Optimized scheduling method for power distribution microgrid

Abstract

The invention discloses an optimal scheduling method for a power distribution network, and the method comprises the steps: dividing the micro-grids into different micro-grid types according to the operation characteristics of a source load, taking the micro-grids of the same type as a whole, carrying out the optimal scheduling of the power distribution network, and effectively considering the power interaction between multiple micro-grids and the power distribution network. And meanwhile, in a first optimal scheduling model for power interaction of the multiple microgrids, power constraints considering power generation behaviors of distributed power supplies and charging lines of electric vehicles are constructed, so that optimal scheduling of power interaction among the multiple microgrids is realized. Besides, when the multi-micro-grid is subjected to optimal scheduling, power coupling between the power distribution network and the multi-micro-grid is considered, so that a second optimal scheduling model of the power distribution network connected with the multi-micro-grid group is constructed, and coupling constraints between the first optimal scheduling model and the second optimal scheduling model are constructed, thereby ensuring safe operation of the power distribution network. And cooperative efficient operation of the micro-grid and the power distribution network is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of microgrid optimization dispatching, and in particular relates to an optimization dispatching method for a distribution microgrid. Background Art

[0002] At present, a large number of distributed energy sources are developing rapidly, and the penetration rate of new energy in new power systems is increasing. Its volatility and intermittency pose huge challenges to the safe operation of power grids. The technology to absorb a high proportion of new energy is becoming increasingly important.

[0003] Microgrids aggregate loads and distributed energy equipment in the vicinity of a region, and improve the energy self-sufficiency rate within the microgrid by increasing the local consumption rate of distributed new energy, reducing the electricity demand of the entire microgrid, thereby alleviating the power supply pressure of the distribution network and improving the safety of distribution network operation. At the same time, the optimized configuration of microgrids can effectively improve the consumption level of distributed new energy and reduce its impact on the power grid.

[0004] With the continued development of distributed energy and the rapid growth of loads, each microgrid has a source-load duality, that is, the installed capacity of new energy sources and the energy characteristics of loads connected to each microgrid are different. This difference makes the energy complementarity between microgrids have great potential for exploration. The multi-microgrid system is formed by the interconnection of multiple microgrids in the same distribution network area, which effectively utilizes the source-load duality of microgrids through power interaction between multiple microgrids. However, although there is power interaction between microgrids, there is still power interaction with the external power grid. This power interaction demand will cause power flow fluctuations in the distribution network. Therefore, when considering the optimal scheduling of microgrids, it is also necessary to consider the need for safe operation of the distribution network. At present, most microgrid scheduling studies only regard the distribution network as a simple energy provider, ignoring the limitations of the power flow transportation capacity of the distribution network grid. Summary of the invention

[0005] In order to solve the problems existing in the prior art, the present invention proposes an optimization scheduling method for a distribution microgrid.

[0006] The technical solution of the present invention is as follows:

[0007] A method for optimizing the dispatching of a distribution microgrid, wherein the distribution microgrid includes a distribution network composed of a plurality of microgrids, and the method includes:

[0008] Based on the source-load operation characteristics of the microgrid source-load data, multiple microgrids are divided into microgrid types;

[0009] Based on the division results, a first optimization scheduling model for interaction between microgrids and a second optimization scheduling model for interaction between microgrids and distribution networks are constructed; the first optimization scheduling model aims to minimize the total operating cost of the Togo microgrid, and takes the power generation behavior of distributed power sources and the power constraints of electric vehicle charging as constraints, and the operation strategy of each microgrid as a decision variable; the second optimization scheduling model aims to minimize the network loss of the distribution network to which each microgrid is connected, and takes the safe operation constraint as a constraint, and takes the operation strategy of the distribution network as a decision variable;

[0010] Construct a coupling constraint between the first optimization scheduling model and the second optimization scheduling model, and iteratively solve the first optimization scheduling model and the second optimization scheduling model under the coupling constraint to obtain the operation strategy of each microgrid and the operation strategy of the distribution network.

[0011] Furthermore, based on the source-load operation characteristics of the microgrid source-load data, a specific method for clustering multiple microgrids related to microgrid types includes:

[0012] The power output data and load data of each microgrid are sorted according to the same rule to generate a power output vector and a load vector, and the Spearman rank correlation coefficient between the power output vector and the load vector is calculated;

[0013] Calculating the Euclidean distance between the power output data and the load data;

[0014] Calculate the source-load matching degree of the source-load data by a weight assignment method according to the Spearman rank correlation coefficient and the Euclidean distance;

[0015] Determine whether the source-load matching degree meets a preset matching value; if so, classify the microgrid type of the microgrid corresponding to the source-load matching degree into a microgrid group; otherwise, classify the microgrid type of the microgrid corresponding to the source-load matching degree into an independent microgrid.

[0016] Furthermore, the objective function of the first optimization scheduling model is:

[0017]

[0018] In the formula, F MG represents the total operating cost of multiple microgrids; F grid,i Represents microgrid MG i The cost of power interaction with the distribution network; if the microgrid MG i is marked as the microgrid group, then F mg,i Represents microgrid MG i The cost of power interaction with other microgrids; if the microgrid MGi is marked as the independent microgrid, then F mg,i is zero; p1(t) and p2(t) represent the price at which the microgrid purchases electricity from the distribution network and the price at which it sells electricity, respectively; P grid,b,i (t) and P grid,s,i (t) represents the microgrid MG i The power purchased from the distribution network and the power sold; Δt represents the dispatching unit time interval; P i,j (t) represents the microgrid MG at time t i With microgrid MG j The interaction power, if P i,j (t) is positive, then the microgrid MG i Microgrid MG j Selling electricity, and vice versa is buying electricity.

[0019] Furthermore, the expression considering the power generation behavior of distributed power sources and the power constraint of electric vehicle charging is:

[0020] P WT,i (t)+P PV,i (t)+X(t)P grid,b,i (t)-P i,j (t)-[1-X(t)]P grid,s,i (t) = P load,i (t)

[0021] Where P WT,i (t) and P PV,i (t) represent the microgrid MG i Wind power output model and photovoltaic power output model at time t; X(t) represents a 0-1 binary variable. i When purchasing electricity from the distribution network, X(t) is 1, otherwise it is 0; P load,i (t) represents the load power at time t, P load,i (t) = P MRC +P FLEX , P MRC represents the charging load model, P FLEX Represents a flexible load model.

[0022] Furthermore, the method for obtaining the charging load model is as follows:

[0023] S41, establishing a charging probability density model based on a normal distribution according to the start time of charging of the electric vehicle;

[0024] S42, establishing a charging stop probability density model based on the log-normal distribution according to the charging stop time of the electric vehicle;

[0025] S43, establishing an expected power probability density model based on the log-normal distribution according to the expected power of the electric vehicle user;

[0026] S44. Calculate the charging load model by using the Monte Carlo sampling method according to the charging probability density model, the charging stop probability density model and the expected power probability density model.

[0027] The method for obtaining the wind power output model and the photovoltaic power output model is as follows:

[0028] Photovoltaic output power model:

[0029]

[0030] Where P PV (t) represents the PV output power at time t; R(t) represents the light intensity at time t; R ste Indicates the light intensity under standard conditions; R c Indicates the light intensity at a certain radiation point; P rc Indicates the PV rated power.

[0031] Fan output power model:

[0032]

[0033] Among them, P WT (t) represents the output power of the fan at time t; P rw represents the rated power of the fan; v(t) represents the wind speed at time t; v r Indicates the rated wind speed of the fan; v on Indicates the fan start-up wind speed; v off Indicates the wind speed when the fan is stopped;

[0034] The flexible load model includes a translatable load model, a transferable load model and a reducible load model;

[0035] The translatable load model:

[0036] ΔP shift (t) = f shift [t+Δt(Δδ shift (t))]-f shift (t);

[0037]

[0038] In the formula, ΔP shift (t) represents the electric power that needs to be shifted during period t, i.e., the shifted load model; f shift (t) represents the initial load curve of the load that can be translated; Δt represents the time period that needs to be translated; Δδshift (t) represents the change in the relevant price during period t, and T is the scheduling period;

[0039] The transferable load model:

[0040] ΔP trans (t) = f trans (Δδ(t),P trans (t), e trans (t))-f trans (t);

[0041]

[0042] In the formula, ΔP trans (t) represents the electric power that needs to be transferred during period t, i.e., the transfer load model; P trans (t) represents the basic transfer load during period t; e trans (t) represents the self-transfer elastic coefficient; f trans (t) initial load curve representing the transferable load;

[0043] The curtailable load model:

[0044] ΔP cut (t) = f cut (P cut (t), Δδ(t), e cut (t))-f cut (t);

[0045] In the formula, ΔP cut (t) represents the electric power that needs to be reduced during period t, i.e., the load reduction model; P cut (t) represents the basic load reduction during period t; e cut (t) represents the reduction elasticity coefficient during period t; f cut (t) represents the initial load curve of the load that can be reduced.

[0046] Furthermore, the objective function of the second optimization scheduling model is:

[0047]

[0048] In the formula, I lm，t Represents the square of the current in line lm at time t; R lm Indicates the resistance of the circuit 1m.

[0049] Furthermore, the safe operation constraints include power flow constraints of distribution network lines, voltage constraints of distribution network operation, line current carrying capacity constraints and substation output power constraints;

[0050] The power distribution network line flow constraint is:

[0051]

[0052]

[0053] Where l represents the starting node of the line with node m as the end point, k represents the end node of the line with node m as the starting point; Ω(m,:) and Ω(:,m) represent the set of lines with m as the starting point and the end point respectively; P lm,t and P km,t They represent the active power of line lm and line km at time t respectively; Q lm,t and Q km,t They represent the reactive power of line lm and line km at time t respectively; l lm,t represents the square of the current of line lm at time t; X lm Indicates the impedance of the line lm; P m,t sub and Q m,t sub Respectively represent the active power and reactive power of the substation at node m; P m,t L0 and Q m,t L0 They represent the active load and reactive load of the user connected to the distribution network node m respectively; P m,t MG represents the power provided by the distribution network to the microgrid at node m;

[0054] The voltage constraint of the distribution network operation is:

[0055]

[0056] Among them, U m,t represents the square of the voltage at node m at time t; U l,t represents the square of the voltage at node l at time t; U m min Indicates the lower limit of the voltage safety amplitude; U m max Indicates the upper limit of the voltage safety amplitude;

[0057] The line current carrying capacity constraint is:

[0058]

[0059] Among them, I lm max Indicates the upper limit of the line's allowable current carrying capacity;

[0060] The substation output power constraint is:

[0061]

[0062] Among them, S m,t sub Represents the rated power of the substation at node m.

[0063] Furthermore, the coupling constraint between the first optimization scheduling model and the second optimization scheduling model is:

[0064]

[0065] Among them, P grid,b,i (t) and P grid,s,i (t) represent the microgrid MG i The power purchased from the distribution network and the power sold; P m,t MG It represents the power provided by the distribution network to the microgrid at node m; MG represents the microgrid; ω(m) represents the set of microgrids connected to the distribution network node m.

[0066] Furthermore, the specific steps of iteratively solving the first optimization scheduling model and the second optimization scheduling model under the coupling constraint to obtain the operation strategy of each microgrid and the operation strategy of the distribution network include:

[0067] Decoupling the first optimization scheduling model from the second optimization scheduling model according to the coupling constraint based on the alternating direction multiplier method, and iteratively solving the decoupled first optimization scheduling model and the second optimization scheduling model respectively;

[0068] Calculating the primal residual and the dual residual between the first optimization scheduling model and the second optimization scheduling model;

[0069] Determining whether the iterative solution result meets the convergence condition according to the original residual and the dual residual;

[0070] If so, the current iterative solution result is output as the optimal scheduling result;

[0071] Otherwise, continue to iterate until the convergence condition is met or the maximum number of iterations is reached.

[0072] Furthermore, based on the alternating direction multiplier method, the first optimization scheduling model and the second optimization scheduling model are decoupled according to the coupling constraint, and the decoupled first optimization scheduling model and the second optimization scheduling model are iteratively solved respectively.

[0073]

[0074] Among them, λ represents the Lagrange multiplier; ρ represents the penalty factor; u represents the current number of iterations; represents the value solved by the current iteration; and and They respectively represent the values ​​of the corresponding variables iteratively solved during the u-1th iteration.

[0075] An electronic device comprises a memory and a processor, wherein the memory stores a computer program, and the processor is used to call and run the computer program stored in the memory to execute any of the methods described above.

[0076] A computer-readable storage medium stores a computer program, wherein the computer program implements the steps of any of the above methods when executed by a processor.

[0077] Compared with the prior art, the present invention has the following beneficial effects:

[0078] The present invention provides an optimization scheduling method for a distribution microgrid. The method divides the microgrid into different microgrid types according to the source-load operation characteristics, and optimizes the scheduling of the distribution network by taking the microgrids of the same type as a whole, effectively considering the power interaction between multiple microgrids and the distribution network. At the same time, in the first optimization scheduling model for power interaction among multiple microgrids, based on the large-scale growth of distributed power sources and electric vehicles, a power constraint considering the power generation behavior of distributed power sources and the charging behavior of electric vehicles is constructed, so as to achieve the optimization scheduling of the power interaction among multiple microgrids. In addition, when optimizing the scheduling of multiple microgrids, considering the power coupling between the distribution network and the multiple microgrids, a second optimization scheduling model of the distribution network connected to the multiple microgrid group is constructed, and the coupling constraints between the first and second optimization scheduling models are constructed, fully considering the impact of the power interaction of multiple microgrids on the safe operation of the distribution network, while making full use of the duality of source and load of the microgrid, ensuring the safe operation of the distribution network, and improving the coordinated and efficient operation of the microgrid and the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 It is a flow chart of the optimization dispatching method of the distribution microgrid in the embodiment;

[0080] Figure 2 It is a solution flow chart of the alternating direction multiplier method in the embodiment;

[0081] Figure 3 A flow chart of constructing a charging load model using the Monte Carlo sampling method in an embodiment; DETAILED DESCRIPTION

[0082] The present invention is further explained below in conjunction with the accompanying drawings and specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading the present invention, various equivalent modifications of the present invention by those skilled in the art all fall within the scope defined by the claims attached to this application.

[0083] Embodiment 1:

[0084] The present invention provides an optimization dispatching method for a distribution microgrid, wherein the distribution microgrid includes a distribution network composed of a plurality of microgrids, such as Figure 1 As shown, the method comprises the following specific steps:

[0085] S1. Obtain the source-load data of each microgrid in the same dispatching cycle to form the source-load operation characteristics of each microgrid, and divide the multiple microgrids into microgrid types based on the source-load operation characteristics of the microgrid source-load data;

[0086] S2. Based on the division results, i.e., the microgrids with microgrid types and their source-load data, a first optimization scheduling model for interaction between microgrids and a second optimization scheduling model for interaction between microgrids and distribution networks are constructed; the first optimization scheduling model aims to minimize the total operating cost of the Togo microgrid, and takes the power generation behavior of distributed power sources and the power constraints of electric vehicle charging as constraints, and takes the operation strategy of each microgrid as a decision variable; the second optimization scheduling model aims to minimize the network loss of the distribution network to which each microgrid is connected, and takes the safe operation constraint as a constraint, and takes the operation strategy of the distribution network as a decision variable;

[0087] S3. Construct a coupling constraint between the first optimization scheduling model and the second optimization scheduling model, and iteratively solve the first optimization scheduling model and the second optimization scheduling model under the coupling constraint to obtain the operation strategy of each microgrid and the operation strategy of the distribution network.

[0088] Embodiment 2:

[0089] This embodiment is further designed on the basis of the first embodiment in that, in this embodiment, based on the source-load operation characteristics of the microgrid source-load data, a specific method for clustering multiple microgrids related to the microgrid type includes:

[0090] In this case, microgrid types include microgrid groups and independent microgrids;

[0091] S21. Generate a power output vector and a load vector according to the same sorting rule for the power output data and load data of each microgrid, and calculate the Spearman rank correlation coefficient between the power output vector and the load vector, specifically:

[0092]

[0093] Where ρ' represents the Spearman rank correlation coefficient between the power output vector and the load vector; r i It represents the numerical difference between two vectors at each moment; D represents the vector dimension, i.e., the selected historical scheduling period.

[0094] S22. Calculate the Euclidean distance between the power output data and the load data, specifically:

[0095]

[0096] Among them, S t represents the Euclidean distance; y 1_t and 2_t Respectively represent the timing values ​​of power output data and load data.

[0097] S23, the source-load matching degree of the source-load data is calculated by the weight assignment method according to the Spearman rank correlation coefficient and the Euclidean distance, the Spearman rank correlation coefficient and the Euclidean distance obtained in step S21 and step S22 are normalized to eliminate the dimension difference, and the source-load matching degree is obtained by the weight assignment method according to formula (20), specifically:

[0098]

[0099] Among them, ε represents the source-charge matching degree; λ' represents the weight coefficient; S min and S max They represent the minimum and maximum values ​​of the Euclidean distance respectively; T is the total scheduling period.

[0100] S24, determining whether the source-load matching degree meets a preset matching value;

[0101] S25, if yes, then dividing the microgrid types of the microgrids corresponding to the source-load matching degree into microgrid groups;

[0102] S26. Otherwise, the microgrid type of the microgrid corresponding to the source-load matching degree is classified as an independent microgrid.

[0103] The microgrid group includes an output group and a receiving group. The output group is set with an output matching value, and the receiving group is set with a receiving matching value, wherein the output matching value is higher than the receiving matching value. When the source-load matching degree is greater than the output matching value, the microgrid is marked as an output group; when the source-load matching degree is less than or equal to the output matching value and greater than the receiving matching value, the microgrid is marked as a receiving group; when the source-load matching degree is less than or equal to the receiving matching value, the microgrid is marked as an independent microgrid. Specifically,

[0104]

[0105] Where Z(i) represents the microgrid type of the i-th microgrid, 2 represents the output group, 1 represents the receiving group, and 0 represents an independent microgrid; ε i represents the source-load matching degree of the i-th microgrid; C2 represents the output matching value, and C1 represents the receiving matching value.

[0106] Embodiment three:

[0107] This embodiment is further designed on the basis of the second embodiment in that the objective function of the first optimization scheduling model in this embodiment is:

[0108]

[0109] In the formula, F MG represents the total operating cost of multiple microgrids; F grid,i Represents microgrid MG i The cost of power interaction with the distribution network; if the microgrid MG i is marked as a microgrid group, then F mg,i Represents microgrid MG i The cost of power interaction with other microgrids; if the microgrid MG i is marked as an independent microgrid, then F mg,i is zero; p1(t) and p2(t) represent the price at which the microgrid purchases electricity from the distribution network and the price at which it sells electricity, respectively; P grid,b,i (t) and P grid,s,i (t) represent the microgrid MG i The power purchased from the distribution network and the power sold; Δt represents the dispatching unit time interval; P i,j (t) represents the microgrid MG at time t i With microgrid MG j The interaction power, if P i,j (t) is positive, then the microgrid MG i Microgrid MG j Selling electricity, and vice versa is buying electricity.

[0110] Embodiment 4:

[0111] This embodiment is further designed on the basis of the third embodiment in that the expression for the power generation behavior of the distributed power source and the power constraint of the charging behavior of the electric vehicle is considered as follows:

[0112] P WT,i (t)+P PV,i (t)+X(t)P grid,b,i (t)-P i,j (t)-[1-X(t)]P grid,s,i (t) = P load,i (t)

[0113] Where PWT,i (t) and P PV,i (t) represent the microgrid MG i Wind power output model and photovoltaic power output model considering the generation behavior of distributed power sources at time t; X(t) represents a 0-1 binary variable. i When purchasing electricity from the distribution network, X(t) is 1, otherwise it is 0; P load,i (t) represents the load power at time t, P load,i (t) = P MRC +P FLEX , P MRC represents the charging load model considering the charging behavior of electric vehicles, P FLEX represents the flexible load model, P FLEX Represents multiple types, then P FLEX Represents the sum of all flexible load models.

[0114] Furthermore, the transmission power between microgrids is limited by the power transmission capacity of the tie line, so it is also necessary to construct the power interaction constraint of the first optimization scheduling model:

[0115]

[0116] Among them, P i,j (t) represents the microgrid MG at time t i With microgrid MG j The interaction power between i,j min and P i,j max Represents microgrid MG i With microgrid MG j The lower and upper limits of the interaction power.

[0117] Embodiment five:

[0118] This embodiment is further designed on the basis of the fourth embodiment in that: Figure 3 As shown in the figure, the method for obtaining the charging load model considering the charging behavior of electric vehicles in this example is as follows:

[0119] S41. Establish a charging probability density model based on normal distribution according to the start time of charging of the electric vehicle. The charging probability density model is specifically:

[0120]

[0121] Among them, f(t) represents the charging probability density model, μ and σ represent the expectation and standard deviation of the normal distribution respectively; t represents the time when the electric vehicle starts charging.

[0122] S42. Establish a charging stop probability density model based on the log-normal distribution according to the charging stop time of the electric vehicle. The charging stop probability density model is specifically:

[0123]

[0124] Among them, f'(t') represents the probability density model of charging stop, μ' and σ' represent the expectation and standard deviation of the log-normal distribution of the charging stop time, respectively; t' represents the time when the electric vehicle stops charging.

[0125] S43. Establish an expected power probability density model based on the log-normal distribution according to the expected power of the electric vehicle user. The expected power probability density model is specifically:

[0126]

[0127] Among them, f s (x) represents the expected power probability density model, μ s , σ s They represent the expectation and standard deviation of the log-normal distribution of expected power respectively; x represents the expected power of the electric vehicle.

[0128] S44. Calculate the charging probability density model, the charging stop probability density model and the expected power probability density model through the Monte Carlo sampling method to obtain a charging load model. Specifically, first initialize the parameters such as the charging start time, the charging stop time and the expected charging time of the electric vehicle to be sampled, and randomly extract the charging start time from the charging probability density model, randomly extract the charging stop time from the charging stop probability density model, and extract the expected power from the expected power probability density model. Superimpose the charging time (the time interval between the charging start time and the charging stop time) and the load (the load required for the expected power) to obtain the charging load model of the electric vehicle.

[0129] The method for obtaining the wind power output model and the photovoltaic power output model is as follows:

[0130] Photovoltaic output power model:

[0131]

[0132] Where P PV (t) represents the PV output power at time t; R(t) represents the light intensity at time t; R ste Indicates the light intensity under standard conditions; R c Indicates the light intensity at a certain radiation point; P rc Indicates the PV rated power.

[0133] Fan output power model:

[0134]

[0135] Among them, P WT (t) represents the output power of the fan at time t; P rw represents the rated power of the fan; v(t) represents the wind speed at time t; v r Indicates the rated wind speed of the fan; v on Indicates the fan start-up wind speed; v off Indicates the wind speed when the fan is stopped;

[0136] Flexible load models include translatable load models, transferable load models, and reducible load models;

[0137] Translatable load model:

[0138] ΔP shift (t) = f shift [t+Δt(Δδ shift (t))]-f shift (t);

[0139]

[0140] In the formula, ΔP shift (t) represents the electric power that needs to be shifted during period t, i.e., the shifted load model; f shift (t) represents the initial load curve of the load that can be translated; Δt represents the time period that needs to be translated; Δδ shift (t) represents the change in the relevant price during period t, and T is the scheduling period;

[0141] Transferable load model:

[0142] ΔP trans (t) = f trans (Δδ(t),P trans (t), e trans (t))-f trans (t);

[0143]

[0144] In the formula, ΔP trans (t) represents the electric power that needs to be transferred during period t, i.e., the transfer load model; P trans (t) represents the basic transfer load during period t; e trans (t) represents the self-transfer elastic coefficient; f trans (t) initial load curve representing the transferable load;

[0145] Load reduction model:

[0146] ΔPcut (t) = f cut (P cut (t), Δδ(t), e cut (t))-f cut (t);

[0147] In the formula, ΔP cut (t) represents the electric power that needs to be reduced during period t, i.e., the load reduction model; P cut (t) represents the basic load reduction during period t; e cut (t) represents the reduction elasticity coefficient during period t; f cut (t) represents the initial load curve of the load that can be reduced.

[0148] Based on this, P FLEX =ΔP shift (t)+ΔP trans (t)+ΔP cut (t).

[0149] Embodiment six:

[0150] This embodiment is further designed on the basis of the fifth embodiment in that the objective function of the second optimization scheduling model in this embodiment is:

[0151]

[0152] In the formula, I lm，t Represents the square of the current in line lm at time t; R lm Indicates the resistance of the circuit 1m.

[0153] Embodiment seven:

[0154] This embodiment is further designed on the basis of the sixth embodiment in that, since the distribution network operation adopts the Distflow equation, its operation optimization goal is set to minimize network loss, and it is necessary to meet the distribution network flow constraints and line current carrying capacity safety constraints, and the power purchase demand of the microgrid for the distribution network node on which it is mounted is used as the boundary condition of the distribution network operation to realize the coupling optimization between the microgrid and the distribution network.

[0155] In this case, the safety operation constraints include power flow constraints of distribution network lines, voltage constraints of distribution network operation, line current carrying capacity constraints, and substation output power constraints;

[0156] Among them, the power flow constraint of the distribution network line is:

[0157]

[0158] Where l represents the starting node of the line with node m as the end point, k represents the end node of the line with node m as the starting point; Ω(m,:) and Ω(:,m) represent the set of lines with m as the starting point and the end point respectively; P lm,t and P km,t They represent the active power of line lm and line km at time t respectively; Q lm,t and Q km,t They represent the reactive power of line lm and line km at time t respectively; l lm,t represents the square of the current of line lm at time t; X lm Indicates the impedance of the line lm; P m,t sub and Q m,t sub Respectively represent the active power and reactive power of the substation at node m; P m,t L0 and Q m,t L0 They represent the active load and reactive load of the user connected to the distribution network node m respectively; P m,t MG represents the power provided by the distribution network to the microgrid at node m;

[0159] The voltage constraint for distribution network operation is:

[0160]

[0161] Among them, U m,t represents the square of the voltage at node m at time t; U l,t represents the square of the voltage at node l at time t; U m min Indicates the lower limit of the voltage safety amplitude; U m max Indicates the upper limit of the voltage safety amplitude;

[0162] The line current carrying capacity constraint is:

[0163]

[0164] Among them, I lm max Indicates the upper limit of the line's allowable current carrying capacity;

[0165] The substation output power constraint is:

[0166]

[0167] Among them, S m,t sub Represents the rated power of the substation at node m.

[0168] Embodiment eight:

[0169] This embodiment is further designed on the basis of the seventh embodiment in that the coupling constraint between the first optimization scheduling model and the second optimization scheduling model in this embodiment is:

[0170]

[0171] Among them, P grid,b,i (t) and P grid,s,i (t) represent the microgrid MG i The power purchased from the distribution network and the power sold; P m,t MG It represents the power provided by the distribution network to the microgrid at node m; MG represents the microgrid; ω(m) represents the set of microgrids connected to the distribution network node m.

[0172] The above P grid,b,i (t) and P grid,s,i The difference between (t) is the microgrid MG i The net amount of electricity purchased from the distribution network is assumed to be i If node m is connected to the distribution network, then P grid,b,i (t) and P grid,s,i (t) is equal to the distribution network at node m to the microgrid MG i The power provided. Therefore, there is a consistent coupling constraint between the microgrid and the distribution network. This coupling constraint is used as a coupling variable between the first optimization scheduling model corresponding to the microgrid and the second optimization scheduling model corresponding to the distribution network, so as to facilitate the subsequent coupling solution of the first optimization scheduling model and the second optimization scheduling model.

[0173] Embodiment nine:

[0174] This embodiment is further designed on the basis of the eighth embodiment in that, in this embodiment, the first optimization scheduling model and the second optimization scheduling model are iteratively solved under the coupling constraint, and the specific steps of obtaining the operation strategy of each microgrid and the operation strategy of the distribution network include:

[0175] S91. Decouple the first optimization scheduling model from the second optimization scheduling model according to the coupling constraint based on the alternating direction multiplier method, and iteratively solve the decoupled first optimization scheduling model and the second optimization scheduling model respectively. S92. Calculate the original residual and the dual residual between the first optimization scheduling model and the second optimization scheduling model, as follows:

[0176]

[0177] Among them, σ P,ADN , σ P,MGrepresents a set of original residuals, which is used to ensure that the coupling variables between connected subsystems are close enough; σ D Represents the dual residual, which is used to ensure the stability of the solution; ε P and ε D denote the primal residual limit and the dual residual limit, respectively.

[0178] S93. Determine whether the iterative solution result meets the convergence condition according to the original residual and the dual residual. For example, when the original residual and the dual residual of the iterative solution result are both less than a given threshold, it is determined that the current iterative solution result meets the convergence condition; otherwise, it is determined that the current iterative solution result does not meet the convergence condition.

[0179] S94, if yes, output the current iterative solution result as the optimal scheduling result;

[0180] S95. Otherwise, continue to iterate and solve until the convergence condition is met or the maximum number of iterations is reached.

[0181] like Figure 2 As shown, before iteratively solving the first optimization scheduling model and the second optimization scheduling model after decoupling based on the alternating direction multiplier method, the average value of the coupling constraint, the Lagrange multiplier, the convergence threshold of the original residual and the dual residual, the maximum solution time and the maximum number of iterations should be initialized. In the process of iterative solution, each subsystem (i.e., the first optimization scheduling model and the second optimization scheduling model after decoupling) first solves the optimization problem of this model in parallel to obtain the value of the coupling constraint in this model; then each subsystem and the connected subsystem exchange the value of the coupling constraint of the latest iteration, and calculate the average value of the coupling constraint respectively, so as to protect the data privacy of each subsystem; then judge whether the iteration result meets the convergence condition through the original residual and the dual residual. If not, continue to iteratively update the auxiliary variables, and solve the optimization problem of this model again to obtain the value of the coupling constraint after the latest iteration in this model, until the convergence condition is met, the iteration times out or the maximum number of iterations is reached, and the current optimized scheduling scheme is output.

[0182] Embodiment ten:

[0183] This embodiment is further designed on the basis of the ninth embodiment in that, in this embodiment, the first optimization scheduling model and the second optimization scheduling model are decoupled according to the coupling constraint based on the alternating direction multiplier method, and the specific steps of iteratively solving the decoupled first optimization scheduling model and the second optimization scheduling model respectively include:

[0184]

[0185] Among them, λ represents the Lagrange multiplier; ρ represents the penalty factor; u represents the current number of iterations; represents the value solved by the current iteration; and and They respectively represent the values ​​of the corresponding variables iteratively solved during the u-1th iteration.

[0186] The alternating direction multiplication method is used to convert the above As auxiliary variables, the coupling constraints are decoupled. The value of each iterative solution is affected by the value of the coupling constraints in each subsystem in the previous iterative solution. Therefore, the auxiliary variables The update process is as follows:

[0187]

[0188] Embodiment eleven:

[0189] An electronic device comprises a memory and a processor, wherein the memory stores a computer program, and the processor is used to call and run the computer program stored in the memory to execute any of the methods described above.

[0190] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of any of the above methods are implemented.

[0191] The above description is only a specific implementation of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by a technician familiar with the technical field within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention.

Claims

1. A method for optimizing the dispatching of a distribution microgrid, wherein the distribution microgrid comprises a distribution network composed of a plurality of microgrids, characterized in that: The method comprises: Based on the source-load operation characteristics of the microgrid source-load data, multiple microgrids are divided into microgrid types; Based on the division results, a first optimization scheduling model for interaction between microgrids and a second optimization scheduling model for interaction between microgrids and distribution networks are constructed; the first optimization scheduling model aims to minimize the total operating cost of the Togo microgrid, and takes the power generation behavior of distributed power sources and the power constraints of electric vehicle charging as constraints, and the operation strategy of each microgrid as a decision variable; the second optimization scheduling model aims to minimize the network loss of the distribution network to which each microgrid is connected, and takes the safe operation constraint as a constraint, and takes the operation strategy of the distribution network as a decision variable; Construct a coupling constraint between the first optimization scheduling model and the second optimization scheduling model, and iteratively solve the first optimization scheduling model and the second optimization scheduling model under the coupling constraint to obtain the operation strategy of each microgrid and the operation strategy of the distribution network.

2. The optimization dispatching method of the distribution microgrid according to claim 1 is characterized in that: The specific method for clustering multiple microgrids related to microgrid types based on the source-load operation characteristics of the microgrid source-load data includes: The power output data and load data of each microgrid are sorted according to the same rule to generate a power output vector and a load vector, and the Spearman rank correlation coefficient between the power output vector and the load vector is calculated; Calculating the Euclidean distance between the power output data and the load data; Calculate the source-load matching degree of the source-load data by a weight assignment method according to the Spearman rank correlation coefficient and the Euclidean distance; Determine whether the source-load matching degree meets a preset matching value; if so, classify the microgrid type of the microgrid corresponding to the source-load matching degree into a microgrid group; otherwise, classify the microgrid type of the microgrid corresponding to the source-load matching degree into an independent microgrid.

3. The optimization dispatching method of the distribution microgrid according to claim 2 is characterized in that: The objective function of the first optimization scheduling model is: In the formula, F MG represents the total operating cost of multi-microgrid; F grid，i Represents microgrid MG i The cost of power interaction with the distribution network; if the microgrid MG i is marked as the microgrid group, then F mg,i Represents microgrid MG i The cost of power interaction with other microgrids; if the microgrid MG i is marked as the independent microgrid, then F mg,i is zero; p1(t) and p2(t) represent the price at which the microgrid purchases electricity from the distribution network and the price at which it sells electricity, respectively; P grid,b，i (t) and P grid,s，i (t) represents the microgrid MG i The power purchased from the distribution network and the power sold; Δt represents the dispatching unit time interval; P i，j (t) represents the microgrid MG at time t i With microgrid MG j The interaction power, if P i，j (t) is positive, then the microgrid MG i Microgrid MG j Selling electricity, and vice versa.

4. The optimization dispatching method of the distribution microgrid according to claim 3 is characterized in that: The expression considering the power generation behavior of the distributed power source and the power constraint of the charging line of the electric vehicle is: P WT，i (t)+P PV,i (t)+X(t)P grid,b,i (t)-P i，j (t)-[1-X(t)]P grid,s，i (t)=P load，i (t) Where P WT，i (t) and P PV，i (t) represents the microgrid MG i Wind power output model and photovoltaic power output model at time t; X(t) represents a 0-1 binary variable. i When purchasing electricity from the distribution network, X(t) is 1, otherwise it is 0; P load,i (t) represents the load power at time t, P load,i (t) = P MRC +P FLEX , P MRC represents the charging load model, P FLEX Represents a flexible load model.

5. The optimization dispatching method of the distribution microgrid according to claim 4 is characterized in that: The method for obtaining the charging load model is as follows: A charging probability density model is established based on normal distribution according to the start time of charging of electric vehicles; According to the stop charging time of electric vehicles, a stop charging probability density model is established based on the log-normal distribution; According to the expected power consumption of electric vehicle users, an expected power probability density model is established based on log-normal distribution; A charging load model is obtained by calculating the charging probability density model, the charging stop probability density model and the expected power probability density model through the Monte Carlo sampling method; The method for obtaining the wind power output model and the photovoltaic power output model is as follows: Photovoltaic output power model: Where P PV (t) represents the PV output power at time t; R(t) represents the light intensity at time t; R ste Indicates the light intensity under standard conditions; R c Indicates the light intensity at a certain radiation point; P rc Indicates the PV rated power; Fan output power model: Among them, P WT (t) represents the output power of the fan at time t; P rw represents the rated power of the fan; v(t) represents the wind speed at time t; v r Indicates the rated wind speed of the fan; v on Indicates the fan start-up wind speed; V off Indicates the wind speed when the fan is stopped; The flexible load model includes a translatable load model, a transferable load model and a reducible load model; The translatable load model: ΔP shift (t)=f shift [t+Δt(Δδ shift (t))]-f shift (t); In the formula, ΔP shift (t) represents the electric power that needs to be shifted during period t, i.e., the shifted load model; f shift (t) represents the initial load curve of the load that can be translated; Δt represents the time period that needs to be translated; Δδ shift (t) represents the change in the relevant price during period t, and T is the scheduling period; The transferable load model: ΔP trans (t)=f trans (Δδ(t),P trans (t),e trans (t))-f trans (t); In the formula, ΔP trans (t) represents the electric power that needs to be transferred during period t, i.e., the transfer load model; P trans (t) represents the basic transfer load during period t; e trans (t) represents the self-transfer elastic coefficient; f trans (t) initial load curve representing the transferable load; The curtailable load model: ΔP cut (t)=f cut (P cut (t),Δδ(t),e cut (t))-f cut (t); In the formula, ΔP cut (t) represents the electric power that needs to be reduced during period t, i.e., the load reduction model; P cut (t) represents the basic load reduction during period t; e cut (t) represents the reduction elasticity coefficient during period t; f cut (t) represents the initial load curve of the load that can be reduced.

6. The optimization dispatching method of the distribution microgrid according to claim 5 is characterized in that: The objective function of the second optimization scheduling model is: In the formula, I lm,t Represents the square of the current in line lm at time t; R lm Indicates the resistance of the circuit 1m.

7. The method for optimizing the dispatching of a distribution microgrid according to claim 6, characterized in that: The safety operation constraints include power distribution network line flow constraints, distribution network operation voltage constraints, line current carrying capacity constraints and substation output power constraints; The power distribution network line flow constraint is: Where l represents the starting node of the line with node m as the end point, k represents the end node of the line with node m as the starting point; Ω(m,:) and Ω(:,m) represent the set of lines with m as the starting point and the end point respectively; P lm,t and P km,t They represent the active power of line lm and line km at time t respectively; Q lm,t and Q km,t They represent the reactive power of line lm and line km at time t respectively; l lm,t represents the square of the current of line lm at time t; X lm Indicates the impedance of the line lm; P m,t sub and Q m,t sub Respectively represent the active power and reactive power of the substation at node m; P m,t L0 and Q m,t L0 They represent the active load and reactive load of the user connected to the distribution network node m respectively; P m,t MG represents the power provided by the distribution network to the microgrid at node m; The voltage constraint of the distribution network operation is: Among them, U m,t represents the square of the voltage at node m at time t; U l,t represents the square of the voltage at node l at time t; U m min Indicates the lower limit of the voltage safety amplitude; U m max Indicates the upper limit of the voltage safety amplitude; The line current carrying capacity constraint is: Among them, I lm max Indicates the upper limit of the line's allowable current carrying capacity; The substation output power constraint is: Among them, S m,t sub Represents the rated power of the substation at node m.

8. The optimization dispatching method of the distribution microgrid according to claim 7, characterized in that: The coupling constraints between the first optimization scheduling model and the second optimization scheduling model are: Among them, P grid,b,i (t) and P grid,s,i (t) represents the microgrid MG i The power purchased from the distribution network and the power sold; P m,t MG It represents the power provided by the distribution network to the microgrid at node m; MG represents the microgrid; ω(m) represents the set of microgrids connected to the distribution network node m.

9. The method for optimizing the dispatching of a distribution microgrid according to claim 8, characterized in that: The specific steps of iteratively solving the first optimization scheduling model and the second optimization scheduling model under the coupling constraint to obtain the operation strategy of each microgrid and the operation strategy of the distribution network include: Decoupling the first optimization scheduling model from the second optimization scheduling model according to the coupling constraint based on the alternating direction multiplier method, and iteratively solving the decoupled first optimization scheduling model and the second optimization scheduling model respectively; Calculating the primal residual and the dual residual between the first optimization scheduling model and the second optimization scheduling model; Determining whether the iterative solution result meets the convergence condition according to the original residual and the dual residual; If so, the current iterative solution result is output as the optimal scheduling result; Otherwise, continue to iterate until the convergence condition is met or the maximum number of iterations is reached.

10. The method for optimizing the dispatching of a distribution microgrid according to claim 9, characterized in that: The alternating direction multiplier method is used to decouple the first optimization scheduling model from the second optimization scheduling model according to the coupling constraint, and the decoupled first optimization scheduling model and the second optimization scheduling model are iteratively solved respectively. Among them, λ represents the Lagrange multiplier; ρ represents the penalty factor; u represents the current number of iterations; represents the value solved by the current iteration; and and They respectively represent the values ​​of the corresponding variables iteratively solved during the u-1th iteration.

Images