A two-level optimization method for short-circuit current and transient stability of virtual synchronous generator systems

Abstract

The present invention belongs to the field of data processing and provides a two-level optimization method for short-circuit current and transient stability of a virtual synchronous generator system. The method includes: determining parameters to be optimized related to short-circuit current in the virtual synchronous generator system; establishing a two-level interactive optimization model for short-circuit current and transient stability with minimizing the short-circuit current of the virtual synchronous generator system as the optimization goal and meeting transient stability standards as a constraint; determining preliminary optimization results corresponding to the parameters to be optimized when the short-circuit current is small based on the two-level interactive optimization model, and correcting the preliminary optimization results using transient stability to obtain a two-level optimization solution for the parameters to be optimized. The solution provided by the present invention can take into account both the short-circuit current and transient stability of the virtual synchronous generator system, achieve efficient optimization of multi-objective parameters, and improve the dynamic stability and operational safety of the virtual synchronous generator system.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a double-layer optimization method for short-circuit current and transient stability of a virtual synchronous generator system. Background Art

[0002] A virtual synchronous generator (VSG) is a power generation unit that simulates the operating characteristics of a traditional synchronous generator based on modern power electronics and advanced control strategies. By implementing the inertia and damping characteristics of a traditional synchronous generator in power electronics, the VSG improves the dynamic stability of the power grid.

[0003] However, under fault conditions, virtual synchronous generators (VSGs) suffer from power angle instability, similar to traditional synchronous generators, and are unable to withstand excessive fault currents. Due to their weak overcurrent capability, a short-circuit current limiting mechanism is typically required. This modification alters the transient characteristics of the VSGs, increasing the complexity of transient analysis.

[0004] Traditional virtual synchronous generator systems have the following deficiencies in short-circuit current and transient stability control optimization:

[0005] First, the ability to coordinately optimize short-circuit current and transient stability is poor. Current research on grid-connected virtual synchronous generator grid-type converter systems mostly focuses on the optimization of single short-circuit current control or transient stability control, resulting in poor robustness and dynamic stability of virtual synchronous generator systems in different fault scenarios.

[0006] Second, there is a lack of efficient strategies for multi-objective parameter optimization. The control parameters of the virtual synchronous generator system (such as inertia parameters and damping coefficient) have a significant impact on the grid's short-circuit current response. However, there is currently a lack of methods that can efficiently and accurately optimize these control parameters. Traditional optimization schemes struggle to provide precise parameter adjustment mechanisms, especially in complex scenarios where multiple parameters are coupled. This results in the virtual synchronous generator system being unable to effectively suppress the rise in short-circuit current during sudden faults, resulting in insufficient system operational safety. Summary of the Invention

[0007] The present invention provides a double-layer optimization method for short-circuit current and transient stability of a virtual synchronous generator system, which is used to solve the defect that traditional optimization schemes cannot ensure the dynamic stability and operational safety of the virtual synchronous generator system.

[0008] The present invention provides a two-layer optimization method for short-circuit current and transient stability of a virtual synchronous generator system, comprising:

[0009] Determine the parameters to be optimized related to short-circuit current in the virtual synchronous generator system;

[0010] Taking minimizing the short-circuit current of the virtual synchronous generator system as the optimization objective and meeting the transient stability standard as the constraint condition, a two-level interactive optimization model of short-circuit current and transient stability is established.

[0011] According to the two-layer interactive optimization model, the preliminary optimization results corresponding to the parameters to be optimized are determined when the short-circuit current is small, and the preliminary optimization results are corrected using transient stability to obtain the two-layer optimization solution results of the parameters to be optimized.

[0012] According to the two-level optimization method for short-circuit current and transient stability of a virtual synchronous generator system provided by the present invention, determining the parameters to be optimized related to the short-circuit current in the virtual synchronous generator system includes:

[0013] Obtain the transient current expression and steady-state current expression of the short-circuit current in the virtual synchronous generator system;

[0014] Extracting multiple parameter components from the transient current expression and the steady-state current expression;

[0015] A key component is determined from the multiple parameter components, and a parameter to be optimized related to the short-circuit current is determined based on the influencing factors corresponding to the amplitude and attenuation coefficient of each key component.

[0016] According to the two-level optimization method for short-circuit current and transient stability of a virtual synchronous generator system provided by the present invention, the key components include: an inherent non-power frequency period attenuation component, an inherent power frequency period attenuation component, and a free power frequency period attenuation component;

[0017] The parameters to be optimized include: reactive integral coefficient, virtual resistance, virtual inductance and power angle parameter.

[0018] According to the two-level optimization method for short-circuit current and transient stability of a virtual synchronous generator system provided by the present invention, minimizing the short-circuit current of the virtual synchronous generator system is taken as the optimization objective, and meeting the transient stability standard is taken as the constraint condition. A two-level interactive optimization model for short-circuit current and transient stability is established, including:

[0019] Based on the genetic algorithm, the main layer optimization model is established with the minimum short-circuit current of the virtual synchronous generator system as the optimization goal;

[0020] An auxiliary layer optimization model is established by using an exponential penalty factor and taking transient stability compliance as a constraint condition.

[0021] The auxiliary layer optimization model is integrated with the main layer optimization model to obtain a two-layer interactive optimization model of short-circuit current and transient stability.

[0022] According to the two-level optimization method for short-circuit current and transient stability of a virtual synchronous generator system provided by the present invention, a main-level optimization model is established based on a genetic algorithm with minimizing the short-circuit current of the virtual synchronous generator system as the optimization goal, including:

[0023] Determine the initial values ​​of the system physical parameters of the virtual synchronous generator system, the initial values ​​of some parameters to be optimized, and the initial parameters related to the genetic algorithm respectively;

[0024] Based on the initial values ​​of the physical parameters of the system, the initial values ​​of some parameters to be optimized and the initial parameters related to the genetic algorithm, population individuals are generated within a set range to establish an initial population;

[0025] Taking the minimization of the short-circuit current of the virtual synchronous generator system as the optimization objective, a fitness function related to the short-circuit current is established.

[0026] Based on the initial values ​​of the system physical parameters, the initial values ​​of some parameters to be optimized, the initial parameters related to the genetic algorithm and the fitness function, the cyclic iteration strategy of the initial population is determined and the main layer optimization model is established.

[0027] According to the two-level optimization method for short-circuit current and transient stability of a virtual synchronous generator system provided by the present invention, determining preliminary optimization results corresponding to the parameters to be optimized when the short-circuit current is small includes:

[0028] Determining the power angle parameters of each population individual based on the initial values ​​of the system physical parameters and the initial values ​​of some parameters to be optimized;

[0029] Determine whether the power angle parameter of each population individual is within the set power angle range, and obtain the power angle judgment result;

[0030] Determining the short-circuit current corresponding to the population individual for which the power angle judgment result is yes, and calculating the fitness value based on the short-circuit current and the fitness function;

[0031] According to the fitness value, a preliminary optimization result corresponding to the parameter to be optimized is obtained when the short-circuit current is small.

[0032] According to the two-level optimization method for short-circuit current and transient stability of a virtual synchronous generator system provided by the present invention, the initial values ​​of the system physical parameters include: virtual internal potential, grid frequency, active power, reactive power and voltage;

[0033] The initial parameters related to the genetic algorithm include: the maximum number of iterations, the number of candidate solutions per generation, and the initial mutation probability.

[0034] According to the double-layer optimization method for short-circuit current and transient stability of a virtual synchronous generator system provided by the present invention, the short-circuit current includes: transient current and steady-state current;

[0035] The fitness function is:

[0036]

[0037] in, represents fitness, and are coefficient values, and , i zt represents the transient current, i wt Represents the steady-state current.

[0038] According to the two-layer optimization method for short-circuit current and transient stability of a virtual synchronous generator system provided by the present invention, an auxiliary layer optimization model is established by using an exponential penalty factor and taking transient stability compliance as a constraint condition, including:

[0039] During the initial population cyclic iteration process in the main layer optimization model, determining whether the transient stability of each population individual meets the standard, and obtaining a stability determination result;

[0040] Determine the target population individuals for which the stability judgment result is negative, and determine the exponential penalty factor corresponding to the target population individuals under the current number of loop iterations;

[0041] The penalty strategy of the target population under different numbers of loop iterations is determined based on an exponential penalty factor, and an auxiliary layer optimization model is established.

[0042] According to the two-level optimization method for short-circuit current and transient stability of a virtual synchronous generator system provided by the present invention, the exponential penalty factor is:

[0043]

[0044] in, represents the exponential penalty factor corresponding to the target population individual under the i-th loop iteration, Represents the penalty coefficient constant.

[0045] The present invention provides a two-level optimization method for short-circuit current and transient stability of a virtual synchronous generator system. By determining the parameters to be optimized related to the short-circuit current in the virtual synchronous generator system, minimizing the short-circuit current of the virtual synchronous generator system is used as the optimization objective, and meeting the transient stability standard is used as a constraint condition, a two-level interactive optimization model for short-circuit current and transient stability is established. Based on the two-level interactive optimization model, preliminary optimization results corresponding to the parameters to be optimized are determined when the short-circuit current is small. The preliminary optimization results are corrected using transient stability to obtain a two-level optimization solution for the parameters to be optimized. Due to the establishment of the two-level interactive optimization model, the optimization process can take into account both the short-circuit current and transient stability of the virtual synchronous generator system, achieving efficient optimization of multi-objective parameters, and thereby improving the dynamic stability and operational safety of the virtual synchronous generator system. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0047] Figure 1 1 is a flow chart of a two-layer optimization method for short-circuit current and transient stability of a virtual synchronous generator system provided by an embodiment of the present invention;

[0048] Figure 2 It is a flow chart of the process of realizing the double-layer optimization of short-circuit current and transient stability by the double-layer interactive optimization model;

[0049] Figure 3 This is a schematic diagram of the output current of the virtual synchronous generator system before a slight voltage sag;

[0050] Figure 4 This is a schematic diagram of the output current of the virtual synchronous generator system after a slight voltage drop.

[0051] Figure 5 This is a schematic diagram of frequency changes before and after a slight voltage drop;

[0052] Figure 6 This is a data diagram of the optimization iterative process of the improved genetic algorithm during the mild voltage sag stage;

[0053] Figure 7 This is a schematic diagram of the output current of the virtual synchronous generator before the moderate voltage drop;

[0054] Figure 8 It is a schematic diagram of the output current of the virtual synchronous generator after a moderate voltage drop;

[0055] Figure 9 This is a comparison diagram of frequency changes before and after a moderate voltage drop;

[0056] Figure 10 This is a data diagram of the optimization iteration process of the improved genetic algorithm during the moderate voltage sag stage;

[0057] Figure 11 This is a schematic diagram of the output current of the virtual synchronous generator before the severe voltage sag;

[0058] Figure 12 This is a schematic diagram of the output current of a virtual synchronous generator after a severe voltage drop;

[0059] Figure 13 This is a comparison diagram of frequency changes before and after a severe voltage drop;

[0060] Figure 14 This is a data diagram of the optimization iteration process of the improved genetic algorithm during the severe voltage sag stage. DETAILED DESCRIPTION

[0061] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0062] The following combination Figures 1 to 14 The detailed scheme of the two-layer optimization method for short-circuit current and transient stability of a virtual synchronous generator system provided by an embodiment of the present invention is described.

[0063] Figure 1 It is a flow chart of a two-layer optimization method for short-circuit current and transient stability of a virtual synchronous generator system provided by an embodiment of the present invention.

[0064] like Figure 1 As shown, the embodiment of the present invention provides a two-level optimization method for short-circuit current and transient stability of a virtual synchronous generator system. The execution subject can be a computer or server with data processing and data transmission and reception capabilities. The above method mainly includes the following steps:

[0065] Step 110: Determine parameters to be optimized related to short-circuit current in the virtual synchronous generator system.

[0066] In this embodiment, the parameters to be optimized refer to parameters that can affect the short-circuit current, and specifically may be multiple parameters that affect the amplitude and attenuation coefficient of the transient component of the short-circuit current.

[0067] Step 120: Taking minimizing the short-circuit current of the virtual synchronous generator system as the optimization objective and meeting the transient stability standard as the constraint condition, a two-layer interactive optimization model of short-circuit current and transient stability is established.

[0068] In this embodiment, the optimization step takes minimizing the short-circuit current of the virtual synchronous generator system as the optimization objective, and takes meeting the transient stability standard as a constraint condition. The resulting two-level interactive optimization model can achieve two-level optimization of short-circuit current and transient stability.

[0069] Step 130: Based on the two-layer interactive optimization model, determine the preliminary optimization results corresponding to the parameters to be optimized when the short-circuit current is small, and correct the preliminary optimization results using transient stability to obtain the two-layer optimization solution results of the parameters to be optimized.

[0070] The solution provided in this embodiment can achieve two-level optimization of the short-circuit current and transient stability of the virtual synchronous generator system through a two-level interactive optimization model. In particular, under fault conditions, it can simultaneously achieve iterative optimization of the short-circuit current and enhancement of transient stability, significantly improving the dynamic response capability of the virtual synchronous generator system. When a power grid fault occurs, the system can recover quickly and stably, ensuring the overall stability and operational safety of the power grid.

[0071] In one embodiment, determining the parameters to be optimized related to the short-circuit current in the virtual synchronous generator system specifically includes:

[0072] The first step is to obtain the transient current expression and steady-state current expression of the short-circuit current in the virtual synchronous generator system.

[0073] It can be understood that the transient current expression of the short-circuit current specifically includes: an expression of the transient current of phase A changing with time, an expression of the transient current of phase B changing with time, and an expression of the transient current of phase C changing with time. Specifically, the expression of the transient current of phase A changing with time is as follows:

[0074] (1)

[0075] In this embodiment, the expression of the transient current of phase B changing with time is as follows:

[0076] (2)

[0077] In this embodiment, the expression of the transient current of phase C changing with time is as follows:

[0078] (3)

[0079] in, It represents the transient current value of phase A at time t, Indicates the transient current value of phase B at time t, represents the transient current value of phase C at time t, and Both represent the correlation coefficient of the system steady-state characteristics, represents the virtual impedance, represents the phase angle associated with the virtual impedance, Indicates the system rated angular frequency, represents the power angle parameter, and are coefficients related to transient processes. represents the angular frequency associated with the transient process, represents the time constant, and d1 represents the coefficient related to transient attenuation.

[0080] Furthermore, the above coefficients 、 、 and The expressions are as follows:

[0081] (4)

[0082] (5)

[0083] (6)

[0084] (7)

[0085] in, represents the virtual internal potential, d1, d2 and d3 are coefficients related to transient attenuation, U and U0 represent the fault voltage and initial voltage respectively, Represents the time constant.

[0086] In this embodiment, the steady-state current expression is as follows:

[0087] (8)

[0088] in, represents the steady-state current value at time t, and Both represent the correlation coefficient of the system steady-state characteristics, represents the virtual impedance, represents the phase angle associated with the virtual impedance, Indicates the system rated angular frequency, Indicates the power angle parameter.

[0089] The second step is to extract multiple parameter components in the transient current expression and the steady-state current expression.

[0090] It can be understood that in the above transient current expression and steady-state current expression, multiple parameter components such as steady-state power frequency component, inherent non-power frequency periodic attenuation component, inherent power frequency periodic attenuation component and free power frequency periodic attenuation component can be extracted, as shown in Table 1 below.

[0091] Table 1 Comparison of parameter components in transient current expression and steady-state current expression

[0092]

[0093] From Table 1 above, we can see that the three-phase fault current (i.e. short-circuit current) has a steady-state power frequency component and three transient components, which are respectively Intrinsic non-power frequency periodic attenuation component of attenuation , with time constant The inherent power frequency periodic attenuation component of the attenuation and the time constant Attenuated free power frequency periodic attenuation component .

[0094] It should be noted that due to the time constant The value of is only related to the internal parameters of the virtual synchronous generator and depends on the inherent properties of the grid-side converter. Therefore, the time constant The dominant attenuation component is called the intrinsic attenuation component. The dominant decay component, due to the time constant The value of depends not only on the internal impedance, but also on the voltage drop amplitude and voltage drop angle, so The dominant decay component is called the free decay component.

[0095] The third step is to determine the key components from the multiple parameter components, and determine the parameters to be optimized related to the short-circuit current based on the influencing factors corresponding to the amplitude and attenuation coefficient of each key component.

[0096] In this embodiment, the above three transient components of the short-circuit current are taken as key components, that is, the key components specifically include: an inherent non-power frequency periodic attenuation component, an inherent power frequency periodic attenuation component and a free power frequency periodic attenuation component.

[0097] The amplitude and attenuation coefficient of the transient component of the short-circuit current are affected by a variety of factors, including fault parameters, control parameters, and power angle parameters. Specifically, among the fault parameters, the fault voltage has a direct impact on the amplitude of the short-circuit current. Control parameters include the reactive integral coefficient, virtual resistance, virtual inductance, etc., which determine the attenuation characteristics and response speed of the short-circuit current. In addition, the power angle parameter is also a key factor that will have a significant impact on the variation trend of the fault current. Therefore, in this embodiment, the parameters to be optimized specifically include: reactive integral coefficient, virtual resistance, virtual inductance, and power angle parameters.

[0098] Among the aforementioned influencing factors, changes in the fault voltage amplitude significantly affect the initial value of the short-circuit current. Higher voltage drops typically result in larger short-circuit current amplitudes, thus affecting the transient response characteristics of the current. The reactive integral coefficient reflects the system's reactive power regulation capability. The larger its value, the more sensitive the system is to voltage disturbances, and the faster the short-circuit current decay process will be. Virtual resistance and virtual inductance, as electrical parameters within the virtual synchronous generator, primarily affect the damping and response speed of the short-circuit current. Virtual resistance primarily controls the decay rate of the short-circuit current, while virtual inductance affects the current phase and waveform characteristics.

[0099] Furthermore, the impact of changes in the power angle parameters on short-circuit current is complex, as it involves not only changes in the system's operating state but also the current phase regulation and dynamic response characteristics. Changes in the power angle parameters are reflected in time-dependent trigonometric functions, and their regulation of current amplitude is nonlinear and dynamic. This makes the transient characteristics of the short-circuit current complex and diverse as the power angle parameters change. Specifically, changes in the power angle parameters cause fluctuations in the amplitude and decay rate of the transient component of the short-circuit current. These fluctuations reflect the differences in the system's response characteristics at different power angles.

[0100] In one embodiment, a two-layer interactive optimization model of short-circuit current and transient stability is established with minimizing the short-circuit current of the virtual synchronous generator system as the optimization objective and meeting the transient stability standard as the constraint condition. Specifically, the model includes:

[0101] Firstly, the main layer optimization model is established based on the genetic algorithm with the minimization of the short-circuit current of the virtual synchronous generator system as the optimization objective.

[0102] It is understandable that short-circuit current is an important indicator affecting the safety of power system equipment and power electronic components. Excessive short-circuit current may cause equipment damage. Therefore, it is necessary to reduce the short-circuit current as much as possible in the control of the virtual synchronous generator system.

[0103] Then, an auxiliary layer optimization model is established through an exponential penalty factor with transient stability as a constraint.

[0104] Whether a virtual synchronous generator system can recover to a stable state after a fault (such as a short circuit) is crucial to the safe operation of the power system. The virtual synchronous generator system needs to have good transient response characteristics to cope with various potential faults in the power grid. Therefore, transient stability must be guaranteed during the optimization process.

[0105] Finally, the auxiliary layer optimization model is integrated with the main layer optimization model to obtain a two-layer interactive optimization model of short-circuit current and transient stability.

[0106] This embodiment constructs a two-level interactive optimization model to optimize the virtual synchronous generator system from two perspectives: minimizing short-circuit current and improving system transient stability. In each optimization iteration, a preliminary optimization result is first obtained by optimizing the short-circuit current. This preliminary optimization result is then corrected and evaluated using transient stability checks. This results in a final two-level optimization solution that not only minimizes short-circuit current but also meets the grid's transient stability requirements for the virtual synchronous generator system, ensuring rapid system recovery in the event of a fault.

[0107] In one embodiment, a main layer optimization model is established based on a genetic algorithm with minimizing the short-circuit current of the virtual synchronous generator system as the optimization objective, including:

[0108] In the first step, the initial values ​​of the system physical parameters of the virtual synchronous generator system, the initial values ​​of some parameters to be optimized, and the initial parameters related to the genetic algorithm are determined respectively.

[0109] This embodiment defines system physical parameters related to the virtual synchronous generator system. Initial values ​​for these parameters include virtual internal potential, virtual resistance, virtual inductance, grid frequency, active power, reactive power, and voltage. These parameters are fundamental components of the virtual synchronous generator system and define the specific structure of the current power system, providing the necessary conditions for subsequent short-circuit current calculations and transient stability analysis.

[0110] On the other hand, this embodiment defines initial parameters related to the genetic algorithm, which specifically include: the maximum number of iterations, the number of candidate solutions per generation, and the initial mutation probability.

[0111] In practical applications, the maximum number of iterations of the genetic algorithm can be set to 100, and the population size, which defines the number of candidate solutions per generation, can be set to 50. Furthermore, this embodiment also sets an initial mutation probability to control the mutation operation in the genetic algorithm. The initial mutation probability can be set to 0.2, and this value can be dynamically adjusted as the number of iterations increases.

[0112] In addition, this embodiment also defines a The matrix is ​​used to store the initial values ​​of some parameters to be optimized for each individual in the population. The initial values ​​of some parameters to be optimized for each individual in the population include the initial values ​​of five key parameters: reactive integral coefficient, virtual resistance, virtual inductance, inertia constant and damping coefficient.

[0113] In the second step, based on the initial values ​​of the system physical parameters, the initial values ​​of some parameters to be optimized and the initial parameters related to the genetic algorithm, population individuals are generated within the set range to establish the initial population.

[0114] Within the set range, the initial population individuals are randomly generated to ensure that the relevant parameters of the population individuals are within the reasonable upper and lower limits. For each population individual, the initial values ​​of the reactive integral coefficient, virtual resistance, virtual inductance, inertia constant, and damping coefficient are randomly generated, and the initial power angle parameters are calculated.

[0115] In this embodiment, the setting ranges corresponding to the above multiple parameters can be seen in Table 2 below.

[0116] Table 2 Parameter setting range comparison table

[0117]

[0118] In the third step, the short-circuit current minimization of the virtual synchronous generator system is taken as the optimization objective, and a fitness function related to the short-circuit current is established.

[0119] The fourth step is to determine the cyclic iteration strategy of the initial population based on the initial values ​​of the system physical parameters, the initial values ​​of some parameters to be optimized, the initial parameters related to the genetic algorithm and the fitness function, and establish the main layer optimization model.

[0120] In one embodiment, determining a preliminary optimization result corresponding to the parameter to be optimized when the short-circuit current is small specifically includes:

[0121] In the first step, the power angle parameters of each population individual are determined based on the initial values ​​of the system physical parameters and the initial values ​​of some parameters to be optimized.

[0122] The second step is to determine whether the power angle parameter of each population individual is within the set power angle range and obtain the power angle judgment result.

[0123] In this embodiment, the power angle parameter specifically includes the actual calculated power angle value and the power angle value of the virtual impedance. The specific calculation formula of the actual calculated power angle value is as follows:

[0124] (9)

[0125] in, Indicates the actual calculated power angle value, P indicates active power, represents the virtual internal potential, Indicates the fault voltage, Indicates virtual impedance.

[0126] In practical applications, the actual calculated power angle value needs to be constrained and checked to ensure that it is within the first set power angle range [30°, 70°]. If it does not meet this set range, it needs to be regenerated.

[0127] In this embodiment, in addition to performing constraint checks on the actual calculated power angle value obtained by the above calculation, it is also necessary to perform constraint checks on the power angle value of the virtual impedance. The power angle value of the virtual impedance can be calculated as follows:

[0128] (10)

[0129] in, Indicates the power angle value of the virtual impedance, Indicates the system rated angular frequency, represents the virtual inductance, Indicates virtual resistance.

[0130] In the constraint check phase, it is necessary to ensure that the power angle value of the virtual impedance is within the second set power angle range [80°, 90°].

[0131] In other words, to determine whether the power angle parameter of each population individual is within the set power angle range, two steps are required: specifically, determining whether the actual calculated power angle value is within the first set power angle range, and determining whether the power angle value of the virtual impedance is within the second set power angle range.

[0132] The third step is to determine the short-circuit current corresponding to the population individual whose power angle judgment result is yes, and calculate the fitness value based on the short-circuit current and the fitness function.

[0133] In this embodiment, the power angle determination result is yes, which means that the actually calculated power angle value is within the first set power angle range, and the power angle value of the virtual impedance is within the second set power angle range.

[0134] It can be understood that the main purpose of defining the fitness function is to calculate the short-circuit current of each individual in the population at different time points, including transient current and steady-state current, and minimize these two current values ​​as the fitness function.

[0135] In this embodiment, the fitness function is defined as the sum of the squares of the transient current and the steady-state current, so that both are controlled simultaneously during the optimization process.

[0136] It should be noted that the smaller short-circuit current refers to a situation where the fitness value is smaller than a preset fitness threshold.

[0137] In one specific implementation, the short-circuit current includes: a transient current and a steady-state current;

[0138] The fitness function is specifically:

[0139] (11)

[0140] in, represents the fitness value, and are coefficient values, and , i zt represents the transient current, i wt Represents the steady-state current.

[0141] The fourth step is to obtain the preliminary optimization results corresponding to the parameters to be optimized when the short-circuit current is small based on the fitness value.

[0142] In this embodiment, when the fitness value is small, it means that the short-circuit current is small. Therefore, a preliminary optimization result corresponding to the parameter to be optimized when the short-circuit current is small can be obtained.

[0143] In one embodiment, an auxiliary layer optimization model is established by using an exponential penalty factor with transient stability as a constraint, specifically including:

[0144] First, during the initial population cyclic iteration process in the main layer optimization model, it is determined whether the transient stability of each individual population meets the standard and the stability judgment result is obtained.

[0145] In practical applications, one or more standard values ​​of transient stability evaluation indicators can be set in advance. Then, by obtaining the actual indicator value corresponding to the transient stability evaluation of each individual population and comparing it with the standard value of the evaluation indicator, it can be determined whether the transient stability of each individual population meets the standard.

[0146] Then, the target population individuals whose stability judgment result is negative are determined, and the exponential penalty factors corresponding to the target population individuals under the current number of loop iterations are determined.

[0147] In this embodiment, the exponential penalty factor is specifically:

[0148] (12)

[0149] in, represents the exponential penalty factor corresponding to the target population individual under the i-th loop iteration, Represents the penalty coefficient constant.

[0150] Finally, the penalty strategy of the target population under different numbers of loop iterations is determined based on the exponential penalty factor, and an auxiliary layer optimization model is established.

[0151] In this embodiment, based on the exponential penalty factor, the expression of the fitness function can be adjusted as follows:

[0152] (13)

[0153] in, represents the adjusted fitness value corresponding to the i-th loop iteration, represents the fitness value corresponding to the i-th loop iteration, Indicates the dynamic coefficient value. When the stability judgment result is negative The value is 1, and the stability judgment result is yes The value is 0, in Represents the exponential penalty factor corresponding to the target population individual under the i-th loop iteration.

[0154] This embodiment uses an exponential penalty factor to deal with population individuals that violate the transient stability constraint. The purpose is to allow population individuals to be more exploratory in the early stages of iteration and to accelerate the elimination of unqualified population individuals in the later stages, thereby improving the optimization convergence speed and solution quality.

[0155] The formula for the exponential penalty factor shows that in the early stages, the exponential penalty factor is small, allowing some individuals that do not meet the requirements to participate in evolution, thereby increasing the diversity of solutions and preventing premature entrapment in local optima. In the later stages, as the number of generations increases, the value of the exponential penalty factor increases exponentially, causing the fitness values ​​of individuals that do not meet the requirements to increase rapidly, ultimately leading to their rapid elimination. Thus, in the early stages of iteration, the system has a strong exploratory nature, while in the later stages of iteration, it quickly converges to the optimal solution.

[0156] Furthermore, this embodiment introduces an elitist algorithm in the optimization phase. This elitist algorithm, also known as the elite retention strategy, is an improved strategy in genetic algorithms that is used to retain the best individuals in a population to ensure they are not lost due to random mutation or crossover during evolution. The introduction of the elitist algorithm addresses the problem of randomness in classic genetic algorithms, which can lead to degradation of solution quality and loss of optimal solutions. This significantly improves the stability and convergence speed of the genetic algorithm.

[0157] In an elitist algorithm, the best performing individuals in each generation are directly retained and copied to the next generation. This means that even if less ideal individuals are generated in the next generation through crossover and mutation, the optimal solution will still be retained, ensuring the global optimal performance of the algorithm.

[0158] It can be seen that the core idea of ​​the elite algorithm can be summarized as follows:

[0159] First, retain the best individuals: In each generation, the individuals with the highest (i.e., best) fitness in the current population are directly copied to the next generation without crossover or mutation. This allows the characteristics of the best individuals to be passed on to future generations, thereby accelerating the evolution of the entire population.

[0160] Second, it reduces the possibility of degradation: Genetic algorithms are inherently randomized global search algorithms that explore the solution space through methods such as crossover and mutation. However, during the evolutionary process, random mutations can degrade or even cause the currently optimal solution to be lost, leading to degradation. The elite retention strategy significantly reduces the possibility of solution degradation by retaining the best individuals in each generation.

[0161] Third, accelerated convergence: By retaining the optimal population individuals, the elite algorithm can accelerate the convergence of the genetic algorithm, especially when the search space of the problem is large and the optimization objective is complex, which enables the elite algorithm to find a near-optimal solution more quickly.

[0162] The main functions of the elite algorithm in this embodiment are as follows:

[0163] First, ensure that the optimal solution is not lost: During the two-level optimization process, the short-circuit current must be minimized while also ensuring the system's transient stability. In this complex optimization problem, the elite retention strategy plays a key role. By retaining the optimal solution directly, it avoids the risk of losing good control parameter combinations due to mutation and crossover operations.

[0164] Second, improve the convergence speed: Combined with the elite algorithm, the system can ensure that there is an optimal solution in each generation of iteration. This can effectively narrow the search space, thereby accelerating the convergence speed and reducing the optimization time.

[0165] Third, prevent premature degeneration: By retaining the best population individuals in each generation, the elite retention strategy can prevent the optimal solution from degenerating during random crossover and mutation. This is especially important in the optimization process of genetic algorithms, because the uncontrollable factors brought about by randomness may cause the quality of the solution to deteriorate.

[0166] Fourth, incorporating an exponential penalty factor: This example incorporates an exponential penalty factor on top of elite retention. This allows for diversity in the early stages and accelerates the elimination of individuals that don't meet the criteria in the later stages. This approach allows elite individuals to gradually lead the population toward the global optimal solution, while the exponential penalty factor ensures that the optimization process is less likely to fall into local optima. This combination of strategies effectively balances the conflict between "exploration" and "exploitation" in genetic algorithms.

[0167] In this embodiment, the process of implementing the short-circuit current and transient stability double-layer optimization using the double-layer interactive optimization model constructed by the improved genetic algorithm can be found in Figure 2 The specific process is as follows:

[0168] Step 201: Initialize the genetic algorithm related parameters to form an initial population.

[0169] Step 202: Set the genetic generation number of the genetic algorithm, that is, the maximum number of iterations.

[0170] Step 203: Calculate the short-circuit current of each generation (population) of individuals and analyze the relevant parameters in the short-circuit current formula (expression), that is, the parameters to be optimized.

[0171] Step 204: Determine whether the virtual impedance angle (power angle parameter) and system stability (transient stability) meet the requirements, that is, determine whether the power angle parameter is within the set power angle range, and determine whether the transient stability meets the standard.

[0172] Step 205: If the above judgment results are all yes, directly calculate the steady-state value and peak value corresponding to the short-circuit current, that is, the maximum value of the steady-state current value and the transient current value of the short-circuit current.

[0173] Step 206: If at least one of the above judgment results is negative, an (exponential) penalty factor is introduced to penalize individuals that do not meet the requirements (i.e., individuals in the target population), and then the steady-state value and peak value corresponding to the short-circuit current are calculated.

[0174] Step 207: Generate the fitness value according to the steady-state value, the peak value, and the penalty value (ie, the exponential penalty factor), that is, calculate the fitness value according to the fitness function.

[0175] Step 208: an elite retention strategy is introduced during the optimization process to retain the best performing individuals in the current generation.

[0176] Step 209: Perform crossover and mutation operations to generate a new population. During the crossover and mutation operations, some high-performing individuals in the population can be selected for crossover and mutation to generate a new population. New individuals are generated by randomly selecting two parent individuals for crossover and setting a crossover ratio coefficient to ensure population diversity. Simultaneously, mutation is performed on top of crossover to further enhance the diversity of the new individuals. Dynamic adjustment of the mutation probability is set to retain high-performing individuals in the early stages of iteration, while increasing the global exploration capability through mutation with a higher probability in the later stages.

[0177] Step 210: Update the population and iterate. During the population update process, the newly generated population can be used as the next generation population for the next round of iteration. Through iteration, individual parameters are continuously optimized to find the control parameter combination that achieves the optimal fitness.

[0178] Step 211: Determine whether the genetic generation is the set value, that is, determine whether the genetic generation has reached the maximum number of iterations. If the judgment result is no, add 1 to the genetic generation and continue the genetic iteration.

[0179] Step 212: If the judgment result of step 211 is yes, the final optimization parameters are output, that is, the double-layer optimization solution results of the parameters to be optimized are output, including data such as reactive integral coefficient, virtual resistance, virtual inductance, inertia constant, damping coefficient, etc., and the genetic iteration process ends.

[0180] After the genetic algorithm iteration process is complete, the final power angle of the virtual impedance can be calculated based on the two-level optimization solution and output for evaluating the system's impedance characteristics. The genetic algorithm's optimization process can then be plotted, and the fitness value can be observed as it changes with the number of iterations to verify the optimization effect. Specifically, a curve chart can be used to display the minimum fitness value for each generation, thereby verifying the optimization process of the genetic algorithm. The algorithm's optimization effect can be determined by observing whether the fitness curve has a significant drop. The final output parameters can then be further verified, for example, by examining their impact on short-circuit current and transient stability to ensure their effectiveness and stability in engineering applications.

[0181] To verify the correctness and superiority of the aforementioned two-level optimization method for short-circuit current and transient stability of a virtual synchronous generator system, this example constructed a model of a virtual synchronous generator integrated into an infinite power grid using Matlab / Simulink simulation software. The parameters to be optimized (i.e., key control parameters before optimization) can be found in Table 3. Three grid fault voltage drops of varying severity were set: mild, moderate, and severe.

[0182] Through the above-mentioned two-level optimization method of short-circuit current and transient stability of the virtual synchronous generator system, the key control parameters before and after optimization are obtained, as shown in Table 3.

[0183] Table 3 Key control parameters before and after optimization

[0184]

[0185] Combined with Table 3 above, the following will compare the output current of renewable energy before and after parameter optimization according to the severity of voltage drop, thereby illustrating the superiority of the proposed short-circuit current suppression method.

[0186] During the mild voltage drop phase, a three-phase symmetrical fault is set in the grid, causing the grid connection point voltage to drop to 0.8pu. The fault occurs at t = 3 seconds and lasts for 0.3 seconds before the fault is cleared. The output current of the virtual synchronous generator system before and after the mild voltage drop is as follows: Figure 3 and Figure 4 The frequency change comparison before and after the slight voltage drop is shown in Figure 5 As shown in this case, the optimization iteration process of the improved genetic algorithm is as follows Figure 6 shown.

[0187] The optimization results show that the genetic algorithm converged to stability around the 70th generation. After parameter optimization, the "shock" value of the short-circuit current decreased significantly: the "shock" value during a fault decreased from 257A to 199A, a 22% decrease; the "shock" value after fault clearance decreased from 287A to 230A, a 20% decrease. However, the "steady-state" component of the short-circuit current did not change significantly. Furthermore, the rotor frequency offset decreased after optimization, indicating enhanced synchronous stability of the system. This indirectly demonstrates that the algorithm not only iteratively optimizes the short-circuit current but also takes into account the system's synchronous stability.

[0188] During the moderate voltage drop phase, a three-phase symmetrical fault is set in the grid, causing the grid connection point voltage to drop to 0.5 pu. The fault occurs at t = 3 seconds and lasts for 0.3 seconds before the fault is cleared. The output currents of the virtual synchronous generator before and after the moderate voltage drop are as follows: Figure 7 and Figure 8 The frequency change before and after the moderate voltage drop is shown in the figure. Figure 9 As shown in this case, the optimization iteration process of the improved genetic algorithm is as follows Figure 10 shown.

[0189] The optimization results show that the genetic algorithm converged to stability around the 130th generation. After parameter optimization, the "shock" value during a fault event decreased from 447A to 271A, a 39% decrease. The "shock" value after fault clearance decreased from 349A to 284A, a 19% decrease. The "steady-state" component of the short-circuit current remained unchanged. Furthermore, the rotor frequency offset was reduced after optimization, indicating enhanced system synchronous stability.

[0190] During the severe voltage drop phase, a three-phase symmetrical fault in the grid is set to cause the grid connection point voltage to drop to 0.2pu. The fault occurs at t = 3 seconds and lasts for 0.3 seconds before the fault is cleared. The output current of the virtual synchronous generator before and after the severe voltage drop is as follows: Figure 11 and Figure 12 The frequency change comparison before and after the severe voltage drop is shown in Figure 13 As shown in this case, the optimization iteration process of the improved genetic algorithm is as follows Figure 14 shown.

[0191] The optimization results show that the genetic algorithm converged to stability around the 70th generation. After parameter optimization, the "shock" value during a fault event decreased from 621A to 323A, a 48% decrease. The "shock" value after the fault was cleared decreased from 332A to 313A, a 6% decrease. The "steady-state" state remained unchanged. Furthermore, the rotor frequency offset was reduced after optimization, indicating enhanced synchronous stability of the system.

[0192] According to the superposition theorem, short-circuit current can be divided into a free component and a forced component. The free component is determined by system parameters and is induced both during fault occurrence and fault clearance. The forced component, on the other hand, is determined by the no-load electromotive force of the virtual synchronous generator. Considering the system's need for reactive power support, the reactive power command remains unchanged during the fault. Therefore, the forced component of the short-circuit current, or the steady-state component, remains unchanged before and after optimization. Short-circuit current optimization primarily focuses on the transient free component. This optimization significantly reduces the impact of this component on the system. In severe fault conditions, the impact can be limited to approximately 50% of the original value.

[0193] At the same time, since the synchronous stability of the system is taken into consideration during the optimization process, it can be ensured that the optimized system parameters can ensure the synchronous stability of the system without power angle instability even in the case of serious faults.

[0194] In summary, the dual-layer optimization method for short-circuit current and transient stability of a virtual synchronous generator system provided by the embodiment of the present invention has at least the following beneficial effects:

[0195] First, by defining a fitness function to evaluate the impact of short-circuit currents on equipment and systems and the stability of each individual population during transients, the team aims to minimize the impact of short-circuit currents on equipment and systems. This addresses the problem of traditional optimization schemes' inaccurate assessment of multiple key system parameters.

[0196] Second, the introduction of an improved genetic algorithm uses an elite strategy to retain the best individuals and introduces an exponential penalty factor, which can better handle constraint violations, improve the overall convergence speed and optimization performance, ensure rapid iteration to the global optimum in the high-dimensional control parameter space, and improve the optimization efficiency and reliability of short-circuit current.

[0197] Third, by performing multi-objective optimization of key control parameters in the virtual synchronous generator system, the dynamic response capability of the virtual synchronous generator system was significantly improved, particularly by simultaneously achieving iterative optimization of short-circuit current and enhanced transient stability under fault conditions. In the event of a grid fault, the system can recover quickly and stably, ensuring the overall stability of the grid.

[0198] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A two-level optimization method for short-circuit current and transient stability of a virtual synchronous generator system, characterized in that: include: Determine the parameters to be optimized related to short-circuit current in the virtual synchronous generator system; A primary optimization model is established based on a genetic algorithm with minimizing the short-circuit current of the virtual synchronous generator system as the optimization objective; an auxiliary optimization model is established using an exponential penalty factor with transient stability compliance as a constraint; and the auxiliary optimization model is integrated with the primary optimization model to obtain a two-layer interactive optimization model for short-circuit current and transient stability. According to the two-layer interactive optimization model, the preliminary optimization results corresponding to the parameters to be optimized are determined when the short-circuit current is small, and the preliminary optimization results are corrected using transient stability to obtain the two-layer optimization solution results of the parameters to be optimized.

2. The two-level optimization method for short-circuit current and transient stability of a virtual synchronous generator system according to claim 1 is characterized in that: Determine the parameters to be optimized related to short-circuit current in the virtual synchronous generator system, including: Obtain the transient current expression and steady-state current expression of the short-circuit current in the virtual synchronous generator system; Extracting multiple parameter components from the transient current expression and the steady-state current expression; A key component is determined from the multiple parameter components, and a parameter to be optimized related to the short-circuit current is determined based on the influencing factors corresponding to the amplitude and attenuation coefficient of each key component.

3. The double-layer optimization method for short-circuit current and transient stability of a virtual synchronous generator system according to claim 2 is characterized in that: The key components include: inherent non-power frequency periodic attenuation component, inherent power frequency periodic attenuation component and free power frequency periodic attenuation component; The parameters to be optimized include: reactive integral coefficient, virtual resistance, virtual inductance and power angle parameter.

4. The double-layer optimization method for short-circuit current and transient stability of a virtual synchronous generator system according to claim 1 is characterized in that: Based on the genetic algorithm, the short-circuit current of the virtual synchronous generator system is minimized as the optimization goal. The main layer optimization model is established, including: Determine the initial values ​​of the system physical parameters of the virtual synchronous generator system, the initial values ​​of some parameters to be optimized, and the initial parameters related to the genetic algorithm respectively; Based on the initial values ​​of the physical parameters of the system, the initial values ​​of some parameters to be optimized and the initial parameters related to the genetic algorithm, population individuals are generated within a set range to establish an initial population; Taking the minimization of the short-circuit current of the virtual synchronous generator system as the optimization objective, a fitness function related to the short-circuit current is established. Based on the initial values ​​of the system physical parameters, the initial values ​​of some parameters to be optimized, the initial parameters related to the genetic algorithm and the fitness function, the cyclic iteration strategy of the initial population is determined and the main layer optimization model is established.

5. The double-layer optimization method for short-circuit current and transient stability of a virtual synchronous generator system according to claim 4 is characterized in that: Determine the preliminary optimization results corresponding to the parameters to be optimized when the short-circuit current is small, including: Determining the power angle parameters of each population individual based on the initial values ​​of the system physical parameters and the initial values ​​of some parameters to be optimized; Determine whether the power angle parameter of each population individual is within the set power angle range, and obtain the power angle judgment result; Determining the short-circuit current corresponding to the population individual for which the power angle judgment result is yes, and calculating the fitness value based on the short-circuit current and the fitness function; According to the fitness value, a preliminary optimization result corresponding to the parameter to be optimized is obtained when the short-circuit current is small.

6. The double-layer optimization method for short-circuit current and transient stability of a virtual synchronous generator system according to claim 4, characterized in that: The initial values ​​of the system physical parameters include: virtual internal potential, grid frequency, active power, reactive power and voltage; The initial parameters related to the genetic algorithm include: the maximum number of iterations, the number of candidate solutions per generation, and the initial mutation probability.

7. The double-layer optimization method for short-circuit current and transient stability of a virtual synchronous generator system according to claim 4 is characterized in that: The short-circuit current includes: transient current and steady-state current; The fitness function is: in, represents fitness, and are coefficient values, and , i zt represents the transient current, i wt Represents the steady-state current.

8. The double-layer optimization method for short-circuit current and transient stability of a virtual synchronous generator system according to claim 3 is characterized in that: An auxiliary layer optimization model is established using an exponential penalty factor with transient stability compliance as a constraint, including: During the initial population cyclic iteration process in the main layer optimization model, determining whether the transient stability of each population individual meets the standard, and obtaining a stability determination result; Determine the target population individuals for which the stability judgment result is negative, and determine the exponential penalty factor corresponding to the target population individuals under the current number of loop iterations; The penalty strategy of the target population under different numbers of loop iterations is determined based on an exponential penalty factor, and an auxiliary layer optimization model is established.

9. The double-layer optimization method for short-circuit current and transient stability of a virtual synchronous generator system according to claim 8, characterized in that: The exponential penalty factor is: in, represents the exponential penalty factor corresponding to the target population individual under the i-th loop iteration, Represents the penalty coefficient constant.

Images