CN104701827B

CN104701827B - Microgrid wide area current protection system and method based on multi-Agent technology

Info

Publication number: CN104701827B
Application number: CN201510116399.4A
Authority: CN
Inventors: 张化光; 刘鑫蕊; 谢志远; 孙秋野; 杨珺; 王智良; 黄博南; 高艺伟
Original assignee: Northeastern University China
Current assignee: Northeastern University China
Priority date: 2015-03-17
Filing date: 2015-03-17
Publication date: 2017-06-20
Anticipated expiration: 2035-03-17
Also published as: CN104701827A

Abstract

A micro-grid wide-area current protection system and method based on multi-Agent technology belongs to the technical field of micro-grid protection. The system is divided into an intelligent electronic device layer, a regional control and protection coordination layer, and a central processing layer; the intelligent electronic device layer is composed of an intelligent electronic device IED configured on each circuit breaker in the microgrid; each IED includes: a state monitoring Agent , measurement Agent, network topology Agent and protection Agent; the regional control and protection coordination layer includes: regional control Agent and protection coordination Agent; the central processing layer includes the central processing Agent; The range is divided, and the taboo algorithm has the advantage of getting rid of local optimization. At the same time, it also makes full use of the advantages of the breadth search algorithm with a wide search range and high efficiency; the method of expanding the current protection range is used to overcome IED refusal or misoperation. Correctly and quickly remove the faults inside the microgrid.

Description

Microgrid wide-area current protection system and method based on multi-Agent technology

Technical Field

The invention belongs to the technical field of microgrid protection, and particularly relates to a microgrid wide-area current protection system and method based on a multi-Agent technology.

Background

The microgrid is a micro low-voltage power distribution system consisting of a series of Distributed Generators (DGs), an energy storage system and loads, and an effective technical means is provided for full application of the distributed generators. Meanwhile, the inside relatively complex electrical characteristics of the microgrid: the tidal current direction in the micro-grid is not fixed, and the micro-grid has two operation modes of grid connection and island. The bidirectional power flow characteristic of the microgrid makes the selectivity of microgrid protection difficult to achieve; grid-connected operation and island operation face short-circuit fault current with large difference. In addition, the microgrid and the distributed power supply have a plug-and-play characteristic, which brings great impact on the voltage and frequency stability of the microgrid, and therefore, a serious challenge is brought to the protection of the microgrid.

With the rapid development of wide area measurement systems and communication networks, wide area current protection realized based on wide area information is concerned by scholars at home and abroad, and meanwhile, the multi-Agent technology is also widely applied to power systems. Therefore, the students tried to combine the multi-Agent technology with the wide-area current protection and obtained some basic research results. However, the research of the microgrid protection methods is limited to a specific microgrid structure model, and the microgrid structure model is not comprehensive, and for wide-area current protection, the larger the range of receiving fault information is, the better the fault information is, the reason is as follows: firstly, the requirement of rapidity of relay protection makes insufficient time for receiving and processing large-scale fault information; secondly, fault disturbance only affects a limited local range, fault information in the range is critical to fault judgment, and information outside the fault is not so critical, so that how to divide the protection range is a problem needing further research. And there is also a problem of how the Intelligent Electronic Device (IED) associated therewith acts when it fails or malfunctions.

Disclosure of Invention

Aiming at the defects of the existing method, the invention provides a microgrid wide-area current protection system and method based on a multi-Agent technology.

The technical scheme of the invention is realized as follows:

a microgrid wide-area current protection system based on a multi-Agent technology is divided into 3 layers including an intelligent electronic equipment layer, a region control and protection coordination layer and a central processing layer; wherein:

the intelligent electronic equipment layer is composed of intelligent electronic equipment IEDs configured on each circuit breaker in the microgrid; the intelligent electronic device layer is responsible for acquiring state information of the microgrid, drawing a real-time microgrid network topological graph and acquiring electric quantity information of the microgrid; when the inter-layer communication is normal, uploading the drawn real-time microgrid network topological graph to a region control and protection coordination layer and a central processing layer, and uploading the electric quantity information of the microgrid to the region control and protection coordination layer; each of the IEDs further includes:

the state monitoring Agent is responsible for monitoring state information of the microgrid, and the state information comprises access positions of the distributed power supplies, switching states of the distributed power supplies and switching states of all circuit breakers in the microgrid;

the measuring Agent is responsible for collecting the electric quantity information of the microgrid, and comprises the following steps: voltage, current, phase angle and frequency of the microgrid;

the network topology Agent is responsible for drawing a real-time microgrid network topology map according to the state information of the microgrid;

the protection Agent is responsible for adaptively adjusting a protection constant value of a pre-written protection criterion according to a microgrid current protection scheme sent by the central processing Agent after analyzing the electric quantity information of the microgrid or autonomously acting to remove faults when interlayer communication is completely interrupted;

the zone control and protection coordination layer comprises:

the regional control Agent is responsible for controlling the active power of the distributed power supply, the reactive power of the distributed power supply, the active power of the energy storage unit, the reactive power of the energy storage unit, the voltage of the microgrid and the frequency of the microgrid and making a control scheme;

the protection coordination Agent determines the optimal microgrid current protection range to be expanded when the main protection IED fails to operate or malfunctions in the microgrid; the system is responsible for carrying out coordination management on each protection Agent, determining how the protection agents of the upper and lower-level lines of the same outgoing line are matched, and making a protection coordination scheme;

the central processing layer includes:

the central processing Agent is responsible for monitoring the working state of each Agent except the central processing Agent; making a microgrid current protection scheme by referring to a control scheme provided by the regional control Agent and a protection coordination scheme provided by the protection coordination Agent, and feeding back the microgrid current protection scheme to the intelligent electronic equipment layer; the method comprises the steps that a microgrid current protection area is divided in real time when interlayer communication is normal;

and the central processing Agent, the area control Agent and the protection coordination Agent are all configured at the low-voltage side outlet end of the transformer substation.

The layers of the system are interconnected by adopting an optical fiber Ethernet; and different agents in the same layer exchange information through a CAN bus.

The method for adopting the microgrid wide-area current protection system based on the multi-Agent technology comprises the following steps:

step 1, initializing each Agent;

step 2, the state monitoring Agent transmits the monitored micro-grid state information to a network topology Agent in real time; the microgrid state information comprises: the switching-in position of the distributed power supply, the switching-out state of the distributed power supply and the switching-on and switching-off state of each circuit breaker in the microgrid; meanwhile, the measurement Agent also transmits the microgrid electrical quantity information to the protection Agent in real time, and the microgrid electrical quantity information comprises: voltage, current, phase angle and frequency of the microgrid;

step 3, if the communication among the layers is normal, executing step 4; if the interlayer communication is completely interrupted, the protection Agent adopts overcurrent protection based on local information to carry out fault removal;

step 4, according to the real-time microgrid state information, the network topology Agent draws a real-time microgrid network topology graph and transmits the real-time microgrid network topology graph to the area control Agent, the protection coordination Agent and the central processing Agent, and meanwhile, the measuring Agent also transmits the microgrid electrical quantity information to the area control Agent in real time;

step 5, according to a real-time microgrid network topological graph drawn by a network topological Agent and microgrid electrical quantity information of a measuring Agent, a region control Agent formulates a control scheme of active power of a distributed power supply, reactive power of the distributed power supply, active power of an energy storage unit, reactive power of the energy storage unit, microgrid voltage and microgrid frequency in a microgrid current protection region, and uploads the control scheme to a central processing Agent; the protection coordination Agent formulates a microgrid protection coordination scheme and uploads the microgrid protection coordination scheme to the central processing Agent;

step 6, according to a real-time microgrid network topological graph drawn by a network topological Agent and a control scheme and a protection coordination scheme made by a region control and protection coordination layer, a central processing Agent introduces a taboo table into a breadth search algorithm to form a fusion taboo breadth search algorithm, and the current protection region of the microgrid is divided in real time by using the fusion taboo breadth search algorithm to obtain a microgrid current protection division region;

and 7, referring to a control scheme and a protection coordination scheme made by the area control and protection coordination layer, and making a microgrid current protection scheme by the central processing Agent: configuring a set of improved pilot current differential protection and a set of over-current protection based on local information in protection agents in each defined microgrid current protection area; the central processing Agent simultaneously feeds the microgrid current protection scheme back to the intelligent electronic equipment layer;

step 8, when the fault information is correct and complete, fault removal is carried out through improved pilot current differential protection;

the method comprises the following steps that a protection Agent analyzes microgrid electric quantity information sent by a measurement Agent, and adaptively adjusts a protection constant value of a pre-written improved pilot current differential protection criterion by combining a microgrid current protection scheme sent by a central processing Agent, and comprises the following steps: setting value I of differential current_set1Coefficient of ratiometric braking K₁、K₂The coefficient of influence K of the unstable current of DG on the improved pilot current differential protection₃And inflection point value I of the rate braking current_INTCollecting the number of line ports;

the improved pilot current differential protection criterion is as follows:

I_r＝I_INT(k₂-k₁)/k₁k₂

in the formula,. DELTA.I_giThe number of the line port collection can be determined according to a network topological graph provided by a network topological Agent for the current quantity collected by each line port, so that the criterion of pilot current differential protection is adaptively adjusted; k₁、K₂To rate-brake coefficient, I_INTIs the knee value of the proportional braking current, I_set1Is a setting value of a differential current, K₃Is the coefficient of the influence of the unstable current of DG on the differential protection of the pilot current;

9, when part of fault information is missing or wrong, the main protection IED fails to operate or malfunction, the protection coordination Agent determines the correlation coefficient and the protection operation characteristic coefficient between the failed or malfunction IED and the adjacent IED in the microgrid, further constructs a current protection operation range output function to obtain a solution set of the microgrid current protection range to be expanded, then an ant colony algorithm is used for optimizing to obtain the optimal microgrid current protection operation range to be expanded, and finally fault removal is carried out through improved longitudinal differential current protection;

according to a real-time network topology diagram provided by a network topology Agent, a protection coordination Agent determines an association coefficient A of each IED and adjacent IEDs thereof_f：

Defining the action characteristic coefficients of each IED:

n, wherein i is 1,2.. No.; k and A_fCorrespondingly, namely k is 1,2, 3; n is the number of IEDs in the microgrid;

establishing a current protection action range expression as follows:

and constructing a current protection action range output function F_out＝{F₁,F₂,F₃In which F₁,F₂,F₃In turn, the priority of the protection actions decreases, F₁Has a solution, F₂And F₃There are multiple solutions; according to F_outThe output of the error detection module can determine that each IED has fault information errors or partial fault information is missing, so that the microgrid current protection action range which needs to be expanded after misoperation or failure is resolved.

In the step 6, the process of obtaining the microgrid current protection partition area by real-time partitioning of the microgrid by using the fusion tabu breadth search algorithm further comprises the following steps:

step 6.1, establishing an objective function for dividing the current protection area of the microgrid, wherein the objective of the function is the minimization of the power failure area in the microgrid, and establishing a constraint condition for the objective function;

the objective function is as follows:

the constraint conditions are as follows:

wherein F (x) represents the total outage area; a. the_iRepresenting the blackout area of each divided area; PS represents a set of measures to minimize the blackout area of each divided region; t is_ijRepresenting the delay time of the fault information; i is_ijRepresenting the communication distance from the node i to the node j in the microgrid; c represents the speed of light; m represents the number of nodes passed by the communication path; t is t_vRepresenting the communication time of a single node; Δ t represents the random jitter delay; t is_ijmA given value representing a delay of the fault information; b represents the fault information acquisition amount; p_iRepresenting the degree of reliability of the communication; p_xiRepresenting the reliability of the communication element; p_yiRepresenting the reliability of the line between the communication elements; p_imA given value representing the reliability of the communication; n, j 1,2.. n, U represents the microgrid bus voltage;

step 6.2, an extent search algorithm is adopted, and an initial solution of extent search, namely an initial microgrid current protection divided region, is given, wherein the initial solution can be generated by a random method or an existing heuristic method;

6.3, searching the optimal solution of the objective function from the initial solution by using a breadth search algorithm with the given constraint condition as a boundary;

and 6.4, putting the searched local optimal solution into a tabu table, updating the tabu table in each iteration, avoiding the local optimal solution in the next search, and searching other spaces until a global optimal solution, namely the optimal microgrid current protection partition area meeting the objective function and the constraint condition is obtained.

The invention has the beneficial effects that:

1. and constructing a layered microgrid wide-area current protection system based on a multi-Agent technology. The protection system makes full use of the autonomy, the interactivity and the cooperative property, the variability, the adaptability and the spontaneity of the Agent, can perform protection control through centralized decision, and can perform independent protection when a communication fault occurs.

2. The improved fusion tabu breadth search algorithm is adopted to divide the micro-grid current protection range, the advantages of the tabu algorithm in getting rid of local optimization are fully exerted, a tabu table is introduced to mark and standardize the local optimal solution obtained by breadth search, the breadth search result is prevented from falling into local optimization, and meanwhile, the search time is saved; the algorithm also fully utilizes the advantages of wide search range, high efficiency and simple algorithm of the breadth search algorithm, so that the optimal solution of the target, namely the optimal microgrid current protection range, can be well obtained by fusing the tabu breadth search algorithm.

3. By adopting the improved pilot current differential protection, the protection coordination Agent can determine the number of the ports of the line to be acquired by the improved pilot current differential protection in the protection Agent according to a network topology provided by the network topology Agent, so that the protection Agent can self-adaptively adjust the protection criterion of the pilot current differential protection, and the influence of DG (distributed generation) on the protection can be overcome.

4. The method for expanding the current protection range is adopted to overcome the defect of failure or misoperation of the IEDs, the correlation coefficient and the protection action characteristic coefficient of each IED are determined, a function is further constructed to determine a solution set of the microgrid current protection action range to be expanded, and finally the optimal microgrid current protection action range to be expanded is determined by an ant colony optimization method. The method can effectively solve the problems caused by the failure or misoperation of the IED, and can accurately and quickly remove the faults in the microgrid.

Drawings

Fig. 1 is a schematic structural diagram of a microgrid wide-area current protection system based on a multi-Agent technology according to an embodiment of the invention;

fig. 2 is a flowchart of microgrid wide-area current protection based on a multi-Agent technology according to an embodiment of the present invention;

FIG. 3 is a flow chart of a fusion tabu breadth search algorithm according to an embodiment of the present invention;

fig. 4 is a diagram of a microgrid wide-area current protection configuration based on a multi-Agent technology according to an embodiment of the present invention;

fig. 5(a) is a current waveform diagram before and after a three-phase short-circuit fault occurs in a line FG under the improved pilot current differential protection according to an embodiment of the present invention; (b) the method is characterized in that a current waveform diagram before and after a single-phase earth fault occurs on a line FG under improved pilot current differential protection in one embodiment of the invention;

fig. 6 is a flowchart for finding an optimal extended current protection range by using an ant colony optimization algorithm according to an embodiment of the present invention.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.

The embodiment builds a layered wide-area current protection system based on multi-Agent technology, as shown in fig. 1.

The system is divided into an intelligent electronic equipment layer, a regional control and protection coordination layer and a central processing layer. And information exchange is carried out between layers through optical fiber Ethernet networking, and an IEC61850 GOOSE model is adopted as an information exchange mechanism. Different agents in the same layer exchange information through a CAN (controller area network) bus; each breaker in the microgrid is provided with a set of intelligent electronic equipment (IED), and all IEDs form an intelligent electronic equipment layer; the intelligent electronic equipment layer is used for acquiring state information in the microgrid, drawing a real-time microgrid network topological graph, acquiring electrical quantity information, uploading the information under the condition that interlayer communication is normal, making a final decision by the central processing layer, and autonomously performing fault information processing and protecting actions when the interlayer communication is completely interrupted; each IED comprises a network topology Agent, a state monitoring Agent, a measurement Agent and a protection Agent. Zone control and protection coordination layer: according to the decision of the central processing layer and the local information provided by the intelligent electronic equipment layer, the active power of the distributed power supply, the reactive power of the distributed power supply, the active power of the energy storage unit, the reactive power of the energy storage unit, the voltage of the microgrid and the frequency of the microgrid are controlled so as to achieve the purposes of stable frequency, normal voltage and balanced power of the microgrid; meanwhile, when partial fault information is missing or fault information is wrong to cause the main protection IED to refuse to operate or malfunction, determining the association coefficient and the protection operation characteristic coefficient of each IED adjacent to the refuse to operate or malfunction IED in the microgrid so as to determine the microgrid current protection range to be expanded, and finally optimizing and determining the optimal microgrid current protection range to be expanded through an ant colony algorithm; carrying out coordination management on each protection Agent in the microgrid, for example, determining how protection agents of upper and lower lines of the same outgoing line are matched; determining the number of the ports of the line to be acquired for improved pilot current differential protection in the protection Agent according to a real-time topological graph drawn by the network topology Agent; the layer comprises a region control Agent and a protection coordination Agent which are both configured at the outlet end of the low-voltage side of the transformer substation. A central processing layer: the microgrid current protection system is used for comprehensively monitoring all agents except the agents, dividing a microgrid current protection area in real time when communication is normal, and making a microgrid current protection scheme, wherein the microgrid current protection scheme comprises a central processing Agent and is also configured at a low-voltage side outlet end of a transformer substation. When the inter-layer communication is normal, the intelligent electronic equipment layer, the area control and protection coordination layer and the central processing layer can transmit commands and feed back information; when the inter-layer communication is completely interrupted, the intelligent electronic equipment layer can fully utilize the self-standing property of the Agent to process fault information and perform current protection action.

The central processing Agent is responsible for monitoring the working state of each Agent except the Agent; the micro-grid protection method comprises the steps that communication is carried out between the intelligent electronic equipment layer and the area control and protection coordination layer, the control schemes for active power of a distributed power supply, reactive power of the distributed power supply, active power of an energy storage unit, reactive power of the energy storage unit, micro-grid voltage and micro-grid frequency provided by an area control Agent are referred, the protection coordination schemes for all protection agents provided by the protection coordination Agent are analyzed to make a micro-grid current protection scheme, and communication and interaction are carried out between the protection coordination Agent and central processing agents of other micro-grids; the system is in charge of information processing of each Agent, integrates all-microgrid fault information sent by each Agent, and stores and displays the information to form a fault record report; and dividing the microgrid current protection area in real time when the communication is normal.

The regional control Agent and the protection coordination Agent are intermediate links, are information intersection points, and perform information interaction with the intelligent electronic device layer and the central processing layer, wherein the regional control Agent is responsible for controlling active power of the distributed power supply, reactive power of the distributed power supply, active power of the energy storage unit, reactive power of the energy storage unit, microgrid voltage and microgrid frequency so as to achieve the purposes of microgrid frequency stabilization, voltage normality and power balance. The protection coordination Agent determines the number of ports of the line to be acquired for improved pilot current differential protection in the protection Agent according to a real-time microgrid network topological graph drawn by the network topological Agent; when partial fault information is missing or fault information is wrong to cause the main protection IED to refuse to operate or malfunction, determining the correlation coefficient and the protection operation characteristic coefficient of each IED adjacent to the refuse to operate or malfunction IED in the microgrid so as to determine the microgrid current protection range to be expanded, and finally optimizing and determining the optimal microgrid current protection range to be expanded through an ant colony algorithm; and the system is responsible for carrying out coordination management on each protection Agent, and for example, determining how protection agents of an upper-level line and a lower-level line of the same outgoing line are matched to make a protection coordination scheme. And uploading the formulated control scheme and protection coordination scheme to a central processing Agent, and making the microgrid current protection scheme by using the central processing Agent as a reference.

In the intelligent electronic equipment layer, the state monitoring Agent is responsible for monitoring the state information of the microgrid, and the microgrid state information comprises: the access position of the distributed power supply, the switching state of the distributed power supply and the switching state of each breaker in the microgrid are communicated with the network topology Agent; the measurement agents realize the collection of the electric quantity required by the current protection of the microgrid, measure the data on a current transformer and a voltage transformer in real time, or measure the amplitude of the voltage, the amplitude of the current, the phase angle and the frequency of the protected area through a synchronous phasor measurement device, directly communicate with the protection agents and also communicate with the area control agents; the main function of the protection Agent is responsible for the microgridAfter the electric quantity information is analyzed, a protection constant value of a pre-written improved pilot current differential protection criterion is adaptively adjusted by combining a microgrid current protection scheme transmitted by a central processing Agent, or faults are removed by adopting overcurrent protection autonomous actions based on local information when interlayer communication is completely interrupted, wherein the protection constant value comprises the following steps: setting value I of differential current_set1Coefficient of ratiometric braking K₁、K₂Coefficient of influence of the unstable current of DG on the differential protection of the pilot current₃And inflection point value I of the rate braking current_INT(ii) a And the network topology Agent draws a network topology graph in real time according to the state information provided by the state monitoring Agent and uploads the network topology graph to the regional control Agent, the protection coordination Agent and the central processing Agent.

The microgrid wide-area current protection method based on the multi-Agent technology in the embodiment is shown in fig. 2.

Step 1, initializing each Agent in the microgrid.

Step 2, the state monitoring Agent transmits the monitored micro-grid state information to a network topology Agent in real time; the microgrid state information comprises: the switching-in position of the distributed power supply, the switching-out state of the distributed power supply and the switching-on and switching-off state of each circuit breaker in the microgrid; meanwhile, the measurement Agent also transmits the microgrid electrical quantity information to the protection Agent in real time, and the microgrid electrical quantity information comprises: voltage magnitude, current magnitude, phase angle and frequency of the microgrid.

and 4, according to the real-time microgrid state information, the network topology Agent draws a real-time microgrid network topology map and transmits the real-time microgrid network topology map to the area control Agent, the protection coordination Agent and the central processing Agent, and meanwhile, the measuring Agent also transmits the microgrid electrical quantity information to the area control Agent in real time.

Step 5, drawing a real-time microgrid network topological graph and microgrid electrical quantity information of a measuring Agent according to a network topological Agent, and making a control scheme of active power of a distributed power supply, reactive power of the distributed power supply, active power of an energy storage unit, reactive power of the energy storage unit, microgrid voltage and microgrid frequency in a microgrid current protection area by a region control Agent and uploading the control scheme to a central processing Agent; the protection coordination Agent formulates a microgrid protection coordination scheme and uploads the microgrid protection coordination scheme to the central processing Agent, and the central processing Agent makes a microgrid current protection scheme by referring to the microgrid protection coordination scheme and the control scheme and feeds the microgrid current protection scheme back to the intelligent electronic equipment layer;

as shown in fig. 4: when PCC is closed and the microgrid operates in a grid-connected mode, distributed power sources DG1 and DG2 in the microgrid are controlled by P-Q (active power-reactive power) to ensure the output of power source power, an energy storage unit is controlled by P-Q, when the voltage or frequency of the microgrid changes, the operating points of DG1 and DG2 change to ensure the invariance of the active power and reactive power output, and the operating points of the energy storage unit also change to perform peak clipping and valley filling to ensure the balance of the active power and reactive power of the whole microgrid; when the PCC is disconnected with the microgrid and operates in an island mode, energy storage devices in the microgrid are controlled by V-F (voltage-frequency) to maintain the stability of the voltage and the frequency of the microgrid, distributed power supplies DG1 and DG2 still adopt P-Q control to ensure the output of power supply power, when the voltage or the frequency of the microgrid changes, energy storage units controlled by V-F can maintain the stability of the voltage and the frequency of the microgrid by adopting a translation operation curve mode, and DG1 and DG2 controlled by P-Q still adopt a mode of changing operation points to ensure that the output of active power and reactive power does not change.

The protection coordination scheme is formulated by taking a line EF as an example, when the line EF breaks down, tripping instructions are sent by protection agents in a main protection IED9 and an IED10 to act on corresponding circuit breakers, tripping information is sent to the protection coordination agents at the same time, after certain time delay, the protection coordination agents detect whether the corresponding circuit breakers are disconnected through the state monitoring agents in the IED9 and the state monitoring agents in the IED10 respectively, if the corresponding circuit breakers are not disconnected due to the fact that the IED9 and the IED10 reject, the protection coordination agents send tripping information to the protection agents in the IED corresponding to an upper-level line AE or the protection agents in the IED corresponding to a lower-level line FG according to the optimal expanded microgrid current protection range, and the protection agents in the corresponding IEDs send tripping instructions to disconnect the circuit breakers to remove the faults.

Step 6, drawing a real-time microgrid network topological graph, a control scheme and a protection coordination scheme made by a region control and protection coordination layer by referring to a network topological Agent, leading a taboo table into a breadth search algorithm by a central processing Agent to form a fusion taboo breadth search algorithm, and dividing a microgrid current protection region in real time by using the fusion taboo breadth search algorithm to obtain a microgrid current protection divided region;

the flow chart of the fusion tabu breadth search algorithm is shown in fig. 3, and comprises the following steps:

the objective function is as follows:

the constraint conditions are as follows:

the delay of transmission of the fault information is less than or equal to a given value of delay of the fault information, i.e.The reliability of communication being greater than or equal to a given reliability value, i.e. P_i＝(P_X1∪P_X2......)∩(P_y1∪P_y2......)≥P_im(ii) a And voltage constraint U_min≤U≤U_maxAnd fault information acquisition amount constraint condition B_min≤B≤B_maxIs a boundary, in which U_maxAnd U_min1.1U and 0.9U, B_maxAnd B_min1.2B and 0.8B respectively; f (x) represents the total outage area；A_iRepresenting the blackout area of each divided area; PS represents a set of measures to minimize the blackout area of each divided region; t is_ijRepresenting the delay time of the fault information; i is_ijRepresenting the communication distance from the node i to the node j in the microgrid; c represents the speed of light; m represents the number of nodes passed by the communication path; t is t_vRepresenting the communication time of a single node; Δ t represents the random jitter delay; t is_ijmA given value representing a delay of the fault information; b represents the fault information acquisition amount; p_iRepresenting the degree of reliability of the communication; p_xiRepresenting the reliability of the communication element; p_yiRepresenting the reliability of the line between the communication elements; p_imA given value representing the reliability of the communication; n, j 1,2.. n, U represents the microgrid bus voltage;

and 6.3, searching the optimal solution of the objective function from the initial solution by using the breadth search algorithm with the given constraint condition as a boundary.

As shown in fig. 4: the optimal microgrid current protection partition area partitioned at a certain moment by utilizing the fusion tabu breadth search algorithm comprises the following steps: the current protection device comprises a current protection region I, a current protection region II, a current protection region III, a current protection region IV, a current protection region V and a current protection region VI. The current protection area I comprises a common coupling Point (PCC) of a transformer substation, a microgrid and a power distribution network; the current protection area II comprises the whole feeder line L1, a LOAD LOAD1 and an energy storage unit; the current protection region iii includes an AC section of the feeder line L2, a LOAD2, and a first distributed power supply DG 1; the current protection region iv includes the remaining part of the feeder line L2 except for the AC section and the LOAD 3; the current protection area V comprises an AE section, an EF section, an FG section, a LOAD LOAD4, a LOAD LOAD5, a LOAD LOAD6 and a second distributed power supply DG2 of a feeder line L3; current protection region vi includes the remainder of feed line L3 except for AE, EF, FG sections and LOAD 7. All the current protection areas can be coordinated and cooperated, so that the purpose of correctly and quickly removing the internal fault of the microgrid is achieved.

Step 7, configuring a set of improved pilot current differential protection and a set of overcurrent protection based on local information in the protection agents in each defined microgrid current protection area;

taking a line FG in a microgrid current protection area v shown in fig. 4 as an example, a protection coordination Agent determines, according to an access state of a DG2 in a network topology, that current amounts of 3-end lines need to be collected for improved longitudinal current differential protection in the protection Agent, and the protection coordination Agent determines, according to the network topology, that an IED11, an IED12, and an IED16 are related to protection of the line FG, allowing measurement agents inside each IED to communicate with the protection Agent, and the protection Agent analyzes electrical quantity information sent from a microgrid by the measurement agents, and adaptively adjusts a protection fixed value of a pre-written improved longitudinal current differential protection criterion according to a microgrid current protection scheme sent by a central processing Agent, including: setting value I of differential current_set1Coefficient of ratiometric braking K₁、K₂The coefficient of influence K of the unstable current of DG on the improved pilot current differential protection₃And inflection point value I of the rate braking current_INT；

The improved pilot current differential protection criterion is as follows:

I_r＝I_INT(k₂-k₁)/k₁k₂

in the formula,. DELTA.I_giThe number of the line port to be collected can be determined according to a network topological graph provided by a network topological Agent for the current quantity collected by each line port, so that the improved criterion of pilot current differential protection is adjusted in a self-adaptive manner, and the influence of DG (distributed generation) plug and play on protection is overcome; k₁、K₂To rate-brake coefficient, I_INTIs the knee value of the proportional braking current, I_set1Is a setting value of a differential current, K₃Is a factor that takes into account the effect of the DG's unstable current on the improved pilot current differential protection;

a microgrid model is built under the PSCAD/EMTDC environment, simulation is carried out when three-phase faults occur in a grid-connected operation state and single-phase ground faults occur in an island operation state on a line FG, and specific parameters in the microgrid model are shown in a table below.

Distributed power supply	Type (B)	Control mode	Rated power	Rated voltage
					DG1	Photovoltaic cell	P-Q control	10KW	400V
DG2	Wind power generator	P-Q control	7.5KW	400V

TABLE 1 distributed Power supply parameters

Line

Length (Km)

R(Ω/km)

X(Ω/km)

AB	3.45	0.126	0.154
				AC	3.40	0.126	0.154
CD	3.40	0.126	0.154
				AE	3.35	0.133	0.167
EF	3.35	0.133	0.167
				FG	3.35	0.133	0.167
GH	3.35	0.133	0.167

TABLE 2 line parameters

Load(s)	Type (B)	Capacity (KVA)	Power factor
				LOAD1	Static load	600	0.991
LOAD2	Static load	600	0.992
				LOAD3	Static load	650	0.992
LOAD4	Static load	600	0.990
				LOAD5	Static load	650	0.993
LOAD6	Static load	650	0.991
				LOAD7	Static load	600	0.992

TABLE 3 load parameters

The capacity of the energy storage unit is set to be 1000KVA, the power factor is 0.991, P-Q control is performed in a grid-connected operation state, and V-F control is performed in an island operation state. In the simulation, all DGs are switched in, three-phase short-circuit faults and single-phase earth faults occur at 0.2S, the duration is 0.05S, and the simulated current waveform curves are respectively shown in FIG. 5(a) and FIG. 5 (b). When a fault starting signal is input at 0.2S, the current on the line FG has a remarkable sudden change, and then the improved pilot current differential protection mentioned in the text starts to act, so that the fault line is cut off. The simulated current waveform curve chart proves that the improved pilot current differential protection can accurately and quickly remove the faults of the internal circuit of the microgrid.

9, when part of fault information is missing or fault information is wrong, the main protection IED rejects or malfunctions, the protection coordination Agent determines the correlation coefficient and the protection action characteristic coefficient of each IED adjacent to the rejected or malfunctioning IED in the microgrid, further constructs a current protection action range output function to obtain a microgrid current protection range solution set which needs to be expanded, then an optimal microgrid current protection action range which needs to be expanded is obtained through ant colony optimization, and finally fault removal is carried out through improved pilot current differential protection;

take the IED9 in the microgrid current protection area v in fig. 4 as an example. First, the IED associated with it is determined: having an association coefficient of 1 with IED9 (1 representing the greatest degree of association) are IED8, IED 10; associated with IED9 by a factor of 2(2 representing a greater degree of association) are IED7, IED 11; associated with IED9 by a factor of 3(3 representing the least associated) is IED 12. The action characteristic coefficients of each IED are-1, 0, 1, i.e. false action, no action, correct action.

Taking the line EF in the microgrid current protection area v in fig. 4 failed, the IED9 is further described as a research object. When the action characteristic coefficient of the IED9 is 1, i.e., correct action, the IED with the associated coefficient of 1, i.e., IED10, is first searched. When the operating characteristic coefficient of this IED is also 1, i.e. F_out＝F₁When the fault line is cut, the fault line can be cut correctly. The protection range between IED9 and IED10 is the optimal current protection range. When the action characteristic coefficient of the IED10 is not 1, the IED is protected from malfunction or failure. Due to failure or error of fault information received by the IED9, the current protection action range is expanded to search for all IEDs with coefficients 2 or 3 associated with the IED9, and the protection characteristic coefficient of the IED9 may not be 1, i.e., protection malfunction or failure. Therefore, the current protection operation range solution set to be expanded includes: (IED9, IED11), (IED9, IED12), (IED8, IED10), (IED7, IED10), (IED8, IED11), (IED8, IED12), (IED7, IED11), and (IED7, IED 12). So that F is now_out＝F₂Or F_out＝F₃There may be multiple protective action scope results. The embodiment then uses the ant colony optimization algorithm to find the optimal current protection action range to be expanded.

As shown in fig. 6, the method for obtaining the optimal microgrid current protection operation range to be expanded by ant colony optimization includes the following steps:

step 9.1, initializing parameters, including: maximum number of iterations N of the ant colony_max300, 1 for pheromone importance factor α, 4 for heuristic function importance factor β, and pheromone volatility factor ρ0.5, 64 total pheromone release amount, and 30 ant colony numbers. The more correct and faster the IED is defined, the greater the pheromone concentration.

Step 9.2, aiming at the determined current protection action scope solution to be expanded (IED9, IED11), (IED9, IED12), (IED8, IED10), (IED7, IED10), (IED8, IED11), (IED8, IED12), (IED7, IED11), (IED7, IED12), randomly placing each ant at different starting points, i.e. setting different initial values for the ant colony algorithm, and setting all the current protection action scopes to be expanded, i.e.: (IED9, IED11), (IED9, IED12), (IED8, IED10), (IED7, IED10), (IED8, IED11), (IED8, IED12), (IED7, IED11), (IED7, IED12) are defined as nodes, and each ant is calculated according to the following formula to determine the node to be accessed next (the node with the highest state transition probability) until all ants have access to all nodes.

Wherein,denotes a state transition probability, τ, of an ant k transitioning from a current node i (i 1,2.... gtn) to a next node j (j 1,2.. gtn)_ij(t) represents the amount of pheromone remaining on the connection between node i and node j at time t η_ij(t) is a heuristic function representing the expected degree of transfer of ants from node i to node j, η_ij(t)＝1/d_ij，d_ijThe distance between the node i and the node j is; all_k(k 1,2 … … m) is the set of current protection action ranges to be accessed by ant k and should be expanded, at the beginning, all_kOf (n-1) elements, i.e. including all nodes except ant k, all over time_kWherein the element is continuously reduced until the element is empty, which means that all the current protection action range to be expanded is completely accessed, α is an pheromone importance degree factor, the larger the value is, the more the element is expressed thatβ is the factor of importance of the heuristic function, the larger the value, the more the heuristic function has, i.e. ants will transfer to the path with the maximum pheromone concentration with a higher probability.

Step 9.3, calculating the path length L of each ant_kAnd (k is 1,2 … … 30), recording an end point corresponding to the path with the maximum pheromone concentration in the current iteration number, namely the current protection action range which is optimally expanded in the current iteration number. Meanwhile, the pheromone concentration on each node connecting path is updated according to the following formula.

τ_ij(t+n)＝(1-ρ)τ_ij(t)+Δτ_ij(t)

τ_ij(t + n) represents the pheromone concentration over n times,indicates the pheromone increment of the kth ant left on the path from the solution i to the solution j in the current iteration, delta tau_ij(t) represents the pheromone increment that all ants left on the path in this iteration.

Step 9.4, if the iteration number N is less than N_maxIf 300, making N equal to N +1, clearing the record table of the path passed by the ant, and returning to step 12.2; otherwise, the calculation is terminated, the optimal solution, namely the path with the maximum pheromone concentration is output, and the corresponding end point of the path is the optimal current protection action range to be expanded. The operation of the ant colony optimization algorithm described above can be used to obtain the current protection operation ranges (IED8, IED11) to be expanded in the present embodiment.

From the above analysis and simulation it can be seen that: the microgrid wide-area current protection method based on the multi-Agent technology can quickly and correctly isolate and remove the internal faults of the microgrid.

Claims

1. A microgrid wide-area current protection method based on multi-Agent technology, characterized in that:

The method is implemented by a micro-grid wide-area current protection system based on multi-agent technology, and the micro-grid wide-area current protection system based on multi-agent technology is divided into three layers: an intelligent electronic device layer, a regional control and protection coordination layer, and a central processing layer. The intelligent electronic device layer is composed of the intelligent electronic device IED configured on each circuit breaker in the micro-grid; the intelligent electronic device layer is responsible for collecting the state information of the micro-grid, and drawing a real-time micro-grid network topology map, and at the same time It is also responsible for collecting the electrical quantity information of the microgrid; and uploading the drawn real-time microgrid network topology map to the regional control and protection coordination layer and the central processing layer when the inter-layer communication is normal, and uploading the electrical quantity information of the microgrid to the regional Control and protection coordination layer; each described IED further includes state monitoring Agent, measurement Agent, network topology Agent and protection agent; area control and protection coordination layer includes area control Agent and protection coordination agent; central processing layer includes central processing Agent; the central processing Agent, regional control Agent and protection coordination Agent are all configured at the outlet end of the low-voltage side of the substation;

The method comprises the steps of:

Step 1. Initialize each Agent;

Step 2. The status monitoring agent transmits the monitored microgrid status information to the network topology agent in real time; the microgrid status information includes: the access location of the distributed power supply, the switching status of the distributed power supply and the status of the microgrid The opening and closing status of each circuit breaker in the circuit; at the same time, the measurement Agent also transmits the electrical quantity information of the microgrid to the protection agent in real time, and the electrical quantity information of the microgrid includes: the voltage quantity, current quantity, phase angle and frequency of the microgrid;

Step 3. If the inter-layer communication is normal, execute step 4; if the inter-layer communication is completely interrupted, the protection agent uses the overcurrent protection based on local information to perform fault removal;

Step 4. According to the real-time micro-grid status information, the network topology agent draws the real-time micro-grid network topology map and transmits it to the regional control agent, protection coordination agent and central processing agent. At the same time, the measurement agent also transmits the micro-grid electrical quantity information to the regional area in real time Control Agent;

Step 5. According to the real-time microgrid network topology map drawn by the network topology Agent and the measured microgrid electrical quantity information of the Agent, the area control Agent formulates the active power of the distributed power supply in the microgrid current protection area, the reactive power of the distributed power supply, The active power of the energy storage unit, the reactive power of the energy storage unit, the control scheme of the microgrid voltage and the microgrid frequency are uploaded to the central processing agent; the protection coordination agent formulates the microgrid protection coordination scheme and uploads it to the central processing agent;

Step 6. According to the real-time microgrid network topology map drawn by the network topology Agent, the control plan and protection coordination plan made by the regional control and protection coordination layer, the central processing Agent introduces the taboo table into the breadth search algorithm to form a fusion tabu breadth search algorithm, using The integrated taboo breadth search algorithm divides the microgrid current protection area in real time, and obtains the division area of microgrid current protection;

Step 7. Referring to the control plan and protection coordination plan made by the regional control and protection coordination layer, the central processing Agent makes a microgrid current protection plan: configure a set of improved protection agents in each defined microgrid current protection area The longitudinal current differential protection and a set of overcurrent protection based on local information; the central processing Agent simultaneously feeds back the microgrid current protection scheme to the intelligent electronic device layer;

Step 8. When the fault information is correct and complete, perform fault removal through the improved longitudinal current differential protection;

The protection agent analyzes the microgrid electrical quantity information sent by the measuring agent, and combines the microgrid current protection scheme sent by the central processing agent to adaptively adjust the protection definition of the pre-written improved longitudinal current differential protection criterion. Values, including: differential current setting value I _set1 , ratio braking coefficients K ₁ , K ₂ , influence coefficient K ₃ of DG unstable current on improved longitudinal current differential protection and ratio braking current inflection point value I _INT , the number of acquisition line ports;

The improved longitudinal current differential protection criterion is as follows:

\{\begin{matrix} {I I}_{d d} > > {I I}_{{set set}_{11}} \\ {I I}_{d d} > > {k k}_{11} {I I}_{z z d d} & {I I}_{d d} \leq \leq {I I}_{I I N N T T} \\ {I I}_{d d} &GreaterEqual; &Greater Equal; {k k}_{22} (({I I}_{z z d d} - - {I I}_{r r})) & {I I}_{d d} > > {I I}_{I I N N T T} \end{matrix}

{I I}_{d d} = = {k k}_{33} | | {Σ Σ}_{i i = = 11}^{n no} {ΔI ΔI}_{g g i i} | |

{I I}_{z z d d} = = | | \sqrt{{Σ Σ}_{i i = = 11}^{n no} {ΔI ΔI}_{g g i i}^{22}} - - Δ Δ {\overset{\cdot &Center Dot;}{I I}}_{g g s the s} | |

Δ Δ {\overset{\cdot \cdot}{I I}}_{g g s the s} = = {Σ Σ}_{i i = = 11}^{n no} {ΔI ΔI}_{g g i i}

I _r =I _INT (k ₂ -k ₁ )/k ₁ k ₂

In the formula, ΔI _gi is the amount of current collected by each line port. The number of collected line ports can be determined according to the network topology map provided by the network topology agent, so as to adaptively adjust the criterion of the longitudinal current differential protection; K ₁ , K ₂ is the ratio braking coefficient, I _INT is the inflection point value of the ratio braking current, I _ser1 is the setting value of the differential current, K ₃ is the coefficient of the influence of the unstable current of DG on the longitudinal current differential protection;

Step 9. When part of the fault information is missing or the fault information is wrong, and the main protection IED refuses to operate or malfunctions, then the protection coordinating agent determines the correlation coefficient and protection Action characteristic coefficient, and then construct the output function of the current protection action range, obtain the solution set of the micro-grid current protection range that should be expanded, and then obtain the optimal micro-grid current protection action range that should be expanded through the ant colony algorithm, and finally through the improved The longitudinal differential current protection performs fault removal;

According to the real-time network topology map provided by the network topology agent, the protection coordination agent determines the correlation coefficient A _f between each IED and its adjacent IEDs:

Define the operating characteristic coefficients for each IED:

Wherein i=1, 2...n; k corresponds to A _f , that is, k=1, 2, 3; n is the number of IEDs in the microgrid;

The expression of the operating range of the current protection is established as follows:

\{\begin{matrix} {A A}_{f f} = = 11 & {F f}_{11} = = (({R R}_{f f I I E E. D D. 11}^{11},, {R R}_{f f I I E E. D D. 22}^{11},, ... ...,, {R R}_{f f I I E E. D D. n no}^{11})) \\ {A A}_{f f} = = 22 & {F f}_{22} = = (({R R}_{f f I I E E. D D. 11}^{22},, {R R}_{f f I I E E. D D. 22}^{22},, ... ...,, {R R}_{f f I I E E. D D. n no}^{22})) \\ {A A}_{f f} = = 33 & {F f}_{33} = = (({R R}_{f f I I E E. D D. 11}^{33},, {R R}_{f f I I E E. D D. 22}^{33},, ... ...,, {R R}_{f f I I E E. D D. n no}^{33})) \end{matrix}

And construct the current protection action range output function F _out = {F ₁ , F ₂ , F ₃ }, where the priority of protection actions of F ₁ , F ₂ , and F ₃ decreases in turn, F ₁ has a solution, and F ₂ and F ₃ There are multiple solutions; according to the output of F _out , the microgrid current protection action range solution set that should be expanded after each IED is misoperated or refused to operate due to fault information error or partial fault information missing can be determined.

2. The micro-grid wide-area current protection method based on multi-Agent technology according to claim 1, characterized in that: in the step 6, the fusion taboo breadth search algorithm is used to divide the micro-grid in real time to obtain the micro-grid current protection division The regional process further includes the following steps:

Step 6.1, establish an objective function for dividing the current protection area of the microgrid, the goal of this function is to minimize the blackout area in the microgrid, and establish constraints for the objective function;

The stated objective function is:

The stated constraints are:

\{\begin{matrix} {T T}_{i i j j} = = \frac{{I I}_{i i j j}}{C C} + + m m {t t}_{v v} + + Δ Δ t t \leq \leq {T T}_{i i j j m m} \\ {B B}_{min min} \leq \leq B B \leq \leq {B B}_{m m a a x x} \\ {P P}_{i i} = = (({P P}_{X x 11} \cup \cup {P P}_{X x 22} ... ... ... ...)) \cap \cap (({P P}_{y the y 11} \cup \cup {P P}_{y the y 22} ... ... ... ...)) &GreaterEqual; &Greater Equal; {P P}_{i i m m} \\ {U u}_{min min} \leq \leq U u \leq \leq {U u}_{max max} \end{matrix}

Among them, F(x) represents the total power outage area; A _i represents the power outage area of each divided area; PS represents the measure set to minimize the power outage area of each divided area; _T _ij represents the delay time of fault information; The communication distance from point i to node j; C represents the speed of light; m represents the number of nodes passed by the communication path; t _v represents the communication time of a single node; Δt represents the random jitter delay; T _ijm represents the given fault information delay B represents the amount of fault information collected; P _i represents the degree of communication reliability; P _xi represents the reliability of communication components; P _yi represents the reliability of lines between communication components; P _im represents the given value of communication reliability; i =1, 2...n, j=1, 2...n, U represents the microgrid bus voltage;

Step 6.2, adopt the breadth search algorithm, and give the initial solution of the breadth search, that is, the initial divided area of the microgrid current protection, the initial solution can be generated by a random method or by an existing heuristic method;

Step 6.3, with the constraints given above as the boundary, use the breadth search algorithm to search for the optimal solution of the above objective function from the initial solution;

Step 6.4, liberate the searched local optimal solution into the tabu table, the tabu table is updated in each iteration, avoid the local optimal solution in the next search, and search other spaces until the global The optimal solution, that is, the optimal microgrid current protection division area that satisfies the above objective function and constraints.