CN104200296A - Wind-solar-stored energy-firewood autonomous micro-grid group cross-domain collaboration energy dispatching and fitting method - Google Patents

Abstract

The invention discloses a cross-domain collaborative energy scheduling and adaptation method of an autonomous micro-grid group for wind, solar and firewood storage. The method constructs the microgrid topology matrix through reasonable generation matrix weights; the improved MST algorithm is used to determine the optimal DG-CL and S-CL power supply correspondence, thereby obtaining the "trunk"; the LMI algorithm is used to select NL, and its It is added to the "trunk" in the form of "leaves" to jointly generate a new autonomous microgrid structure. The present invention uses the IEEE33-bus network as a test system to provide a detailed algorithm description, and through a series of experiments to prove the effectiveness of the proposed strategy in improving the overall power utilization rate of the system and the continuous and safe power supply to important loads.

Description

A Cross-Domain Cooperative Energy Scheduling and Adaptation Method for Autonomous Microgrid Group of Solar Energy Storage and Firewood

technical field

The invention relates to the field of distributed power generation and energy storage equipment of a microgrid system, in particular to an autonomous microgrid group cross-domain coordinated energy scheduling and adaptive optimization cooperative operation method including wind, solar, storage, and diesel power sources.

Background technique

With the maturity of the autonomous micro-grid system, this kind of power network system that can operate independently or grid-connected has adopted more renewable energy technologies. However, although there are relatively stable and reliable power supply systems such as internal combustion generators and gas turbine generators among the currently commonly used distributed power sources, their scale and power supply will be limited to a certain extent because they need to consume traditional energy for power supply, and cannot fully meet The power consumption requirements of all power loads in the autonomous microgrid, and solar and wind power generation systems are more infiltrated into the autonomous microgrid system because of their clean and renewable characteristics. However, due to the influence of weather, environment and other factors, the latter has the characteristics of intermittent and fluctuating power supply, so it cannot provide continuous and stable power supply. This will reduce the operating efficiency of the autonomous microgrid system from two aspects: on the one hand, when the distributed power supply (DG) is sufficient, a large amount of electric energy will not be effectively utilized. Although the energy storage device can absorb part of the excess power at this time, its effect is limited and a large number of energy storage units are required, which greatly increases the investment and maintenance cost of the autonomous microgrid system; on the other hand, when the DG power supply is insufficient, the load The loss caused by the lack of sufficient power supply will be more serious, especially when the power supply of the important load (CL) in the system is insufficient.

The present invention aims at the "tertiary level" in the autonomous microgrid system and considers the monitoring equipment at the connection switch of the power system, such as multi-agents, to obtain system status information and use it to control the contact switch, and then from the logical level Reconstruct the network topology, re-optimize and combine the power supply and consumption systems of multiple autonomous micro-grid systems, and realize the optimal dispatch of electric energy through collaboration. The "stem and leaf generation strategy" mainly includes two parts of the algorithm, that is, the minimum spanning tree (MST) searches for the optimal DG-CL power supply relationship to determine the basic system of the network; the linear matrix inequality (LMI) determines the non-important load (NL) node Add or delete to ensure the maximum utilization of DG electric energy. Its innovation and technical contributions are mainly reflected in the following aspects: (1) Considering multiple factors to construct network topology matrix weights, and organically unifying these low-correlation metric values, thus greatly reducing the number of individual constraints. The number laid the foundation for the simplification of the algorithm; (2) Combine the MST search algorithm and the LMI optimization algorithm to construct the "trunk" and "leaves" under the new autonomous microgrid topology, which not only utilizes the respective advantages of the algorithm but also It is convenient to combine and split the algorithm at any time (it can be re-divided according to the system state or only increase or decrease the "leaf" nodes), thus avoiding the re-running of the algorithm after each sampling, greatly simplifying the calculation amount and improving the Algorithm efficiency; (3) Reasonably arrange the charging and discharging state of the energy storage unit, so that the strategy can optimize the arrangement of electric energy from two dimensions of space and time, so that the utilization rate of electric energy can be improved to a greater extent; at the same time, minimize The number of charging and discharging times of the energy storage device is also avoided, and the charging and discharging action from the energy storage device (S) to the energy storage device (S) is avoided, thereby reducing the use cost of the energy storage unit and prolonging its service life. A series of simulation structures under the International Institute of Electrical and Electronics Engineers (IEEE) 33-node standard test platform show that this strategy effectively ensures the power consumption of CL, improves the efficiency of DG power use, and greatly guarantees the power consumption requirements of NL .

Contents of the invention

In view of the deficiencies in the prior art, the purpose of the present invention is to provide a method for cross-domain collaborative energy scheduling and adaptation of an autonomous microgrid group for wind, solar and firewood storage.

The object of the present invention is achieved by the following technical means, and concrete implementation steps are as follows:

Step (1), take the time interval Δt as the sampling time, periodically monitor the real-time output and power consumption of DG and load in the system, and obtain the working status information of the system at the current moment:

I) When the output power of DG in the system is greater than the power demand of all loads, that is , keep the current autonomous micro-grid structure, and perform step (4); wherein N _DG is the total number of DG, N _CL is the total number of CL, and N _NL is the total number of NL;

All the loads mentioned include important load (CL) and non-important load (NL);

II) When the output power of DG in the system is greater than the power demand of all important loads, but cannot meet the requirements of all loads, that is $(\underset{i\_\mathrm{CL}=1}{\overset{N_{\mathrm{CL}}}{\mathrm{\Σ}}}{\mathrm{CL}}_{i\_\mathrm{CL}}+\underset{i\_\mathrm{NL}=1}{\overset{N_{\mathrm{NL}}}{\mathrm{\Σ}}}{\mathrm{NL}}_{i\_\mathrm{nl}})>\underset{i\_\mathrm{DG}=1}{\overset{N_{\mathrm{DG}}}{\mathrm{\Σ}}}{\mathrm{DG}}_{i\_\mathrm{DG}}>\underset{i=1}{\overset{N_{\mathrm{CL}}}{\mathrm{\Σ}}}{\mathrm{CL}}_{i\_\mathrm{CL}},$ Then execute step (2);

III) When the output power of DG in the system cannot meet the power demand of all important loads, that is When , delete some important loads to ensure that as many important loads CL as possible can be supplied with power supply, and at the same time ensure that the output load of DG is fully utilized, and then perform step (6);

Step (2), when judging that the system is in state II), whether the change ratio of i_AMG (DG-CL) in each autonomous microgrid _{i_AMG} exceeds the set division trigger threshold θ, namely i_AMG=1,...,N _AMG (t-Δt), if so, re-divide the autonomous microgrid structure, that is, execute step (3); otherwise, add or delete some non-important loads NL, that is, execute step (17); where N _AMG (t-Δt) is the number of autonomous microgrids existing at (t-Δt) time, and are the power difference between DG and CL in the autonomous microgrid i_AMG at time (t-Δt) and time t, respectively;

Step (3), calculate the real-time network loss value and risk coefficient of each branch line in the current state of the micro-grid, and weight the normalized value of the two to form the branch weight value of the system, and then use this weight as a matrix element , to construct the network topology matrix A(t) at the current moment t:

Assuming that a load j in the autonomous microgrid has active power and reactive power P _j and Q _j respectively, and its upstream power supply node is i, then the total resistance and total resistance of the power supply line from node i to node j The reactances are R _ij and X _ij respectively; assuming that the voltage of node j remains U _j , the real-time network loss value of the line on the power supply line transmitted from node i to node j can be expressed as:

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; loss</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; =</span>\frac{{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{{\mathrm{<span\; class="notranslate">\; Q</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{{{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

The real-time network loss value of the line between any two nodes in the network can be obtained by using the expression (1).

In the process of power system reliability evaluation, the unavailability K of lines and components is a commonly used measure, which is determined by the annual line fault frequency f and the line repair time r, that is

$\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; r</span>}}{\mathrm{<span\; class="notranslate">\; 8760</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

In addition to calculating the unavailability using the statistical values of the fault frequency of the line and the repair time of the line, the expert evaluation value of the line is also an important factor. Therefore, combining the two comprehensively evaluates the risk coefficient of each line of the system

${\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; risk</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; \&eta;</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&eta;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; E.</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$

Among them, K _ij is the unavailability of the line between node i and node j, obtained by expression (2); E _ij is the expert evaluation value of the line between node i and node j; η is the adjustment factor, which can be adjusted The proportion of the unavailability of the line and the expert evaluation value of the line in the risk assessment process.

In order to better unify the two into one metric, we first need to normalize the two.

Assuming that the line between node i and node j is L _ij , the normalized real-time network loss value of the line and the risk coefficient of the line are respectively:

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; the\; norm</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; loss</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; loss</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}}{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ;</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&NotEqual;</span>\mathrm{<span\; class="notranslate">\; j</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; loss</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

${\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; the\; norm</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; risk</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; risk</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}}{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ;</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&NotEqual;</span>\mathrm{<span\; class="notranslate">\; j</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; risk</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 5</span>}<span\; class="notranslate">\; )</span>$

Among them, N is the number of nodes in the entire network system;

Using the normalized line real-time network loss value and line risk coefficient, the final branch weight of the line L _ij is obtained as:

${\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; \&beta;</span>}<span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; the\; norm</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; loss</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&beta;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; the\; norm</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; risk</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

β is an adjustment factor, which can be adjusted according to the actual situation.

Assuming that there is a direct connection between any two nodes i and j in the network, the weight of the two in the network topology matrix A(t) is obtained by expression (6); otherwise, if the _two If nodes i and j are not directly connected, their weights in the network topology matrix A(t) are a _ij =0; in addition, the diagonal elements in the network topology matrix A(t) are defined as a _ii =0 .

Taking a _ij as the matrix weight between node i and node j, the weight between any two points of the system can be obtained; using it as a matrix element, the topological relationship matrix A ^Δt of the system at any time can be obtained, and then the steps ( 5);

Step (4), when it is judged that the system is in state I), that is, when the output power of the DG in the system is greater than the power demand of all loads, all energy storage devices in the system are used as power loads, and then according to their own state of charge (SOC) for further judgment, execute step (7);

Step (5), judging that the energy storage device is regarded as a load (load) or a power supply (generator), further judgment needs to be made according to its own state of charge (SOC), and steps (7) and (8) are performed at the same time ;

Step (6), when it is judged that the system is in state III), that is, the total power supply load of the system DG is less than the power consumption load of all electrical equipment in the system, then all energy storage equipment in the system are regarded as power sources, and need to be based on their own loads Electric state (SOC) is further judged, and execution step (8);

Step (7), judging whether the SOC state of each energy storage device is greater than the maximum capacity threshold (SOC _max ), if so, perform step (9); otherwise, perform step (10);

Step (8), judging whether the SOC state of each energy storage device is less than the minimum capacity threshold (SOC _min ), if so, perform step (9); otherwise, perform step (11);

Step (9), if the state of charge of an energy storage device satisfies SOC>SOC _max or SOC _min >SOC, then the energy storage device does not operate, neither as a power supply nor as a power consumption device, and execute the step (12);

Step (10), if a certain energy storage device is not greater than its maximum capacity threshold (SOC _max ), and at this time the system meets the electricity demand that the output power of DG is less than the full load, then the energy storage device is used as a power consumption load, And execute step (12);

Step (11), if a certain energy storage device is not less than its minimum capacity threshold (SOC _min ), and at this time the system meets the power demand that the output power of DG is greater than the full load, then the energy storage device is used as a power source, and Execute step (12);

Step (12), take the node where each power supply (including DG and the energy storage device determined in step (11)) in the autonomous microgrid is located as the root node, search for the minimum spanning tree; calculate each minimum spanning tree, from the root node The minimum weight sum to each critical load (CL) node;

With the minimum weight sum as the goal, it is decided to assign each important load to each power supply and be responsible for supplying power to it;

Step (13), select the power supply corresponding to the minimum value among all the "minimum weight sums" of a certain important load (CL) as the power supply node of the CL;

Step (14), according to step (13), determine the set of power supply relationships corresponding to all "power supply (G)-important load (CL)";

Step (15), judging whether the power supply load of the power supply in the power supply relationship set corresponding to each "power supply (G)-important load (CL)" is greater than the power consumption requirement of the important load, if so, perform step (17), if not, perform step (16);

Step (16), select the power supply corresponding to the second smallest value in the "minimum weight sum" of the CL as the power supply node of the CL and perform step (17);

Step (17), based on the principle that the electric energy generated by the power supply is fully utilized to the maximum extent, using the LMI algorithm to determine the number of non-important loads NL of various types between the power supply (G) and the important load (CL);

Step (18), select the non-important load by the number determined in step (17) from various types of non-important loads (NL) between the power supply (G) and the important load (CL), and select the non-important load with the power supply (G) Together with the important load (CL), a new autonomous microgrid is formed;

Step (19), assign all nodes in the system to different new autonomous microgrids according to the above steps, thereby forming a new autonomous microgrid power supply system.

The beneficial effects of the present invention are: the present invention has the characteristics of volatility, randomness and intermittent in distributed power supply; Unified, thus greatly reducing the number of constraints and laying the foundation for the simplification of the algorithm; (2) Combining the MST search algorithm and the LMI optimization algorithm to use the "trunk" and "trunk" under the new autonomous microgrid topology respectively to construct "leaves", which not only makes use of the respective advantages of the algorithm, but also facilitates the combination and splitting of the algorithm at any time (it can be re-divided according to the system state or only increase or decrease the "leaf" nodes), thus avoiding the need to update the algorithm after each sampling Re-running greatly simplifies the amount of calculation and improves the efficiency of the algorithm; (3) rationally arrange the charge and discharge states of the energy storage unit, so that the strategy can optimize the arrangement of electric energy from two dimensions of space and time, so that the utilization rate of electric energy At the same time, the charging and discharging times of the energy storage device are reduced as much as possible and the charging and discharging action from S to S is avoided, thereby reducing the use cost of the energy storage unit and prolonging its service life. It should be noted that (1) in order to make the energy storage device have a longer service life, there is usually a certain margin for charging and discharging. This paper selects the charging and discharging range as 20%-80% of the maximum storage capacity. (2) According to the overall power supply situation in the system, this strategy regards all energy storage devices as "power consumption" or "discharge" devices, which has certain practical significance: this strategy can effectively avoid the transfer from one energy storage device to Another action of charging the energy storage device, thus avoiding the repeated charge and discharge "shock" between the batteries. (3) This strategy first ensures that all CLs and NLs are fully supplied with power when charging the energy storage device, that is, according to the power supply priority of CL>NL>S, which can minimize the charging of the energy storage device Discharge action, thereby prolonging the service life of the battery and reducing operating costs. This scheduling strategy can effectively realize the matching of input and output power under the condition of insufficient output power, and realize the coordination of unified power consumption of multi-autonomous microgrids. On the basis of ensuring sufficient power supply for important loads, it also improves the safety of power consumption for important loads in the entire system to a certain extent, and at the same time can realize the effective utilization of electric energy among multi-autonomous microgrids.

Description of drawings

Fig. 1 is the network topology matrix method;

Figure 2 is the working area of energy storage equipment;

Fig. 3 is a flow chart of the dispatching strategy method of the present invention;

Figure 4 shows the IEEE33 node system test topology and initial network division;

Figure 5 is the DG and load characteristic curve; where (a) is the DG and load characteristic curve, (b) is the load characteristic curve

Figure 6(a) is the MST generated with DG ₁ as the root node at 00:00;

Figure 6(b) is the MST generated with DG ₂ as the root node at 00:00;

Figure 6(c) is the MST generated with DG ₃ as the root node at 00:00;

Figure 7(a) is the MST generated with DG ₁ as the root node at 05:00;

Figure 7(b) shows the MST generated with DG ₂ as the root node at 05:00;

Figure 7(c) shows the MST generated with DG ₃ as the root node at 05:00;

Figure 7(d) is the MST generated with _S1 as the root node at 05:00;

Figure 7(e) is the MST generated with _S2 as the root node at 05:00;

Figure 7(f) is the MST generated with S ₃ as the root node at 05:00;

Figure 8(a) is the result of collaborative scheduling at 00:00;

Figure 8(b) is the result of collaborative scheduling at 05:00;

Figure 9 shows the comparison of CLs24h power supply under the initial structure and scheduling structure;

Figure 10 is a comparison curve of DG utilization;

Figure 11(a) is a diagram of the load satisfaction status without using the strategy;

Figure 11(b) is a diagram of the load satisfaction status of using this strategy but without energy storage equipment in the system;

Figure 11(c) is a diagram of the load satisfaction status using this strategy;

Figure 12 shows the results of 24-hour collaborative scheduling.

Detailed ways

Below in conjunction with accompanying drawing, the present invention is described in further detail:

What Fig. 3 shows is the processing flowchart of this invention. Its specific implementation will be described as follows in conjunction with specific examples. The specific steps for dividing the network will be described below using the IEEE33 node topology, and its topology structure is shown in FIG. 4 .

As can be seen from Figure 4, there are 3 DGs in the entire network, where it is assumed that DG ₁ is photovoltaic (0-624.205MW), DG ₂ and DG ₃ are wind energy (82.01-419.50MW), and its power characteristic curve comes from the Belgian transmission operation Shang Elias (Belgian electricity transmission operator Elias) (May13 ^th ,2014), as shown in Figure 5(a). Assuming that each energy storage unit has a maximum capacity of 900MWh, the maximum total capacity of the energy storage system in the system shown in Figure 4 is 2700MWh. In addition, the system shown in Figure 4 contains 6 important loads and 21 non-important loads, and its typical 24h operating characteristic curve is shown in Figure 5(b). The connection types of each node are shown in Table I.

Table I

IEEE33 Node Characteristics Table

Step (1), taking the time interval Δt as the sampling time, periodically monitor the real-time output and power consumption of DG and loads in the system, obtain the working status information of the system at the current moment, and determine its status classification according to its working status:

I) When the output power of DG in the system is greater than the power demand of all loads, that is , keep the current autonomous micro-grid structure, and perform step (4); wherein N _DG is the total number of DG, N _CL is the total number of CL, and N _NL is the total number of NL;

All the loads mentioned include important load (CL) and non-important load (NL);

II) When the output power of DG in the system is greater than the power demand of all important loads, but cannot meet the requirements of all loads, that is $(\underset{i\_\mathrm{CL}=1}{\overset{N_{\mathrm{CL}}}{\mathrm{\&Sigma;}}}{\mathrm{CL}}_{i\_\mathrm{CL}}+\underset{i\_\mathrm{NL}=1}{\overset{N_{\mathrm{NL}}}{\mathrm{\&Sigma;}}}{\mathrm{NL}}_{i\_\mathrm{nl}})>\underset{i\_\mathrm{DG}=1}{\overset{N_{\mathrm{DG}}}{\mathrm{\&Sigma;}}}{\mathrm{DG}}_{i\_\mathrm{DG}}>\underset{i=1}{\overset{N_{\mathrm{CL}}}{\mathrm{\&Sigma;}}}{\mathrm{CL}}_{i\_\mathrm{CL}},$ Then execute step (2);

III) When the output power of DG in the system cannot meet the power demand of all important loads, that is When , delete some important loads to ensure that as many CLs as possible can be supplied with power, and at the same time ensure that the output load of DG is fully utilized, then perform step (6);

Step (2), when judging that the system is in state II), whether the change ratio of i_AMG (DG-CL) in each autonomous microgrid _{i_AMG} exceeds the set division trigger threshold θ, namely i_AMG=1,...,N _AMG (t-Δt), if so, re-divide the autonomous microgrid structure, that is, execute step (3); otherwise, add or delete some non-important loads NL, that is, execute step (17); among them, N _AMG (t-Δt) is the number of autonomous microgrids existing at the time (t-Δt); and are the power difference between DG and CL in the autonomous microgrid i_AMG at time (t-Δt) and time t, respectively;

Step (3), calculate the network loss and risk coefficient of each branch in the current state of the microgrid, and weight the normalized value of the two to form the branch weight of the system, and then use the weight as a matrix element to construct the current The network topology matrix A(t) at time t, the specific steps are as follows:

Assuming that a load j in the autonomous microgrid has active power and reactive power P _j and Q _j respectively, and its upstream power supply node is i, then the total resistance and total resistance of the power supply line from node i to node j The reactances are R _ij and X _ij respectively; assuming that the voltage of node j remains U _j , the real-time network loss value of the line on the power supply line transmitted from node i to node j can be expressed as:

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; loss</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; =</span>\frac{{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{{\mathrm{<span\; class="notranslate">\; Q</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{{{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

The real-time network loss value of the line between any two nodes in the network can be obtained by using the expression (1).

In the process of power system reliability evaluation, the unavailability K of lines and components is a commonly used measure, which is determined by the annual fault frequency f of lines and the repair time r of lines, that is

$\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; r</span>}}{\mathrm{<span\; class="notranslate">\; 8760</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

In addition to calculating the unavailability using the statistical values of the fault frequency of the line and the repair time of the line, the expert evaluation value of the line is also an important factor. Therefore, combining the two comprehensively evaluates the risk coefficient of each line of the system

${\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; risk</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; \&eta;</span>}<span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&eta;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; E.</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$

Among them, K _ij is the unavailability of the line between node i and node j, obtained by expression (2); E _ij is the expert evaluation value of the line between node i and node j; η is the adjustment factor, which can be adjusted The proportion of the unavailability of the line and the expert evaluation value of the line in the risk assessment process.

In order to better unify the two into one metric, we first need to normalize the two.

Assuming that the line between node i and node j is L _ij , the normalized real-time network loss value of the line and the risk coefficient of the line are respectively:

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; the\; norm</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; loss</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; loss</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}}{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ;</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&NotEqual;</span>\mathrm{<span\; class="notranslate">\; j</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; loss</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

${\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; the\; norm</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; risk</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; risk</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}}{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ;</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&NotEqual;</span>\mathrm{<span\; class="notranslate">\; j</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; risk</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 5</span>}<span\; class="notranslate">\; )</span>$

Among them, N is the number of nodes in the entire network system;

Using the normalized line real-time network loss value and line risk coefficient, the final weight of the line L _ij is obtained as:

${\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; \&beta;</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; the\; norm</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; loss</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&beta;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; the\; norm</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; risk</span>}}^{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

β is an adjustment factor, which can be adjusted according to the actual situation.

Assuming that there is a direct connection between any two nodes i and j in the network, the weight of the two in the network topology matrix A(t) is obtained by expression (6); otherwise, if the _two If nodes i and j are not directly connected, their weights in the network topology matrix A(t) are a _ij =0; in addition, the diagonal elements in the network topology matrix A(t) are defined as a _ii =0 . Accordingly, an example of constructing a network topology matrix is shown in Figure 1:

In Figure 1, if nodes 1 and 2 are connected, their weight is obtained from (1)-(6), which is a ₁₂ (or a ₂₁ , a ₁₂ = a ₂₁ ); while nodes 3 and 4 are not connected, then Its weight a ₁₃ =a ₃₁ =0; in addition, the diagonal element a ₁₁ =a ₂₂ =a ₃₃ =a ₄₄ =a ₅₅ =a ₆₆ =0 in FIG. 1 .

Taking a _ij as the matrix weight between node i and node j, the weight between any two points of the system can be obtained; using it as a matrix element, the topological relationship matrix A ^Δt of the system at any time can be obtained, and then the steps ( 5);

Step (4), when it is judged that the system is in state I), that is, when the output power of the DG in the system is greater than the power demand of all loads, all energy storage devices in the system are used as power loads, and then according to their own state of charge (SOC) for further judgment, execute step (7);

Step (5), judging that the energy storage device is regarded as a load (load) or a power supply (generator), further judgment needs to be made according to its own state of charge (SOC), and steps (7) and (8) are performed at the same time ;

Step (6), when it is judged that the system is in state III), that is, the total power supply load of the system DG is less than the power consumption load of all electrical equipment in the system, then all energy storage equipment in the system are regarded as power sources, and need to be based on their own loads Electric state (SOC) is further judged, and execution step (8);

Step (7), judging whether the SOC state of each energy storage device is greater than the maximum capacity threshold (SOC _max =80%, as shown in Figure 2), if so, execute step (9); otherwise execute step (10);

Step (8), judging whether the SOC state of each energy storage device is less than the minimum capacity threshold (SOC _min =20%, as shown in Figure 2), if so, execute step (9); otherwise execute step (11);

Step (9), if the state of charge of an energy storage device satisfies SOC>SOC _max (in this paper, SOC _max =80%, as shown in Figure 2) or SOC _min >SOC (in this paper, SOC _min =20%, as shown in Figure 2 2), then the energy storage device does not act, neither as a power supply nor as a power consumption device, and execute step (12);

Step (10), if a certain energy storage device is not greater than its maximum capacity threshold (SOC _max = 80%, as shown in Figure 2), and at this time the system meets the power demand that the output power of the DG is less than the full load, then the The energy storage device is used as an electric load, and step (12) is performed;

Step (11), if a certain energy storage device is not less than its minimum capacity threshold (SOC _min = 20%, as shown in Figure 2), and at this time the system meets the power demand that the output power of the DG is greater than the full load, then the The energy storage device is used as a power source, and step (12) is performed;

Step (12), take the node where each power supply (including DG and the energy storage device determined in step (11)) in the autonomous microgrid is located as the root node, search for the minimum spanning tree; calculate each minimum spanning tree, from the root node The minimum weight sum to each critical load (CL) node;

With the minimum weight sum as the goal, it is decided to assign each important load to each power source and be responsible for supplying power to it.

Step (13), select the power supply corresponding to the minimum value among all the "minimum weight sums" of a certain important load (CL) as the power supply node of the CL;

Step (14), according to step (13), determine the set of power supply relationships corresponding to all "power supply (G)-important load (CL)";

Step (15), judging whether the power supply load of the power supply in the power supply relationship set corresponding to each "power supply (G)-important load (CL)" is greater than the power consumption requirement of the important load, if so, perform step (17), if not, perform step (16);

Step (16), select the power supply corresponding to the second smallest value in the "minimum weight sum" of the CL as the power supply node of the CL and perform step (17);

Step (17), based on the principle that the electric energy generated by the power supply is fully utilized to the maximum extent, using the LMI algorithm to determine the number of non-important loads NL of various types between the power supply (G) and the important load (CL);

Step (18), select the non-important load by the number determined in step (17) from various types of non-important loads (NL) between the power supply (G) and the important load (CL), and select the non-important load with the power supply (G) Together with the important load (CL), a new autonomous microgrid is formed;

Step (19), assign all nodes in the system to different new autonomous microgrids according to the above steps, thereby forming a new autonomous microgrid power supply system.

Therefore, it can be seen from the DG load curve shown in Figure 4 that at 00:00, the system That is, the total output is greater than the total demand. Therefore, at this time, all energy storage devices in the system are treated as electric loads and decide whether to charge them according to their own SOC conditions (according to step 9), and the system only uses DG as the root node to search MST to determine the power supply of CL. The MST from DG to CL obtained according to the network topology matrix A (00:00) at 00:00 is shown in Fig. 6 . According to the MST shown in Fig. 6, calculate the sum of weights from each DG to each CL, as shown in Table II. The bold font indicates the smallest weight sum among the three root nodes from a certain CL, that is, the power supply root node of the CL. However, it should be pointed out that the power supply output power of DG ₁ is 0 at 00:00, therefore, according to the proposed strategy, the nodes 7, 8 and 21CL responsible for power supply follow the principle of the next smallest weight (according to step 16), The other two DGs are respectively arranged to be responsible for the power supply, as shown in the underlined value in Table II. Accordingly, the 00:00 network system is reconstructed into two autonomous micro-grid subsystems, and according to the power supply capacity of DG ₂ and DG ₃ and the load requirements of other NLs at this time, with the goal of making full use of the remaining power of DG, according to the LMI The algorithm (per step 17) determines the non-essential load nodes to add to each sub-autonomous microgrid. The result of collaborative scheduling at 00:00 is shown in Figure 8(a).

Table II

The sum of weights from DGS to CLS

Different from 00:00, at 05:00 the system At this time, according to the proposed strategy, all energy storage devices (S) in the system are used as power sources and decide whether to discharge according to their own SOC status (step 10). According to the network topology matrix A (05:00) at 05:00, the MST with DGs and Ss as root nodes is shown in Figure 7. At this time, the sum of weights from each root node to each CL is shown in Table III. According to the results of Table III, the network system was reconstructed into five subsystems at 05:00, and according to the power supply capacity of DG ₂ , DG ₃ , S ₁ -S ₃ and the load requirements of other NL at this time, according to the LMI algorithm (step 17 ) to add "leaf" non-important loads to each autonomous microgrid, so as to maximize the use of excess power of the power supply to achieve full utilization of power. The result of collaborative scheduling at 05:00 is shown in Figure 8(b).

Table III

The sum of weights from DGS to CLS

Assume that the system consists of three autonomous microgrid systems in the initial state, as shown in Figure 4. Each autonomous microgrid has a DG, an energy storage unit and two CLs. Under the initial structure and the scheduling proposed in this paper, the comparison of the overall power supply of the system CL is shown in Figure 9. From the results shown in Figure 9, it can be seen that the satisfaction rate of CL in the system under the unified scheduling of the cooperative strategy is significantly higher than that of CL under the original autonomous microgrid system structure. Since CL load has greater significance and value to the system than NL, this proves that this strategy has effective economic value from one aspect.

Figure 10 shows the comparison curve of system DG power generation utilization. From Figure 10, it can be clearly seen that the utilization rate of DG power generation under the action of the autonomous microgrid cooperative scheduling strategy proposed in this paper is higher than that without the cooperative scheduling strategy (the small window in Figure 10 shows the DG power utilization efficiency percentage). By reconstructing the architecture of the autonomous microgrid system and reasonably scheduling the charging and discharging actions of the energy storage devices, the electric energy generated by the DG is utilized by the electric load or stored by the energy storage devices, and released when the system lacks power supply. The system can effectively optimize the use of power from two dimensions of space and time, thus improving the efficiency of DG power utilization, which is of great significance in the actual power grid.

Figure 11 shows the curve of the overall load satisfaction rate comparison. It can be seen from Figure 11 that when the system includes an energy storage unit, the algorithm can guarantee the power demand of the entire load of the system most of the time, as shown in Figure 11(a). However, if the energy storage equipment in the system has a large enough capacity, combined with a reasonable autonomous microgrid cooperative scheduling strategy, it can ensure the power consumption of all loads in the whole time period. Secondly, even if there is no energy storage device in the system, due to the coordination and dispatching effect of this strategy, the electricity demand of all loads can be satisfied in most of the time period, but only in part of the time when the total power output of the system When it is less than the total demand for electric energy, some NLs will not receive power supply, as shown in Figure 11(b). In stark contrast, even if there are energy storage devices in the system, without reasonable scheduling, the system still cannot meet the power consumption requirements of the entire load most of the time, as shown in Figure 11(c). Figure 11 fully illustrates the significance and value of reasonable scheduling of autonomous microgrids, and at the same time shows that a reasonable cooperative scheduling strategy is more important than simply adding energy storage devices to the system for the optimal use of power resources.

Figure 12 shows the results of the cooperative scheduling strategy proposed in this paper for all the power supply and load power consumption of the system in a day. From the results shown in Figure 12, it can be seen that through the reasonable charging and discharging scheduling of energy storage equipment, the energy generated by DG in the time period of 00:00-04:30, 07:30-08:30 and 11:00-13:00 The excess power is fully absorbed, so that the insufficient power demand from 05:00-07:00, 09:00-10:30 and 18:30-21:00 is effectively supplemented; at the same time, at 00:00-22: In the 30 time period, all important load power consumption has been fully utilized, and only the important load power demand during 22:30-24:00 has no power output due to DG and energy storage, that is, the total power supply in the system is less than CL total demand, so no matter how the scheduling can not meet and have to give up part of the CL power supply. But in actual operation, this part of the power gap can be compensated by purchasing power from the public grid. In addition, for non-important loads, through the cooperative scheduling in this paper, the NL power demand in the system can be fully satisfied in most time periods. Only when the total power demand of the system is greater than the total supply, it cannot be met through scheduling, and This part of electricity can still be obtained by buying electricity.

Claims (2)

1. the cross-domain collaborative energy scheduling of the autonomous microgrid group of wind-light storage bavin and an adaptation method, is characterized in that the method comprises the following steps:

Step (1), taking time interval Δ t as the sampling time, periodically the exerting oneself in real time and electricity consumption situation of DG and load in supervisory system, obtain the work state information of current time system:

I) in the time that the output power of DG in system is greater than the need for electricity of whole loads, time, keep current autonomous microgrid structure, and execution step (4); Wherein N _dGfor total number of DG, N _cLfor total number of important load, N _nLfor total number of non-important load;

Described whole loads comprise important load, non-important load;

II) be greater than the need for electricity of whole important loads when the output power of DG in system, but cannot meet whole burden requirement time, $(\underset{i\_\mathrm{CL}=1}{\overset{N_{\mathrm{CL}}}{\mathrm{\&Sigma;}}}{\mathrm{CL}}_{i\_\mathrm{CL}}+\underset{i\_\mathrm{NL}=1}{\overset{N_{\mathrm{NL}}}{\mathrm{\&Sigma;}}}{\mathrm{NL}}_{i\_\mathrm{nl}})>\underset{i\_\mathrm{DG}=1}{\overset{N_{\mathrm{DG}}}{\mathrm{\&Sigma;}}}{\mathrm{DG}}_{i\_\mathrm{DG}}>\underset{i=1}{\overset{N_{\mathrm{CL}}}{\mathrm{\&Sigma;}}}{\mathrm{CL}}_{i\_\mathrm{CL}},$ Execution step (2);

III) in the time that the output power of DG in system cannot meet whole important load need for electricity, time, part important load is done to delete and process, ensure that as far as possible many important loads obtain power supply and meet, ensure that the output load of DG is fully utilized, i.e. execution step (6) simultaneously;

Step (2), when judgement system is in state I I) time, in each autonomous microgrid i_AMG (DG-CL) _{i_AMG}variation than the division trigger gate limit value θ that whether exceedes setting, i_AMG=1 ..., N _aMG(t-Δ t), if repartition autonomous microgrid structure, performs step (3); Add if not or the non-important load of the non-important load of deletion, i.e. execution step (17); Wherein N _aMG(t-Δ be t) (t-Δ is the number of autonomous microgrid that exists of moment t), with be respectively that (t-Δ is the power of DG and important load poor in autonomous microgrid i_AMG of moment and t moment t);

Real time power loss value and the risk factor of each branch road circuit under step (3), calculating microgrid current state, and the two normalized value of weighting, the branch road weights of formation system, and then taking these weights as matrix element, build the network topology matrix A (t) in current t moment, then execution step (5);

Step (4), when judgement system is in state I) time, be when in system, the output power of DG is greater than the need for electricity of whole loads, using whole energy storage devices in system as power load, then according to himself further judgement of state-of-charge do, execution step (7);

Step (5), judge that energy storage device is regarded as power load or power supply, need to do further judgement according to himself state-of-charge, simultaneously execution step (7) and step (8);

Step (6), when judgement system is in state I II) time, it is the power load that total supply load of system DG is less than all consumers in system, in system, whole energy storage devices is regarded power supply as, need to make further judgement, execution step (8) according to himself state-of-charge;

Step (7), judge whether the SOC state of each energy storage device is greater than max cap. threshold value SOC _maxif, execution step (9); Perform step if not (10);

Step (8), judge whether the SOC state of each energy storage device is less than minimum capacity threshold value SOC _minif, execution step (9); Perform step if not (11);

Step (9) is if the state-of-charge of certain energy storage device meets SOC > SOC _maxor SOC _min> SOC, this energy storage device is failure to actuate, neither as Power supply, also not as consumer electricity consumption, execution step (12);

Step (10) is not if certain energy storage device is greater than its max cap. threshold value (SOC _max), and the output power that now system meets DG is less than the need for electricity of whole loads, and this energy storage device is used as power load, and execution step (12);

Step (11) is not if certain energy storage device is less than its minimum capacity threshold value (SOC _min), and the output power that now system meets DG is greater than the need for electricity of whole loads, and this energy storage device is used as power supply, and execution step (12);

Step (12), taking the node at each power supply place in autonomous microgrid as root node, search minimum spanning tree; Calculate in every minimum spanning tree, the minimum weights from root node to each important load node and; Incorporate each important load into each power supply taking minimum weights with as target decision, be responsible for its power supply;

Described power supply comprises DG and the definite energy storage device of step (11);

Step (13), select the corresponding power supply of minimum value in whole " the minimum weights and " of certain important load supply node as this important load;

Step (14), determine all " power supply-important load " corresponding power supply set of relationship according to step (13);

Step (15), judge whether the supply load of power supply in each " power supply-important load " corresponding power supply set of relationship is greater than the electricity consumption requirement of important load, if execution step (17), performs step (16) if not;

Step (16), select the corresponding power supply of sub-minimum in " the minimum weights and " of this important load supply node the execution step (17) as this important load;

Step (17), made full use of to greatest extent as principle taking the electric energy of the power generation of powering, adopted LMI algorithm, determined the number of all types of non-important loads between power supply and important load;

Step (18), all types of non-important load between power supply and important load, select non-important load by the definite number of step (17), and and power supply and important load form a new autonomous microgrid;

Step (19), by above step by system all nodes be assigned in different new autonomous microgrids, thereby form new autonomous microgrid electric power system.

2. the cross-domain collaborative energy scheduling of the autonomous microgrid group of a kind of wind-light storage bavin as claimed in claim 1 and adaptation method, it is characterized in that real time power loss value and the risk factor of the each branch road circuit of step (3), and the two normalized value of weighting, the branch road weights of formation system, and then taking these weights as matrix element, build the network topology matrix A (t) in current t moment;

Real time power loss value calculate by expression formula (1):

$P_{\mathrm{loss}}^{\mathrm{ij}}=\frac{{P_j}^2+{Q_j}^2}{{U_j}^2}\&CenterDot;R_{\mathrm{ij}}---\left(1\right)$

Wherein P _jfor the active power of a certain power load j in autonomous microgrid, Q _jfor the reactive power of a certain power load j in autonomous microgrid, R _ijfor the all-in resistance the power supply branch road circuit from node i to node j, X _ijfor the total reactance the power supply branch road circuit from node i to node j, U _jfor the voltage of node j;

Degree of unavailability K by year circuit failure-frequency f and r repair time of circuit determine,

$K=\frac{f\&CenterDot;r}{8760}---\left(2\right)$

The risk factor of each circuit

$K_{\mathrm{risk}}^{\mathrm{ij}}=\mathrm{\&eta;}\&CenterDot;K_{\mathrm{ij}}+(1-\mathrm{\&eta;})\&CenterDot;E_{\mathrm{ij}}---\left(3\right)$

Wherein K _ijfor the degree of unavailability of circuit between node i and node j, tried to achieve by expression formula (2); E _ijfor the expert assessment and evaluation value of circuit between node i and node j, η is regulatory factor;

After normalization, the real time power loss value of circuit and the risk factor of circuit are respectively:

$P_{\mathrm{norm}\_\mathrm{loss}}^{\mathrm{ij}}=\frac{P_{\mathrm{loss}}^{\mathrm{ij}}}{\underset{i,j=1;i\&NotEqual;j}{\overset{N}{\mathrm{\&Sigma;}}}{P}_{\mathrm{loss}}^{\mathrm{ij}}}---\left(4\right)$

$K_{\mathrm{norm}\_\mathrm{risk}}^{\mathrm{ij}}=\frac{K_{\mathrm{risk}}^{\mathrm{ij}}}{\underset{i,j=1;i\&NotEqual;j}{\overset{N}{\mathrm{\&Sigma;}}}{K}_{\mathrm{risk}}^{\mathrm{ij}}}---\left(5\right)$

Wherein N is the nodes in whole network system,

Utilize circuit real time power loss value and circuit risk factor after normalization, obtain circuit L _ijfinal branch road weights be:

$a_{\mathrm{ij}}=\mathrm{\&beta;}\&CenterDot;P_{\mathrm{norm}\_\mathrm{loss}}^{\mathrm{ij}}+(1-\mathrm{\&beta;})\&CenterDot;K_{\mathrm{norm}\_\mathrm{risk}}^{\mathrm{ij}}---\left(6\right)$

β is regulatory factor;

With a _ijfor the matrix weights between node i and node j, obtain the weights between system any two points; Set it as matrix element, can obtain the topological relation matrix A of any time system ^{Δ t}.