CN103259265B - Power network partitioning and detached island identification method of intelligent power distribution network based on reachability matrix - Google Patents

Abstract

The invention relates to a power network partitioning and detached island identification method of an intelligent power distribution network based on a reachability matrix. The method has good power network partitioning capability, and can quickly and accurately detect the quantity and a range of detached islands, and a specific parallel-in switch. The method comprises the following steps that 1, a branch-node diagram is established according to the intelligent power distribution network and used for describing a structure of the intelligent power distribution network; 2, an adjacent matrix A is established and used for describing connection relations between various branches and nodes in the branch-node diagram; 3, each idempotent matrix of the adjacent matrix A is calculated, and a partitioning situation of the network is obtained; and 4, a switch property vector K is defined, a block vector Pi for describing network partitioning information is obtained according to the partitioning situation and value situations of elements in a reachability matrix P; corresponding elements in the vector K and the vector Pi are subjected to logic and calculation; vectors for describing properties of various areas of the network are obtained; whether the detached islands exist is judged; and the quantity of the detached islands, and power supplies corresponding to each detached island are judged.

Description

Based on intelligent distribution network sub area division and the island identification method of reachability matrix

Technical field

The present invention relates to a kind of sub area division and island detection method, especially a kind of intelligent distribution network sub area division based on reachability matrix and island identification method.

Background technology

Power distribution network node is numerous, various DG throws and moves back frequently, cause the operational mode of intelligent distribution network flexible and changeable, especially when some switch in electrical network disconnects, electrical network can be divided into some distinct, separate subnets, if the load in subnet is powered by DG or microgrid, so this subnet just runs in the mode of isolated island.The appearance of isolated island brings considerable influence to the scheduling of intelligent distribution network, automated system, relaying protection system and automatic reclosing etc. on the one hand, and the safety and stability of isolated island self-operating also exists problems on the other hand.

Current island detection method can be divided into passive type to detect and active detection two class generally.Passive type detection method refers to that the change of the power output by directly measuring distributed generation system inverter or the change of PCC place voltage or frequency judge isolated island situation.Passive detection method principle is simple, is easy to realize, can not the quality of power supply of influential system and stability.But the threshold value of these class methods is all more difficult to be determined, and whether they mate with the power of load the power output of inverter and have strict requirement, there is certain check frequency.Active detection method is that when distributed generation system and electrical network are normally incorporated into the power networks, microvariations are not enough to have an impact to system by exporting the little disturbance of generation one to the inverter of distributed generation system; When isolated island produces, the microvariations initiatively produced will cause the change that island state is large.The check frequency of active detecting method has had certain improvement, but owing to introducing disturbance, have certain influence to the transient response of system, in multiple stage grid-connected inverters situation, the consistency of disturbance is also ubiquitous problem.

Current islanding detect and discrimination method study mainly based on and the local information at site place, if using overall for power distribution network as research object, the position of the switch state information in network is utilized in conjunction with network topology structure, electric network composition situation of change can be obtained fast and accurately, and identification can be carried out to isolated island reliably.

Have a kind of based on normalization spectral clustering and the two benches Active Splitting method retraining spectral clustering at present.The method utilizes normalization spectrum to birds of the same feather flock together to identify Coherent Generator Group in the first phase, and utilizes these Coherent Generator Groups to be embodied as constraint; Second stage utilization constraint spectral clustering finds satisfied constraint in pairs, has the off-the-line section of minimum effective power flow impact.When system needs to be split as multiple isolated island, recursive bisection is used to realize.The method relates to generating and impacts two factors with mediation effective power flow, and algorithm reliability has much room for improvement.

In addition based on the parallel network active alone island detection system that M sequence is modulated and variance judges and method.This detection method is the pseudo-random signal M sequence of mechanics of communication, modulates inverter modulating wave according to the regularity of distribution of M sequence, and with the characteristic making output also have M sequence, thus output signal also has the feature of random signal.A variance threshold values is set, when detecting that variance exceedes threshold value and can be judged as isolated island during detection.This algorithm is simple, cost is low, and little to the harmonic effects exporting electric energy, islanding detect blind area is little, but amount of calculation is high, cannot meet the demands completely for large scale system.

Summary of the invention

The object of the invention is for overcoming above-mentioned the deficiencies in the prior art, a kind of intelligent distribution network sub area division based on reachability matrix and island identification method are provided, the method is not by network topology structure variable effect, clear principle, highly sensitive, there is good sub area division ability, the quantity of isolated island, the scope of isolated island and concrete grid-connected switch can be detected fast and accurately.

For achieving the above object, the present invention adopts following technical proposals:

Based on intelligent distribution network sub area division and the island identification method of reachability matrix, comprise the following steps:

The first step: be corresponding node by each switch equivalent in intelligent distribution network, the line equivalent between switch is branch road, sets up a road – node diagram and describes intelligent power distribution web frame;

Second step: set up adjacency matrix A to describe the annexation of each branch road and node in Zhi Lu – node diagram, if certain branch road is directly connected with certain node, then corresponding in matrix A element value is got " 1 ", otherwise gets " 0 ";

3rd step: each the power matirx calculating adjacency matrix A, carries out logic summation operation to corresponding element in each power matirx, obtains reachability matrix P, analyzes the value rule of element in reachability matrix P, obtains the partitioning scenario of network;

4th step: be power supply, circuit or load according to what be connected with switch, definition switching behavior vector K, according to the value condition of element in partitioning scenario and reachability matrix P, obtains the block vector P describing net partition information _i, to vectorial K and P _iin corresponding element carry out logical AND calculating, obtain the vector describing network each region character, and then judge whether there is isolated island, which power supply corresponding in the quantity of isolated island and each isolated island.

The define method of each element of adjacency matrix A in described second step is:

Reachability matrix computational methods in the 3rd described step are: each the power matirx first obtaining adjacency matrix A: A, A ², A ³..., A ⁿ, wherein n is the nodes of Zhi Lu – node diagram, and then reachability matrix P is obtained in calculating, and computational methods are: P=A+A ²+ A ³+ ... + A ⁿ, addition involved is here logic add (∨), i.e. 0 ∨ 0=0,0 ∨ 1=1,1 ∨ 0=1,1 ∨ 1=1.

Switching behavior vector K=[k is defined in the 4th described step ₁k ₂k _n], element k in K _ivalue is defined as follows:

If a road – node diagram is divided into some regions, block vector P can be used in each region ₁, P ₂p _mrepresent, will judge whether to comprise in each region the character that power supply and institute comprise power supply further, need with switching behavior vector K to the vectorial P of block _irevise:

p _i'＝k _i·p _i

Analyze vectorial P _i' in element value, just can judge the character of each isolated area fast, thus determine whether there is isolated island, the quantity of isolated island and the scope of isolated island.

The present invention adopts reachability matrix to carry out sub area division and the isolated island identification of intelligent distribution network, and tool has the following advantages compared with existing analogous algorithms:

1. clear principle, information needed is few, is easy to realize;

2. matrix degree of rarefication is high, and amount of calculation is little, is convenient to adopt centralized structure;

3. can define by flexible expansion switchtype as requested, meet the sub area division requirement of different object;

4. can detect the quantity of isolated island, the scope of isolated island and concrete grid-connected switch fast and accurately.

Accompanying drawing explanation

Fig. 1 is the structural representation of intelligent distribution network;

Fig. 2 is node-branch road schematic diagram that intelligent distribution network simplifies;

Embodiment

Below in conjunction with drawings and Examples, the present invention is further described.

1. based on the intelligent distribution network subregion of reachability matrix

In Fig. 1, S represents backbone switch and branch switch, corresponds to node in fig. 2; K is interconnection switch; D represents offside line switching.

When switch S 2 disconnects, adjacency matrix corresponding to the system shown in Fig. 2 is:

$A=\left[\begin{array}{cccccccccc}0& 0& 0& 0& 0& 1& 1& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 1& 0& 0& 0& 1& 1& 0\\ 0& 0& 1& 0& 1& 0& 0& 0& 1& 1\\ 0& 0& 0& 1& 0& 0& 0& 0& 0& 1\\ 1& 0& 0& 0& 0& 0& 1& 0& 0& 0\\ 1& 0& 0& 0& 0& 1& 0& 0& 0& 0\\ 0& 0& 1& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 1& 1& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 1& 1& 0& 0& 0& 0& 0\end{array}\right]$

Can calculate each power of adjacency matrix A, and then the reachability matrix P trying to achieve its correspondence is:

$P=\left[\begin{array}{cccccccccc}1& 0& 0& 0& 0& 1& 1& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 1& 1& 1& 0& 0& 1& 1& 1\\ 0& 0& 1& 1& 1& 0& 0& 1& 1& 1\\ 0& 0& 1& 1& 1& 0& 0& 1& 1& 1\\ 1& 0& 0& 0& 0& 1& 1& 0& 0& 0\\ 1& 0& 0& 0& 0& 1& 1& 0& 0& 0\\ 0& 0& 1& 1& 1& 0& 0& 1& 1& 1\\ 0& 0& 1& 1& 1& 0& 0& 1& 1& 1\\ 0& 0& 1& 1& 1& 0& 0& 1& 1& 1\end{array}\right]$

The value analyzing element in reachability matrix P is visible, 1st, 6,7 row (column) element values are consistent, and the 3rd, 4,5,8,9,10 row (column) element values are consistent, and the 2nd row (column) element is complete " 0 ", therefore known switch S 2 disconnects, now Zhi Lu – node diagram is divided into 2 relatively independent regions.First region comprises the branch road of switch S 1, S6 and S7 and correspondence, and corresponding element value is " 1 ", shows can reach between S1, S6 and S7; Two Areas comprises switch S 3, the branch road of S4, S5, S8, S9 and S10 and correspondence, and can reach each other.Because S2 disconnects, therefore not any with other node is communicated with.The division result obtained by calculating reachability matrix P and the connected relation of each node, completely the same with the practical structures of the Lu of Zhi shown in Fig. 2 – node diagram.

2. Grid type judges and isolated island identification

After the switch S 2 in Fig. 2 disconnects, electrical network is divided into two relatively independent regions, and corresponding block vector is designated as P1 and P2 respectively, and based on the result of calculation of reachability matrix P, P1 and P2 obtained is respectively:

P ₁＝[1 0 0 0 0 1 1 0 0 0]

P ₂＝[0 0 1 1 1 0 0 1 1 1]

According to the switchtype vector K that system configuration and running status obtain be:

K＝[1 0 0 0 0 -1 -2 -2 -2 -2]

Utilize vectorial K to carry out correction to vectorial P1 and P2 respectively to obtain:

P ₁'＝[1 0 0 0 0 -1 -2 0 0 0]

P ₂'＝[0 0 0 0 0 0 0 -2 -2 -2]

At P ₁' in, owing to comprising the element that value is " 1 ", therefore can judge in this region, have switch (i.e. switch S 1) to be connected with system power supply, belong to region of being incorporated into the power networks; At P ₂' in, except the element that value is " 0 ", only comprise the element of value for "-2 ", therefore can judge that this region is islet operation region, and be respectively S8 with the switch that DG is directly connected, S9 and S10.

Claims (4)

1., based on intelligent distribution network sub area division and the island identification method of reachability matrix, it is characterized in that, comprise the following steps:

The first step: be corresponding node by each switch equivalent in intelligent distribution network, the line equivalent between switch is branch road, sets up a road – node diagram and describes intelligent power distribution web frame;

Second step: set up adjacency matrix A to describe the annexation of each branch road and node in Zhi Lu – node diagram, if certain branch road is directly connected with certain node, then corresponding in matrix A element value is got " 1 ", otherwise gets " 0 ";

3rd step: each the power matirx calculating adjacency matrix A, carries out logic summation operation to corresponding element in each power matirx, obtains reachability matrix P, analyzes the value rule of element in reachability matrix P, obtains the partitioning scenario of network;

4th step: be power supply, circuit or load according to what be connected with switch, definition switching behavior vector K, according to the value condition of element in partitioning scenario and reachability matrix P, obtains the block vector P describing net partition information _i, to vectorial K and P _iin corresponding element carry out logical AND calculating, obtain the vector describing network each region character, and then judge whether there is isolated island, the quantity of isolated island and each isolated island which power supply corresponding.

2., as claimed in claim 1 based on intelligent distribution network sub area division and the island identification method of reachability matrix, it is characterized in that, the define method of each element of adjacency matrix A in described second step is:

3., as claimed in claim 1 based on intelligent distribution network sub area division and the island identification method of reachability matrix, it is characterized in that, the reachability matrix computational methods in the 3rd described step are: each the power matirx first obtaining adjacency matrix A: A, A ², A ³..., A ⁿ, wherein n is the nodes of Zhi Lu – node diagram, and then reachability matrix P is obtained in calculating, and computational methods are: P=A+A ²+ A ³+ ... + A ⁿ, addition involved is here logic add (∨), i.e. 0 ∨ 0=0,0 ∨ 1=1,1 ∨ 0=1,1 ∨ 1=1.

4. as claimed in claim 1 based on intelligent distribution network sub area division and the island identification method of reachability matrix, it is characterized in that, in the 4th described step, define switching behavior vector K=[k ₁k ₂k _n], element k in K _ivalue is defined as follows:

If a road – node diagram is divided into some regions, block vector P can be used in each region ₁, P ₂p _mrepresent, will judge whether to comprise in each region the character that power supply and institute comprise power supply further, need with switching behavior vector K to the vectorial P of block _irevise:

p _i′＝k _i·p _i

Analyze P _ielement value in ' vector, just can judge the character of each isolated area fast, thus determine whether there is isolated island, the quantity of isolated island and the scope of isolated island.