CN110212536A - A kind of medium voltage distribution network interconnection switch state identification method - Google Patents

Abstract

The invention discloses a method for identifying the state of a tie switch in a medium-voltage distribution network, which includes the following steps: step S1, calculating a feasible topology, the voltage at the head end of the medium-voltage distribution network, and the active and reactive power of each node injected into the load are known, The line impedance between nodes is known, and the state of the switch on the tie switch branch is calculated separately, and a variety of tie switch state combinations are obtained, and the topology analysis is quickly performed on the tie switch state combination, and the combination that forms a loop is eliminated to obtain a feasible tie switch group; Step S2, for the feasible state of the tie switch group in step S1, or a certain feasible topology strategy, carry out the power flow calculation based on the forward-back generation; step S3, carry out the posterior probability calculation based on the recursive Bayesian method based on the power flow calculation result . The BRA method of the present invention based on the calculation of the power flow of the forward and future generations has advantages over the BRA method based on the WLS state estimation in terms of the number of iterations and calculation time.

Description

A method for state identification of tie switches in medium voltage distribution network

technical field

The invention belongs to the technical field of medium-voltage power distribution systems, and in particular relates to a method for identifying the state of a distribution network tie switch in a medium-voltage distribution network based on load measurement data and forward and subsequent power flow calculations.

Background technique

The medium voltage of the power system refers to the 22/10/6.3kV voltage level; the medium voltage distribution network refers to the medium voltage line and related switches drawn from the medium voltage busbar of the high voltage substation (voltage level is 220/110/35kV) The network connected to the device and the user's electrical equipment.

The medium-voltage distribution network adopts a closed-loop design and an open-loop operation mode. By installing a large number of line section switches, branch switches, contact switches or setting up switching stations and ring stations, the distribution network can be operated according to maintenance, failure, and economical operation. Various open-loop operation modes are adopted to ensure the quality and reliability of power supply; the closing and unlinking operations of the distribution network are very frequent, accounting for 48.1% of the daily operation items of dispatching.

The normally open switches of switching stations and ring stations are also collectively referred to as tie switches, so these tie switches become the most important control objects to change the operation mode of the medium-voltage distribution network, and the state of the tie switch becomes the determining factor of the dynamic topology of the distribution network.

The static topology of the medium-voltage distribution network refers to the connection relationship between all the nodes in which all the contact switches are in the normally open state according to the design state; the dynamic topology refers to the connection relationship between the nodes formed by all the switches according to the real switch state; A feasible dynamic topology is that the actual connection relationship formed cannot form a loop.

At present, the load nodes (also terminal nodes) of the medium-voltage 10kV distribution network, including dedicated distribution transformers, public distribution transformers, and distributed power sources connected to large user lines, are all equipped with data collection systems including smart meters. terminal, thereby providing electrical and power data collected at 15-minute intervals. However, due to the lag of distribution automation, it is difficult to fully obtain the switch status (including segment, branch, contact switch or switch station, and switch position in the ring network) that determines the dynamic topology of the medium-voltage distribution network. , so for a long period of time in the future, the medium-voltage 10kV distribution network can only obtain all the injected power and part of the switch position, and the lack of dynamic topology makes it difficult for a large number of advanced applications to be applied in actual engineering, so based on It has become an urgent problem to be solved urgently by collecting data such as load power measurement and incomplete network switch position to identify the dynamic topology of medium voltage distribution network.

Contents of the invention

The technical problem to be solved by the present invention is to provide a method for identifying the state of tie switches in a medium-voltage distribution network based on load measurement data and calculation of forward and subsequent power flows to identify the opening and closing states of the remaining switches that cannot be automatically monitored.

In order to solve the above technical problems, the present invention adopts the following technical solution: a method for identifying the state of a tie switch in a medium-voltage distribution network, comprising the following steps:

Step S1, calculate the feasible topology, the head-end voltage in the medium-voltage distribution network, the active and reactive power injected into each node are known, the line impedance between nodes is known, and the states of the switches on the tie switch branch are calculated respectively, and multiple A kind of tie switch state combination, quickly carry out topology analysis to the tie switch state combination, eliminate the combination that forms a loop, obtain the feasible tie switch combination, a certain topology strategy in the feasible tie switch combination is expressed as η _i ;

Step S2, for the feasible state of the tie switch group in step S1, or a certain feasible topology strategy η _i , carry out the power flow calculation based on the forward-back generation, first make an assumption, assuming that the voltage of the whole network is the rated voltage, according to the load power by Calculate from the end to the beginning step by step, only calculate the power loss in each component without calculating the node voltage, obtain the current and power loss on each branch, and obtain the power at the start end accordingly, which is a back-substitution process; then according to the given The initial voltage and the obtained initial power are calculated step by step from the initial end to the end voltage drop, and the voltage of each node is obtained. This is the forward process, and the above process is repeated until the power deviation of each node meets the allowable conditions;

Step S3, performing posterior probability calculation based on the recursive Bayesian method based on the power flow calculation results.

Optionally, the specific process of step S2 is as follows:

1) First search for the end nodes of each branch, and use this as a starting point to calculate the current on the branch according to Kirchhoff’s current law:

2) Find each node with the terminal node as the parent node, reversely push back the current of each branch one by one, according to Kirchhoff’s current law, the injected current of the node is equal to formula (2) and the outgoing current of the node The sum can be expressed as:

3) The branch currents of all branches can be obtained from equations (2) and (3), and then the known voltage of the root node is used. At this point, the parent node and the child node exchange positions, and each The voltage at the node can be expressed as:

Among them, i is the new parent node, j is the child node, and Z is the impedance of the branch between i and j. So far, a complete process of pushing forward and descending is completed.

4) Calculate the voltage difference of each node after each iteration:

Then the maximum value of the voltage difference is

5) The conditions for judging convergence are:

Among them, k is the number of iterations. When formula (6) is satisfied, the cycle ends and the voltage calculation result is output, otherwise, steps 1)-5) are repeated until the convergence condition is reached.

Optionally, step S3 is specifically as follows:

Among them, N _models is the number of topological policies in the policy library, k is the number of iterations, is the set of error vectors for each topological strategy, p(η _i丨ε ^k-1 ) is the prior probability, is the error probability of the i-th topological strategy under the condition that the given η _i topological strategy is correct,

The prior probability can be expressed as:

Among them, C _f,i is a diagonal matrix, and its diagonal elements are the inverse of the variance of each element of the error vector, and the error vector It can be obtained by the following formula:

where z _real is the real measurement vector, is the state estimate of the i-th topology policy after the k-th iteration,

In the solution algorithm, by default, the initial probability of each topology strategy is evenly distributed as 1/N _models , and N _models are the total number of possible topological strategies. The algorithm uses (8)(9 ) formula, constantly find out the probability corresponding to the new topology strategy, and finally determine the required target topology according to the probability.

Optionally, it is stipulated that the topology strategy with probability p _i >0.5 is determined as the required target topology.

The present invention utilizes the known static topology, and all load nodes have measurement data, and the opening and closing states of some switches can be automatically monitored to identify the opening and closing states of the remaining switches that cannot be automatically monitored.

Adopting the BRA method based on the calculation of forward and future power flow, the BRA method based on the calculation of forward and future power flow has advantages over the BRA method based on WLS state estimation in terms of the number of iterations and calculation time.

The specific technical solutions and beneficial effects of the present invention will be described in detail in the following specific embodiments with reference to the accompanying drawings.

Description of drawings

The present invention will be further described below in conjunction with accompanying drawing and specific embodiment:

Figure 1 is an IEEE33 node distribution network as a case.

Detailed ways

The purpose of the present invention is to identify the opening and closing states of the remaining switches that cannot be automatically monitored when the known static topology is used, all load nodes have measurement data, and the opening and closing states of some switches can be automatically monitored.

A method for identifying the state of tie switches in medium-voltage distribution networks based on load measurement data and forward-generation power flow calculations, referred to as "BRA method based on forward-generation power flow calculations" for short.

Specific steps are as follows.

Step 1. Calculate Feasible Topology

The distribution network only allows a short-lived ring state, so only topological combinations without rings are considered.

It can be seen from Figure 1 that each tie switch will have a loop when it is closed, and only a unique loop will be formed when the other tie switches are open, the loop L ₁ -L marked in the figure ₅ . When actually formulating the operation mode, when a tie switch is closed, there must be a line segment or branch switch to be opened, so that the loop is unlooped, and all nodes are still charged; assuming that only one in the loop is associated with the tie switch The dual switch of , and assumed to be S' _i (i=1～5), that is:

Where K represents the state of the switch, the above means that when K(S _i )=1, K(S' _i )=0; when K(S _i )=0, K(S' _i )=1. According to this assumption, there are 10 switches in Figure 1, but the state of all dual switches is determined by the state of their corresponding tie switches, so the state of the tie switch determines the topology of the distribution network.

The present invention selects the IEEE33 node distribution network in Fig. 1 as an example, wherein the lines of five tie switches are marked by dotted lines (S1-S5). In this medium-voltage distribution network, the head-end voltage (node 0) is 10.5kV, the active and reactive power injected into each node is known, and the line impedance between nodes is known. state.

There are 5 tie switches in Fig. 1, theoretically there are 2 ⁵ = 32 kinds of tie switch state combinations; however, when 2 tie switches are closed, a new loop may be formed; therefore, for 32 combinations, it is possible to Quickly carry out topology analysis, eliminate the combinations that form loops, and obtain the feasible tie switch combinations as shown in Table 1.

The feasible state of the tie switch group in Table 1 is called the feasible topological strategy library (model bank), in which a certain topology strategy is expressed as η _i , and N _models is the total number of topology strategies in the strategy bank, that is, the feasible state of the tie switch group the number of groups.

Table 1 The state topology strategy table of the tie switch group without loops in the IEEE33 node distribution network

Step 2. Forward and backward power flow calculations for all feasible topologies

For the above feasible state of the tie switch group, or a certain feasible topology strategy η _i , the power flow calculation based on forward and backward generation is performed.

First make assumptions, assuming that the voltage of the whole network is the rated voltage, calculate the load power step by step from the end to the beginning, only calculate the power loss in each component without calculating the node voltage, and obtain the current and power loss on each branch, and According to this, the initial power is obtained, which is the back-substitution process; then, according to the given initial voltage and the obtained initial power, the voltage drop is calculated step by step from the initial to the end, and the voltage of each node is obtained, which is the forward process. The above process is repeated until the power deviation of each node satisfies the allowable condition.

Specific steps are as follows:

1) First search for the end nodes of each branch, and use this as a starting point to calculate the current on the branch according to Kirchhoff’s current law:

2) Find each node with the terminal node as the parent node, reversely push back the current of each branch one by one, according to Kirchhoff’s current law, the injected current of the node is equal to formula (2) and the outgoing current of the node The sum can be expressed as:

3) The branch currents of all branches can be obtained from equations (2) and (3), and then the known voltage of the root node is used. At this point, the parent node and the child node exchange positions, and each The voltage at the node can be expressed as:

Among them, i is the new parent node, j is the child node, and Z is the impedance of the branch between i and j. So far, a complete process of pushing forward to offspring has been completed.

4) Calculate the voltage difference of each node after each iteration:

Then the maximum value of the voltage difference is

5) The conditions for judging convergence are:

Among them, k is the number of iterations. When formula (6) is satisfied, the cycle ends and the voltage calculation result is output, otherwise, steps 1)-5) are repeated until the convergence condition is reached.

Step 3. Topology identification based on forward-back generation power calculation results

For the feasible state of the tie switch group in Table 1, or a certain feasible topology strategy η _i , the posterior probability calculation based on the recursive Bayesian method (Recursive Bayesian Approach, RBA) is performed based on the power flow calculation results, specifically:

Among them, N _models is the number of topological policies in the policy library, k is the number of iterations, is the set of error vectors for each topological strategy, p(η _i丨ε ^k-1 ) is the prior probability, is the error probability of the i-th topological strategy under the condition that the given η _i topological strategy is correct.

The prior probability can be expressed as:

Among them, C _f,i is a diagonal matrix, and its diagonal elements are the inverse of the variance of each element of the error vector, and the error vector It can be obtained by the following formula:

where z _real is the real measurement vector, is the estimated value of the state of the i-th topology strategy after the k-th iteration. In the solution algorithm, by default, the initial probability of each topological strategy is evenly distributed as 1/N _models , and N _models are the total number of various possible topological strategies. The algorithm uses equations (8) and (9) in an iterative way to continuously find out the probability corresponding to the new topological strategy. Finally, it is stipulated that the topological strategy that exceeds 50%, that is, the probability p _i >0.5, is the required goal topology.

Take the IEEE33 node distribution network as an example. Take the feasible topology combination state or topology strategy in Table 1 as the correct line topology for each test, use the power flow calculation method to obtain the estimated value of each iteration, and use the measured value and the estimated value to obtain the error according to formula (9). Then calculate the prior probability under this error according to formula (8).

The similar method that adopts technical solution step 3 of the present invention recursive Bayesian method commonly adopts state estimation based on weighted least squares method (WLS), and uses each step state estimation error as formula (9), and then obtains similar judgment; Method is called for short "BRA method based on WLS state estimation".

Take the IEEE33 node distribution network as an example. Taking the feasible topology combination state or topology strategy in Table 1 as the correct line topology for each experiment, the probability difference of all topology groups in the first few iterations is not obvious, but in the next few iterations, the probability is compared with Other topological strategies increase more obviously, and finally tend to 1, indicating that this combination is the correct combination of tie switches.

Table 2 gives the BRA method based on the calculation of power flow forward and backward.

Table 2. Comparison results of RBA based on forward power flow calculation and RBA based on WLS state estimation

From Table 2, it can be concluded that the BRA method based on the calculation of forward and future generation power flow has advantages over the BRA method based on WLS state estimation, both in terms of the number of iterations and calculation time.

The above description is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Those skilled in the art should understand that the present invention includes but is not limited to the content described in the above specific embodiment. Any modifications that do not depart from the functional and structural principles of the present invention will be included in the scope of the claims.

Claims (4)

1. A method for identifying the state of a tie switch in a medium-voltage distribution network, characterized in that it comprises the steps: Step S1, calculate the feasible topology, the head-end voltage in the medium-voltage distribution network, the active and reactive power injected into each node are known, the line impedance between nodes is known, and the states of the switches on the tie switch branch are calculated respectively, and multiple A kind of tie switch state combination, quickly carry out topology analysis to the tie switch state combination, eliminate the combination that forms a loop, obtain the feasible tie switch combination, a certain topology strategy in the feasible tie switch combination is expressed as η _i ; Step S2, for the feasible state of the tie switch group in step S1, or a certain feasible topology strategy η _i , carry out the power flow calculation based on the forward-back generation, first make an assumption, assuming that the voltage of the whole network is the rated voltage, according to the load power by Calculate from the end to the beginning step by step, only calculate the power loss in each component without calculating the node voltage, obtain the current and power loss on each branch, and obtain the power at the start end accordingly, which is a back-substitution process; then according to the given The initial voltage and the obtained initial power are calculated step by step from the initial end to the end voltage drop, and the voltage of each node is obtained. This is the forward process, and the above process is repeated until the power deviation of each node meets the allowable conditions; Step S3, performing posterior probability calculation based on the recursive Bayesian method based on the power flow calculation results. 2. A method for identifying the state of a tie switch in a medium-voltage distribution network according to claim 1, wherein the specific process of step S2 is as follows: 1) First search for the end nodes of each branch, and use this as a starting point to calculate the current on the branch according to Kirchhoff’s current law: 2) Find each node with the terminal node as the parent node, reversely push back the current of each branch one by one, according to Kirchhoff’s current law, the injected current of the node is equal to formula (2) and the outgoing current of the node The sum can be expressed as: 3) The branch currents of all branches can be obtained from equations (2) and (3), and then the known voltage of the root node is used. At this point, the parent node and the child node exchange positions, and each The voltage at the node can be expressed as: Among them, i is the new parent node, j is the child node, and Z is the impedance of the branch between i and j. So far, a complete process of pushing forward and descending is completed. 4) Calculate the voltage difference of each node after each iteration: Then the maximum value of the voltage difference is 5) The conditions for judging convergence are: Among them, k is the number of iterations. When formula (6) is satisfied, the cycle ends and the voltage calculation result is output, otherwise, steps 1)-5) are repeated until the convergence condition is reached. 3. A method for identifying the state of a tie switch in a medium-voltage distribution network according to claim 1, wherein step S3 is specifically as follows: Among them, N _models is the number of topological policies in the policy library, k is the number of iterations, is the set of error vectors for each topological strategy, p(η _i丨ε ^k-1 ) is the prior probability, is the error probability of the i-th topological strategy under the condition that the given η _i topological strategy is correct, The prior probability can be expressed as: Among them, C _f,i is a diagonal matrix, and its diagonal elements are the inverse of the variance of each element of the error vector, and the error vector It can be obtained by the following formula: where z _real is the real measurement vector, is the state estimate of the i-th topology policy after the k-th iteration, In the solution algorithm, by default, the initial probability of each topology strategy is evenly distributed as 1/N _models , and N _models are the total number of possible topological strategies. The algorithm uses (8)(9 ) formula, constantly find out the probability corresponding to the new topology strategy, and finally determine the required target topology according to the probability. 4. A method for identifying the state of a tie switch in a medium-voltage distribution network according to claim 3, characterized in that it is specified that a topology strategy with probability p _i >0.5 is determined as the required target topology.