A wind farm power optimization control method and system based on distributed control

Abstract

The present invention discloses a method and system for optimizing the power control of a wind farm based on distributed control. The present invention includes establishing an incremental state space model for fatigue loads, determining the active power of the wind turbine as the main optimization objective based on the incremental state space model of fatigue loads, coordinating the control of reactive power and node voltage as the secondary optimization objective for the multi-objective function, and determining the constraint conditions; determining the centralized control model representation of the objective function and its constraint conditions; performing distributed iterative control model decomposition based on the centralized algorithm representation of the objective function and its constraint conditions to obtain a distributed iterative control model, and executing the distributed iterative control model through a distributed controller to converge to obtain a global optimal solution. The present invention optimizes the power control of the wind farm based on the distributed iterative control model, enabling the wind farm to have better robustness and stability, and being able to reduce the fatigue load of the wind turbine, reduce the deviation of the wind turbine voltage and maintain voltage stability, and reduce network losses.

Description

Technical Field

[0001] The present invention relates to the power control technology of a wind farm, and particularly to a method and system for optimizing the power control of a wind farm based on distributed control. Background Art

[0002] Wind power generation is an emerging renewable energy technology with the most mature technology, basically realized commercialization and the greatest development potential at present. Developing wind power conforms to the development strategy of building a clean, low-carbon, safe and efficient energy supply system in China. The power optimization control of large-scale wind farms can ensure the efficient and safe power supply of wind farms.

[0003] In traditional wind farm control, there is no coupling between the active power and reactive power controllers, and the active and reactive powers can be controlled separately. When the wind farm is in a load reduction operation mode, dynamically optimizing and adjusting the active power reference command of the fan can effectively reduce the fatigue load of the fan. The wind farm can effectively reduce the voltage deviation of the fan access node by dynamically optimizing the available reactive power. The centralized control method is the main control method of the current wind farm. This method needs to collect all parameter information inside the wind farm, and then process the control algorithms of all fans through a central controller and issue the reference commands of all fans. The control effect of the centralized control method is too dependent on the communication quality, and processing the control short hair of all fans leads to a large calculation burden, resulting in insufficient robustness of this control method. The distributed control method decomposes the calculation tasks of the central controller into multiple subtasks and calculates them simultaneously by multiple distributed processing units, which is very suitable for the implementation of optimization algorithms for large-scale wind farms. Summary of the Invention

[0004] The technical problem to be solved by the present invention: Aiming at the above problems of the prior art, a method and system for optimizing the power control of a wind farm based on distributed control are provided. The present invention optimizes the power control of the wind farm based on a distributed iterative control model, so that the wind farm has better robustness and stability, and can reduce the fatigue load of the fan, reduce the voltage deviation of the fan and maintain voltage stability, and reduce network losses.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0006] A method for optimizing the power control of a wind farm based on distributed control, comprising:

[0007] 1) Establish an incremental state space model for fatigue load, and determine the first part of the control target Obj1 of the fan active power based on the incremental state space model for fatigue load; determine the second part of the control target Obj2 of the fan active power through the measurement of the available reactive power of the fan and the voltage deviation at the access point; determine the third part of the control target Obj3 of the fan active power by optimizing the radiative power flow of the wind farm;

[0008] 2) Determine the total control objective Obj of the wind farm based on the first - part control objective Obj1, the second - part control objective Obj2, and the third - part control objective Obj3 total as the objective function, and determine the constraint conditions;

[0009] 3) Determine the centralized control model representation of the objective function and its constraint conditions;

[0010] 4) Based on the centralized algorithm representation of the objective function and its constraint conditions, disassemble the distributed iterative control model to obtain the distributed iterative control model, and execute the distributed iterative control model through the distributed controller to converge and obtain the global optimal solution.

[0011] Optionally, the functional expression of the incremental state - space model regarding fatigue load established in step 1) is:

[0012]

[0013] In the above formula, ΔT S is the incremental model of the bearing torque T S , η g is the generator efficiency, J t is the equivalent inertia, J g is the generator inertia, J r is the impeller - rotor inertia, J t = J r + J g , K ωT , K βT , K βF and K ωF are the coefficients of the first - order Taylor expansion of the bearing torque T S and the wind thrust F t at the initial time t0 point, Δβ is the increment of the pitch angle, Δω g is the increment of the generator speed, is the value of the generator output power at time t0, is the value of the filtered rotor speed at the initial time t0, is the increment of the generator power reference value, ΔF t is the increment model of the wind thrust; The functional expression for determining the first - part control objective Obj1 of the active power of the wind turbine based on the incremental state - space model regarding fatigue load in step 1) is:

[0014]

[0015] In the above formula, N f is the number of branches of the wind farm, N w is the number of wind turbines on one branch, ΔTs.i (k) is the bearing torque T S The incremental equation of, ΔF t.i (k) is the wind thrust F t The incremental equation of is the bearing torque T S The weight coefficient of is the wind thrust F t The weight coefficient of, and the optimization objective corresponds to the independent variable P which is the local variable of the wind turbine ij where P ij is the active power output by the j-th wind turbine on the i-th branch

[0016] Optionally, the functional expression of the second part of the control objective Obj2 of the active power of the wind turbine determined by measuring the available reactive power of the wind turbine and the voltage deviation at the access point in step 1) is:

[0017]

[0018] In the above formula, N f is the number of branches of the wind farm, N w is the number of wind turbines on a branch, Q j_i is the reactive power output by the wind turbines on a branch is the reference value of the reactive power ratio distribution of the j-th wind turbine on the i-th branch is the terminal voltage of the j-th wind turbine on the i-th branch, U rated is the rated voltage of the wind farm, Q q is the weight coefficient of the reactive voltage, Q u is the weight coefficient of the voltage deviation, and the optimization objective corresponds to the independent variables Q which is the local variable of the wind turbine ij and the local variable where Q ij is the reactive power output by the j-th wind turbine on the i-th branch is the terminal voltage of the j-th wind turbine on the i-th branch

[0019] Optionally, the functional expression of the third part of the control objective Obj3 of the active power of the wind turbine determined by optimizing the radiative power flow of the wind farm in step 1) is:

[0020]

[0021] In the above formula, N f is the number of branches of the wind farm, N w is the number of wind turbines on a branch is the line resistance of the j-th wind turbine on the i-th branch is the local variable of the wind turbine, representing the active power flowing into the access point; It is a local variable of the fan, representing the reactive power flowing into the access point; V s is the standard voltage, Q loss is the line loss of the wind farm. The independent variable corresponding to the optimization objective is the power flow information of the access point of the fan, including the global variables P ij and Q ij , where P ij is the active power flowing into the access point of the j-th fan on the i-th branch, Q ij is the reactive power flowing into the access point of the j-th fan on the i-th branch.

[0022] Optionally, the functional expression for determining the total control objective Obj total of the wind farm in step 2) is:

[0023] Obj total = Obj1 + Obj2 + Obj3,

[0024] The functional expressions for determining the constraint conditions in step 2) include some or all of the constraints in the following functional equations:

[0025]

[0026] In the above formula, P j_i is a local variable of the fan, representing the output active power; is the available active power of the fan; is the available reactive power of the fan; Q j_i is a local variable of the fan, representing the output reactive power; is a local variable of the fan, representing the active power flowing into the access point; is a local variable of the fan, representing the active power flowing out of the access point; is a local variable of the fan, representing the reactive power flowing into the access point; is a local variable of the fan, representing the reactive power flowing out of the access point; is a local variable of the fan, representing the square of the voltage at the front end of the access point; is a local variable of the fan, representing the square of the voltage at the end of the access point; is the line resistance of the j-th fan on the i-th branch; is the line reactance of the j-th fan on the i-th branch; δ is a preset error coefficient; V s is the rated voltage of the wind farm.

[0027] Optionally, the functional expression represented by the centralized control model of the objective function and its constraint conditions determined in step 3) is:

[0028]

[0029] And its independent variable is:

[0030]

[0031] In the above formula, is the distributed iterative control model representation of the objective function and its constraints; x is the local variable matrix, y is the dual variable matrix, and z is the global variable matrix; ΔT s.i (k) is the incremental equation of the bearing torque T S The incremental equation of ΔF t.i (k) is the incremental equation of the wind thrust F t The incremental equation of is the weight coefficient of the bearing torque T S The weight coefficient of is the weight coefficient of the wind thrust F t The weight coefficient of Q q is the weight coefficient of the reactive voltage, Q j_i is the reactive power output by the number of wind turbines on a branch is the reference value of the reactive power ratio distribution of the jth wind turbine on the ith branch; Q u is the weight coefficient of the voltage deviation, U j_i is the terminal voltage of the jth wind turbine on the ith branch, U rated is the rated voltage of the wind farm; Q loss is the line loss of the wind farm; is the line resistance of the jth wind turbine on the ith branch; is the local variable of the wind turbine, representing the active power flowing into the connection point; is the local variable of the wind turbine, representing the reactive power flowing into the connection point; N w is the number of wind turbines on a branch; is the local variable of the active power flowing out of the connection point; is the global variable of the active power flowing into the connection point; is the global variable of the active power flowing out of the connection point; is the local variable of the reactive power flowing out of the connection point; is the global variable of the reactive power flowing into the connection point; is the global variable of the reactive power flowing out of the connection point; is the local variable of the voltage at the end of the connection point; is the global variable of the voltage at the front end of the connection point; is the global variable of the voltage at the end of the connection point; is the dual variable of the active power flowing into the connection point; is the dual variable of the active power flowing out of the connection point; is the dual variable of the reactive power flowing into the connection point; is the dual variable of the reactive power flowing out of the access point; is the dual variable of the voltage at the front end of the access point; is the dual variable of the voltage at the end of the access point.

[0032] Optionally, the functional expression of the distributed iterative control model obtained in step 4) is:

[0033]

[0034] In the above formula, k represents the number of steps of iteration, is the k-th step iteration value of the global variable of the active power flowing into the access point; is the k-th step iteration value of the global variable of the active power flowing out of the access point; is the k-th step iteration value of the local variable of the reactive power flowing out of the access point; is the k-th step iteration value of the global variable of the reactive power flowing into the access point; is the k-th step iteration value of the de-global variable of the reactive power flowing out of the access point; is the k-th step iteration value of the local variable of the voltage at the end of the access point; is the k-th step iteration value of the global variable of the voltage at the front end of the access point; is the k-th step iteration value of the global variable of the voltage at the end of the access point; is the (k + 1)-th step iteration value of the global variable of the active power flowing into the access point; is the (k + 1)-th step iteration value of the global variable of the active power flowing out of the access point; is the (k + 1)-th step iteration value of the local variable of the reactive power flowing out of the access point; is the (k + 1)-th step iteration value of the global variable of the reactive power flowing into the access point; is the (k + 1)-th step iteration value of the de-global variable of the reactive power flowing out of the access point; is the (k + 1)-th step iteration value of the local variable of the voltage at the end of the access point; is the (k + 1)-th step iteration value of the global variable of the voltage at the front end of the access point; is the (k + 1)-th step iteration value of the global variable of the voltage at the end of the access point; and the functional expression for the update of the dual variable is:

[0035] y [k+1] = δ(x [k+1] - z [k+1] ),

[0036] In the above formula, x [k+1] is the local variable matrix of the (k + 1)-th step iteration, y [k+1] is the dual variable matrix of the (k + 1)-th step iteration, z [k+1]is the global variable matrix for the (k + 1)-th iteration step, and δ is the penalty coefficient.

[0037] Optionally, the expression of the convergence condition function for executing the distributed iterative control model convergence by the distribution controller in step 4) is:

[0038] x [k+1] -x [k] ≤ ε1,

[0039] x [k+1] -z [k+1] ≤ ε1,

[0040]

[0041] In the above formula, x [k+1] is the local variable matrix for the (k + 1)-th iteration step, x [k] is the local variable matrix for the k-th iteration step, z [k+1] is the global variable matrix for the (k + 1)-th iteration step, is the value of the objective function for the (k + 1)-th iteration step, is the value of the objective function for the k-th iteration step, and ε1, ε2 are error coefficients.

[0042] In addition, the present invention also provides a wind farm power optimization control system based on distributed control, including a microprocessor and a memory connected to each other, and the microprocessor is programmed or configured to execute the steps of the wind farm power optimization control method based on distributed control.

[0043] In addition, the present invention also provides a computer-readable storage medium, in which a computer program for being executed by a computer device to implement the wind farm power optimization control method based on distributed control is stored.

[0044] Compared with the prior art, the present invention mainly has the following advantages:

[0045] 1. The present invention includes establishing an incremental state space model regarding fatigue load, and determining the first part of the control target Obj1 of the active power of the wind turbine based on the incremental state space model regarding fatigue load; determining the second part of the control target Obj2 of the active power of the wind turbine by measuring the available reactive power of the wind turbine and the voltage deviation at the access point; determining the third part of the control target Obj3 of the active power of the wind turbine by optimizing the radiative power flow of the wind farm; and determining the total control target Obj of the wind farm according to the first part of the control target Obj1, the second part of the control target Obj2, and the third part of the control target Obj3 totalTake it as the objective function and determine the constraint conditions; determine the centralized control model representation of the objective function and its constraint conditions; perform distributed iterative control model decomposition based on the centralized algorithm representation of the objective function and its constraint conditions to obtain a distributed iterative control model, and execute the distributed iterative control model through a distributed controller to converge and obtain the global optimal solution. The present invention is based on the distributed iterative control model for wind farm power optimization control, which can reduce the fatigue load of the fan, reduce the deviation of the fan voltage and maintain voltage stability, and reduce network losses.

[0046] 2. The present invention performs distributed iterative control model decomposition based on the centralized algorithm representation of the objective function and its constraint conditions to obtain a distributed iterative control model, and executes the distributed iterative control model through a distributed controller to converge and obtain the global optimal solution. Executing the distributed iterative control model through a distributed controller can make the wind farm have better robustness and stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic diagram of the basic process of the method of the embodiment of the present invention.

[0048] Figure 2 It is a typical structure configuration diagram of a radial power flow network in the embodiment of the present invention.

[0049] Figure 3 It is a control block diagram of the method of the embodiment of the present invention.

[0050] Figure 4 It is a schematic diagram of the multi-objective optimization algorithm in the embodiment of the present invention.

[0051] Figure 5 It is a simulation comparison schematic diagram of the fan fatigue load - torque bearing torque T of the method (OPT) of this embodiment and the existing proportional distribution algorithm (PD). s Optimization effect simulation comparison schematic diagram.

[0052] Figure 6 It is a simulation comparison schematic diagram of the fan fatigue load - wind thrust F of the method (OPT) of this embodiment and the existing proportional distribution algorithm (PD). t Optimization effect simulation comparison schematic diagram.

[0053] Figure 7 It is a simulation diagram of the optimization effect of the feeder head fan voltage of the wind farm of the method (OPT) of this embodiment and the existing proportional distribution algorithm (PD).

[0054] Figure 8 It is a simulation diagram of the optimization effect of the feeder end fan voltage of the wind farm of the method (OPT) of this embodiment and the existing proportional distribution algorithm (PD).

[0055] Figure 9It is a simulation diagram of the optimization effect of the wind farm network loss according to the method of this embodiment. Detailed implementation manners

[0056] As Figure 1 shown, the wind farm power optimization control method based on distributed control in this embodiment includes:

[0057] 1) Establish an incremental state - space model for fatigue load, and determine the first - part control target Obj1 of the fan active power based on the incremental state - space model for fatigue load; determine the second - part control target Obj2 of the fan active power through the measurement of the available reactive power of the fan and the voltage deviation at the access point; determine the third - part control target Obj3 of the fan active power by optimizing the radiative power flow of the wind farm;

[0058] 2) Determine the total control target Obj of the wind farm according to the first - part control target Obj1, the second - part control target Obj2, and the third - part control target Obj3 total as the objective function, and determine the constraint conditions;

[0059] 3) Determine the centralized - control model representation of the objective function and its constraint conditions;

[0060] 4) Based on the centralized - algorithm representation of the objective function and its constraint conditions, perform distributed iterative - control model decomposition to obtain a distributed iterative - control model, and execute the distributed iterative - control model through a distributed controller to converge to obtain the global optimal solution.

[0061] Figure 2 This is a typical structure configuration diagram of the radiative - power - flow network in this embodiment. Referring to Figure 2 it can be seen that multiple groups of fans are connected to the collector feeder of the wind farm, and the collector feeder is connected to the AC power grid through a transformer. Each fan is equipped with an independent fan local controller, and there is also a collector - feeder controller, which is respectively connected to each fan local controller to realize the coordination of the distributed control of each fan local controller and data transmission. The "distributed controller" in step 4) is that each fan is equipped with an independent fan local controller, but step 4) does not depend on this implementation method.

[0062] In this embodiment, the functional expression of the incremental state - space model for fatigue load established in step 1) is:

[0063]

[0064] In the above formula, ΔT S is the incremental model of the bearing torque T S , η g is the generator efficiency, J tis the equivalent inertia, J g is the generator inertia, J r is the impeller rotor inertia, J t = J r + J g , K ωT , K βT , K βF and K ωF are respectively the coefficients of the first-order Taylor expansion of the bearing torque T S and the wind thrust F t at the initial time point t0, Δβ is the increment of the pitch angle, Δω g is the increment of the generator speed, is the value of the generator output power at the time t0, is the value of the filtered rotor speed at the initial time t0, is the increment of the generator power reference value, ΔF t is the increment model of the wind thrust; K ωT , K βT are the coefficients of the first-order Taylor expansion of the bearing torque T S at the initial time point t0, K βF and K ωF are the coefficients of the first-order Taylor expansion of the wind thrust F t at the initial time point t0. The functional expression of the first part of the control target Obj1 of the active power of the wind turbine determined based on the incremental state space model regarding the fatigue load in step 1) is:

[0065]

[0066] In the above formula, N f is the number of branches of the wind farm, N w is the number of wind turbines on one branch, ΔT s.i (k) is the incremental equation of the bearing torque T S ΔF t.i (k) is the incremental equation of the wind thrust F t is the weight coefficient of the bearing torque T S is the weight coefficient of the wind thrust F t and the independent variable corresponding to the optimization target is the local variable P of the wind turbine ij , where P ij is the active power output by the jth wind turbine on the ith branch.

[0067] In this embodiment, the functional expression of the second part of the control target Obj2 of the active power of the wind turbine determined by measuring the available reactive power of the wind turbine and the voltage deviation at the connection point in step 1) is: ​​

[0068]

[0069] In the above formula, N f is the number of branches of the wind farm, N w is the number of wind turbines on one branch, Q j_i is the reactive power output by the wind turbines on one branch, is the reference value of the reactive power ratio distribution of the j-th wind turbine on the i-th branch, is the terminal voltage of the j-th wind turbine on the i-th branch, U rated is the rated voltage of the wind farm, Q q is the weight coefficient of the reactive power voltage, Q u is the weight coefficient of the voltage deviation. The independent variable corresponding to the optimization objective is the local variable Q of the wind turbine ij and the local variable where Q ij is the reactive power output by the j-th wind turbine on the i-th branch, is the terminal voltage of the j-th wind turbine on the i-th branch.

[0070] In this embodiment, the function expression of the third part of the control objective Obj3 of the active power of the wind turbine determined by optimizing the radiative power flow of the wind farm in step 1) is:

[0071]

[0072] In the above formula, N f is the number of branches of the wind farm, N w is the number of wind turbines on one branch, is the line resistance of the j-th wind turbine on the i-th branch, is the local variable of the wind turbine, representing the active power flowing into the connection point; is the local variable of the wind turbine, representing the reactive power flowing into the connection point; V s is the standard voltage, Q loss is the line loss of the wind farm. The independent variable corresponding to the optimization objective is the power flow information of the connection point of the wind turbine, including the global variables P ij and Q ij where P ij is the active power flowing into the connection point of the j-th wind turbine on the i-th branch, Q ij is the reactive power flowing into the connection point of the j-th wind turbine on the i-th branch.

[0073] In this embodiment, the function expression for determining the total control objective Obj of the wind farm in step 2) is: total is:

[0074] Obj total = Obj1 + Obj2 + Obj3,

[0075] In step 2), the function expressions for determining the constraint conditions include some or all of the constraints in the following functional expressions:

[0076]

[0077] In the above formula, P j_i is a local variable of the fan, representing the output active power; is the available active power of the fan; is the available reactive power of the fan; Q j_i is a local variable of the fan, representing the output reactive power; is a local variable of the fan, representing the active power flowing into the access point; is a local variable of the fan, representing the active power flowing out of the access point; is a local variable of the fan, representing the reactive power flowing into the access point; is a local variable of the fan, representing the reactive power flowing out of the access point; is a local variable of the fan, representing the square of the voltage at the front end of the access point; is a local variable of the fan, representing the square of the voltage at the end of the access point; is the line resistance of the j-th fan on the i-th branch; is the line reactance of the j-th fan on the i-th branch; δ is a preset error coefficient; V s is the rated voltage of the wind farm.

[0078] In this embodiment, the function expression represented by the centralized control model of the objective function and its constraint conditions determined in step 3) is:

[0079]

[0080]

[0081] And its independent variables are:

[0082]

[0083] In the above formula, is the representation of the distributed iterative control model of the objective function and its constraint conditions; x is the local variable matrix, y is the dual variable matrix, and z is the global variable matrix; ΔT s.i (k) is the increment equation of the bearing torque T S t.i t.i (k) is the increment equation of the wind thrust F t t is the bearing torque T S S is the wind thrust Ft The weight coefficient, Q q is the weight coefficient of reactive power and voltage, Q j_i is the reactive power output by the number of wind turbines on a branch is the reference value of the reactive power proportion distribution of the j-th wind turbine on the i-th branch; Q u is the weight coefficient of voltage deviation, U j_i is the terminal voltage of the j-th wind turbine on the i-th branch, U rated is the rated voltage of the wind farm; Q loss is the line loss of the wind farm is the line resistance of the j-th wind turbine on the i-th branch is a local variable of the wind turbine, representing the active power flowing into the connection point is a local variable of the wind turbine, representing the reactive power flowing into the connection point; N w is the number of wind turbines on a branch is the local variable of the active power flowing out of the connection point is the global variable of the active power flowing into the connection point is the global variable of the active power flowing out of the connection point is the local variable of the reactive power flowing out of the connection point is the global variable of the reactive power flowing into the connection point is the de-global variable of the reactive power flowing out of the connection point is the local variable of the voltage at the end of the connection point is the global variable of the voltage at the front end of the connection point is the global variable of the voltage at the end of the connection point is the dual variable of the active power flowing into the connection point is the dual variable of the active power flowing out of the connection point is the dual variable of the reactive power flowing into the connection point; is; is the dual variable of the reactive power flowing out of the connection point is the dual variable of the voltage at the front end of the connection point is the dual variable of the voltage at the end of the connection point

[0084] Figure 3 This is the control block diagram of the method in this embodiment. Among them, the constraints of the active and reactive power dispatching instructions of the wind farm come from the dispatching requirements of the wind farm. The collector feeder controller and the power flow distribution of each feeder are for the dispatching requirements. The optimization objective of the objective function is the local variable P of the wind turbine i_j to reduce fatigue loads; the local variable Q of the wind turbine i_j to be used to reduce voltage deviation and optimize the available reactive power capacity, the local variable of the wind turbine and To reduce network losses. Figure 4 It is a schematic diagram of the multi-objective optimization algorithm in the embodiment of the present invention. The collector feeder controller is used to meet the scheduling instructions and optimize the node voltage. The power flow is used to reduce the line losses. The active power of the fan is used to reduce the fatigue load. The reactive power of the fan is used to reduce the voltage deviation. The variables of all optimization objectives are formed into an objective control function through the weight coefficient, and then the minimum value is obtained to obtain the optimization solutions of each variable.

[0085] In this embodiment, the function expression of the distributed iterative control model obtained in step 4) is:

[0086]

[0087] In the above formula, k represents the number of iteration steps, is the k-step iteration value of the global variable of the active power flowing into the access point; is the k-step iteration value of the global variable of the active power flowing out of the access point; is the k-step iteration value of the local variable of the reactive power flowing out of the access point; is the k-step iteration value of the global variable of the reactive power flowing into the access point; is the k-step iteration value of the de-global variable of the reactive power flowing out of the access point; is the k-step iteration value of the local variable of the voltage at the end of the access point; is the k-step iteration value of the global variable of the voltage at the front end of the access point; is the k-step iteration value of the global variable of the voltage at the end of the access point; is the (k + 1)-step iteration value of the global variable of the active power flowing into the access point; is the (k + 1)-step iteration value of the global variable of the active power flowing out of the access point; is the (k + 1)-step iteration value of the local variable of the reactive power flowing out of the access point; is the (k + 1)-step iteration value of the global variable of the reactive power flowing into the access point; is the (k + 1)-step iteration value of the de-global variable of the reactive power flowing out of the access point; is the (k + 1)-step iteration value of the local variable of the voltage at the end of the access point; is the (k + 1)-step iteration value of the global variable of the voltage at the front end of the access point; is the (k + 1)-step iteration value of the global variable of the voltage at the end of the access point; and the function expression for the update of the dual variable is:

[0088] y [k+1] =δ(x [k+1] -z [k+1] ),

[0089] In the above formula, x [k+1]is the local variable matrix for the (k + 1)-th iteration, y [k+1] is the dual variable matrix for the (k + 1)-th iteration, z [k+1] is the global variable matrix for the (k + 1)-th iteration, and δ is the penalty coefficient.

[0090] In this embodiment, the expression of the convergence condition function for the distributed iterative control model to converge in step 4) by the distribution controller is:

[0091] x [k+1] -x [k] ≤ε1,

[0092] x [k+1] -z [k+1] ≤ε1,

[0093]

[0094] In the above formula, x [k+1] is the local variable matrix for the (k + 1)-th iteration, x [k] is the local variable matrix for the k-th iteration, z [k+1] is the global variable matrix for the (k + 1)-th iteration, is the value of the objective function for the (k + 1)-th iteration, is the value of the objective function for the k-th iteration, and ε1, ε2 are error coefficients.

[0095] In order to verify the performance of the method in this embodiment, the existing proportional distribution algorithm (PD) is used in this embodiment for simulation as a comparison with the method (OPT) in this embodiment, and the obtained results are as Figures 5 to 9 shown. Figure 5 is the schematic diagram of the simulation comparison of the optimization effect of the wind turbine fatigue load - torque bearing torque T between the method (OPT) in this embodiment and the existing proportional distribution algorithm (PD) s In the figure, Figure 6 is the schematic diagram of the simulation comparison of the optimization effect of the wind turbine fatigue load - wind thrust F between the method (OPT) in this embodiment and the existing proportional distribution algorithm (PD) t In the figure. Comparing Figure 5 and Figure 6 in the simulation result diagrams, it can be seen that compared with the proportional distribution algorithm, the control scheme proposed by the method in this embodiment for the torque bearing torque T s and the wind thrust F t have significantly reduced fluctuations, indicating that this control scheme can better reduce the fatigue load of the wind turbine. Figure 7 is the simulation diagram of the optimization effect of the feeder - end wind turbine voltage of the wind farm between the method (OPT) in this embodiment and the existing proportional distribution algorithm (PD), Figure 8This is a simulation diagram of the optimization effect of the feeder-end fan voltage of the method (OPT) in this embodiment and the existing proportional distribution algorithm (PD) for a wind farm. Comparison Figure 7 and Figure 8 From the simulation result diagrams, it can be seen that compared with the proportional distribution algorithm, the feeder-end fan voltage of the method in this embodiment is more stable, and the voltage of the end fan is smaller, indicating that this control scheme can more effectively reduce the deviation of the node voltage so as to achieve a lower end voltage and prevent the risk of over-line. Figure 9 This is a simulation diagram of the optimization effect of the network loss of the wind farm according to the method in this embodiment. From the simulation results, it can be seen that by optimizing the power flow, the loss of the line is smaller, indicating that the control effect of this control scheme in optimizing the line loss is more superior.

[0096] In summary, this embodiment discloses a wind farm power optimization control method based on a distributed algorithm, which is applicable to the active power and reactive power coordination control scheme of a wind farm with a radial power flow network. The optimization objectives include reducing the fatigue load of the fans, the transmission network loss, and minimizing the voltage deviation. The method in this embodiment first dynamically optimizes and adjusts the active power reference command of each fan according to the real-time fatigue load borne by each fan to achieve the effect of reducing the fatigue load. Secondly, it optimizes the reactive power reference command of each fan according to the available reactive power capacity to minimize the node voltage deviation at the fan connection point. Finally, it optimizes the power flow to achieve the effect of reducing the wind farm network loss. In addition, the method in this embodiment converts the adopted optimization control algorithm into a distributed control algorithm through the alternating direction multiplier algorithm, and performs wind farm power optimization control based on the distributed control algorithm, enabling the wind farm to have better robustness and stability, and being able to reduce the fatigue load of the fans, reduce the deviation of the fan voltage and maintain voltage stability, and reduce the network loss.

[0097] In addition, this embodiment also provides a wind farm power optimization control system based on distributed control, including a microprocessor and a memory connected to each other. The microprocessor is programmed or configured to execute the steps of the aforementioned wind farm power optimization control method based on distributed control.

[0098] In addition, this embodiment also provides a computer-readable storage medium, in which a computer program for being executed by a computer device to implement the aforementioned wind farm power optimization control method based on distributed control is stored.

[0099] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) that contain computer-usable program code. The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for realizing the functions specified in the flow Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks. These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device that realizes the functions specified in the flow Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks. These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for realizing the functions specified in the flow Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0100] The above are only the preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the idea of the present invention belong to the protection scope of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements should also be regarded as within the protection scope of the present invention.

Claims

1. A wind farm power optimization control method based on distributed control, characterized in that Including: 1) Establish an incremental state - space model for fatigue load, and determine the first - part control objective Obj1 of the fan active power based on the incremental state - space model for fatigue load; determine the second - part control objective Obj2 of the fan active power through the measurement of the fan available reactive power and the access - point voltage deviation; determine the third - part control objective Obj3 of the fan active power by optimizing the radiative power flow of the wind farm; 2) Determine the total control objective Obj of the wind farm based on the first - part control objective Obj1, the second - part control objective Obj2, and the third - part control objective Obj3 total as the objective function, and determine the constraint conditions; 3) Determine the centralized control - model representation of the objective function and its constraints; 4) Based on the centralized - algorithm representation of the objective function and its constraints, perform a distributed - iterative control - model decomposition to obtain a distributed - iterative control model, and execute the distributed - iterative control model through a distributed controller to converge to a global optimal solution; The functional expression of the distributed - iterative control model obtained in step 4) is: In the above formula, k represents the number of iteration steps, is the k-step iteration value of the global variable of the active power flowing into the access point; is the k-step iteration value of the global variable of the active power flowing out of the access point; is the k-step iteration value of the local variable of the reactive power flowing out of the access point; is the k-step iteration value of the global variable of the reactive power flowing into the access point; is the k-step iteration value of the de-global variable of the reactive power flowing out of the access point; is the k-step iteration value of the local variable of the voltage at the end of the access point; is the k-step iteration value of the global variable of the voltage at the front end of the access point; is the k-step iteration value of the global variable of the voltage at the end of the access point; is the (k + 1)-step iteration value of the global variable of the active power flowing into the access point; is the (k + 1)-step iteration value of the global variable of the active power flowing out of the access point; is the (k + 1)-step iteration value of the local variable of the reactive power flowing out of the access point; is the (k + 1)-step iteration value of the global variable of the reactive power flowing into the access point; is the (k + 1)-step iteration value of the de-global variable of the reactive power flowing out of the access point; is the (k + 1)-step iteration value of the local variable of the voltage at the end of the access point; is the (k + 1)-step iteration value of the global variable of the voltage at the front end of the access point; is the (k + 1)-step iteration value of the global variable of the voltage at the end of the access point; and the functional expression for updating the dual variable is: y [k+1] = δ(x [k+1] - z [k+1] ), In the above formula, x [k+1] is the local variable matrix of the (k + 1)-th iteration, y [k+1] is the dual variable matrix of the (k + 1)-th iteration, z [k+1] is the global variable matrix of the (k + 1)-th iteration, and δ is the penalty coefficient.

2. The wind farm power optimization control method based on distributed control according to claim 1, wherein The functional expression of the incremental state - space model for fatigue load established in step 1) is: In the above formula, ΔT S is the incremental model of the bearing torque T S , η g is the generator efficiency, J t is the equivalent inertia, J g is the generator inertia, J r is the impeller rotor inertia, J t = J r + J g , K ωT , K βT , K βF and K ωF are the coefficients of the first-order Taylor expansion at the initial time t0 of the bearing torque T S and the wind thrust F t , Δβ is the increment of the pitch angle, Δω g is the increment of the generator speed, is the value of the generator output power at the time t0, is the value of the filtered rotor speed at the initial time t0, is the increment of the generator power reference value, ΔF t is the incremental model of the wind thrust; The functional expression of the first part of the control target Obj1 of the active power of the wind turbine determined based on the incremental state space model regarding the fatigue load in step 1) is: In the above formula, N f is the number of branches of the wind farm, and N w is the number of wind turbines on one branch. ΔT s.i (k) is the incremental equation of the bearing torque T S . ΔF t.i (k) is the incremental equation of the wind thrust F t . Q Ts is the weight coefficient of the bearing torque T S . Q Ft is the weight coefficient of the wind thrust F t , and the independent variable corresponding to the optimization objective is the local variable P of the wind turbine ij , where P ij is the active power output by the j-th wind turbine on the i-th branch.

3. The wind farm power optimization control method based on distributed control according to claim 2, wherein, The functional expression of the second - part control objective Obj2 of the fan active power determined through the measurement of the fan available reactive power and the access - point voltage deviation in step 1) is: In the above formula, N f is the number of branches in the wind farm, and N w is the number of wind turbines on a branch. Q j_i is the reactive power output by the wind turbines on a branch. is the reference value of the reactive power ratio distribution of the j-th wind turbine on the i-th branch. is the terminal voltage of the j-th wind turbine on the i-th branch. U rated is the rated voltage of the wind farm. Q q is the weight coefficient of the reactive power voltage. Q u is the weight coefficient of the voltage deviation. The independent variables corresponding to the optimization objective are the local variables Q ij and the local variable U j i , where Q ij is the reactive power output by the j-th wind turbine on the i-th branch. is the terminal voltage of the j-th wind turbine on the i-th branch.

4. The method for optimizing the power control of a wind farm based on distributed control according to claim 3, characterized in that, The functional expression of the third - part control objective Obj3 of the fan active power determined by optimizing the radiative power flow of the wind farm in step 1) is: In the above formula, N f is the number of branches in the wind farm, and N w is the number of wind turbines on a branch. is the line resistance of the j-th wind turbine on the i-th branch. is a local variable of the wind turbine, representing the active power flowing into the connection point. is a local variable of the wind turbine, representing the reactive power flowing into the connection point. V s is the standard voltage, and Q loss is the line loss of the wind farm. The independent variable corresponding to the optimization objective is the power flow information of the connection points of the wind turbines, including the global variables P ij and Q ij , where P ij is the active power flowing into the connection point of the j-th wind turbine on the i-th branch, and Q ij is the reactive power flowing into the connection point of the j-th wind turbine on the i-th branch.

5. The method for optimizing the power control of a wind farm based on distributed control according to claim 4, wherein Determine the total control objective Obj of the wind farm in step 2) total The functional expression of which is: Obj total = Obj1 + Obj2 + Obj3, The functional expression for determining the constraints in step 2) includes some or all of the constraints in the following functional formulas: In the above formula, P j_i is a local variable of the fan, representing the output active power; is the available active power of the fan; is the available reactive power of the fan; Q j_i is a local variable of the fan, representing the output reactive power; is a local variable of the fan, representing the active power flowing into the connection point; is a local variable of the fan, representing the active power flowing out of the connection point; is a local variable of the fan, representing the reactive power flowing into the connection point; is a local variable of the fan, representing the reactive power flowing out of the connection point; is a local variable of the fan, representing the square of the voltage at the front end of the connection point; is a local variable of the fan, representing the square of the voltage at the end of the connection point; is the line resistance of the j-th fan on the i-th branch; is the line reactance of the j-th fan on the i-th branch; δ is a preset error coefficient; V s is the rated voltage of the wind farm.

6. The method for optimizing the power control of a wind farm based on distributed control according to claim 5, characterized in that The functional expression of the centralized control - model representation of the objective function and its constraints determined in step 3) is: And its independent variable is: In the above formula, is the distributed iterative control model representation of the objective function and its constraints; x is the local variable matrix, y is the dual variable matrix, and z is the global variable matrix; ΔT s.i (k) is the increment equation of the bearing torque T S ΔF t.i (k) is the increment equation of the wind thrust F t Q Ts is the weight coefficient of the bearing torque T S ; is the weight coefficient of the wind thrust F t ; Q q is the weight coefficient of the reactive voltage; Q j_i is the reactive power output by the number of wind turbines on a branch, is the reference value of the reactive power ratio distribution of the j-th wind turbine on the i-th branch; Q u is the weight coefficient of the voltage deviation, U j_i is the terminal voltage of the j-th wind turbine on the i-th branch, U rated is the rated voltage of the wind farm; Q loss is the line loss of the wind farm; is the line resistance of the j-th wind turbine on the i-th branch; is the local variable of the wind turbine, representing the active power flowing into the connection point; is the local variable of the wind turbine, representing the reactive power flowing into the connection point; N w is the number of wind turbines on a branch; is the local variable of the active power flowing out of the connection point; is the global variable of the active power flowing into the connection point; is the global variable of the active power flowing out of the connection point; is the local variable of the reactive power flowing out of the connection point; is the global variable of the reactive power flowing into the connection point; is the global variable of the reactive power flowing out of the connection point; is the local variable of the voltage at the end of the connection point; is the global variable of the voltage at the front end of the connection point; is the global variable of the voltage at the end of the connection point; is the dual variable of the active power flowing into the connection point; is the dual variable of the active power flowing out of the connection point; is the dual variable of the reactive power flowing into the connection point; is the dual variable of the reactive power flowing out of the connection point; is the dual variable of the voltage at the front end of the connection point; is the dual variable of the voltage at the end of the access point.

7. The power optimization control method for a wind farm based on distributed control according to claim 1, characterized in that The convergence - condition functional expression for the distributed - iterative control model to converge through the distributed controller in step 4) is: x [k+1] -x [k] ≤ε1, x [k+1] -z [k+1] ≤ε1, In the above formula, x [k+1] is the local variable matrix at the (k + 1)-th iteration, x [k] is the local variable matrix at the k-th iteration, z [k+1] is the global variable matrix at the (k + 1)-th iteration, is the value of the objective function at the (k + 1)-th iteration, is the value of the objective function at the k-th iteration, and ε1, ε2 are error coefficients.

8. A wind farm power optimization control system based on distributed control, comprising a microprocessor and a memory connected to each other, characterized in that, The microprocessor is programmed or configured to execute the steps of the wind - farm power optimization control method based on distributed control according to any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that, The computer - readable storage medium stores a computer program for being executed by a computer device to implement the wind - farm power optimization control method based on distributed control according to any one of claims 1 to 7.

Images