the limited configuration range of the distribution network measurement devices in the related technology, it is not possible to obtain the voltage phase angle information of the 10kV bus in the distribution network, and the voltage amplitude and phase angle difference between the two sides of the contact switch fracture. The real-time load data of each load point on the feeder is also unavailable, which limits the usefulness of these studies.

The present application addresses the problem that the voltage phase angle data of the bus bar in the distribution network cannot be obtained, and adopts a state estimation method for the high-voltage distribution network to obtain the bus phase angle information. Aiming at the problem that the real-time load data of the feeder load points required for the calculation of the combined loop current cannot be obtained, the probability distribution characteristics of the load value at each load point are first analyzed based on the historical load data, and then the half-step variable method is used to obtain the combination based on the probability flow theory. The cumulative distribution curve of the loop current, and finally, the safety of the loop closing operation is evaluated by obtaining the probability that the loop closing current exceeds the limit.

As shown in Figure 1, the system's power transmission network (220kV and above voltage level power grid) operates in a ring structure, and 110kV and below voltage level distribution network operates in open loop. Q1 and Q2 are the outlet circuit breakers of the feeder on both sides of the contact switch. The two 10kV feeders are connected through a tie switch Q3; when the system is operating normally, the tie switch Q3 is opened; when the operating mode is adjusted or emergency handling is performed, the tie switch Q3 can be closed and the load can be overturned by closing the loop.

As shown in FIG. 2, a method for assessing the operation safety of a closed loop of a medium voltage distribution network provided in this application includes:

S10. Based on the obtained topology structure, equipment parameters, and real-time operating data of the loop network before loop closing, perform a state estimation on the high-voltage distribution network, and obtain the voltage amplitudes of the 10kV busbars at the ends of the two loop feeders. Phase angle.

S20. Select as input the ratio of the active load at each load point on each combined feeder to the active power at the feeder end and the ratio of the reactive load at each load point on the combined feeder to the reactive power at the feeder end. Variable, based on the historical load data of each load point, a multi-order semi-invariant of each input variable is calculated.

S30. Based on the voltage amplitudes and phase angles of the first 10kV buses of the two feeder loop feeders, perform a deterministic load flow calculation and a feeder loop current calculation at the reference operating point to obtain the benchmark loop system system state variable reference value and the feeder loop current reference. A conversion matrix based on the reference value of the state loop network system variable and the reference value of the loop current.

S40. Calculate a multi-order semi-invariant of a closing current according to the multi-order semi-invariant of each input variable and the conversion matrix.

S50. Obtain a cumulative probability distribution of the closing current according to a multi-order semi-invariant of the closing current.

S60. Using the maximum allowable current carrying capacity of the feeder and the current protection setting value as the limit values, calculating the limit crossing probability of each loop current based on the cumulative probability distribution, in a case where multiple of the limit crossing probability are less than a preset threshold Make sure that the loop closing operation is safe.

In an embodiment, based on the obtained topology structure, device parameters, and real-time operating data of the ring network before the ring is combined, the state of the high-voltage power distribution network is estimated to obtain the heads of the two ring feeders. 10kV bus voltage amplitude and phase angle, including:

Based on the obtained topology structure, equipment parameters, and real-time operating data of the loop network before the loop, the weighted least squares criterion was used to estimate the state of the high-voltage distribution network, and the first 10 kV buses of the two loop feeders were obtained. Voltage amplitude and phase angle, and at the same time identify and correct bad measurement data.

In an embodiment, the ratio of the active load at each load point on the combined feeder line to the active power at the head of the feeder and the ratio of the reactive load at each load point on the combined feeder line to the reactive power at the head of the feeder are selected. As input variables, a multi-order semi-invariant of each input variable is calculated based on the historical load data of each load point, including:

Let the input variable k _P be the ratio of the active load at any load point to the active power at the head of the feeder; from the annual load historical data of 96 points throughout the year at the load point and the outgoing line, sort out the year-round dispersion of the input variable. Measure the data, and construct a multi-order semi-invariant sample set S for obtaining the input variable k _P ;

The sample set S is divided into a plurality of sub-sample sets according to the season and period of the ring-turning moment. By analyzing any one of the sub-sample sets, the seasonal sum of the input variable k _P at the ring-turning moment is obtained. Multi-order semi-invariants over time.

In an embodiment, the step of obtaining a multi-order semi-invariant of the input variable k _P in the season and time period in which the loop moment is located includes:

In response to the N discrete historical data sets {k _P1 , k _P2 , k _P3 , ..., k _PN } in the sub-sample set corresponding to the time of the loop, calculate a multi-order origin moment α _v :

According to the relationship between the semi-invariant and the origin moment, calculate the multi-order semi-invariant γ _v :

Among them, α ₁ and α _j are the origin moments when v = 1, v = j, and v represents the order;

Is the number of different combinations of j elements from v elements.

In an embodiment, based on the voltage amplitudes and phase angles of the 10kV busbars at the ends of the two feeders, the deterministic power flow calculation and the feeder current calculation are performed at the reference operating point to obtain the feeder network system state variables. The reference value and the closing loop current reference value, based on the closing loop network system state variable reference value and the closing loop current reference value, obtain a conversion matrix, including:

Based on the voltage amplitudes and phase angles of the 10kV busbars at the ends of the two loop feeders, deterministic power flow calculations are performed at the reference operating point K ₀ according to the loop network system power flow equation to obtain the loop network system state variable reference value X ₀ and the Jacobian matrix J ₀ ;

Calculate the closing current according to the closing current equation at the reference value X ₀ of the closing network system state, and obtain the closing current reference value Z ₀ and the coefficient matrix G ₀ ;

The Taylor series expansion of the closed loop network system power flow equation and the closed loop current equation are omitted and higher-order terms are omitted to obtain the random disturbance of the reference operating point K ₀ and the reference value of the closed loop network system state. A linear relationship between the random perturbation of X _{0 and} the random perturbation of the reference loop current reference value Z ₀ , and a conversion matrix T _{0 is} obtained according to the linear relationship;

Wherein, the loop current equation is Z = g (X), Z is a loop current variable, and X is a state variable of the loop network system; the loop flow system equation is W = f (X), where W is Injected power at each load point in a closed loop network system, W ₀ = AK ₀ , A is a diagonal matrix composed of active power and reactive power at the ends of the two feeders.

In an embodiment, the process of obtaining the loop current equation includes:

Let the voltage difference between the two sides of the contact switch fracture be

The total impedance of the closed loop is Z _∑ , then the closed loop steady-state circulation current

for:

Assume

with

The initial currents at the front ends of the feeders on both sides of the loop, respectively. The loop network system state variable X = [θ ₁ , V ₁ , θ ₂ , V ₂ , ..., θ _n , V _n ] ^T , according to the superposition theorem, The effective values of the steady-state currents I ′ ₁ and I ′ ₂ at the ends of the feeders on both sides after the loop are:

Let I ₁ and I ₂ be

with

The effective values of the maximum inrush current I _1M and I _2M appearing at the ends of the feeders on both sides during the transient state of the closed loop are:

I _1M = I ₁ + 1.51I _c = g ₃ (X)

I _2M = I ₂ + 1.51I _c = g ₄ (X)

Assuming that the loop closing current variable Z = [I ′ ₁ , I ′ ₂ , I _1M , I _2M ] ^T , the loop closing current equation is:

Z = g (X)

The process of obtaining the power flow equation of the loop network system includes: setting K as an input variable of the probabilistic power flow calculation; W represents the injected power of each node in the loop network system, then W = AK; the power flow equation of the loop network system can be expressed for:

W = f (X)

Wherein, the input variable K is a random variable, which can be expressed as K ₀ + ΔK, K ₀ is a reference value of the random variable K, and is a reference operating point of a closed loop network system; ΔK is a random disturbance; and a closed loop network system state variable X It can be expressed as X ₀ + ΔX, X ₀ is the reference value of the state variable of the loop network system, ΔX is the random perturbation; the injected power W of each node can be expressed as W ₀ + ΔW, W ₀ is the reference value of the node injected power, and ΔW is Random perturbation corresponding to ΔX; closed loop current variable Z can be expressed as Z ₀ + ΔZ, Z ₀ is the reference value of closed loop current variable, and ΔZ is a random perturbation corresponding to ΔX;

The linear relationship between the ΔZ and the ΔK:

Of which, the Jacobian matrix

Coefficient matrix

In an embodiment, the calculating the multi-order semi-invariant of the closing current based on the multi-order semi-invariant of each input variable and the conversion matrix includes:

Let ΔK ^{(v) be} the v-order semi-invariant of the input variable,

A matrix of coefficients representing the power of v of each element in T ₀ , then the property of semi-invariants

Find the v-order semi-invariant ΔZ ^{(v) of} the combined loop current variable.

In an embodiment, the step of obtaining the cumulative probability distribution of the closing current according to the multi-order semi-invariant of the closing current includes:

Use the Cornish-Fisher series to find the cumulative probability distribution of the closed loop current, including:

Let the cumulative distribution function of the closed loop current variable Z be F (z) and the standard normal distribution function be Φ (z). The α quantiles of F (z) and Φ (z) can be expressed as z (α) and

That is, z (α) = F ^-1 (α),

Then z (α) and

Meet the following relationships:

Among them, g _v is a v-order normalized semi-invariant of the closed-loop current variable Z, v = 1, 2, 3 ..., and the v-order semi-invariant of the closed-loop current variable Z is γ _v with a standard deviation of σ,

A cumulative probability distribution function of the closing current is obtained based on the multi-order semi-invariant of the closing current.

In an embodiment, the maximum allowable ampacity of the feeder and the current protection setting value are used as the limit values, and the limit crossing probability of each loop current is calculated based on the cumulative probability distribution. In the case of a preset threshold, it is determined that the loop closing operation is safe, including:

The closing loop current variable Z includes the steady-state current effective values I ′ ₁ and I ′ ₂ of the feeder heads on both sides after the closing loop and the maximum rush current effective values I _1M and I _2M appearing on the feeder ends on both sides;

Let the cumulative distribution functions of the variables I ′ ₂ , I ′ ₂ , I _1M , and I _2M be F ₁ (x), F ₂ (x), F ₃ (x), and F ₄ (x), on both sides of the junction The maximum allowable current carrying capacity of the feeder are I _{max, 1} and I _{max, 2} respectively _, and the setting values of the current I protection on both sides are I _{setI, 1} and I _{setI, 2} , respectively.

P ₁ = P (I ′ ₁ ≥I _{max, 1} ) = 1-F ₁ (I _{max, 1} )

P ₂ = P (I ′ ₂ ≥I _{max, 2} ) = 1-F ₂ (I _{max, 2} )

P ₃ = P (I _1M ≥I _{setI, 1} ) = 1-F ₃ (I _{setI, 1} )

P ₄ = P (I _2M ≥I _{setI, 2} ) = 1-F ₄ (I _{setI, 2} )

Quantitatively evaluate the safety of the closed-loop operation of the medium-voltage distribution network according to the magnitude of the over-limit probability of the multiple closing currents: in response to the over-limit probability of the multiple closing currents being less than the preset threshold, determine The loop closing operation is safe; in response to that the limit crossing probabilities of the multiple loop closing currents are not all less than the preset threshold, it is determined that the loop closing operation is not safe.

As shown in FIG. 3, the method provided in this application for evaluating the operation safety of the medium-voltage distribution network in a closed loop includes:

Step A: State estimation of the high-voltage distribution network;

In step A, based on the topology structure, equipment parameters, and real-time operating data of the ring network before the ring is closed, the state of the high-voltage distribution network is estimated, and the 10kV busbars at the ends of the two ring feeders are obtained. Voltage amplitude and phase angle.

specifically:

Since the 10kV bus node at the end of the feeder is generally selected as the reference node when calculating the combined current of the medium-voltage distribution network, the voltage amplitude and phase angle of the 10kV bus on both sides of the contact switch must be known. At present, the measurement system of the medium voltage distribution network can collect the voltage amplitude, but its phase angle information cannot be obtained. In view of this, based on the topology, equipment parameters, and real-time operating data of the transmission network and high-voltage distribution network in the ring network, the weighted least squares (WLS) rule is used for high-voltage power distribution. State estimation of the network. The weighted least squares method has the advantages of simple model and small calculation. Therefore, through the state estimation of the high-voltage distribution network, the voltage amplitude and phase angle of the 10kV bus at the leading end of the combined feeder can be obtained, and at the same time, the bad measurement data can be identified and corrected.

Step B: Calculate the multi-order semi-invariants of the input variables;

In step B, the ratio of the active load at each load point of each combined feeder to the active power at the feeder end and the ratio of the reactive load at each load point at the combined feeder to the reactive power at the feeder end are selected as The input variables are calculated based on the historical load data of the 96-point daily load of each year at each load point, and calculate the multi-order semi-invariants of each input variable in the season and period in which the loops are located.

Specifically, before the medium voltage distribution network is closed, the real-time load at each load point is generally difficult to obtain. In view of this, the load value at each load point is regarded as a random variable, and the ratio of the active load at the load point to the active power at the feeder end and the reactive load at the load point at the combined loop feeder and the reactive power at the feeder end are selected. The power ratio is used as an input variable for probabilistic power flow calculations. When obtaining the multi-order semi-invariants of the input variables, the traditional method is to use numerical methods to obtain the variable's probability distribution function. However, the distribution functions of these input variables are unknown in engineering. To this end, this application uses a method of obtaining a multi-order semi-invariant of input variables based on historical load data. The description of this method is as follows:

Let the input variable k _P be the ratio of the active load at a load point to the active power at the head of the feeder. From the load point and the outgoing line's annual full-time 96-day daily load historical data, sort out the year-round discrete measured data of the input variable, and construct a multi-order semi-invariant sample set for obtaining the input variable k _P S. Considering that the load may show different characteristics in different seasons and different periods of the day, the sample set S is seasonal (for example, spring, summer, autumn, winter) and time period (for example, 00: 00-08: 00, 08: 00-18: 00, 18: 00-24: 00) is divided into 12 (4 × 3 = 12) sub-sample sets (S ₁ , S ₂ , ..., S ₁₂ ). Through the analysis of a certain sub-sample set, the multi-order semi-invariants of the input variable k _P in the season and period in which the loops are combined are obtained.

If there are N discrete historical data in the sub-sample set corresponding to the time of the loop, it can be expressed as {k _P1 , k _P2 , k _P3 , ..., k _PN }. First calculate its multi-order origin moment α _v :

Then from the relationship between the semi-invariant and the origin moment, calculate its multi-order semi-invariant γ _v :

among them,

Is the number of different combinations of j elements from v elements; α ₁ and α _j are the origin moments when v = 1 and v = j, respectively.

Step C: deterministic power flow calculation and loop current calculation;

In step C, combined with the topology information and operating data of the loop network, deterministic power flow calculation and loop current calculation are performed at the reference operating point to obtain the loop network system state variable reference value, loop current reference value, and coefficient matrix.

Specifically, the voltage difference between the two sides of the contact switch fracture is set as

for:

Assume

with

The initial currents at the front ends of the feeders on both sides of the loop, X = [θ ₁ , V ₁ , θ ₂ , V ₂ , ..., θ _n , V _n ] ^T are the state variables of the loop network system. According to the superposition theorem, the The effective values of the steady-state currents I ′ ₁ and I ′ _{2 at the} front ends of the feeders on both sides of the loop are:

Let I ₁ and I ₂ be

with

The effective values of the maximum inrush current I _1M and I _2M that may appear at the ends of the feeders on both sides of the feeder during the transient state of the loop are:

I _1M = I ₁ + 1.51I _c = g ₃ (X) (6)

I _2M = I ₂ + 1.51I _c = g ₄ (X) (7)

Let Z = [I ′ ₁ , I ′ ₂ , I _1M , I _2M ] ^T be the loop current variable, then the loop current equation is:

Z = g (X): (8)

Let K be the input variable of the probabilistic load flow calculation, which represents the ratio of the active and reactive power of each node to the active and reactive power of the line head; W represents the injected power of each node in the closed loop network system, then W = AK, where A is a diagonal matrix composed of active and reactive power at the ends of the two feeders. The power flow equation of a closed loop network system can be expressed as:

W = f (X): (9)

The input variable K is a random variable, which can be expressed as K ₀ + ΔK, where K ₀ is the reference value (or expected value) of the random variable K, which is the reference operating point of the closed loop network system, and ΔK is a random disturbance. Similarly, the loop system state variable X can be expressed as X ₀ + ΔX, X ₀ is the expected value of the loop network system state variable, and ΔX is random disturbance. The injected power W of each node can be expressed as W ₀ + ΔW, W ₀ is the expected value of the node injected power, and ΔW is a random perturbation corresponding to ΔX. The closed loop current variable Z can be expressed as Z ₀ + ΔZ, Z ₀ is the expected value of the closed loop current variable, and ΔZ is a random perturbation corresponding to ΔX.

The closed loop network system power flow equation (9) and closed loop current equation (8) are expanded by Taylor series at X ₀ and the higher-order terms are omitted. The linear relationship between ΔZ and ΔK is obtained:

Of which, the Jacobian matrix

Coefficient matrix

In this step C, first perform a deterministic power flow calculation at the reference operating point K ₀ (w ₀ = AK ₀ at this time) according to formula (9), and obtain the loop network system state variable X ₀ and the Jacobian matrix J ₀ ; Then, the closing current calculation is performed at X ₀ according to formula (8), and the closing current variable Z ₀ and the coefficient matrix G ₀ are obtained; finally, the conversion matrix T _{0 is obtained} .

Step D: Calculate a multi-order semi-invariant of the closing current;

In step D, the multi-order semi-invariant of the closing current is calculated from the multi-order semi-invariant of the input variable obtained in step B and the transformation matrix obtained in step C.

Specifically, in step B, a multi-order semi-invariant of the input variable has been obtained. Let ΔK ^(v) represent the v-order semi-invariant of the input variable.

Represents the coefficient matrix composed of the power of v of each element in T ₀ , then in this step D, the property of semi-invariant

Step E: Obtain a probability distribution of the closed loop current;

In step E, a Cornish-Fisher series expansion method is used to obtain the cumulative probability distribution of the closed-loop current from the multi-order semi-invariants of the closed-loop current.

Specifically, in this step E, the method for obtaining the cumulative probability distribution of the closed loop current by using the Cornish-Fisher series is as follows:

Let the cumulative distribution function of a closed loop current variable Z be F (z) and the standard normal distribution function be Φ (z). with

That is, z (α) = F ^-1 (α),

Then z (α) and

Meet the following relationships:

Where g _v is the v-order normalized semi-invariant of the closed-loop current variable Z, that is, if the v-order semi-invariant of the random variable Z is γ _v and the standard deviation is σ, then

From formula (11), the cumulative probability distribution function of the closed-loop current can be obtained from the multi-order semi-invariants.

Step F: Assess the overrun probability of multiple closing currents;

In step F, according to the safe closing conditions, the maximum allowable current carrying capacity of the feeder and the current protection setting value are respectively used as the limit values, and the probability of exceeding the limit of each closing current is calculated, and the safety of the closing operation is evaluated.

Specifically, the closing loop current variable Z includes: the steady-state current effective values I ′ ₁ and I ′ ₂ of the feeder heads on both sides after the closing loop, and the maximum inrush current effective values I _1M and I _2M that may occur at the ends of the feeder lines on both sides. Let the cumulative distribution functions of the variables I ′ ₁ , I ′ ₂ , I _1M , and I _2M be F ₁ (x), F ₂ (x), F ₃ (x), and F ₄ (x), on both sides of the junction The maximum allowable current carrying capacity of the feeder are I _{max, 1} and I _{max, 2} respectively _, and the setting values of the current I protection on both sides are I _{setI, 1} and I _{setI, 2} , respectively.

P ₁ = P (I ′ ₁ ≥I _{max, 1} ) = 1-F ₁ (I _{max, 1} )

P ₂ = P (I ′ ₂ ≥I _{max, 2} ) = 1-F ₂ (I _{max, 2} )

P ₃ = P (I _1M ≥I _{setI, 1} ) = 1-F ₃ (I _{setI, 1} )

P ₄ = P (I _2M ≥I _{setI, 2} ) = 1-F ₄ (I _{setI, 2} )

Finally, according to the above four types of threshold crossing probability, the safety of the loop closing operation of the medium-voltage distribution network can be quantitatively evaluated. If the threshold crossing probability is less than 5%, the safety of the loop closing operation can be considered high. The safety of ring operation cannot be guaranteed, and it is recommended to consider the ring closing after adjusting the corresponding operating mode.

In summary, the present application provides a method for evaluating the safety of the closing operation of a medium-voltage distribution network. The safety of the closing operation is quantified by calculating the out-of-limit probability of the steady-state current and the transient inrush current of the closing circuit. The evaluation solves the problem that the real-time load data at the load point cannot be obtained in practice. Applying this method to actual production can quickly evaluate the safety of closing operation before closing the ring, and provide support for the operator to make the decision of closing the ring.

This application considers the load point load on the feeder loop feeder as a random variable. Based on the probabilistic load flow theory, the out-of-limit probability of the steady-state current and the transient inrush current is obtained and the safety of the joint loop operation is evaluated. The safety of closing operation of the medium-voltage distribution network; solving the problem that the real-time load data of each load point in the feeder cannot be obtained in practice; and solving the probability distribution of the traditional input method that requires a known input variable to obtain a semi-invariant Problems with functions.

Compared with related technologies, the beneficial effects of this application are:

(1) This application can quantitatively evaluate the safety of the close-loop operation of a medium-voltage distribution network by calculating the close-loop steady-state current and transient inrush current generated by the close-loop operation. The evaluation results have a theoretical basis and can be The operator's loop closing operation provides a reference.

(2) This application uses the load point load on the feeder loop as a random variable, and obtains the probability distribution characteristics of the steady-state current and transient inrush current based on the probability flow theory, which solves the problem that real-time load data at the load point cannot be obtained. .

(3) The present application obtains a multi-order semi-invariant of the input variable by analyzing the historical load data, and solves the problem that the traditional numerical method needs to know the probability distribution function of the input variable when obtaining the semi-invariant.

Claims

A method for assessing the operation safety of the medium-voltage distribution network in a closed loop operation, including:

Based on the obtained topology and equipment parameters of the loop-to-loop network, and real-time operating data of the loop-to-loop network before looping, the state of the high-voltage distribution network is estimated, and the voltage amplitudes of the 10kV buses at the ends of the two loop-feeding feeders are obtained With phase angle

Selecting as input variables the ratio of the active load at each load point of each combined loop feeder to the active power of the feeder end and the ratio of the reactive load at each load point of the combined loop feeder to the reactive power at the feeder end. Calculate the multi-order semi-invariants of each input variable based on the historical load data of each load point;

Based on the voltage amplitudes and phase angles of the 10kV busbars at the ends of the two loop feeders, perform deterministic power flow calculation and loop current calculation at the reference operating point to obtain the loop network system state variable reference value and loop current reference value. Obtaining a conversion matrix based on the reference value of the state loop network system variable and the reference value of the loop current;

Calculating a multi-order semi-invariant of a loop current according to the multi-order semi-invariant of each input variable and the conversion matrix;

Obtaining a cumulative probability distribution of the closing current according to a multi-order semi-invariant of the closing current;

The maximum allowable current carrying capacity of the feeder and the setting value of the current protection are used as limit values, and the limit crossing probability of each loop current is calculated based on the cumulative probability distribution, and determined when a plurality of the limit crossing probabilities are less than a preset threshold. Closed loop operation is safe.
The method according to claim 1, wherein the state estimation of the high-voltage distribution network is obtained based on the obtained topology structure, equipment parameters of the ring-combined network, and real-time operation data of the ring-combined network before the ring-combination, and obtains The voltage amplitude and phase angle of the 10kV bus bars at the ends of the two feeder loops include:

Based on the obtained topology structure, equipment parameters, and real-time operating data of the combined network before it is combined, the weighted least squares criterion is used to estimate the state of the high-voltage distribution network, and the two ends of the combined feeder are obtained. 10kV bus voltage amplitude and phase angle, and identify and correct the bad measurement data.
The method according to claim 1, wherein the ratio of the active load at each load point to the active power at the head of the feeder and the reactive load at each load point on the feeder and feeder are selected. The ratio of the reactive power at the head end is used as the input variable. Based on the historical load data of each load point, a multi-order semi-invariant of each input variable is calculated, including:

Let the input variable k P be the ratio of the active load at any load point to the active power at the head of the feeder; from the annual load historical data of 96 points throughout the year at the load point and the outgoing line, sort out the year-round dispersion of the input variable. Measure the data, and construct a multi-order semi-invariant sample set S for obtaining the input variable k P ;

The sample set S is divided into a plurality of sub-sample sets according to the season and period of the ring-turning moment. By analyzing any one of the sub-sample sets, the seasonal sum of the input variable k P at the ring-turning moment is obtained. Multi-order semi-invariants over time.
The method according to claim 3, wherein the obtaining a multi-order semi-invariant of the input variable k P in a season and a time period at which the loops are closed comprises:

In response to the N discrete historical data sets {k P1 , k P2 , k P3 , ..., k PN } in the sub-sample set corresponding to the time of the loop, calculate a multi-order origin moment α v :

According to the relationship between the semi-invariant and the origin moment, calculate the multi-order semi-invariant γ v :

Among them, α 1 and α j are the origin moments when v = 1, v = j, and v represents the order;
Is the number of different combinations of j elements from v elements.
The method according to claim 1, wherein, based on the voltage amplitudes and phase angles of the 10kV busbars at the leading ends of the two closed loop feeders, a deterministic power flow calculation and a closed loop current calculation are performed at a reference operating point to obtain a closed loop current The loop network system state variable reference value and the loop closing current reference value. Based on the loop closing network system state variable reference value and the loop closing current reference value, a conversion matrix is obtained, including:

Based on the voltage amplitudes and phase angles of the 10kV busbars at the ends of the two loop feeders, deterministic power flow calculations are performed at the reference operating point K 0 according to the loop network system power flow equation to obtain the loop network system state variable reference value X 0 and the Jacobian matrix J 0 ;

Calculate the closing current according to the closing current equation at the reference value X 0 of the closing network system state, and obtain the closing current reference value Z 0 and the coefficient matrix G 0 ;

The Taylor series expansion of the closed loop network system power flow equation and the closed loop current equation are omitted and higher-order terms are omitted to obtain the random disturbance of the reference operating point K 0 and the reference value of the closed loop network system state. A linear relationship between the random perturbation of X 0 and the random perturbation of the reference loop current reference value Z 0 , and a conversion matrix T 0 is obtained according to the linear relationship;

Wherein, the loop current equation is Z = g (X), Z is a loop current variable, and X is a state variable of the loop network system; the loop flow system equation is W = f (X), where W is Injected power at each load point in a closed loop network system, W 0 = AK 0 , A is a diagonal matrix composed of active power and reactive power at the ends of the two feeders.
The method according to claim 5, wherein the process of obtaining the loop current equation comprises:

Let the voltage difference between the two sides of the contact switch fracture be
The total impedance of the closed loop is Z ∑ , then the closed loop steady-state circulation current
for:

Assume
with
The initial currents at the front ends of the feeders on both sides of the loop, respectively. The loop network system state variable X = [θ 1 , V 1 , θ 2 , V 2 , ..., θ n , V n ] T , according to the superposition theorem, The effective values of the steady-state currents I ′ 1 and I ′ 2 at the ends of the feeders on both sides after the loop are:

Let I 1 and I 2 be
with
The effective values of the maximum inrush current I 1M and I 2M appearing at the ends of the feeders on both sides during the transient state of the closed loop are:

I 1M = I 1 + 1.51I c = g 3 (X)

I 2M = I 2 + 1.51I c = g 4 (X)

Assuming that the loop closing current variable Z = [I ′ 1 , I ′ 2 , I 1M , I 2M ] T , the loop closing current equation is:

Z = g (X)

The process of obtaining the power flow equation of the loop network system includes: setting K as an input variable of the probabilistic power flow calculation; W represents the injected power of each node in the loop network system, then W = AK; the power flow equation of the loop network system can be expressed for:

W = f (X)

Wherein, the input variable K is a random variable, which can be expressed as K 0 + ΔK, K 0 is a reference value of the random variable K, and is a reference operating point of a closed loop network system; ΔK is a random disturbance; and a closed loop network system state variable X It can be expressed as X 0 + ΔX, X 0 is the reference value of the state variable of the loop network system, ΔX is the random perturbation; the injected power W of each node can be expressed as W 0 + ΔW, W 0 is the reference value of the node injected power, and ΔW is Random perturbation corresponding to ΔX; closed loop current variable Z can be expressed as Z 0 + ΔZ, Z 0 is the reference value of closed loop current variable, and ΔZ is a random perturbation corresponding to ΔX;

The linear relationship between the ΔZ and the ΔK:

Of which, the Jacobian matrix
Coefficient matrix
The method according to claim 1, wherein said calculating a multi-order semi-invariant of a loop current based on said multi-order semi-invariant of each input variable and said conversion matrix comprises:

Let ΔK (v) be the v-order semi-invariant of the input variable,
A matrix of coefficients representing the power of v of each element in T 0 , then the property of semi-invariants
Find the v-order semi-invariant ΔZ (v) of the combined loop current variable.
The method according to claim 1, wherein the obtaining the cumulative probability distribution of the closing current according to the multi-order semi-invariant of the closing current comprises:

Use the Cornish-Fisher series to find the cumulative probability distribution of the closed loop current, including:

Let the cumulative distribution function of the loop current variable Z be F (z) and the standard normal distribution function be Φ (z).
which is

Then z (α) and
Meet the following relationships:

Among them, g v is a v-order normalized semi-invariant of the closed-loop current variable Z, v = 1, 2, 3 ..., and the v-order semi-invariant of the closed-loop current variable Z is γ v with a standard deviation of σ,
A cumulative probability distribution function of the closing current is obtained based on the multi-order semi-invariant of the closing current.
The method according to claim 1, wherein the maximum allowable ampacity of the feeder and the current protection setting value are used as limit values, and the probability of exceeding the limit of each loop current is calculated based on the cumulative probability distribution. When the probability of exceeding the limit is less than the preset threshold, the loop closure operation is determined to be safe, including:

The closing loop current variable Z includes the steady-state current effective values I ′ 1 and I ′ 2 of the feeder heads on both sides after the closing loop and the maximum rush current effective values I 1M and I 2M appearing on the feeder ends on both sides;

Let the cumulative distribution functions of the variables I ′ 1 , I ′ 2 , I 1M , and I 2M be F 1 (x), F 2 (x), F 3 (x), and F 4 (x), on both sides of the junction The maximum allowable current carrying capacity of the feeder are I max, 1 and I max, 2 respectively , and the setting values of the current I protection on both sides are I setI, 1 and I setI, 2 , respectively.

P 1 = P (I ′ 1 ≥I max, 1 ) = 1-F 1 (I max, 1 )

P 2 = P (I ′ 2 ≥I max, 2 ) = 1-F 2 (I max, 2 )

P 3 = P (I 1M ≥I setI, 1 ) = 1-F 3 (I setI, 1 )

P 4 = P (I 2M ≥I setI, 2 ) = 1-F 4 (I setI, 2 )

Quantitatively evaluate the safety of the closed-loop operation of the medium-voltage distribution network according to the magnitude of the over-limit probability of the multiple closing currents: in response to the over-limit probability of the multiple closing currents being less than the preset threshold, determine The loop closing operation is safe; in response to that the multiple crossing currents of the multiple loop closing currents are not less than the preset threshold, it is determined that the loop closing operation is not safe.