CN113131477A - Power grid calculation method based on feeder calculation - Google Patents

Abstract

The invention discloses a power grid calculation method based on feeder calculation, which comprises the following steps: and processing all islands according to the feeder line adjacent queue and the zone queue to form a calculation network, optimizing in stages based on the calculation network, and outputting a calculation result. The invention relates to a power grid calculation method based on feeder calculation, which uses a section as a minimum unit, arranges islands by an optimal subsection switch or a suboptimal subsection switch, can obtain a calculation network formed by all the islands after the feeder queue and the section queue are processed, processes different operation modes aiming at the calculation network, and outputs a calculation result. The method can be used for calculating the power grid efficiently and accurately, and particularly has a good processing effect on the looped network.

Description

Power grid calculation method based on feeder calculation

Technical Field

The invention relates to the field of power grid optimization, in particular to a power grid calculation method based on feeder calculation.

Background

The optimization calculation of the power grid is a multivariable and multi-constraint nonlinear programming problem, which is always a hotspot and a difficulty of research, and related calculation methods are dozens of types, and an accurate solution is not provided.

Aiming at the optimization calculation of the 10kV radiation type power distribution network, firstly, the reactive compensation of the low-voltage side of the distribution transformer and secondly, the pole-on compensation of the feeder line are realized, the two aspects are mutually influenced, and the final effect of the optimization calculation of the power grid is determined.

Disclosure of Invention

In order to solve the defects and shortcomings in the prior art, the invention provides a power grid calculation method based on feeder calculation, which comprises the following steps.

A. All islands are processed according to feeder adjacency queues and sector queues to form a computational network.

B. Optimizing in stages based on the computing network, and outputting a computing result.

Preferably, all islands are processed according to feeder adjacency queues and zone queues to form a computational network. Comprises the following steps.

(1) And acquiring a calculation feeder set consisting of the feeder lines and the junctor thereof.

(2) A feeder adjacency queue and a segment queue are initialized.

(3) Calculating the distribution of the feeder line moisture collection flow; and (4) counting the total load and the transfer margin of each feeder line.

(4) And taking a feeder line, adding all trunk line sections of the feeder line into the section queue, and adding all adjacent feeder lines of the feeder line into the adjacent queue.

(5) A block is fetched from the block queue and the fetched block is removed from the block queue.

(6) And opening the downstream boundary switch of the extracted section, and taking the downstream residual network formed by the downstream boundary switch as a subset of the whole feeder line.

(7) And searching a downstream residual network to obtain all interconnection switches and all load points connected in the network, respectively storing the interconnection switches and all load points into a switch structure array and a load structure array, and sequencing the switch structure arrays, wherein the margin is from high to low. And simultaneously storing the downstream residual network into a computation island queue.

(8) And (5) taking one island from the computation island queue for computation, and if all the islands are processed, turning to (15).

(9) And sequentially taking two interconnection switches from the switch structure array as an interconnection switch combination, and calculating the optimal disconnection section switch. Two tie switches are taken in proper order as tie switch combination in the follow switch structure array, calculate the optimal disconnection section switch, include: and closing the interconnection switch combination, and determining the section switch for splitting the island as an optimal disconnection section switch by calculating the optimal ring current of two interconnection switches in the interconnection switch combination.

(10) If there are no feasible sectionalizers as the optimal disconnection sectionalizers, removing the tie switch combination from the switch structure array, and turning to (9); if all the interconnection switches are circulated completely, turning to (12); otherwise, the process goes to (11).

(11) Opening the found optimal section switch, respectively carrying out power flow verification on the two split islands, and stopping searching and utilizing the interconnection switch combination to carry out switching if no out-of-limit and low voltage exist; or other interconnection switch combinations can be continuously searched, the transfer indexes are compared finally, the switch combination with the highest transfer index is selected as the optimal transfer path, the id of the optimal section switch is recorded, and the switch (15) is transferred.

(12) The section switch is recorded into a suboptimal section switch sequence by changing the opening position when the section switch is opened, calculating the switch index.

(13) And if all the switch structure arrays finish the circulation, selecting the section switch with the highest transfer index as the suboptimal section switch.

(14) And opening the suboptimum section switch to form two new islands, adding the new islands into the computation island queue, removing the original islands, and turning to (8).

(15) The computational network consisting of all islands is output.

Preferably, the calculation result is output based on the calculation network staged optimization. Comprises the following steps.

(16) And calculating initial load flow of the feeder based on the calculation feeder set, judging whether the calculation network is a looped network with more than one power supply point, and performing reactive power optimization solution on the looped power distribution network (24).

(17) Taking a feeder line from the feeder line set, and turning to (25) if all the feeder line sections are completely circulated; otherwise, setting the maximum operation mode for calculation.

(18) And entering the first stage of optimization. And taking all load points on the feeder line, and calculating the power factor of the load points. And if the power factor of the load point does not reach the preset range, calculating the low-voltage reactive compensation quantity of the load point and writing the result into a reactive compensation result array.

(19) And (5) performing load flow calculation, if no voltage out-of-limit exists, turning to (21), and otherwise, entering the second-stage optimization.

(20) And optimizing in the second stage. And calculating the secondary reactive moment or the net loss micro-increment rate. And selecting the maximum reactive moment for compensation, and calculating the compensation capacity.

(21) And calculating by taking the minimum operation mode.

(22) And calculating the load flow, and judging whether reactive power is reversely delivered.

(23) If reactive power is fed back, reducing the reactive compensation capacity, and turning to (22); otherwise go to (17).

(24) And carrying out reactive power optimization solution.

(25) And outputting a calculation result.

Preferably, the performing reactive power optimization solution includes.

S1, data are recorded. Inputting network parameters of the dynamic power topology model, wherein the network parameters comprise node information, branch information and the like; and recording the real-time telemetering remote communication quantity, including the active and reactive output values of the generator nodes, the active and reactive load values of the load nodes and the current compensation value of the reactive compensation node.

And S2, calculating the initial power flow. And calculating the current, the network loss and the voltage qualification rate before optimization to determine the initial running state of the power system.

And S3, generating an initial individual. And calling a random function to randomly generate a group of populations within the upper limit and the lower limit of the control variable.

And S4, carrying out load flow calculation on each individual in the population. And performing iterative calculation, and solving the load flow result of each individual to obtain the active load flow, the reactive load flow, the voltage amplitude value of each node and the phase angle information of the line.

And S5, calculating an adaptive function value. And calculating an adaptive function value of each individual according to the result of the load flow calculation, wherein the objective function is a function which minimizes the network loss of the power grid, and the smaller the calculated fitness function value is, the closer the corresponding individual is to the optimal solution.

And S6, evaluating and selecting excellent individuals. And calculating the individual selection pressure according to the individual adaptive function values, sequencing the individual selection pressure by adopting a sequencing method, and reserving a plurality of preset individuals with low selection pressure.

And S7, forming a new generation of individuals. And carrying out updating operation on the selected individuals to form a new generation of individuals.

And S8, judging whether a convergence condition is met. If so, proceed to the next step, otherwise go to S4 to continue.

And S9, outputting a result. And decoding the updated and reserved optimal individuals, and outputting a decoding result, a network load flow result and a network loss result together.

Preferably, the calculation method further comprises calculating a center point of the power grid, including specifically.

a. Initializing a minimum path array; array length n = 0.

b. And judging whether all the load points are circulated completely, if so, turning to the step e, and otherwise, taking one load point.

c. Backtracking from the load point up to the power point.

d. Adding the minimum path of the load point into the minimum path array, increasing n by 1, and converting b.

e. And obtaining a minimum path set of all load points, and transposing all paths of the minimum path array to obtain a transposed array.

f. And circulating the minimum paths in the transposed array, sequentially taking one node in each minimum path from the transposed array from a power supply point, comparing every two nodes, judging whether the node numbers are equal, and if unequal node numbers exist, taking a sequence formed by the front equal nodes in the unequal node numbers in the minimum paths as a pivot point sequence.

The invention relates to a power grid calculation method based on feeder calculation, which uses a section as a minimum unit, arranges islands by an optimal subsection switch or a suboptimal subsection switch, can obtain a calculation network formed by all the islands after the feeder queue and the section queue are processed, processes different operation modes aiming at the calculation network, and outputs a calculation result. The method can be used for calculating the power grid efficiently and accurately, and particularly has a good processing effect on the looped network.

Drawings

Fig. 1 is a flowchart of a power grid calculation method based on feeder calculation according to the present invention.

Detailed Description

As shown in fig. 1, the present invention relates to a method for calculating a power grid based on feeder calculation, which includes the following steps.

A. All islands are processed according to feeder adjacency queues and sector queues to form a computational network.

(1) And acquiring a calculation feeder set consisting of the feeder lines and the junctor thereof.

(2) A feeder adjacency queue and a segment queue are initialized.

(3) Calculating the distribution of the feeder line moisture collection flow; and (4) counting the total load and the transfer margin of each feeder line.

(4) And taking a feeder line, adding all trunk line sections of the feeder line into the section queue, and adding all adjacent feeder lines of the feeder line into the adjacent queue.

(5) A block is fetched from the block queue and the fetched block is removed from the block queue.

(6) And opening the downstream boundary switch of the extracted section, and taking the downstream residual network formed by the downstream boundary switch as a subset of the whole feeder line.

(7) And searching a downstream residual network to obtain all interconnection switches and all load points connected in the network, respectively storing the interconnection switches and all load points into a switch structure array and a load structure array, and sequencing the switch structure arrays, wherein the margin is from high to low. And simultaneously storing the downstream residual network into a computation island queue.

(8) And (5) taking one island from the computation island queue for computation, and if all the islands are processed, turning to (15).

(9) And sequentially taking two interconnection switches from the switch structure array as an interconnection switch combination, and calculating the optimal disconnection section switch. Two tie switches are taken in proper order as tie switch combination in the follow switch structure array, calculate the optimal disconnection section switch, include: and closing the interconnection switch combination, and determining the section switch for splitting the island as an optimal disconnection section switch by calculating the optimal ring current of two interconnection switches in the interconnection switch combination.

(10) If there are no feasible sectionalizers as the optimal disconnection sectionalizers, removing the tie switch combination from the switch structure array, and turning to (9); if all the interconnection switches are circulated completely, turning to (12); otherwise, the process goes to (11).

(11) Opening the found optimal section switch, respectively carrying out power flow verification on the two split islands, and stopping searching and utilizing the interconnection switch combination to carry out switching if no out-of-limit and low voltage exist; or other interconnection switch combinations can be continuously searched, the transfer indexes are compared finally, the switch combination with the highest transfer index is selected as the optimal transfer path, the id of the optimal section switch is recorded, and the switch (15) is transferred.

(12) The section switch is recorded into a suboptimal section switch sequence by changing the opening position when the section switch is opened, calculating the switch index.

(13) And if all the switch structure arrays finish the circulation, selecting the section switch with the highest transfer index as the suboptimal section switch.

(14) And opening the suboptimum section switch to form two new islands, adding the new islands into the computation island queue, removing the original islands, and turning to (8).

(15) The computational network consisting of all islands is output.

B. Optimizing in stages based on the computing network, and outputting a computing result.

(16) And calculating initial load flow of the feeder based on the calculation feeder set, judging whether the calculation network is a looped network with more than one power supply point, and performing reactive power optimization solution on the looped power distribution network (24).

(17) Taking a feeder line from the feeder line set, and turning to (25) if all the feeder line sections are completely circulated; otherwise, setting the maximum operation mode for calculation.

(18) And entering the first stage of optimization. And taking all load points on the feeder line, and calculating the power factor of the load points. And if the power factor of the load point does not reach the preset range, calculating the low-voltage reactive compensation quantity of the load point and writing the result into a reactive compensation result array.

(19) And (5) performing load flow calculation, if no voltage out-of-limit exists, turning to (21), and otherwise, entering the second-stage optimization.

(20) And optimizing in the second stage. And calculating the secondary reactive moment or the net loss micro-increment rate. And selecting the maximum reactive moment for compensation, and calculating the compensation capacity.

(21) And calculating by taking the minimum operation mode.

(22) And calculating the load flow, and judging whether reactive power is reversely delivered.

(23) If reactive power is fed back, reducing the reactive compensation capacity, and turning to (22); otherwise go to (17).

(24) And carrying out reactive power optimization solution.

(25) And outputting a calculation result.

And performing reactive power optimization solving, including.

S1, data are recorded. Inputting network parameters of the dynamic power topology model, wherein the network parameters comprise node information, branch information and the like; and recording the real-time telemetering remote communication quantity, including the active and reactive output values of the generator nodes, the active and reactive load values of the load nodes and the current compensation value of the reactive compensation node.

And S2, calculating the initial power flow. And calculating the current, the network loss and the voltage qualification rate before optimization to determine the initial running state of the power system.

And S3, generating an initial individual. And calling a random function to randomly generate a group of populations within the upper limit and the lower limit of the control variable.

And S4, carrying out load flow calculation on each individual in the population. And performing iterative calculation, and solving the load flow result of each individual to obtain the active load flow, the reactive load flow, the voltage amplitude value of each node and the phase angle information of the line.

And S5, calculating an adaptive function value. And calculating an adaptive function value of each individual according to the result of the load flow calculation, wherein the objective function is a function which minimizes the network loss of the power grid, and the smaller the calculated fitness function value is, the closer the corresponding individual is to the optimal solution.

And S6, evaluating and selecting excellent individuals. And calculating the individual selection pressure according to the individual adaptive function values, sequencing the individual selection pressure by adopting a sequencing method, and reserving a plurality of preset individuals with low selection pressure.

And S7, forming a new generation of individuals. And carrying out updating operation on the selected individuals to form a new generation of individuals.

The update operation is: child individual = n parent individual 1+ (1-n) parent individual 2; where n is a scale factor and can be generated by random numbers uniformly distributed over [0,1], and the sub-individuals are new generation individuals.

And S8, judging whether a convergence condition (such as reaching the maximum updating times) is met. If so, proceed to the next step, otherwise go to S4 to continue.

And S9, outputting a result. And decoding the updated and reserved optimal individuals, and outputting a decoding result, a network load flow result and a network loss result together.

Preferably, the calculation method further includes calculating a center point of the power grid, and the calculation of the center point of the power grid may be performed after step B or simultaneously with step B, specifically including.

a. Initializing a minimum path array; array length n = 0.

b. And judging whether all the load points are circulated completely, if so, turning to the step e, and otherwise, taking one load point.

c. Backtracking from the load point up to the power point.

d. Adding the minimum path of the load point into the minimum path array, increasing n by 1, and converting b.

e. And obtaining a minimum path set of all load points, and transposing all paths of the minimum path array to obtain a transposed array.

f. And circulating the minimum paths in the transposed array, sequentially taking one node in each minimum path from the transposed array from a power supply point, comparing every two nodes, judging whether the node numbers are equal, and if unequal node numbers exist, taking a sequence formed by the front equal nodes in the unequal node numbers in the minimum paths as a pivot point sequence.

For example, in the case that there are two load points 10 and 12, tracing back the two load points upwards to obtain two minimum paths {10,9,8,2} and {12,9,8,2}, where the node 2 is a power point, the minimum path array is [ {10,9,8,2}, {12,9,8,2} ], the transposed array after transposing the minimum path array is [ {2,8,9,10}, {2,8.9,12} ], starting from the power point 2, taking one node from each minimum path in sequence to compare the node numbers, finding that the node numbers of the fourth nodes 10 and 12 are not equal, and then taking the sequence {2,8,9} formed by the front equal nodes of the nodes 10 and 12 as the pivot point sequence.

Before initializing the feeder adjacency queue and the segment queue, the method further comprises dividing the feeder segments, and particularly comprises the following steps.

1) All nodes in the feeder are set to "unprocessed".

2) And acquiring a power supply point of the feeder line, and expanding by taking the power supply point as a selected node.

3) The selected node is obtained and set to "processed".

4) Finding the load of the node from the node load relation table, if the node is connected with the load, switching to a load processing flow, and processing the load point; if there is no load point, go to 5).

5) Finding the switch connected with the node from the node switch relation table, if the node is connected with the switch, switching to a switch processing flow, and processing the switch; if the switch is not connected, go to 6).

6) Finding the feeder line segment connected with the node from the node line segment relation table, and processing all line segments connected with the node if the node is connected with the feeder line segment; the method comprises the following steps: taking out the first line segment to obtain a node on the opposite side of the feeder line segment, and taking out a next line segment if the node is processed; if not, recording the node at the opposite side of the feeder line section as an expanded switch node into a section switch relation table, and simultaneously recording the node at the opposite side to prepare for expansion of the next node; processing each connecting feeder segment in turn; if no feeder segment is connected, go to 7).

7) After the load, the switch and the feeder line section connected with the nodes are processed, the nodes or the load points connected with the nodes are used as the nodes in the section, the section attributes are obtained, and the section attributes and the nodes in the section are inserted into a section table.

8) And sequentially selecting the expanded switch nodes in the section switch relation table, and repeating the steps 3) -7) until all the expanded switch nodes in the section switch relation table are processed.

9) And selecting an unprocessed node in the feeder line, and repeating the steps 3) -8) until all the feeder line nodes are processed, namely all the feeder line nodes are processed, and at the moment, all sections of the feeder line are divided.

The invention relates to a power grid calculation method based on feeder calculation, which uses a section as a minimum unit, arranges islands by an optimal subsection switch or a suboptimal subsection switch, can obtain a calculation network formed by all the islands after the feeder queue and the section queue are processed, processes different operation modes aiming at the calculation network, and outputs a calculation result. The method can be used for calculating the power grid efficiently and accurately, and particularly has a good processing effect on the looped network.

The above description is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, several modifications and variations can be made without departing from the technical principle of the present invention, and these modifications and variations should also be regarded as the protection scope of the present invention.

Claims (5)

1. A power grid computing method based on feeder computing is characterized by comprising the following steps:

A. processing all islands according to the feeder line adjacent queue and the zone queue to form a computing network;

B. optimizing in stages based on the computing network, and outputting a computing result.

2. The method of claim 1, wherein processing all islands according to feeder adjacency queues and segment queues forms a computational network comprising:

(1) acquiring a calculation feeder set consisting of a feeder line and a tie line thereof;

(2) initializing a feeder line adjacent queue and a section queue;

(3) calculating the distribution of the feeder line moisture collection flow; counting the total load and the transfer margin of each feeder line;

(4) taking a feeder line, adding all trunk line sections of the feeder line into a section queue, and adding all adjacent feeder lines of the feeder line into an adjacent queue;

(5) fetching a segment from the segment queue and removing the fetched segment from the segment queue;

(6) opening a downstream boundary switch of the retrieved section, taking the downstream residual network formed thereby as a subset of the entire feeder;

(7) searching a downstream residual network to obtain all interconnection switches and all load points connected in the network, respectively storing the interconnection switches and all load points into a switch structure array and a load structure array, and sequencing the switch structure arrays, wherein the margin is from high to low; simultaneously storing the downstream residual network into a computation island queue;

(8) taking one island from the computation island queue for computation, and if all the islands are processed, turning to (15);

(9) sequentially taking two interconnection switches from the switch structure array as an interconnection switch combination, and calculating an optimal disconnection section switch; two tie switches are taken in proper order as tie switch combination in the follow switch structure array, calculate the optimal disconnection section switch, include: closing the interconnection switch combination, and determining a section switch for splitting the island as an optimal disconnection section switch by calculating the optimal ring current of two interconnection switches in the interconnection switch combination;

(10) if there are no feasible sectionalizers as the optimal disconnection sectionalizers, removing the tie switch combination from the switch structure array, and turning to (9); if all the interconnection switches are circulated completely, turning to (12); otherwise, entering (11);

(11) opening the found optimal section switch, respectively carrying out power flow verification on the two split islands, and stopping searching and utilizing the interconnection switch combination to carry out switching if no out-of-limit and low voltage exist; or other interconnection switch combinations can be continuously searched, the transfer indexes are compared finally, the switch combination with the highest transfer index is selected as the optimal transfer path, the id of the optimal section switch is recorded, and the switch (15) is transferred;

(12) calculating a transfer index by changing an opening position when the section switch is opened, and recording the section switch into a suboptimal section switch sequence;

(13) if all the switch structure arrays finish the circulation, selecting the section switch with the highest transfer index as a suboptimal section switch;

(14) opening the suboptimal section switch to form two new islands, adding the new islands into a calculation island queue, removing the original islands, and turning to (8);

(15) the computational network consisting of all islands is output.

3. The method of claim 1, wherein the computing-network-based phased optimization to output the computation results comprises:

(16) calculating initial load flow of a feeder based on the calculation feeder set, judging whether the calculation network is a looped network with more than one power supply point, and performing reactive power optimization solution on the looped power distribution network (24);

(17) taking a feeder line from the feeder line set, and turning to (25) if all the feeder line sections are completely circulated; otherwise, setting a maximum operation mode for calculation;

(18) entering a first stage of optimization; taking all load points on the feeder line, and calculating the power factor of the load points; if the power factor of the load point does not reach the preset range, calculating the low-voltage reactive compensation quantity of the load point and writing the result into a reactive compensation result array;

(19) load flow calculation, if no voltage out-of-limit exists, turning to (21), otherwise, entering the second stage optimization;

(20) optimizing in the second stage; calculating a secondary reactive moment or a network loss micro-increment rate; selecting the maximum reactive moment for compensation, and calculating compensation capacity;

(21) calculating by taking a minimum operation mode;

(22) calculating the load flow, and judging whether reactive power reverse transmission exists or not;

(23) if reactive power is fed back, reducing the reactive compensation capacity, and turning to (22); otherwise go to (17);

(24) performing reactive power optimization solution;

(25) and outputting a calculation result.

4. The method of claim 3, wherein the performing a reactive power optimization solution comprises:

s1, inputting data; inputting network parameters of the dynamic power topology model, wherein the network parameters comprise node information, branch information and the like; inputting real-time remote-measuring remote communication quantity, wherein the real-time remote-measuring remote communication quantity comprises an active and reactive output value of a generator node, an active and reactive load value of a load node and a current compensation value of a reactive compensation node;

s2, calculating an initial power flow; calculating the current, network loss and voltage qualification rate before optimization to determine the initial running state of the power system;

s3, generating an initial individual; calling a random function to randomly generate a group of populations within the upper limit range and the lower limit range of the control variable;

s4, carrying out load flow calculation on each individual in the population; iterative calculation is carried out, and the load flow result of each individual is solved to obtain the active load flow, the reactive load flow, the voltage amplitude value of each node and the phase angle information of the line;

s5, calculating an adaptive function value; calculating an adaptive function value of each individual according to the result of the load flow calculation, wherein the objective function is a function which enables the grid loss of the power grid to be minimum, and the smaller the calculated adaptive function value is, the closer the corresponding individual is to the optimal solution;

s6, evaluating and selecting excellent individuals; calculating the individual selection pressure according to the individual adaptive function values, sequencing the individual selection pressure by adopting a sequencing method, and reserving a plurality of preset individuals with low selection pressure;

s7, forming a new generation of individuals; updating the selected individuals to form a new generation of individuals;

s8, judging whether a convergence condition is met; if yes, go to the next step, otherwise go to S4 to continue;

s9, outputting a result; and decoding the updated and reserved optimal individuals, and outputting a decoding result, a network load flow result and a network loss result together.

5. The method according to claim 1, wherein the calculating method further comprises calculating a center point of the power grid, and specifically comprises:

a. initializing a minimum path array; array length n = 0;

b. judging whether all the load points are circulated completely, if so, turning to e, and otherwise, taking one load point;

c. backtracking from the load point to the power point;

d. adding the minimum path of the load point into a minimum path array, increasing n by self by 1, and turning b;

e. obtaining a minimum path set of all load points, and transposing all paths of the minimum path array to obtain a transposed array;

f. and circulating the minimum paths in the transposed array, sequentially taking one node in each minimum path from the transposed array from a power supply point, comparing every two nodes, judging whether the node numbers are equal, and if unequal node numbers exist, taking a sequence formed by the front equal nodes in the unequal node numbers in the minimum paths as a pivot point sequence.

Images