Island micro-grid economic frequency adaptive control method and device

Abstract

The invention provides an island micro-grid economic frequency adaptive control method and device, and belongs to the technical field of electric power. The method comprises the following steps: collecting measurement information and a controller control instruction from an island micro-grid internal network construction type converter machine end, and carrying out dimension raising mapping on the measurement information to obtain a corresponding extended observation vector; when the updating cycle of the micro-grid model arrives, constructing a measurement matrix, a control instruction matrix and a dimension raising data matrix by using measurement information of a plurality of recent sampling cycles, a controller control instruction and an extended observation vector; wherein a dominant dynamic mode is extracted from the dimension raising data matrix, then the microgrid model is updated, and then a control model giving consideration to grid frequency recovery and power generation cost is constructed and solved to obtain an optimal active power set value of the grid construction type converter so as to realize frequency control of the island microgrid. The method has the advantages of being high in calculation efficiency and capable of being applied to online control of the micro-grid, and the multi-target control requirement of the micro-grid can be met.

Description

Technical Field

[0001] The present invention belongs to the field of electric power technology, and in particular relates to an economic frequency adaptive control method and device for an island microgrid. Background Art

[0002] Frequency stability is a core issue for the safe operation of isolated microgrids. Traditional control methods employ a layered architecture to achieve multi-timescale coordination. Primary frequency regulation, based on active power-frequency droop control, achieves rapid power balancing on a microsecond timescale, but introduces steady-state frequency deviations. Secondary frequency regulation coordinates distributed generation (DGs) and load setpoints to recover frequency deviations caused by primary frequency regulation on a second-to-minute scale. Tertiary frequency regulation, building on secondary frequency regulation, considers load information, DG generation forecasts, and other information to reduce system operating costs.

[0003] Existing frequency regulation strategies can be categorized as model-based and model-free. Traditional model-based control strategies, such as model predictive control, rely on accurate microgrid dynamic models. However, in scenarios with a high proportion of power electronic equipment connected, frequent switching of microgrid topology leads to model parameter mismatch. The nonlinear characteristics of devices such as virtual synchronous machines further exacerbate linearization errors, causing control commands to deviate from actual requirements. Among model-free approaches, while traditional proportional-integral control offers strong robustness, fixed gain parameters make it difficult to reconcile the conflicting goals of frequency restoration and maximum renewable energy consumption. Data-driven approaches, such as neural network control, suffer from black-box characteristics, resulting in poor strategy interpretability and difficulty in safety verification. While the existing extended dynamic mode decomposition (EDMD) method based on the Koopman operator can construct high-precision linear prediction models through dimensionality-increasing mapping, its reliance on offline training data leads to two key bottlenecks: first, small disturbances under normal operating conditions cannot fully excite all system dynamic modes, resulting in the presence of unobserved subspaces in the model; second, when the control input excitation is insufficient, the data matrix is prone to ill-conditioned condition numbers, leading to accumulated control matrix identification errors. Summary of the Invention

[0004] The present invention aims to overcome the shortcomings of existing technologies by proposing an economic frequency adaptive control method and device for an islanded microgrid. This method offers the advantages of high computational efficiency and applicability for online microgrid control. It can also meet the multi-objective control requirements of microgrids, and is particularly suitable for simultaneously controlling frequency and improving system economics.

[0005] The first embodiment of the present invention provides an economic frequency adaptive control method for an island microgrid, comprising:

[0006] Collecting measurement information and controller control instructions from the grid-type converter in the island microgrid according to a set sampling period interval, wherein the corresponding extended observation vector is obtained by performing dimensional mapping on the measurement information;

[0007] When a preset microgrid model update cycle arrives, a measurement matrix, a control instruction matrix, and an increased-dimensional data matrix are constructed using the measurement information of the most recent multiple sampling cycles, the controller control instructions, and the extended observation vectors; wherein the dominant dynamic mode is extracted from the increased-dimensional data matrix to update the microgrid model;

[0008] Based on the updated microgrid model, a control model that takes into account both grid frequency recovery and power generation costs is constructed and solved to obtain the optimal active power setting value of the grid-forming converter to achieve frequency control of the island microgrid.

[0009] In a specific embodiment of the present invention, it also includes:

[0010] The sampling period interval is T s ; The current sampling period is t, and the collected measurement data include: the phase angle δ of the grid-type converter in the t-th sampling period t , the frequency f of the network converter in the tth sampling period t , the output power p of the grid-type converter in the tth sampling period t , the terminal voltage v of the grid-type converter in the tth sampling period t ; The collected measurement data are combined into the machine-side measurement information vector x of the t-th sampling period t =[δ t T ,f t T ,p t T ,v t T ] T The collected controller control instruction is the active power setting value of the grid-type converter in the tth sampling period, that is, u t =p ref，t , where u t is the controller control instruction of the tth sampling period, p ref，t is the active power setting value of the grid-type converter in the tth sampling period.

[0011] In a specific embodiment of the present invention, the extended observation vector is expressed as follows:

[0012]

[0013] Among them, y t is the extended observation vector of the t-th sampling period; Δδ t is the phase angle difference of the grid-type converter in the tth sampling period, which is δ t The difference between different elements within; is the Kronecker product.

[0014] In a specific embodiment of the present invention, updating the microgrid model includes:

[0015] 1) When the preset microgrid model update cycle arrives, obtain the x of the latest N sampling cycles t 、y t and u t , respectively forming the measurement data matrix X, the dimension-increased data matrix Y and the control instruction matrix U; where N is greater than the number of elements in the extended observation vector; the model update cycle interval T u Greater than T s ;

[0016] The expressions of the measurement data matrix X, the dimension-increased data matrix Y, and the control instruction matrix U are as follows:

[0017] X=[x t-N+1 ,...,x t ], (2)

[0018] U=[u t-N+1 ,...,u t-1 ], (3)

[0019] Y=[y t-N+1 ,...,y t ] (4)

[0020] 2) extracting the dominant dynamic mode from the dimension-increased data matrix Y obtained in step 1) to obtain the dominant modal data matrix Z, and then dividing the dominant modal data matrix Z and the measurement data matrix X in chronological order;

[0021] Among them, the principal component analysis is performed on the dimension-increased data matrix Y:

[0022]

[0023] Where Cov(Y) is the covariance matrix of Y, Q is the orthogonal projection matrix; diag(Λ) is the main diagonal matrix with eigenvalues as elements, and its main diagonal elements are arranged in descending order from large to small; retain the projection vectors corresponding to the first k largest eigenvalues in Q to obtain the projection matrix W, so that the dominant modal data matrix Z = WY satisfies the continuous excitation condition Where I is the identity matrix, and the parameter α is greater than the system measurement noise level;

[0024] According to the time sequence of data vector sampling, the dominant modal data matrix Z and the measurement data matrix X are divided into the following categories:

[0025] Z t =[z t-N+2 ,...,zt ],Z t-1 =[z t-N+1 ,...,z t-1 ], X t =[x t-N+2 ,...,x t ],X t-1 =[x t-N+1 ,...,x t-1 ];

[0026] Among them, z t =Wy t is the dominant mode vector of the t-th sampling period;

[0027] 3) Based on the matrix partitioning result of step 2), fit the microgrid model;

[0028] Among them, the linear microgrid model is constructed as shown in formula (5):

[0029]

[0030] By solving the least squares problem shown in (6) and (7), we can obtain matrices A, B, C, and D, where A is the state transfer matrix of the microgrid model, B is the control matrix of the microgrid model, C is the output matrix of the microgrid model, and D is the direct transfer matrix of the microgrid model.

[0031]

[0032] Afterwards, the model shown in formula (5) is tested according to the preset screening conditions: if the screening conditions are met, the microgrid model is output as the microgrid model of the current model update cycle; otherwise, go to step 4);

[0033] Among them, the screening conditions include controllability conditions and positive correlation conditions; the controllability conditions require the matrix [B, AB, A 2 B,...,A n-1 B] full rank; the positive correlation condition is based on the grid-type converter frequency f and the control instruction active reference value p ref The positive correlation between The main diagonal elements of are greater than 0, where the subscript f represents the rows of the matrix related to frequency, and the integer K represents the number of prediction steps, which is greater than T. u / T s ;

[0034] 4) Utilize the characteristics of the grid-type converter to assist in the design of the control matrix and update the microgrid model through the auxiliary model;

[0035] Among them, for the network-type converter there are:

[0036]

[0037] Among them, Equation (9) corresponds to droop control, and Equation (10) corresponds to the virtual synchronous machine;

[0038] Subscript i represents the converter number; δ i is the phase angle of the i-th converter, is δ i The differential of ω B is the base frequency value; f i is the frequency of the ith converter, is f i The differential of p i The active power of the i-th converter, f ref,i is the frequency set point of the i-th converter, p ref,i is the active power set point of the i-th converter, m P,i is the droop coefficient of the i-th converter, J r,i is the virtual inertia of the ith converter, D p,i is the virtual damping of the i-th converter;

[0039] The continuous-time state-space model of the microgrid with the state quantity of the grid-type converter as the state quantity is organized into the form of formula (10), where is the differential of the state quantity x, g(x) is the function related to the state quantity x in the system model, B p is the control matrix, D p is the frequency-dependent direct transfer matrix;

[0040]

[0041] Let B = T s WB p ,D=D p , substitute into equations (7) and (8) to obtain matrices A and C;

[0042] Afterwards, the model shown in formula (5) is tested according to the screening conditions: if the screening conditions are met, the microgrid model is output as the microgrid model of the current model update cycle; otherwise, the microgrid model is not updated in the current model update cycle, and the microgrid model of the previous model update cycle is kept as the microgrid model of the current model update cycle.

[0043] In a specific embodiment of the present invention, the control model that takes both grid frequency recovery and power generation cost into consideration is expressed as follows:

[0044]

[0045] Among them, the objective function Obj(t) To minimize the frequency deviation, f ref is the frequency reference value, and the per-unit value is 1; N ctr Represents the number of steps of model predictive control, N ctr Greater than T u / T s ;The symbol ‖·‖ represents the two-norm; is the cost function related to the output of the grid-type converter, the parameter c is the weight term, and c is greater than 0;

[0046] In the constraint condition g(x t ,u t ) is an inequality constraint, h(x t ,u t ) is an equality constraint.

[0047] A second embodiment of the present invention provides an economic frequency adaptive control device for an island microgrid, comprising:

[0048] A data sampling module is used to collect measurement information and controller control instructions from the grid-type converter in the island microgrid according to a set sampling period interval, wherein the corresponding extended observation vector is obtained by performing dimensional mapping on the measurement information;

[0049] a model updating module configured to, when a preset microgrid model update period arrives, construct a measurement matrix, a control instruction matrix, and an increased-dimensional data matrix using the measurement information, the controller control instructions, and the extended observation vectors of the most recent multiple sampling periods; extract a dominant dynamic mode from the increased-dimensional data matrix, and thereby update the microgrid model;

[0050] A control module is used to construct and solve a control model that takes into account both grid frequency recovery and power generation cost based on the updated microgrid model, so as to obtain the optimal active power setting value of the grid-forming converter to achieve frequency control of the island microgrid.

[0051] In a specific embodiment of the present invention, it also includes:

[0052] The sampling period interval is T s ; The current sampling period is t, and the collected measurement data include: the phase angle δ of the grid-type converter in the t-th sampling period t , the frequency f of the network converter in the tth sampling period t , the output power p of the grid-type converter in the tth sampling period t , the terminal voltage v of the grid-type converter in the tth sampling period t ; The collected measurement data are combined into the machine-side measurement information vector x of the t-th sampling period t =[δ tT ,f t T ,p t T ,v t T ] T The collected controller control instruction is the active power setting value of the grid-type converter in the tth sampling period, that is, u t =p ref，t , where u t is the controller control instruction of the tth sampling period, p ref，t is the active power setting value of the grid-type converter in the tth sampling period.

[0053] In a specific embodiment of the present invention, the extended observation vector is expressed as follows:

[0054]

[0055] Among them, y t is the extended observation vector of the t-th sampling period; Δδ t is the phase angle difference of the grid-type converter in the tth sampling period, which is δ t The difference between different elements within; is the Kronecker product.

[0056] In a specific embodiment of the present invention, updating the microgrid model includes:

[0057] 1) When the preset microgrid model update cycle arrives, obtain the x of the latest N sampling cycles t 、y t and u t , respectively forming the measurement data matrix X, the dimension-increased data matrix Y and the control instruction matrix U; where N is greater than the number of elements in the extended observation vector; the model update cycle interval T u Greater than T s ;

[0058] The expressions of the measurement data matrix X, the dimension-increased data matrix Y, and the control instruction matrix U are as follows:

[0059] X=[x t-N+1 ,...,x t ], (2)

[0060] U=[u t-N+1 ,...,u t-1 ], (3)

[0061] Y=[y t-N+1 ,...,y t ] (4)

[0062] 2) extracting the dominant dynamic mode from the dimension-increased data matrix Y obtained in step 1) to obtain the dominant modal data matrix Z, and then dividing the dominant modal data matrix Z and the measurement data matrix X in chronological order;

[0063] Among them, the principal component analysis is performed on the dimension-increased data matrix Y:

[0064]

[0065] Where Cov(Y) is the covariance matrix of Y, Q is the orthogonal projection matrix; diag(Λ) is the main diagonal matrix with eigenvalues as elements, and its main diagonal elements are arranged in descending order from large to small; retain the projection vectors corresponding to the first k largest eigenvalues in Q to obtain the projection matrix W, so that the dominant modal data matrix Z = WY satisfies the continuous excitation condition Where I is the identity matrix, and the parameter α is greater than the system measurement noise level;

[0066] According to the time sequence of data vector sampling, the dominant modal data matrix Z and the measurement data matrix X are divided into the following categories:

[0067] Z t =[z t-N+2 ,...,z t ],Z t-1 =[z t-N+1 ,...,z t-1 ], X t =[x t-N+2 ,...,x t ],X t-1 =[x t-N+1 ,...,x t-1 ];

[0068] Among them, z t =Wy t is the dominant mode vector of the t-th sampling period;

[0069] 3) Based on the matrix partitioning result of step 2), fit the microgrid model;

[0070] Among them, the linear microgrid model is constructed as shown in formula (5):

[0071]

[0072] By solving the least squares problem shown in (6) and (7), we can obtain matrices A, B, C, and D, where A is the state transfer matrix of the microgrid model, B is the control matrix of the microgrid model, C is the output matrix of the microgrid model, and D is the direct transfer matrix of the microgrid model.

[0073]

[0074] Afterwards, the model shown in formula (5) is tested according to the preset screening conditions: if the screening conditions are met, the microgrid model is output as the microgrid model of the current model update cycle; otherwise, go to step 4);

[0075] Among them, the screening conditions include controllability conditions and positive correlation conditions; the controllability conditions require the matrix [B, AB, A 2 B,...,A n-1 B] full rank; the positive correlation condition is based on the grid-type converter frequency f and the control instruction active reference value p ref The positive correlation between The main diagonal elements of are greater than 0, where the subscript f represents the rows of the matrix related to frequency, and the integer K represents the number of prediction steps, which is greater than T. u / T s ;

[0076] 4) Utilize the characteristics of the grid-type converter to assist in the design of the control matrix and update the microgrid model through the auxiliary model;

[0077] Among them, for the network-type converter there are:

[0078]

[0079] Among them, Equation (9) corresponds to droop control, and Equation (10) corresponds to the virtual synchronous machine;

[0080] Subscript i represents the converter number; δ i is the phase angle of the i-th converter, is δ i The differential of ω B is the base frequency value; f i is the frequency of the ith converter, is f i The differential of p i The active power of the i-th converter, f ref,i is the frequency set point of the i-th converter, p ref,i is the active power set point of the i-th converter, m P,i is the droop coefficient of the i-th converter, J r,i is the virtual inertia of the ith converter, D p,i is the virtual damping of the i-th converter;

[0081] The continuous-time state-space model of the microgrid with the state quantity of the grid-type converter as the state quantity is organized into the form of formula (10), where is the differential of the state quantity x, g(x) is the function related to the state quantity x in the system model, B p is the control matrix, D p is the frequency-dependent direct transfer matrix;

[0082]

[0083] Let B = T s WB p ,D=D p , substitute into equations (7) and (8) to obtain matrices A and C;

[0084] Afterwards, the model shown in formula (5) is tested according to the screening conditions: if the screening conditions are met, the microgrid model is output as the microgrid model of the current model update cycle; otherwise, the microgrid model is not updated in the current model update cycle, and the microgrid model of the previous model update cycle is kept as the microgrid model of the current model update cycle.

[0085] In a specific embodiment of the present invention, the control model that takes both grid frequency recovery and power generation cost into consideration is expressed as follows:

[0086]

[0087] Among them, the objective function Obj(t) To minimize the frequency deviation, f ref is the frequency reference value, and the per-unit value is 1; N ctr Represents the number of steps of model predictive control, N ctr Greater than T u / T s ;The symbol ‖·‖ represents the two-norm; is the cost function related to the output of the grid-type converter, the parameter c is the weight term, and c is greater than 0;

[0088] In the constraint condition g(x t ,u t ) is an inequality constraint, h(x t ,u t ) is an equality constraint.

[0089] A third embodiment of the present invention provides an electronic device, including:

[0090] at least one processor; and a memory communicatively coupled to the at least one processor;

[0091] The memory stores instructions that can be executed by the at least one processor, and the instructions are configured to execute the above-mentioned economic frequency adaptive control method for an island microgrid.

[0092] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the above-mentioned economic frequency adaptive control method of an island microgrid.

[0093] Features and beneficial effects of the present invention:

[0094] The present invention performs dynamic adaptive modeling of isolated island microgrids based on Koopman operator theory. By extracting the dominant dynamic modes online and combining the control matrix with the control affine structure to reduce the amount of data required for model updates, the present invention achieves the goal of updating the model using real-time data. The present invention overcomes the difficulty of online application of traditional EDMD algorithms, thereby overcoming the difficulties in maintaining microgrid models and the low efficiency and even instability of model-based methods caused by parameter uncertainty. Furthermore, based on this online updated model, the present invention utilizes model predictive control to achieve the goals of rapid frequency recovery and minimized operating costs in isolated island microgrids. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] Figure 1 This is a flow chart of an economic frequency adaptive control method for an island microgrid according to an embodiment of the present invention. DETAILED DESCRIPTION

[0096] The present invention proposes an economic frequency adaptive control method and device for an island microgrid, which is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0097] The first embodiment of the present invention provides an economic frequency adaptive control method for an island microgrid, comprising:

[0098] Collecting measurement information and controller control instructions from the grid-type converter in the island microgrid according to a set sampling period interval, wherein the corresponding extended observation vector is obtained by performing dimensional mapping on the measurement information;

[0099] When a preset microgrid model update cycle arrives, a measurement matrix, a control instruction matrix, and an increased-dimensional data matrix are constructed using the measurement information of the most recent multiple sampling cycles, the controller control instructions, and the extended observation vectors; wherein the dominant dynamic mode is extracted from the increased-dimensional data matrix to update the microgrid model;

[0100] Based on the updated microgrid model, a control model that takes into account both grid frequency recovery and power generation costs is constructed and solved to obtain the optimal active power setting value of the grid-forming converter to achieve frequency control of the island microgrid.

[0101] In a specific embodiment of the present invention, the economic frequency adaptive control method of an isolated microgrid is as follows: Figure 1 As shown, the symbol convention is: italic bold letters represent column vectors, positive bold letters represent matrices, and non-bold letters represent scalars; the method includes the following steps:

[0102] 1) Collecting measurement information and controller control instructions from the grid-type converter in the island microgrid at set sampling intervals, wherein the corresponding extended observation vector is obtained by performing dimensional mapping on the measurement information; the specific steps are as follows:

[0103] 1-1) Collect the measurement information and controller control instructions from the grid-connected converter in the island microgrid system at the set sampling period interval.

[0104] In this embodiment, the data sampling period interval T s In this embodiment, the current sampling period is t, and the collected measurement data include: the phase angle δ of the grid-type converter in the t-th sampling period t , the frequency f of the network converter in the tth sampling period t , the output power p of the grid-type converter in the tth sampling period t , the terminal voltage v of the grid-type converter in the tth sampling period t The collected measurement data are combined into the machine-side measurement information vector x of the t-th sampling period t =[δ t T ,f t T ,p t T ,v t T ] T The collected controller control instruction is the active power setting value of the grid-type converter in the tth sampling period, that is, u t =p ref,t , where u t is the controller control instruction of the tth sampling period, p ref，t is the active power setting value of the grid-type converter in the tth sampling period.

[0105] 1-2) Performing dimensional mapping on the networked converter machine-side measurement information collected in step 1-1) to obtain an extended observation vector.

[0106] In this embodiment, x t Perform dimensional mapping to obtain the extended observation vector y of the t-th sampling period t as follows:

[0107]

[0108] Among them, Δδ t is the phase angle difference of the grid-type converter in the tth sampling period, which is δ t The difference between different elements within; is the Kronecker product.

[0109] 1-3) The machine-side measurement information, controller control instructions and extended observation vectors obtained in each sampling period are stored in the database.

[0110] 2) Based on the result of step 1), when the preset microgrid model update cycle arrives, update the microgrid model online; the specific steps are as follows

[0111] 2-1) When the preset microgrid model update cycle arrives, the x values of the most recently stored N sampling cycles are obtained from the database. t 、y t and u t , respectively forming the measurement data matrix X, the dimension-increased data matrix Y and the control instruction matrix U. Where N should be greater than the number of elements in the extended observation vector, and in a specific embodiment of the present invention is 250; the model update cycle interval T u It should be greater than the data sampling period interval and less than 500ms, and in a specific embodiment of the present invention, it is 200ms.

[0112] In this embodiment, the expressions of the measurement data matrix X, the dimension-increased data matrix Y, and the control instruction matrix U are respectively as follows:

[0113] X=[x t-N+1 ,...,x t ], (2)

[0114] U=[u t-N+1 ,...,u t-1 ], (3)

[0115] Y=[y t-N+1 ,...,y t ] (4)

[0116] 2-2) Extract the dominant dynamic mode from the dimension-increased data matrix Y obtained in step 2-1) to obtain the dominant modal data matrix Z, and then divide the dominant modal data matrix Z and the measurement data matrix X in chronological order.

[0117] In this embodiment, principal component analysis is performed on the dimension-increased data matrix Y:

[0118]

[0119] Where Cov(Y) is the covariance matrix of Y, Q is the orthogonal projection matrix; diag(Λ) is the main diagonal matrix with eigenvalues as elements, and its main diagonal elements are arranged in descending order from large to small. The projection vectors corresponding to the first k largest eigenvalues in Q are retained to obtain the projection matrix W, so that the dominant modal data matrix Z = WY satisfies the continuous excitation condition (The value of k is automatically determined based on this inequality.) Where I is the identity matrix, and the parameter α should be greater than the system measurement noise level, which is determined by historical measurement statistics. In a specific embodiment of the present invention, it is 5e -5 .

[0120] At the same time, for the needs of subsequent model fitting, according to the time sequence of data vector sampling, the dominant mode data matrix Z and the measurement data matrix X are divided into the following categories:

[0121] Z t =[z t-N+2 ,...,z t ],Z t-1 =[z t-N+1 ,...,z t-1 ], X t =[x t-N+2 ,...,x t ],X t-1 =[x t-N+1 ,...,x t-1 ];

[0122] where z t =Wy t is the dominant mode vector of the tth sampling period, Z t ,Z t-1 To split the dominant modal data matrix Z in time order, each column of the former corresponds to the data of the next sampling period of each column of the latter, X t ,X t-1 Similarly, the measurement data matrix X is divided in time order.

[0123] 2-3) Based on the matrix partitioning result of step 2-2), fit the microgrid model.

[0124] In this embodiment, a linear microgrid model is constructed as shown in formula (5):

[0125]

[0126] By solving the least squares problem shown in (6) and (7), we can obtain matrices A, B, C, and D, where A is the state transfer matrix of the microgrid model, B is the control matrix of the microgrid model, C is the output matrix of the microgrid model, and D is the direct transfer matrix of the microgrid model.

[0127]

[0128] Furthermore, the model shown in formula (5) is tested according to the preset screening conditions to avoid numerical problems: if the screening conditions are met, the microgrid model is output as the microgrid model of the current model update cycle, and then step 3) is entered; otherwise, step 2-4) is entered.

[0129] In this embodiment, the screening conditions include controllability conditions and positive correlation conditions. Among them, the controllability condition requires the matrix [B, AB, A 2 B,...,A n-1 B] row full rank, ensuring that the model can be used for the control of island microgrids; the positive correlation condition is based on the grid-type converter frequency f and the control command active reference value p ref The positive correlation between The main diagonal elements of are greater than 0, where the subscript f represents the rows of the matrix related to frequency, and the integer K represents the number of prediction steps, which should be greater than T. u / T s , in a specific embodiment of the present invention, it is 25.

[0130] 2-4) Utilize the characteristics of the grid-type converter to assist in the design of the control matrix, and update the microgrid model through the auxiliary model.

[0131] The grid-type converter includes the droop control shown in Equation (9) and the virtual synchronous machine shown in Equation (10). i is the phase angle of the i-th converter, is its differential, ω B is the base frequency value 20πrad / s, f i is the frequency of the ith converter, is its differential, p i The active power of the i-th converter, f ref,i is the frequency set point of the i-th converter, p ref,i is the active power set point of the i-th converter, m P,i is the droop coefficient of the i-th converter, J r,i is the virtual inertia of the ith converter, D p,i is the virtual damping of the i-th converter (these control parameters are set by the network device owner and uploaded to the microgrid controller and are system parameters).

[0132]

[0133] Both types of converters have a control affine structure, that is, there is a linear relationship between the state quantity and the control input. Based on this feature, the continuous-time state space model of the microgrid with the state quantity of the meshed converter as the state quantity can be organized into the form of formula (10), where is the differential of the state quantity x, g(x) is the function related to the state quantity x in the system model, where the control matrix B p , the frequency-dependent direct transfer matrix D p are all known quantities, so according to the discrete time form shown in formula (5), let B = T s WB p ,D=D p , substitute into equations (7) and (8) to obtain matrices A and C.

[0134]

[0135] Furthermore, the model shown in formula (5) is tested according to the screening conditions in steps 2-3) to avoid numerical problems: if the screening conditions are met, the microgrid model is output as the microgrid model of the current model update cycle, and then step 3) is entered; otherwise, the microgrid model is not updated in the current model update cycle, and the microgrid model of the previous model update cycle is maintained as the microgrid model of the current model update cycle.

[0136] It should be noted that if no microgrid model meeting the screening conditions is output in the first model update cycle, this embodiment uses a small signal model obtained by linearization at the current operating equilibrium point as the microgrid model for the first model update cycle.

[0137] 3) Based on the microgrid model obtained in step 2), the model predictive control technology is applied to construct a control model that takes into account both grid frequency recovery and power generation costs, and the optimal control instructions are obtained.

[0138] In this embodiment, the control model expression for balancing grid frequency recovery and power generation cost is as follows:

[0139]

[0140] In formula (11), the objective function Obj(t) consists of two parts: the first part To minimize the frequency deviation, f ref is the frequency reference value, which is 50 Hz in this embodiment and 1 per unit; N ctr Represents the number of steps of model predictive control, N ctr Should be greater than T u / T s ; The symbol ‖·‖ represents the two-norm; the second part is the cost function related to the output of the grid-type converter, where ut =p ref,t , the cost function is not unique, and the parameter c (greater than 0) is a weight term that can be used to adjust the priority of the two objectives. t ,u t ) is an inequality constraint, h(x t ,u t ) is an equality constraint. The objective function and the constraint conditions are not unique, and C(u t ) is a quadratic function and the constraints are linear constraints, which makes Equation (11) a second-order cone programming problem that can be quickly solved using commercial solvers.

[0141] This embodiment solves the model shown in formula (11) to obtain the optimal active power setting value p of the grid-type converter ref,t,opt And send it to the converter for execution to achieve frequency control of the island microgrid and improve the economy of system operation.

[0142] To implement the above embodiment, a second embodiment of the present invention provides an economic frequency adaptive control device for an island microgrid, comprising:

[0143] A data sampling module is used to collect measurement information and controller control instructions from the grid-type converter in the island microgrid according to a set sampling period interval, wherein the corresponding extended observation vector is obtained by performing dimensional mapping on the measurement information;

[0144] a model updating module configured to, when a preset microgrid model update period arrives, construct a measurement matrix, a control instruction matrix, and an increased-dimensional data matrix using the measurement information, the controller control instructions, and the extended observation vectors of the most recent multiple sampling periods; extract a dominant dynamic mode from the increased-dimensional data matrix, and thereby update the microgrid model;

[0145] A control module is used to construct and solve a control model that takes into account both grid frequency recovery and power generation cost based on the updated microgrid model, so as to obtain the optimal active power setting value of the grid-forming converter to achieve frequency control of the island microgrid.

[0146] In a specific embodiment of the present invention, it also includes:

[0147] The sampling period interval is T s ; The current sampling period is t, and the collected measurement data include: the phase angle δ of the grid-type converter in the t-th sampling period t , the frequency f of the network converter in the tth sampling period t , the output power p of the grid-type converter in the tth sampling period t , the terminal voltage v of the grid-type converter in the tth sampling period t; The collected measurement data are combined into the machine-side measurement information vector x of the t-th sampling period t =[δ t T ,f t T ,p t T ,v t T ] T The collected controller control instruction is the active power setting value of the grid-type converter in the tth sampling period, that is, u t =p ref，t , where u t is the controller control instruction of the tth sampling period, p ref,t is the active power setting value of the grid-type converter in the tth sampling period.

[0148] In a specific embodiment of the present invention, the extended observation vector is expressed as follows:

[0149]

[0150] Among them, y t is the extended observation vector of the t-th sampling period; Δδ t is the phase angle difference of the grid-type converter in the tth sampling period, which is δ t The difference between different elements within; is the Kronecker product.

[0151] In a specific embodiment of the present invention, updating the microgrid model includes:

[0152] 1) When the preset microgrid model update cycle arrives, obtain the x of the latest N sampling cycles t 、y t and u t , respectively forming the measurement data matrix X, the dimension-increased data matrix Y and the control instruction matrix U; where N is greater than the number of elements in the extended observation vector; the model update cycle interval T u Greater than T s ;

[0153] The expressions of the measurement data matrix X, the dimension-increased data matrix Y, and the control instruction matrix U are as follows:

[0154] X=[x t-N+1 ,...,x t ], (2)

[0155] U=[u t-N+1 ,...,u t-1 ], (3)

[0156] Y=[y t-N+1 ,...,y t ] (4)

[0157] 2) extracting the dominant dynamic mode from the dimension-increased data matrix Y obtained in step 1) to obtain the dominant modal data matrix Z, and then dividing the dominant modal data matrix Z and the measurement data matrix X in chronological order;

[0158] Among them, the principal component analysis is performed on the dimension-increased data matrix Y:

[0159]

[0160] Where Cov(Y) is the covariance matrix of Y, Q is the orthogonal projection matrix; diag(Λ) is the main diagonal matrix with eigenvalues as elements, and its main diagonal elements are arranged in descending order from large to small; retain the projection vectors corresponding to the first k largest eigenvalues in Q to obtain the projection matrix W, so that the dominant modal data matrix Z = WY satisfies the continuous excitation condition Where I is the identity matrix, and the parameter α is greater than the system measurement noise level;

[0161] According to the time sequence of data vector sampling, the dominant modal data matrix Z and the measurement data matrix X are divided into the following categories:

[0162] Z t =[z t-N+2 ,...,z t ],Z t-1 =[z t-N+1 ,...,z t-1 ], X t =[x t-N+2 ,...,x t ],X t-1 =[x t-N+1 ,...,x t-1 ];

[0163] Among them, z t =Wy t is the dominant mode vector of the t-th sampling period;

[0164] 3) Based on the matrix partitioning result of step 2), fit the microgrid model;

[0165] Among them, the linear microgrid model is constructed as shown in formula (5):

[0166]

[0167] By solving the least squares problem shown in (6) and (7), we can obtain matrices A, B, C, and D, where A is the state transfer matrix of the microgrid model, B is the control matrix of the microgrid model, C is the output matrix of the microgrid model, and D is the direct transfer matrix of the microgrid model.

[0168]

[0169] Afterwards, the model shown in formula (5) is tested according to the preset screening conditions: if the screening conditions are met, the microgrid model is output as the microgrid model of the current model update cycle; otherwise, go to step 4);

[0170] Among them, the screening conditions include controllability conditions and positive correlation conditions; the controllability conditions require the matrix [B, AB, A 2 B,...,A n-1 B] full rank; the positive correlation condition is based on the grid-type converter frequency f and the control instruction active reference value p ref The positive correlation between The main diagonal elements of are greater than 0, where the subscript f represents the rows of the matrix related to frequency, and the integer K represents the number of prediction steps, which is greater than T. u / T s ;

[0171] 4) Utilize the characteristics of the grid-type converter to assist in the design of the control matrix and update the microgrid model through the auxiliary model;

[0172] Among them, for the network-type converter there are:

[0173]

[0174] Among them, Equation (9) corresponds to droop control, and Equation (10) corresponds to the virtual synchronous machine;

[0175] Subscript i represents the converter number; δ i is the phase angle of the i-th converter, is δ i The differential of ω B is the base frequency value; f i is the frequency of the ith converter, is f i The differential of p i The active power of the i-th converter, f ref,i is the frequency set point of the i-th converter, p ref,i is the active power set point of the i-th converter, m P,i is the droop coefficient of the i-th converter, J r,i is the virtual inertia of the ith converter, D p,iis the virtual damping of the i-th converter;

[0176] The continuous-time state-space model of the microgrid with the state quantity of the grid-type converter as the state quantity is organized into the form of formula (10), where is the differential of the state quantity x, g(x) is the function related to the state quantity x in the system model, B p is the control matrix, D p is the frequency-dependent direct transfer matrix;

[0177]

[0178] Let B = T s WB p ,D=D p , substitute into equations (7) and (8) to obtain matrices A and C;

[0179] Afterwards, the model shown in formula (5) is tested according to the screening conditions: if the screening conditions are met, the microgrid model is output as the microgrid model of the current model update cycle; otherwise, the microgrid model is not updated in the current model update cycle, and the microgrid model of the previous model update cycle is kept as the microgrid model of the current model update cycle.

[0180] In a specific embodiment of the present invention, the control model that takes both grid frequency recovery and power generation cost into consideration is expressed as follows:

[0181]

[0182] Among them, the objective function Obj(t) To minimize the frequency deviation, f ref is the frequency reference value, and the per-unit value is 1; N ctr Represents the number of steps of model predictive control, N ctr Greater than T u / T s ;The symbol ‖·‖ represents the two-norm; is the cost function related to the output of the grid-type converter, the parameter c is the weight term, and c is greater than 0;

[0183] In the constraint condition g(x t ,u t ) is an inequality constraint, h(x t ,u t ) is an equality constraint.

[0184] This approach reduces the amount of data required for model updates by extracting the dominant dynamic modes online and designing the control matrix based on the control affine structure, enabling real-time model updates. Based on this online updated model, model predictive control can then be used to achieve rapid frequency restoration and minimize operating costs in islanded microgrids.

[0185] To implement the above embodiment, a third aspect of the present invention provides an electronic device, including:

[0186] at least one processor; and a memory communicatively coupled to the at least one processor;

[0187] The memory stores instructions that can be executed by the at least one processor, and the instructions are configured to execute the above-mentioned economic frequency adaptive control method for an island microgrid.

[0188] To implement the above embodiment, a fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the above-mentioned economic frequency adaptive control method of an island microgrid.

[0189] It should be noted that the computer-readable medium mentioned above in the present disclosure may be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or component, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, device, or component. In the present disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted using any suitable medium, including but not limited to wires, optical cables, RF (radio frequency), etc., or any suitable combination thereof.

[0190] The computer-readable medium may be included in the electronic device or may exist independently without being incorporated into the electronic device. The computer-readable medium carries one or more programs. When executed by the electronic device, the one or more programs cause the electronic device to execute the economic frequency adaptive control method for an islanded microgrid according to the above-described embodiment.

[0191] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on the remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0192] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and features of different embodiments or examples without contradiction.

[0193] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of such features. Throughout the description of this application, "plurality" means at least two, for example, two, three, etc., unless otherwise specifically defined.

[0194] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0195] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and a portable compact disc read-only memory (CDROM). Furthermore, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or otherwise processing it in a suitable manner if necessary, and then storing it in a computer memory.

[0196] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0197] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0198] In addition, the functional units in the various embodiments of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into a module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0199] The storage medium mentioned above may be a read-only memory, a magnetic disk, or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present application. Persons skilled in the art may make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. An economic frequency adaptive control method for an island microgrid, characterized in that: include: Collecting measurement information and controller control instructions from the grid-type converter in the island microgrid according to a set sampling period interval, wherein the corresponding extended observation vector is obtained by performing dimensional mapping on the measurement information; When a preset microgrid model update cycle arrives, a measurement matrix, a control instruction matrix, and an increased-dimensional data matrix are constructed using the measurement information of the most recent multiple sampling cycles, the controller control instructions, and the extended observation vectors; wherein the dominant dynamic mode is extracted from the increased-dimensional data matrix to update the microgrid model; Based on the updated microgrid model, a control model that takes into account both grid frequency recovery and power generation costs is constructed and solved to obtain the optimal active power setting value of the grid-forming converter to achieve frequency control of the island microgrid.

2. The method according to claim 1, characterized in that Also includes: The sampling period interval is T s ; The current sampling period is t, and the collected measurement data include: the phase angle δ of the grid-type converter in the t-th sampling period t , the frequency f of the network converter in the tth sampling period t , the output power p of the grid-type converter in the tth sampling period t , the terminal voltage v of the grid-type converter in the tth sampling period t ; The collected measurement data are combined into the machine-side measurement information vector x of the t-th sampling period t =[δ t T ,f t T ,p t T ,v t T ] T The collected controller control instruction is the active power setting value of the grid-type converter in the tth sampling period, that is, u t =p ref，t , where u t is the controller control instruction of the tth sampling period, p ref，t is the active power setting value of the grid-type converter in the tth sampling period.

3. The method according to claim 2, characterized in that The extended observation vector expression is as follows: Among them, y t is the extended observation vector of the t-th sampling period; Δδ t is the phase angle difference of the grid-type converter in the tth sampling period, which is δ t The difference between different elements within; is the Kronecker product.

4. The method according to claim 3, characterized in that The updating of the microgrid model includes: 1) When the preset microgrid model update cycle arrives, obtain the x of the latest N sampling cycles t 、y t and u t , respectively forming the measurement data matrix X, the dimension-increased data matrix Y and the control instruction matrix U; where N is greater than the number of elements in the extended observation vector; the model update cycle interval T u Greater than T s ; The expressions of the measurement data matrix X, the dimension-increased data matrix Y, and the control instruction matrix U are as follows: X=[x t-N+1 ,...,x t ], (2) U=[u t-N+1 ,...,in t-1 ], (3) And=[and t-N+1 ,...,and t ] (4) 2) extracting the dominant dynamic mode from the dimension-increased data matrix Y obtained in step 1) to obtain the dominant modal data matrix Z, and then dividing the dominant modal data matrix Z and the measurement data matrix X in chronological order; Among them, the principal component analysis is performed on the dimension-increased data matrix Y: Where Cov(Y) is the covariance matrix of Y, Q is the orthogonal projection matrix; diag(Λ) is the main diagonal matrix with eigenvalues as elements, and its main diagonal elements are arranged in descending order from large to small; retain the projection vectors corresponding to the first k largest eigenvalues in Q to obtain the projection matrix W, so that the dominant modal data matrix Z = WY satisfies the continuous excitation condition Where I is the identity matrix, and the parameter α is greater than the system measurement noise level; According to the time sequence of data vector sampling, the dominant modal data matrix Z and the measurement data matrix X are divided into the following categories: Z t =[z t-N+2 ,...,z t ],Z t-1 =[z t-N+1 ,...,z t-1 ],X t =[x t-N+2 ,...,x t ],X t-1 =[x t-N+1 ,...,x t-1 ]; Among them, z t =Wy t is the dominant mode vector of the t-th sampling period; 3) Based on the matrix partitioning result of step 2), fit the microgrid model; Among them, the linear microgrid model is constructed as shown in formula (5): By solving the least squares problem shown in (6) and (7), we can obtain matrices A, B, C, and D, where A is the state transfer matrix of the microgrid model, B is the control matrix of the microgrid model, C is the output matrix of the microgrid model, and D is the direct transfer matrix of the microgrid model. Afterwards, the model shown in formula (5) is tested according to the preset screening conditions: if the screening conditions are met, the microgrid model is output as the microgrid model of the current model update cycle; otherwise, go to step 4); Among them, the screening conditions include controllability conditions and positive correlation conditions; the controllability conditions require the matrix [B, AB, A 2 B,...,A n-1 B] full rank; the positive correlation condition is based on the grid-type converter frequency f and the control instruction active reference value p ref The positive correlation between The main diagonal elements of are greater than 0, where the subscript f represents the rows of the matrix related to frequency, and the integer K represents the number of prediction steps, which is greater than T. u / T s ; 4) Utilize the characteristics of the grid-type converter to assist in the design of the control matrix and update the microgrid model through the auxiliary model; Among them, for the network-type converter there are: Among them, Equation (9) corresponds to droop control, and Equation (10) corresponds to the virtual synchronous machine; Subscript i represents the converter number; δ i is the phase angle of the i-th converter, is δ i The differential of ω B is the base frequency value; f i is the frequency of the ith converter, is f i The differential of p i The active power of the i-th converter, f ref,i is the frequency set point of the i-th converter, p ref,i is the active power set point of the i-th converter, m P,i is the droop coefficient of the i-th converter, J r,i is the virtual inertia of the ith converter, D p,i is the virtual damping of the i-th converter; The continuous-time state-space model of the microgrid with the state quantity of the grid-type converter as the state quantity is organized into the form of formula (10), where is the differential of the state quantity x, g(x) is the function related to the state quantity x in the system model, B p is the control matrix, D p is the frequency-dependent direct transfer matrix; Let B = T s WB p ,D=D p , substitute into equations (7) and (8) to obtain matrices A and C; Afterwards, the model shown in formula (5) is tested according to the screening conditions: if the screening conditions are met, the microgrid model is output as the microgrid model of the current model update cycle; otherwise, the microgrid model is not updated in the current model update cycle, and the microgrid model of the previous model update cycle is kept as the microgrid model of the current model update cycle.

5. The method according to claim 4, characterized in that The control model that takes into account both grid frequency recovery and power generation cost is expressed as follows: Among them, the objective function Obj(t) To minimize the frequency deviation, f ref is the frequency reference value, and the per-unit value is 1; N ctr Represents the number of steps of model predictive control, N ctr Greater than T u / T s ;The symbol ‖·‖ represents the two-norm; is the cost function related to the output of the grid-type converter, the parameter c is the weight term, and c is greater than 0; In the constraint condition g(x t ,u t ) is an inequality constraint, h(x t ,u t ) is an equality constraint.

6. An economic frequency adaptive control device for an island microgrid, characterized in that: include: A data sampling module is used to collect measurement information and controller control instructions from the grid-type converter in the island microgrid according to a set sampling period interval, wherein the corresponding extended observation vector is obtained by performing dimensional mapping on the measurement information; a model updating module configured to, when a preset microgrid model update period arrives, construct a measurement matrix, a control instruction matrix, and an increased-dimensional data matrix using the measurement information, the controller control instructions, and the extended observation vectors of the most recent multiple sampling periods; extract a dominant dynamic mode from the increased-dimensional data matrix, and thereby update the microgrid model; A control module is used to construct and solve a control model that takes into account both grid frequency recovery and power generation cost based on the updated microgrid model, so as to obtain the optimal active power setting value of the grid-forming converter to achieve frequency control of the island microgrid.

7. The device according to claim 6, characterized in that Also includes: The sampling period interval is T s ; The current sampling period is t, and the collected measurement data include: the phase angle δ of the grid-type converter in the t-th sampling period t , the frequency f of the network converter in the tth sampling period t , the output power p of the grid-type converter in the tth sampling period t , the terminal voltage v of the grid-type converter in the tth sampling period t ; The collected measurement data are combined into the machine-side measurement information vector x of the t-th sampling period t =[δ t T ,f t T ,p t T ,v t T ] T The collected controller control instruction is the active power setting value of the grid-type converter in the tth sampling period, that is, u t =p ref，t , where u t is the controller control instruction of the tth sampling period, p ref，t is the active power setting value of the grid-type converter in the tth sampling period.

8. The device according to claim 7, characterized in that The extended observation vector expression is as follows: Among them, y t is the extended observation vector of the t-th sampling period; Δδ t is the phase angle difference of the grid-type converter in the tth sampling period, which is δ t The difference between different elements within; is the Kronecker product.

9. The device according to claim 8, characterized in that The updating of the microgrid model includes: 1) When the preset microgrid model update cycle arrives, obtain the x of the latest N sampling cycles t 、y t and u t , respectively forming the measurement data matrix X, the dimension-increased data matrix Y and the control instruction matrix U; where N is greater than the number of elements in the extended observation vector; the model update cycle interval T u Greater than T s ; The expressions of the measurement data matrix X, the dimension-increased data matrix Y, and the control instruction matrix U are as follows: X=[x t-N+1 ,...,x t ], (2) U=[u t-N+1 ,...,in t-1 ], (3) And=[and t-N+1 ,...,and t ] (4) 2) extracting the dominant dynamic mode from the dimension-increased data matrix Y obtained in step 1) to obtain the dominant modal data matrix Z, and then dividing the dominant modal data matrix Z and the measurement data matrix X in chronological order; Among them, the principal component analysis is performed on the dimension-increased data matrix Y: Where Cov(Y) is the covariance matrix of Y, Q is the orthogonal projection matrix; diag(Λ) is the main diagonal matrix with eigenvalues as elements, and its main diagonal elements are arranged in descending order from large to small; retain the projection vectors corresponding to the first k largest eigenvalues in Q to obtain the projection matrix W, so that the dominant modal data matrix Z = WY satisfies the continuous excitation condition Where I is the identity matrix, and the parameter α is greater than the system measurement noise level; According to the time sequence of data vector sampling, the dominant modal data matrix Z and the measurement data matrix X are divided into the following categories: Z t =[z t-N+2 ,...,z t ],Z t-1 =[z t-N+1 ,...,z t-1 ],X t =[x t-N+2 ,...,x t ],X t-1 =[x t-N+1 ,...,x t-1 ]; Among them, z t =Wy t is the dominant mode vector of the t-th sampling period; 3) Based on the matrix partitioning result of step 2), fit the microgrid model; Among them, the linear microgrid model is constructed as shown in formula (5): By solving the least squares problem shown in (6) and (7), we can obtain matrices A, B, C, and D, where A is the state transfer matrix of the microgrid model, B is the control matrix of the microgrid model, C is the output matrix of the microgrid model, and D is the direct transfer matrix of the microgrid model. Afterwards, the model shown in formula (5) is tested according to the preset screening conditions: if the screening conditions are met, the microgrid model is output as the microgrid model of the current model update cycle; otherwise, go to step 4); Among them, the screening conditions include controllability conditions and positive correlation conditions; the controllability conditions require the matrix [B, AB, A 2 B,...,A n-1 B] full rank; the positive correlation condition is based on the grid-type converter frequency f and the control instruction active reference value p ref The positive correlation between The main diagonal elements of are greater than 0, where the subscript f represents the rows of the matrix related to frequency, and the integer K represents the number of prediction steps, which is greater than T. u / T s ; 4) Utilize the characteristics of the grid-type converter to assist in the design of the control matrix and update the microgrid model through the auxiliary model; Among them, for the network-type converter there are: Among them, Equation (9) corresponds to droop control, and Equation (10) corresponds to the virtual synchronous machine; Subscript i represents the converter number; δ i is the phase angle of the i-th converter, is δ i The differential of ω B is the base frequency value; f i is the frequency of the ith converter, is f i The differential of p i The active power of the i-th converter, f ref,i is the frequency set point of the i-th converter, p ref,i is the active power set point of the i-th converter, m P,i is the droop coefficient of the i-th converter, J r,i is the virtual inertia of the ith converter, D p,i is the virtual damping of the i-th converter; The continuous-time state-space model of the microgrid with the state quantity of the grid-type converter as the state quantity is organized into the form of formula (10), where is the differential of the state quantity x, g(x) is the function related to the state quantity x in the system model, B p is the control matrix, D p is the frequency-dependent direct transfer matrix; Let B = T s WB p ,D=D p , substitute into equations (7) and (8) to obtain matrices A and C; Afterwards, the model shown in formula (5) is tested according to the screening conditions: if the screening conditions are met, the microgrid model is output as the microgrid model of the current model update cycle; otherwise, the microgrid model is not updated in the current model update cycle, and the microgrid model of the previous model update cycle is kept as the microgrid model of the current model update cycle.

10. The device according to claim 9, characterized in that The control model that takes into account both grid frequency recovery and power generation cost is expressed as follows: Among them, the objective function Obj(t) To minimize the frequency deviation, f ref is the frequency reference value, and the per-unit value is 1; N ctr Represents the number of steps of model predictive control, N ctr Greater than T u / T s ;The symbol ‖·‖ represents the two-norm; is the cost function related to the output of the grid-type converter, the parameter c is the weight term, and c is greater than 0; In the constraint condition g(x t ,u t ) is an inequality constraint, h(x t ,u t ) is an equality constraint.

Images