Network construction type new energy power reserve and droop coefficient optimization method, system and device and storage medium

Abstract

The invention discloses a network construction type new energy power reserve and droop coefficient optimization method, system and device and a storage medium. The method comprises the steps of collecting basic data and short-time power disturbance characteristics of a to-be-optimized power system under a certain working condition; constructing a system frequency response model considering frequency emergency control and network construction type new energy, and obtaining a time domain expression of system frequency by using Laplace inverse transformation; determining an initial value of network construction type new energy power reserve and an initial value of a droop coefficient, and calculating upper and lower boundaries of emergency control effectiveness at the moment; calculating an initial value of an emergency control effective boundary area; determining an iteration step length; iteratively calculating the emergency control upper and lower boundaries and the emergency control validity boundary area when the power reserve and droop coefficient of the network-forming type new energy is a new value, and obtaining the optimized power reserve and droop coefficient of the new energy; according to the invention, aiming at the problem of poor adaptability of frequency emergency control under short-time power disturbance, each parameter of the network-forming type new energy is optimized, so that the frequency safety level of the system is improved.

Description

Technical Field

[0001] The present invention relates to a method, system, device and storage medium for optimizing power standby and droop coefficient, and in particular to a method, system, device and storage medium for optimizing power standby and droop coefficient of a grid-connected new energy, belonging to the field of power system and automation technology. Background Art

[0002] With the continuous expansion of the scale of new energy installed capacity, the equivalent inertia level and anti-interference ability of the power system will be significantly reduced, which has triggered a series of frequency safety incidents. For example, from 2021 to 2022, three power grid accidents occurred in the Texas power system in the United States. New energy continued to enter the low voltage ride-through state, and eventually a large number of them were disconnected from the grid. When wind power was generated in the northern Hebei power grid in China, new energy stations repeatedly entered and exited the low voltage ride-through, and the power continued to fluctuate. It can be seen that in low-inertia systems, the short-term power shock caused by the low voltage ride-through of new energy caused by AC and DC faults has become a new type of transient problem that threatens the safety and stability of the system.

[0003] Compared with the permanent power disturbance caused by a single fault, the low voltage ride-through of large-scale renewable energy causes the system to bear a large capacity power shock in a short time, resulting in a sharp drop in frequency. The frequency drop speed is faster. The basic frequency regulation means such as primary frequency regulation and secondary frequency regulation of traditional synchronous units are insufficient to offset the power shortage of the system in a short time due to the response delay. It is necessary to consider the frequency control means to deal with the expected faults of the power grid - frequency emergency control, but improper setting of frequency emergency control measures may cause high-frequency or low-frequency problems in the system. In addition, the unreasonable setting of grid-type renewable energy power reserve and droop coefficient will also deteriorate the frequency emergency control boundary. These have posed new challenges to the effectiveness of frequency emergency control under short-term power disturbances. It is urgent to study new grid-type renewable energy power reserve and droop coefficient optimization methods to solve the adaptability problem of frequency emergency control under short-term power disturbances.

[0004] The existing literature mainly uses the frequency index in the frequency response model as a constraint condition for the optimization of grid-type renewable energy power reserve and droop coefficient, such as frequency extreme value, etc. However, most of the current frequency response models fail to comprehensively consider the impact of frequency emergency control and synchronous machine speed regulator limiter on system frequency index. In addition, as the frequency change speed in low-inertia systems accelerates and the disturbance form becomes more complex, the frequency modulation effect of the frequency emergency control after the disturbance has a greater impact on the frequency index. The existing optimization method fails to combine the active support capability of renewable energy with frequency emergency control, and only optimizes the renewable energy frequency modulation parameters themselves separately. Summary of the invention

[0005] Purpose of the invention: The purpose of the present invention is to provide a method, system, device and storage medium for optimizing grid-connected renewable energy power reserve and droop coefficient with a more accurate and comprehensive iterative model.

[0006] Technical solution: A grid-type new energy power reserve and droop coefficient optimization method described in the present invention includes:

[0007] Collect basic data of the power system to be optimized under certain working conditions;

[0008] Determine the short-term power disturbance characteristics caused by a large amount of renewable energy entering low voltage ride-through when N-1 fault occurs in the power system to be optimized;

[0009] Based on basic data and short-term power disturbance characteristics, a system frequency response model considering frequency emergency control and grid-connected new energy is constructed, and the time domain expression of the system frequency is obtained using the inverse Laplace transform.

[0010] Determine the initial value ΔP of grid-connected new energy power reserve max (0), initial value of droop coefficient K w (0), and calculate the upper and lower boundaries of the effectiveness of emergency control at this time;

[0011] Based on the upper and lower boundaries of emergency control effectiveness, the upper and lower boundary curves of emergency control are fitted according to the least squares method, and the initial value of the emergency control effective boundary area is calculated;

[0012] Determine the iterative step length of grid-connected renewable energy power reserve and droop coefficient according to the trapezoidal rule;

[0013] Iteratively calculate the upper and lower boundaries of emergency control and the area of ​​the emergency control effectiveness boundary when the grid-type new energy power reserve and the droop coefficient are new values ​​until the area of ​​the emergency control effectiveness boundary is maximized, and obtain the optimized new energy power reserve and droop coefficient.

[0014] Furthermore, the collecting of basic data of the power system to be optimized under a certain working condition includes:

[0015] The equivalent inertia M of the generator of the power system to be optimized, the equivalent damping D, the system equivalent turbine characteristic coefficient α, the turbine equivalent time constant T and the regulation rate R of the synchronous machine governor are determined by the weighted average method;

[0016] The determination of the short-term power disturbance characteristics caused by a large number of new energy sources entering low voltage ride-through when the system has N-1 faults includes: determining the active power shortage ΔP of the new energy sources entering low voltage ride-through fc , low voltage ride-through time t1 and low voltage ride-through recovery time t2.

[0017] Furthermore, based on the basic data and the characteristics of short-term power disturbance, a system frequency response model considering frequency emergency control and grid-forming new energy is constructed, including:

[0018] The system frequency is a piecewise function. When the speed regulator is not saturated, the dynamic s-domain expression of the system frequency is:

[0019]

[0020] in:

[0021]

[0022] When the speed regulator and grid-connected new energy reserves are both saturated, the dynamic s-domain expression of the system frequency is:

[0023]

[0024] Where K ec is the emergency control quantity proportional coefficient, 0≤K ec ≤1, τ0 is the emergency control delay, α is the system equivalent turbine characteristic coefficient, T is the turbine equivalent time constant, R is the adjustment rate of the synchronous machine speed regulator, x sg is the capacity proportion of the synchronous machine in the system, x = 1-x sg , M is the equivalent inertia of the system generator, D is the equivalent damping, T j is the virtual inertia coefficient of the networked new energy, K w is the grid-type new energy droop control coefficient, ΔP fc is the active power shortage of renewable energy when it enters low voltage ride-through, t1 is the low voltage ride-through time, t2 is the low voltage ride-through recovery time, ΔP max1 is the speed regulator limit, ΔP max2 is the maximum value of grid-building new energy reserve, Δω0 and Δω0′ are non-zero initial state values;

[0025] The time domain expression of the system frequency is obtained by using the inverse Laplace transform, including:

[0026] When the speed regulator is not saturated, the time domain expression of the system frequency is obtained by Laplace inverse transform:

[0027]

[0028] in:

[0029]

[0030] Among them, u() is a step function, and t is the current time. When the speed regulator and the grid-forming new energy are saturated at the same time, the system frequency time domain expression is obtained by inverse Laplace transform:

[0031]

[0032] in:

[0033]

[0034] Furthermore, the initial value ΔP of the grid-connected new energy power reserve is determined max (0), initial value of droop coefficient K w (0), and calculate the upper and lower limits of the effectiveness of emergency control at this time, including:

[0035] Set the initial value x of the synchronous machine ratio sg (0) = 0, initial value of the emergency control proportional coefficient K ec (0) = 0, when the proportion of synchronous machines is fixed, the emergency control amount proportional coefficient is increased with a preset fixed step size until the following iteration target is met:

[0036]

[0037] Output: At this time, the proportion of synchronous machines is the initial value x sg (0) The two emergency control quantity boundary values ​​K ec1 and K ec2 , The proportion of synchronous machines is the initial value x sg (0), ε is the accuracy value, Δω is the system frequency time domain after Laplace inverse transformation, the first iteration is over, and the proportion of synchronous machines is increased with a fixed step size. The two emergency control boundary values ​​K corresponding to this time are iteratively calculated. ec1 and K ec2 ;

[0038] When the system frequency is at its lowest point of 49 Hz, the required emergency control amount proportional coefficient is the minimum value. For the series of minimum values ​​K corresponding to different proportions of new energy, ec1 The curve is the lower boundary of emergency control; the required emergency control amount proportional coefficient is the maximum value when the system frequency is 50.5Hz. For the series of maximum values ​​K corresponding to different proportions of new energy ec2 The curve formed is the upper boundary of emergency control.

[0039] Furthermore, the upper and lower boundaries of the emergency control effectiveness are based on the least square method to fit the upper and lower boundary curves of the emergency control, and calculate the initial value of the emergency control effective boundary area, including:

[0040] The intersection of the two curves is determined by fitting the curve using the least squares method, and the step interval [x sg (i),x sg(i+1)], and use the trapezoidal rule to calculate the area of ​​each cell:

[0041]

[0042] Where A(0) is the initial value of the emergency control effectiveness boundary, A i (0) is the boundary area of ​​each small step, i is the number of iterations of the synchronous machine ratio, K ec1 and K ec2 is the emergency control value boundary value, x sg is the capacity proportion of synchronous machines in the system.

[0043] Furthermore, the iterative step length of determining the grid-connected new energy power reserve and droop coefficient according to the trapezoidal rule includes:

[0044] Determine the grid-connected new energy power reserve ΔP based on the gradient descent method max , droop coefficient K w The iteration value of

[0045]

[0046] Where η1 and η2 are learning rates, k is the number of iterations, and Δω is the system frequency domain after inverse Laplace transform, which controls the step size of parameter update in each iteration.

[0047] Further, the iterative calculation of the grid-type new energy power reserve and the emergency control upper and lower boundaries and the emergency control effectiveness boundary area when the droop coefficient is a new value, until the area of ​​the emergency control effectiveness boundary is the largest, and the optimized new energy power reserve and droop coefficient are obtained, including:

[0048] According to the new grid-type renewable energy power reserve and droop coefficient parameter values, a series of emergency control boundary data under different proportions are re-iterated, and the upper and lower boundary curves of emergency control are fitted by the least squares method. If the curves have no intersection, the emergency control under the grid-type power reserve and droop coefficient has no effective boundary, and this parameter is not the optimal parameter. If the curves have an intersection, the effective boundary area of ​​the emergency control is recalculated until A(k+1)≥A(k) is no longer satisfied, and the optimal parameter value at this time is output.

[0049] Based on the same inventive concept, the present invention also provides a grid-type new energy power reserve and droop coefficient optimization system, comprising:

[0050] A data processing module is used to collect basic data of the power system to be optimized under a certain working condition; determine the short-term power disturbance characteristics caused by a large number of new energy sources entering low voltage ride-through when N-1 fault occurs in the power system to be optimized;

[0051] The frequency model building module is used to build a system frequency response model that takes into account frequency emergency control and grid-connected new energy based on basic data and short-term power disturbance characteristics, and obtain the time domain expression of the system frequency using the inverse Laplace transform;

[0052] Parameter optimization module, used to determine the initial value ΔP of grid-connected new energy power reserve max (0), initial value of droop coefficient K w (0), and calculate the upper and lower boundaries of the emergency control effectiveness at this time; fit the upper and lower boundary curves of the emergency control according to the least squares method, and calculate the initial value of the emergency control effective boundary area; determine the iteration step of the grid-forming new energy power reserve and the droop coefficient according to the trapezoidal rule; iteratively calculate the upper and lower boundaries of the emergency control and the emergency control effectiveness boundary area when the grid-forming new energy power reserve and the droop coefficient are new values, until the area of ​​the emergency control effective boundary is maximized, and the optimized new energy power reserve and droop coefficient are obtained.

[0053] Based on the same inventive concept, the present invention also provides a computing device, comprising: one or more processors, one or more memories and one or more programs, wherein the programs are stored in the memories and configured to be executed by the processors, and when the programs are loaded into the processors, the steps of the grid-based new energy power backup and droop coefficient optimization method according to any one of the above items are implemented.

[0054] Based on the same inventive concept, the present invention also provides a storage medium, which stores a computer program, and the computer program includes program instructions. When the program instructions are executed by a processor, the processor executes the steps of the grid-based new energy power backup and droop coefficient optimization method according to any of the above items.

[0055] Beneficial effects: Compared with the prior art, the present invention aims at the problem of short-term power disturbance caused by low-voltage crossing of new energy, constructs an extended frequency response model that takes into account frequency emergency control and grid-forming new energy, so that the model is more accurate and comprehensive, and can provide a more practical iterative model for the parameters to be optimized; the present invention effectively characterizes the advantages and disadvantages of the power reserve and droop coefficient of the grid-forming new energy by comparing the effective boundary area values ​​of the frequency emergency control under different parameters; the present invention determines the optimal values ​​of the power reserve and droop coefficient of the grid-forming new energy through continuous iterative calculation, which can effectively solve the problem that the frequency emergency control is not adaptable or has poor adaptability under short-term power disturbances. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 is a flow chart of a method according to an embodiment of the present invention;

[0057] Figure 2 A schematic diagram of a method and an iterative process of an embodiment of the present invention;

[0058] Figure 3 A schematic diagram of an area formed by calculating an effective boundary of frequency emergency control according to an embodiment of the present invention. DETAILED DESCRIPTION

[0059] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0060] As attached Figure 1 As shown, the grid-connected new energy power reserve and droop coefficient optimization method of this embodiment includes:

[0061] Step 1: Collect basic data of the power system to be optimized under a certain working condition;

[0062] Step 2: determining the short-term power disturbance characteristics caused by a large amount of renewable energy entering low voltage ride-through when N-1 fault occurs in the power system to be optimized;

[0063] Step 3: Based on the basic data and short-term power disturbance characteristics, a system frequency response model considering frequency emergency control and grid-connected new energy is constructed, and the time domain expression of the system frequency is obtained using the inverse Laplace transform;

[0064] Step 4: Determine the initial value ΔP of the grid-connected new energy power reserve max (0), initial value of droop coefficient K w (0), and calculate the upper and lower boundaries of the effectiveness of emergency control at this time;

[0065] Step 5: Based on the upper and lower boundaries of emergency control effectiveness, the upper and lower boundary curves of emergency control are fitted according to the least squares method, and the initial value of the emergency control effective boundary area is calculated;

[0066] Step 6: Determine the iteration step length of the grid-connected new energy power reserve and droop coefficient according to the trapezoidal rule;

[0067] Step 7: Iteratively calculate the upper and lower boundaries of emergency control and the area of ​​the emergency control effectiveness boundary when the grid-type new energy power reserve and the droop coefficient are new values, until the area of ​​the emergency control effectiveness boundary is the largest, and the optimized new energy power reserve and droop coefficient are obtained.

[0068] Specifically, in step 1, basic data of the power system to be optimized under a certain working condition is collected to provide necessary model initial values ​​for the system frequency response model in step 3;

[0069] The system generator equivalent inertia M, equivalent damping D, system equivalent turbine characteristic coefficient α, turbine equivalent time constant T and synchronous machine speed regulator regulation rate R are obtained by weighted average method.

[0070] Step 2: determine the short-term power disturbance characteristics caused by a large number of renewable energy sources entering low voltage ride-through when N-1 fault occurs in the power system to be optimized;

[0071] Determine the active power deficit ΔP of renewable energy when it enters low voltage ride-through fc , low voltage ride-through time t1, low voltage ride-through recovery time t2.

[0072] Step 3: construct a system frequency response model that takes into account frequency emergency control and grid-connected new energy, and use the inverse Laplace transform to obtain the time domain expression of the system frequency;

[0073] The system frequency is a multivariable piecewise function. When the speed regulator is not saturated, the frequency dynamic s-domain expression can be obtained according to the system transfer function:

[0074]

[0075] The frequency-time domain expression obtained by inverse Laplace transform is:

[0076]

[0077] in:

[0078]

[0079] When the speed regulator and grid-connected new energy reserve are both saturated, the system frequency dynamic s-domain expression is:

[0080] Similarly, the time domain expression of the system frequency is obtained according to the inverse Laplace transform:

[0081]

[0082] in:

[0083]

[0084] Where K ec is the emergency control proportional coefficient, the value is between 0 and 1, τ0 is the emergency control delay, α is the equivalent turbine characteristic coefficient, T is the turbine equivalent time constant, R is the adjustment rate of the synchronous machine speed regulator, x sg is the capacity proportion of the synchronous machine in the system, x = 1-x sg , M is the equivalent inertia of the system generator, D is the equivalent damping of the system, T j is the virtual inertia coefficient of the networked new energy, K w is the grid-type new energy droop control coefficient, ΔP max1 is the speed regulator limit, ΔP max2To construct the maximum value of the grid-type new energy reserve, Δω0 and Δω0′ are non-zero initial state values, s is the complex frequency domain, the signal is represented by the complex frequency domain of the system block diagram, and then it is converted into the time domain form according to the Laplace transform.

[0085] Step 4: Determine the initial value ΔP of the grid-connected new energy power reserve max (0), initial value of droop coefficient K w (0), and calculate the upper and lower boundaries of the effectiveness of emergency control at this time;

[0086] When the synchronous machine ratio is the initial value x sg (0), the emergency control amount proportional coefficient is increased with a fixed step size of 0.001 until the following iteration target is met:

[0087]

[0088] Output: At this time, the proportion of synchronous machines is the initial value x sg (0) The two emergency control quantity boundary values ​​K ec1 , K ec2 , The proportion of synchronous machines is the initial value x sg (0) corresponds to the effective action interval of emergency control, ε is the precision value, and the first iteration ends. Continue to increase the proportion of synchronous machines with a fixed step size of 0.001, and recalculate the two corresponding emergency control boundary values ​​K at this time ec1 , K ec2 .

[0089] Step 5: Fit the upper and lower boundary curves of emergency control using the least squares method, and calculate the initial value of the area of ​​the effective boundary of emergency control, such as Figure 3 As shown;

[0090] Each iteration value of the synchronous machine ratio in step 4 will correspond to two emergency control quantity boundary values ​​K ec1 , K ec2 , the proportion of synchronous machines is x sg With K ec1 Each iteration data of is fitted by the least square method to obtain the emergency control lower boundary curve; similarly, the proportion of synchronous machines x sg With K ec2 The data of each iteration is fitted by the least squares method to obtain the emergency control upper boundary curve.

[0091] Method for calculating the initial value of the area consisting of the emergency control effective boundary:

[0092] The intersection of the two upper and lower boundary curves is determined based on the least squares fitting curve, and the step interval [x sg (i),x sg(i+1)], and use the trapezoidal rule to calculate the area of ​​each cell:

[0093]

[0094] Where A(0) is the initial value of the emergency control effectiveness boundary, A i (0) is the boundary area of ​​each small step, and i is the number of iterations.

[0095] Step 6, determining the iterative step length of the grid-type new energy power reserve and droop coefficient according to the trapezoidal rule;

[0096] The iterative values ​​of the grid-connected new energy power reserve and droop coefficient are determined according to the gradient descent method.

[0097]

[0098] Among them, η1 and η2 are learning rates, which control the step size of parameter update in each iteration.

[0099] Step seven, iteratively calculate the upper and lower boundaries of emergency control and the effective boundary area when the grid-type new energy power reserve and the droop coefficient are new values, until the area of ​​the emergency control effective boundary is maximized, and the optimized new energy power reserve and droop coefficient are obtained.

[0100] According to the new parameter value in step 6, repeat steps 4 and 5: that is, iteratively determine the upper and lower boundaries of emergency control, and fit the upper and lower boundary curves of emergency control by the least squares method. If the curve has no intersection, it means that the emergency control under the grid-type power reserve and droop coefficient has no effective boundary, and this parameter is not the optimal parameter. If the curve has an intersection, recalculate the effective boundary area of ​​emergency control until A(k+1)≥A(k) is not satisfied and then output the optimal parameter value at this time. The method and iterative process of this embodiment are as follows Figure 2 shown.

[0101] Based on the same inventive concept, this embodiment also provides a grid-type new energy power reserve and droop coefficient optimization system, including:

[0102] A data processing module is used to collect basic data of the power system to be optimized under a certain working condition; determine the short-term power disturbance characteristics caused by a large amount of renewable energy entering low voltage ride-through when N-1 fault occurs in the system;

[0103] The frequency model building module is used to build a system frequency response model that takes into account frequency emergency control and grid-connected new energy based on basic data and short-term power disturbance characteristics, and obtain the time domain expression of the system frequency using the inverse Laplace transform;

[0104] Parameter optimization module, used to determine the initial value ΔP of grid-connected new energy power reserve max(0), initial value of droop coefficient K w (0), and calculate the upper and lower boundaries of the emergency control effectiveness at this time; fit the upper and lower boundary curves of the emergency control according to the least squares method, and calculate the initial value of the emergency control effective boundary area; determine the iteration step of the grid-forming new energy power reserve and the droop coefficient according to the trapezoidal rule; iteratively calculate the upper and lower boundaries of the emergency control and the emergency control effectiveness boundary area when the grid-forming new energy power reserve and the droop coefficient are new values, until the area of ​​the emergency control effective boundary is maximized, and the optimized new energy power reserve and droop coefficient are obtained.

[0105] In the data processing module, the system generator equivalent inertia M, equivalent damping D, system equivalent turbine characteristic coefficient α, turbine equivalent time constant T and synchronous machine speed regulator regulation rate R are determined by weighted average method.

[0106] In the data processing module, the active power deficit ΔP of the new energy source entering low voltage ride-through is determined fc , low voltage ride-through time t1, low voltage ride-through recovery time t2.

[0107] In the frequency model building module, the system frequency is a multi-variable piecewise function. When the speed regulator is not saturated, the frequency dynamic s-domain expression can be obtained according to the system transfer function:

[0108]

[0109] The frequency-time domain expression obtained by inverse Laplace transform is:

[0110]

[0111] in:

[0112]

[0113] When the speed regulator and grid-connected new energy reserve are both saturated, the system frequency dynamic s-domain expression is:

[0114] Similarly, the time domain expression of the system frequency is obtained according to the inverse Laplace transform:

[0115]

[0116] in:

[0117]

[0118] Where K ecis the emergency control proportional coefficient, the value is between 0 and 1, τ0 is the emergency control delay, α is the turbine characteristic coefficient, T is the turbine equivalent time constant, R is the adjustment rate of the synchronous machine speed regulator, x sg is the capacity proportion of the synchronous machine in the system, x = 1-x sg , M is the system equivalent inertia time constant, D is the system equivalent damping, T j is the virtual inertia coefficient of the networked new energy, K w is the grid-type new energy droop control coefficient, ΔP max1 is the speed regulator limit, ΔP max2 is the maximum value of the grid-building new energy reserve, and Δω0 and Δω0′ are non-zero initial state values.

[0119] In the parameter optimization module, according to the initial value x of the synchronous machine ratio sg (0) Initial value of the emergency control quantity proportional coefficient K ec (0), when the proportion of synchronous machines is fixed, the emergency control amount proportional coefficient is increased with a fixed step size of 0.001 until the following iteration target is met:

[0120]

[0121] Output: At this time, the proportion of synchronous machines is the initial value x sg (0) The two emergency control quantity boundary values ​​K ec1 , K ec2 , The proportion of synchronous machines is the initial value x sg (0) corresponds to the effective action interval of emergency control, ε is the precision value, and the first iteration ends. Continue to increase the proportion of synchronous machines with a fixed step size of 0.001, and recalculate the two corresponding emergency control boundary values ​​K at this time ec1 , K ec2 .

[0122] In the parameter optimization module, the required emergency control amount proportional coefficient is the minimum value when the system frequency is at the lowest point of 49Hz. For the above series of minimum values ​​K corresponding to different proportions of new energy ec1 The curve is the lower boundary of emergency control; the required emergency control amount proportional coefficient is the maximum value when the system frequency is 50.5Hz. For the above series of maximum values ​​K corresponding to different proportions of new energy ec2 The curve formed is the upper boundary of emergency control.

[0123] In the parameter optimization module, the intersection of the two curves is determined based on the curve fitted by the least squares method, and the step interval [x sg (i),x sg (i+1)], and use the trapezoidal rule to calculate the area of ​​each cell:

[0124]

[0125] Where A(0) is the initial value of the emergency control effectiveness boundary, A i (0) is the boundary area of ​​each small step, and i is the number of iterations of the synchronous machine ratio.

[0126] In the parameter optimization module, the iterative values ​​of the grid-type renewable energy power reserve and droop coefficient are determined according to the gradient descent method.

[0127]

[0128] Among them, η1 and η2 are learning rates, which control the step size of parameter update in each iteration.

[0129] In the parameter optimization module, according to the new grid-type renewable energy power reserve and droop coefficient parameter values, a series of emergency control boundary data under different proportions are re-iterated, and the upper and lower boundary curves of emergency control are fitted by the least squares method. If the curves have no intersection, it means that the emergency control under the grid-type power reserve and droop coefficient has no effective boundary, and this parameter is not the optimal parameter. If the curves have an intersection, the effective boundary area of ​​the emergency control is recalculated until A(k+1)≥A(k) is no longer satisfied, and the optimal parameter value at this time is output.

[0130] Based on the same inventive concept, this embodiment also provides a computing device, including: one or more processors, one or more memories and one or more programs, wherein the programs are stored in the memories and configured to be executed by the processors, and when the programs are loaded into the processors, the steps of the grid-based new energy power backup and droop coefficient optimization method according to any one of the above items are implemented.

[0131] Based on the same inventive concept, this embodiment also provides a storage medium, which stores a computer program, and the computer program includes program instructions. When the program instructions are executed by a processor, the processor executes the steps of the grid-based new energy power backup and droop coefficient optimization method according to any of the above items.

[0132] The storage media may include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage, flash memory, or any other medium that can be used to store the desired program code in the form of instructions or data structures and that can be accessed by a computer.

[0133] The processor is used to execute the computer program stored in the memory to implement each step of the method involved in the above embodiment.

Claims

1. A method for optimizing grid-connected renewable energy power reserve and droop coefficient, characterized in that: include: Collect basic data of the power system to be optimized under certain working conditions; Determine the short-term power disturbance characteristics caused by a large amount of renewable energy entering low voltage ride-through when N-1 fault occurs in the power system to be optimized; Based on basic data and short-term power disturbance characteristics, a system frequency response model considering frequency emergency control and grid-connected new energy is constructed, and the time domain expression of the system frequency is obtained using the inverse Laplace transform. Determine the initial value ΔP of grid-connected new energy power reserve max (0), initial value of droop coefficient K w (0), and calculate the upper and lower boundaries of the effectiveness of emergency control at this time; Based on the upper and lower boundaries of emergency control effectiveness, the upper and lower boundary curves of emergency control are fitted according to the least squares method, and the initial value of the emergency control effective boundary area is calculated; Determine the iterative step length of grid-connected renewable energy power reserve and droop coefficient according to the trapezoidal rule; Iteratively calculate the upper and lower boundaries of emergency control and the area of ​​the emergency control effectiveness boundary when the grid-type new energy power reserve and the droop coefficient are new values ​​until the area of ​​the emergency control effectiveness boundary is maximized, and obtain the optimized new energy power reserve and droop coefficient.

2. The grid-connected new energy power reserve and droop coefficient optimization method according to claim 1 is characterized in that: The basic data of the power system to be optimized under a certain working condition is collected, including: The equivalent inertia M of the generator of the power system to be optimized, the equivalent damping D, the system equivalent turbine characteristic coefficient α, the turbine equivalent time constant T and the regulation rate R of the synchronous machine governor are determined by the weighted average method; The determination of the short-term power disturbance characteristics caused by a large number of renewable energy sources entering low voltage ride-through when N-1 faults occur in the power system to be optimized includes: determining the active power shortage ΔP of renewable energy sources entering low voltage ride-through fc , low voltage ride-through time t1 and low voltage ride-through recovery time t2.

3. The grid-connected new energy power reserve and droop coefficient optimization method according to claim 1 is characterized in that: The system frequency response model considering frequency emergency control and grid-building new energy is constructed based on basic data and short-term power disturbance characteristics, including: The system frequency is a piecewise function. When the speed regulator is not saturated, the dynamic s-domain expression of the system frequency is: in: When the speed regulator and grid-connected new energy reserve are both saturated, the dynamic s-domain expression of the system frequency is: Where K ec is the emergency control quantity proportional coefficient, 0≤K ec ≤1, τ0 is the emergency control delay, α is the system equivalent turbine characteristic coefficient, T is the turbine equivalent time constant, R is the adjustment rate of the synchronous machine speed regulator, x sg is the capacity proportion of the synchronous machine in the system, x = 1-x sg , M is the equivalent inertia of the system generator, D is the equivalent damping, T j is the virtual inertia coefficient of the networked new energy, K w is the grid-type new energy droop control coefficient, ΔP fc is the active power shortage of renewable energy when it enters low voltage ride-through, t1 is the low voltage ride-through time, t2 is the low voltage ride-through recovery time, ΔP max1 is the speed regulator limit, ΔP max2 is the maximum value of grid-building new energy reserve, Δω0 and ω0′ are non-zero initial state values; The time domain expression of the system frequency is obtained by using the inverse Laplace transform, including: When the speed regulator is not saturated, the time domain expression of the system frequency is obtained by Laplace inverse transform: in: Among them, u() is a step function, and t is the current time. When the speed regulator and the grid-forming new energy are saturated at the same time, the system frequency time domain expression is obtained by inverse Laplace transform: in:

4. The grid-connected new energy power reserve and droop coefficient optimization method according to claim 1 is characterized in that: The initial value ΔP of the grid-connected new energy power reserve is determined max (0), initial value of droop coefficient K w (0), and calculate the upper and lower limits of the effectiveness of emergency control at this time, including: Set the initial value x of the synchronous machine ratio sg (0) = 0, initial value of the emergency control proportional coefficient K ec (0) = 0, when the proportion of synchronous machines is fixed, the emergency control amount proportional coefficient is increased with a preset fixed step size until the following iteration target is met: Output: At this time, the proportion of synchronous machines is the initial value x sg (0) The two emergency control quantity boundary values ​​K ec1 and K ec2 , The proportion of synchronous machines is the initial value x sg (0), ε is the accuracy value, Δω is the system frequency time domain after Laplace inverse transformation, the first iteration is over, and the proportion of synchronous machines is increased with a fixed step size. The two emergency control boundary values ​​K corresponding to this time are iteratively calculated. ec1 and K ec2 ; When the system frequency is at its lowest point of 49 Hz, the required emergency control amount proportional coefficient is the minimum value. For the series of minimum values ​​K corresponding to different proportions of new energy, ec1 The curve is the lower boundary of emergency control; the required emergency control amount proportional coefficient is the maximum value when the system frequency is 50.5Hz. For the series of maximum values ​​K corresponding to different proportions of new energy ec2 The curve formed is the upper boundary of emergency control.

5. The grid-connected new energy power reserve and droop coefficient optimization method according to claim 1 is characterized in that: The upper and lower boundaries of the emergency control effectiveness are fitted according to the least square method to obtain the upper and lower boundary curves of the emergency control, and the initial value of the emergency control effective boundary area is calculated, including: The intersection of the two curves is determined by fitting the curve using the least squares method, and the step interval [x sg (i),x sg (i+1)], and use the trapezoidal rule to calculate the area of ​​each cell: Where A(0) is the initial value of the emergency control effectiveness boundary, A i (0) is the boundary area of ​​each small step, i is the number of iterations of the synchronous machine ratio, K ec1 and K ec2 is the emergency control value boundary value, x sg is the capacity proportion of synchronous machines in the system.

6. The grid-connected new energy power reserve and droop coefficient optimization method according to claim 1 is characterized in that: The iterative step length of determining the grid-connected new energy power reserve and droop coefficient according to the trapezoidal rule includes: Determine the grid-connected new energy power reserve ΔP based on the gradient descent method max , droop coefficient K w The iteration value of Where η1 and η2 are learning rates, k is the number of iterations, and Δω is the system frequency domain after inverse Laplace transform, which controls the step size of parameter update in each iteration.

7. The grid-connected new energy power reserve and droop coefficient optimization method according to claim 1 is characterized in that: The iterative calculation of the grid-type new energy power reserve and the emergency control upper and lower boundaries and the emergency control effectiveness boundary area when the droop coefficient is a new value, until the area of ​​the emergency control effectiveness boundary is the largest, and the optimized new energy power reserve and droop coefficient are obtained, including: According to the new grid-type renewable energy power reserve and droop coefficient parameter values, a series of emergency control boundary data under different proportions are re-iterated, and the upper and lower boundary curves of emergency control are fitted by the least squares method. If the curves have no intersection, the emergency control under the grid-type power reserve and droop coefficient has no effective boundary, and this parameter is not the optimal parameter. If the curves have an intersection, the effective boundary area of ​​the emergency control is recalculated until A(k+1)≥A(k) is no longer satisfied, and the optimal parameter value at this time is output.

8. A grid-type new energy power reserve and droop coefficient optimization system, characterized in that: include: Data processing module, used to collect basic data of the power system to be optimized under certain working conditions; Determine the short-term power disturbance characteristics caused by a large amount of renewable energy entering low voltage ride-through when N-1 fault occurs in the power system to be optimized; The frequency model building module is used to build a system frequency response model that takes into account frequency emergency control and grid-connected new energy based on basic data and short-term power disturbance characteristics, and obtain the time domain expression of the system frequency using the inverse Laplace transform; Parameter optimization module, used to determine the initial value ΔP of grid-connected new energy power reserve max (0), initial value of droop coefficient K w (0), and calculate the upper and lower boundaries of the emergency control effectiveness at this time; fit the upper and lower boundary curves of the emergency control according to the least squares method, and calculate the initial value of the emergency control effective boundary area; determine the iteration step of the grid-forming new energy power reserve and the droop coefficient according to the trapezoidal rule; iteratively calculate the upper and lower boundaries of the emergency control and the emergency control effectiveness boundary area when the grid-forming new energy power reserve and the droop coefficient are new values, until the area of ​​the emergency control effective boundary is maximized, and the optimized new energy power reserve and droop coefficient are obtained.

9. A computing device, characterized in that include: One or more processors, one or more memories and one or more programs, wherein the programs are stored in the memories and configured to be executed by the processors, and when the programs are loaded into the processors, the steps of the grid-connected new energy power standby and droop coefficient optimization method according to any one of claims 1 to 7 are implemented.

10. A storage medium, characterized in that: The storage medium stores a computer program, which includes program instructions. When the program instructions are executed by a processor, the processor executes the steps of the grid-based new energy power backup and droop coefficient optimization method according to any one of claims 1 to 7.

Images