Over-current suppression method and device for black start, electronic equipment and storage medium

Abstract

The invention discloses a black-start overcurrent suppression method and device, electronic equipment and a storage medium, and relates to the field of artificial intelligence or other related technical fields, and the method comprises the steps: collecting electric power data of an optical storage grid-connected system, carrying out the preprocessing of the electric power data, and extracting an electric power feature vector; the electric power feature vector is input into a virtual impedance coefficient optimization model, a target virtual impedance coefficient is output, and the virtual impedance coefficient optimization model integrates an attractor trend strategy, a coupling interference strategy and an information projection strategy for dynamic optimization; and the power data and the target virtual impedance coefficient are substituted into a virtual impedance calculation formula, a virtual impedance value is calculated, a reference signal of an inverter control loop is adjusted based on the virtual impedance value, and overcurrent caused by black start is suppressed. According to the invention, the technical problem that the overcurrent suppression stability is low when the overcurrent is suppressed based on a fixed virtual impedance control strategy in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence or other related technical fields, and in particular to a black start overcurrent suppression method and device, electronic equipment, and storage medium. Background Art

[0002] Electricity supply is closely linked to people's daily lives, and power outages can cause widespread inconvenience and potential risks. Therefore, when environmental or human factors cause a power outage on the receiving grid, the grid requires the sending power source to have black start capabilities to restore power as quickly as possible. Black start, as a power system recovery strategy, aims to restore power without external grid support after a large-scale power outage using self-starting power sources. Black start capability is a key indicator of power system resilience and self-healing capabilities. Especially in emergency situations where natural disasters such as earthquakes, typhoons, and extreme weather events cause power grid collapse, black start solutions can be quickly initiated and gradually restore power system operations, which is of great significance for ensuring the normal operation of social and economic activities.

[0003] During a black start, the power system is vulnerable. The sudden addition of loads, especially devices with high starting currents such as transformers and induction motors, can easily trigger overcurrent. Overcurrent occurs when the current in a circuit exceeds its rated or preset value. In a power system, this can be caused by sudden load changes, short circuits, or improper control strategies. In a black start scenario, overcurrent can not only damage critical equipment (such as converters and inverters), but can also affect system stability and voltage control, hindering the subsequent smooth connection and recovery of power sources and loads. Therefore, suppressing overcurrent during a black start is essential for ensuring a stable and safe power system recovery process.

[0004] In related technologies, a fixed virtual impedance control strategy is typically used to suppress overcurrent during a black start. While this strategy can provide a certain degree of current suppression, its fixed impedance value has limited applicability under varying load and grid conditions, and it cannot intelligently adjust to complex and changing system conditions. This results in low overcurrent suppression efficiency and stability.

[0005] To address the above-mentioned problems, no effective solutions have been proposed so far. Summary of the Invention

[0006] The embodiments of the present invention provide a black start overcurrent suppression method and device, electronic device and storage medium, so as to at least solve the technical problem in the related art that when suppressing overcurrent based on a fixed virtual impedance control strategy, the overcurrent suppression stability is low.

[0007] According to one aspect of an embodiment of the present invention, a method for suppressing overcurrent during black start is provided, comprising: collecting power data of a photovoltaic energy storage grid-connected system, preprocessing the power data, and extracting a power characteristic vector, wherein the power data includes at least voltage data, current data, and frequency data; inputting the power characteristic vector into a virtual impedance coefficient optimization model, and outputting a target virtual impedance coefficient, wherein the virtual impedance coefficient optimization model integrates an attractor trend strategy, a coupling interference strategy, and an information projection strategy, and performs iterative calculation in a dynamic optimization manner to obtain the target virtual impedance coefficient; substituting the power data and the target virtual impedance coefficient into a virtual impedance calculation formula, calculating a virtual impedance value, and adjusting a reference signal of an inverter control loop based on the virtual impedance value to suppress overcurrent caused by black start.

[0008] Furthermore, the power characteristic vector is input into the virtual impedance coefficient optimization model, and the step of outputting the target virtual impedance coefficient includes: the virtual impedance coefficient optimization model generates a group of initial neuron populations based on the power characteristic vector, wherein the initial neuron population includes N neuron individuals, each neuron individual represents a virtual impedance coefficient, and N is a positive integer; the virtual impedance coefficient optimization model calculates the fitness value of each neuron individual in the initial neuron population based on the objective function, wherein the objective function is constructed based on overcurrent, frequency deviation and voltage harmonic distortion rate; the virtual impedance coefficient optimization model iteratively updates the initial neuron population based on the fitness value until the iteration number threshold is reached and the iteration is stopped; the virtual impedance coefficient optimization model selects the neuron individual with the largest fitness value as the target virtual impedance coefficient based on the updated neuron population.

[0009] Furthermore, the step of iteratively updating the initial neuron population based on the fitness value of the virtual impedance coefficient optimization model includes: step 1, the virtual impedance coefficient optimization model optimizes the initial neuron population based on the fitness value of the individual neuron and the attractor trend strategy to obtain a neuron population based on attractor optimization; step 2, the virtual impedance coefficient optimization model optimizes the initial neuron population based on the coupling interference strategy to obtain a neuron population based on coupling interference optimization; step 3, the virtual impedance coefficient optimization model optimizes the initial neuron population based on the information projection strategy and combines the neuron population based on attractor and the neuron population based on coupling interference. The neuron population is optimized by the meta-population to obtain a neuron population optimized based on information projection; in step four, the virtual impedance coefficient optimization model calculates the fitness value of each individual neuron in the neuron population optimized based on information projection through a pre-established objective function, and screens the neuron population optimized based on information projection according to the fitness value to obtain the screened neuron population, and mutates the screened neuron population through a mutation strategy to generate a new neuron population, and uses the new neuron population as the initial neuron population for the next round of iteration; repeat steps one to four to iteratively update the neuron population.

[0010] Furthermore, the virtual impedance coefficient optimization model optimizes the initial neuron population based on the attractor trend strategy, and the step of obtaining the attractor-based neuron population includes: the virtual impedance coefficient optimization model sorts the neuron individuals according to the fitness values ​​of the neuron individuals to obtain a sorted list; calculates the number of attractors to be selected based on the attractor ratio and the total number of the neuron individuals in the initial neuron population; selects attractors from the sorted list based on the number of attractors to obtain an attractor set; optimizes the initial neuron population based on the attractors in the attractor set to obtain a neuron population optimized based on attractors, wherein, when optimizing the initial neuron population based on the attractors in the attractor set, the neuron individuals other than the attractors in the initial neuron population are controlled to randomly converge to any of the attractors.

[0011] Furthermore, the virtual impedance coefficient optimization model optimizes the initial neuron population based on the coupling interference strategy, and the steps of obtaining the neuron population based on coupling interference include: calculating the average value of other neuron populations except the initial neuron population, and optimizing the initial neuron population according to the average value to obtain a first optimized neuron population; calculating the average value of the difference between the initial neuron population and the other neuron populations, and optimizing the initial neuron population according to the average value of the difference to obtain a second optimized neuron population; generating a neuron population based on coupling interference optimization based on the first optimized neuron population, the second optimized neuron population and the global scaling factor.

[0012] Furthermore, the virtual impedance coefficient optimization model is based on an information projection strategy, and combines an attractor-based neuron population and a coupled interference-based neuron population to optimize the initial neuron population. The steps of obtaining a neuron population based on information projection include: generating an adjacency matrix and a communication intensity matrix, and performing secondary optimization on the attractor-optimized neuron population according to the adjacency matrix and the communication intensity matrix to obtain a third optimized neuron population; performing secondary optimization on the coupled interference-optimized neuron population according to the adjacency matrix, the communication intensity matrix and the ratio of the number of iterations to obtain a fourth optimized neuron population; and generating a neuron population based on information projection based on the initial neuron population, the third optimized neuron population and the fourth optimized neuron population.

[0013] Furthermore, the virtual impedance calculation formula is expressed as: Among them, Z v (t) is the virtual impedance, E is the output voltage amplitude of the converter, V PCC is the voltage amplitude of the common connection point of the switching load, I max is the maximum allowable current of the converter, k r (t) is the virtual impedance coefficient, I(t) is the current value of the overcurrent control system, I ref It is the current reference value of the overcurrent control system.

[0014] According to another aspect of an embodiment of the present invention, a black start overcurrent suppression device is also provided, including: an acquisition unit, used to collect power data of a photovoltaic storage grid-connected system, and preprocess the power data to extract a power characteristic vector, wherein the power data at least includes voltage data, current data and frequency data; an output unit, used to input the power characteristic vector into a virtual impedance coefficient optimization model, and output a target virtual impedance coefficient, wherein the virtual impedance coefficient optimization model integrates an attractor trend strategy, a coupling interference strategy and an information projection strategy, and performs iterative calculation in a dynamic optimization manner to obtain the target virtual impedance coefficient; a suppression unit, used to substitute the power data and the target virtual impedance coefficient into a virtual impedance calculation formula, calculate the virtual impedance value, and adjust the reference signal of the inverter control loop based on the virtual impedance value to suppress the overcurrent caused by the black start.

[0015] Furthermore, the output unit includes: a first generating subunit, for the virtual impedance coefficient optimization model to generate a group of initial neuron populations based on the power characteristic vector, wherein the initial neuron population includes N neuron individuals, each neuron individual represents a virtual impedance coefficient, and N is a positive integer; a first calculating subunit, for the virtual impedance coefficient optimization model to calculate the fitness value of each neuron individual in the initial neuron population based on the objective function, wherein the objective function is constructed based on overcurrent, frequency deviation and voltage harmonic distortion rate; a first iterative subunit, for the virtual impedance coefficient optimization model to iteratively update the initial neuron population based on the fitness value until the iteration number threshold is reached and the iteration is stopped; a first selecting subunit, for the virtual impedance coefficient optimization model to select the neuron individual with the largest fitness value as the target virtual impedance coefficient based on the updated neuron population.

[0016] Furthermore, the first iterative subunit includes: a first optimization module, used for step one, the virtual impedance coefficient optimization model optimizes the initial neuron population based on the fitness value of the individual neuron and the attractor trend strategy to obtain the neuron population based on attractor optimization; a second optimization module, used for step two, the virtual impedance coefficient optimization model optimizes the initial neuron population based on the coupling interference strategy to obtain the neuron population based on coupling interference optimization; a third optimization module, used for step three, the virtual impedance coefficient optimization model optimizes the neuron population based on the information projection strategy and combines the neuron population based on attractor and the neuron population based on coupling interference to optimize the neuron population. The virtual impedance coefficient optimization model calculates the fitness value of each individual neuron in the neuron population based on information projection through a pre-established objective function, and screens the neuron population based on information projection according to the fitness value to obtain the screened neuron population, and performs a mutation operation on the screened neuron population through a mutation strategy to generate a new neuron population, and uses the new neuron population as the initial neuron population for the next round of iteration; the first repetition module is used to repeat steps one to four above to iteratively update the neuron population.

[0017] Furthermore, the first optimization module includes: a first sorting submodule, which is used for the virtual impedance coefficient optimization model to sort the neuron individuals according to the fitness values ​​of the neuron individuals to obtain a sorted list; a first calculation submodule, which is used to calculate the number of attractors to be selected based on the attractor ratio and the total number of the neuron individuals in the initial neuron population; a first selection submodule, which is used to select attractors from the sorted list based on the number of attractors to obtain an attractor set; a first optimization submodule, which is used to optimize the initial neuron population based on the attractors in the attractor set to obtain a neuron population optimized based on the attractors, wherein, when optimizing the initial neuron population based on the attractors in the attractor set, the neuron individuals other than the attractors in the initial neuron population are controlled to randomly converge to any of the attractors.

[0018] Furthermore, the second optimization module includes: a second calculation submodule, used to calculate the average value of other neuron populations except the initial neuron population, and optimize the initial neuron population according to the average value to obtain a first optimized neuron population; a third calculation submodule, used to calculate the difference average value between the initial neuron population and the other neuron populations, and optimize the initial neuron population according to the difference average value to obtain a second optimized neuron population; a first generation submodule, used to generate a neuron population based on coupled interference optimization based on the first optimized neuron population, the second optimized neuron population and the global scaling factor.

[0019] Furthermore, the third optimization module includes: a second generation submodule, used to generate an adjacency matrix and a communication intensity matrix, and perform secondary optimization on the neuron population based on attractor optimization according to the adjacency matrix and the communication intensity matrix to obtain a third optimized neuron population; a second optimization submodule, used to perform secondary optimization on the neuron population based on coupling interference optimization according to the adjacency matrix and the communication intensity matrix and the ratio of the number of iterations to obtain a fourth optimized neuron population; a third generation submodule, used to generate a neuron population based on information projection based on the initial neuron population, the third optimized neuron population and the fourth optimized neuron population.

[0020] Furthermore, the virtual impedance calculation formula is expressed as: Among them, Z v (t) is the virtual impedance, E is the output voltage amplitude of the converter, V PCC is the voltage amplitude of the common connection point of the switching load, I max is the maximum allowable current of the converter, k r (t) is the virtual impedance coefficient, I(t) is the current value of the overcurrent control system, I ref It is the current reference value of the overcurrent control system.

[0021] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is further provided, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any of the above-mentioned black start overcurrent suppression methods.

[0022] According to another aspect of an embodiment of the present invention, an electronic device is further provided, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement any of the above-mentioned black start overcurrent suppression methods.

[0023] In this application, the following steps are taken: collecting power data of the photovoltaic storage grid-connected system, preprocessing the power data, extracting the power characteristic vector, wherein the power data includes at least voltage data, current data and frequency data, and then inputting the power characteristic vector into the virtual impedance coefficient optimization model, outputting the target virtual impedance coefficient, wherein the virtual impedance coefficient optimization model integrates the attractor trend strategy, the coupling interference strategy and the information projection strategy, and performs iterative calculation in a dynamic optimization manner to obtain the target virtual impedance coefficient, and finally substitutes the power data and the target virtual impedance coefficient into the virtual impedance calculation formula to calculate the virtual impedance value, and adjusts the reference signal of the inverter control loop based on the virtual impedance value to suppress the overcurrent caused by the black start.

[0024] In the present application, a virtual impedance coefficient optimization model is pre-constructed to optimize the virtual impedance coefficient according to the actual situation of grid-connected operation, so as to provide the optimal virtual impedance parameters in the current state according to the real-time operating state of the power grid system, so as to suppress the overcurrent generated by the black start, thereby achieving the purpose of dynamic optimization of the virtual resistance parameters and obtaining the technical effect of improving the stability of overcurrent suppression, thereby solving the technical problem in the related technology of low overcurrent suppression stability when suppressing overcurrent based on a fixed virtual impedance control strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0026] Figure 1 is a flow chart of an optional black start overcurrent suppression method according to an embodiment of the present invention;

[0027] Figure 2 is a schematic diagram of an optional virtual impedance coefficient optimization process according to an embodiment of the present invention;

[0028] Figure 3 is a schematic diagram of an optional black start overcurrent suppression device according to an embodiment of the present invention;

[0029] Figure 4 The figure is a hardware structure block diagram of an electronic device (or mobile device) that performs an overcurrent suppression method for a black start according to an embodiment of the present invention. DETAILED DESCRIPTION

[0030] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0031] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0032] It should be noted that the black start overcurrent suppression method and device in the present application can be used in the field of artificial intelligence when the overcurrent generated by the black start is suppressed based on artificial intelligence, and can also be used in any field other than the field of artificial intelligence when the overcurrent generated by the black start is suppressed based on artificial intelligence. The application of the black start overcurrent suppression method and device in the present application does not limit the application field.

[0033] It should be noted that the relevant information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data comply with the relevant laws, regulations and standards of the relevant regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse. For example, an interface is set up between this system and the relevant users or institutions. Before obtaining relevant information, it is necessary to send an acquisition request to the aforementioned user or institution through the interface, and obtain the relevant information after receiving the consent information fed back by the aforementioned user or institution.

[0034] It should be noted that in this application, when collecting and analyzing customer information, corresponding operation entrances are provided for users to choose to agree or reject the automated decision-making results; if the user chooses to reject, the expert decision-making process will be entered.

[0035] The following embodiments of the present invention can be applied to various black start overcurrent suppression systems, applications, and devices. This invention proposes a black start overcurrent suppression strategy using an adaptive virtual resistor based on a neural population dynamic optimization algorithm. By introducing an adaptive virtual resistor, this strategy suppresses overcurrent generated when special loads such as transformers are connected during the black start process, thereby extending the life of the converter and enhancing the intelligent features of the photovoltaic storage system.

[0036] The present invention will be described in detail below with reference to various embodiments.

[0037] Example 1

[0038] According to an embodiment of the present invention, an embodiment of a method for suppressing overcurrent during black start is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0039] Figure 1 FIG. 1 is a flow chart of an optional black start overcurrent suppression method according to an embodiment of the present invention. Figure 1 As shown, the method includes the following steps:

[0040] Black start capability is key to the power grid's ability to respond to emergencies. It represents the power system's ability to autonomously restore power after a major outage or natural disaster-induced blackout. A black start is essentially the process of gradually returning the power grid to normal operation from a complete blackout. During this process, a specific power source (such as a hydropower station, thermal power station, or renewable energy power station) is selected as the starting power source. Through self-starting or with the help of external small power sources, key nodes and links in the grid are gradually restored, ultimately achieving full recovery. A black start not only tests the grid's design redundancy and flexibility but also demonstrates the power system's ability to self-repair and restore power in extreme situations.

[0041] Overcurrent is a common but extremely dangerous problem in power systems, especially during black start. Overcurrent refers to the instantaneous current in a circuit exceeding its rated current during normal operation. This situation usually occurs in the event of a sudden load change, a short circuit, or improper control strategy. During the black start phase, as the system gradually recovers, a large number of devices and loads are connected to the grid in a short period of time, especially devices containing large electromagnetic induction elements (such as transformers and motors). These devices will absorb several times the current during normal operation at the moment of startup, resulting in severe overcurrent. Overcurrent not only damages power electronic equipment such as converters and inverters, shortening their service life, but also causes voltage and frequency fluctuations in the grid, threatening the stability of the entire system and the safe recovery process.

[0042] Grid-connected photovoltaic (PV) and energy storage systems play an increasingly important role in modern power systems, particularly in terms of black start capabilities and overcurrent suppression. Grid-connected PV and energy storage systems combine photovoltaic power generation and energy storage technologies, are connected to the grid via inverters, and can provide stable voltage and frequency support without external power supply support. During the black start process, the grid-connected PV and energy storage system's grid-connected control characteristics enable it to serve as an important starting power source, providing initial energy support for grid recovery. However, the PV and energy storage system's rapid response and stable output in the early stages of a black start may also bring the risk of overcurrent. This is because in the early stages of system recovery, the grid's inertia and damping characteristics are low, and coupled with the uncertainty of load access, current surges are likely to occur. Therefore, grid-connected PV and energy storage systems must be equipped with an effective overcurrent suppression strategy to ensure their safe operation during the black start process, while avoiding damage to power electronic equipment and improving the recovery efficiency and stability of the entire grid.

[0043] Step S101 : collecting power data of the photovoltaic energy storage grid-connected system, pre-processing the power data, and extracting power feature vectors.

[0044] It should be noted that a grid-connected photovoltaic (PV)-storage system is an integrated power system that combines photovoltaic cells with energy storage devices (such as battery packs). It converts direct current (DC) electricity into alternating current (AC) via an inverter, which is then fed into the grid. This system's advantage lies in its ability to leverage the clean power generation capabilities of photovoltaic cells and, combined with energy storage devices, to store and release energy, providing stable and reliable power support to the grid. This system plays a crucial role in specific conditions, such as black start operations.

[0045] In the above step S101, when there is an overcurrent phenomenon in the photovoltaic storage grid-connected system, the primary task is to collect the power data of the photovoltaic storage grid-connected system. These data include but are not limited to voltage data, current data and frequency data, which respectively reflect the voltage level, current state and frequency stability of the system operation. Voltage data and current data are important parameters for measuring the operating status of the system, while frequency data is directly related to system synchronization and stability. The collection of these data is usually achieved through sensors, such as voltage transformers, current transformers and frequency measurement equipment. Through real-time data collection, the system status can be monitored in real time and accurately to achieve real-time suppression in different overcurrent scenarios.

[0046] Collected power data may contain noise, outliers, or discontinuous measurement points, all of which can affect the accuracy and stability of subsequent control strategies. Therefore, preprocessing of the power data is necessary. This includes steps such as data cleaning, data standardization, and feature extraction. This involves selecting and constructing, from the preprocessed power data, quantities that reflect the system's dynamic characteristics, such as voltage amplitude and phase, current magnitude and rate of change, and frequency deviation. These feature vectors are input into the virtual impedance coefficient optimization model to assess the system's current state and predict the optimal virtual resistance value to effectively suppress black start overcurrent.

[0047] Step S102: input the power characteristic vector into a virtual impedance coefficient optimization model, and output a target virtual impedance coefficient.

[0048] It should be noted that the present application proposes an adaptive virtual impedance, which can adjust the resistance of the virtual resistor according to the real-time operating status of the power system, and then adjust the current flowing through the electronic equipment. When the real-time power data of the power system is determined, the final virtual impedance value can be determined by determining the virtual impedance coefficient.

[0049] Specifically, the calculation formula of the adaptive virtual impedance is expressed as:

[0050]

[0051] Among them, k r (t) Virtual impedance coefficient, V PCC is the voltage amplitude of the common connection point of the switching load, E is the converter output voltage, I max is the maximum allowable current of the converter, which is set to 1.5 pu (per unit) of the converter rated current. I(t) is the current value of the overcurrent control system. I ref is the current reference value of the overcurrent control system. Ideally, the converter output voltage E and V PCC Roughly equal, that is, under stable operation, the virtual resistance has no effect.

[0052] Therefore, in the above step S102, the virtual impedance coefficient is dynamically optimized according to the operating characteristics of the grid-connected system through the virtual impedance coefficient optimization model to find the optimal virtual impedance coefficient. The virtual impedance coefficient is the basis for determining the virtual impedance value. The virtual impedance coefficient optimization model integrates the attractor trend strategy, the coupling interference strategy and the information projection strategy, and performs iterative calculations in a dynamic optimization manner. Through the iterative calculations of these strategies, the optimal virtual impedance coefficient can be efficiently explored and determined.

[0053] Among them, the coupled interference strategy introduces interactions between groups into the algorithm, simulating the coupling effects between neuronal groups. By adding random perturbations during the optimization process, this strategy can prompt the algorithm to escape the local optimal solution and explore a wider solution space, thereby increasing the possibility of finding the global optimal solution. This strategy is particularly suitable for the dynamic environment during the black start process, ensuring that the optimization process of the virtual impedance coefficient has sufficient exploration capabilities to cope with overcurrent challenges in different situations. The information projection strategy controls the information exchange between neuronal groups, realizing information transfer and knowledge sharing during the optimization process. By generating an adjacency matrix and a communication strength matrix, it determines which groups will transfer information and the weight of the transferred information. This strategy helps the algorithm find a balance between exploration and exploitation, maintaining the diversity of the optimization process while ensuring rapid convergence to the optimal solution, improving the stability and reliability of the optimization process.

[0054] After the power characteristic vector is input into the virtual impedance coefficient optimization model, the model dynamically adjusts the virtual impedance coefficient through iterative calculations integrating the three strategies described above. In each iteration, the model evaluates the suitability of the current virtual impedance coefficient for overcurrent suppression, uses an attractor trend strategy to approach the optimal solution, utilizes a coupled interference strategy to increase solution diversity, and optimizes communication between groups through an information projection strategy, ultimately outputting the target virtual impedance coefficient. This process is iteratively executed until the termination criteria are met, such as reaching the maximum number of iterations or no significant improvement in the optimization effect.

[0055] Through the dynamic optimization process of the virtual impedance coefficient optimization model, the optimal virtual impedance coefficient can be determined in real time and efficiently to set the optimal virtual impedance value. This virtual impedance effectively suppresses overcurrent during the black start process while maintaining the system's dynamic stability and energy efficiency. Compared to static or preset virtual impedance control strategies, this significantly improves the adaptability and control accuracy of the photovoltaic and energy storage grid-connected system during the black start phase, avoids the risk of overload of power electronic equipment, extends equipment life, and enhances the resilience and overall security of the grid.

[0056] Furthermore, the power characteristic vector is input into the virtual impedance coefficient optimization model, and the step of outputting the target virtual impedance coefficient includes: the virtual impedance coefficient optimization model generates a set of initial neuron populations based on the power characteristic vector, wherein the initial neuron population contains N neuron individuals, each neuron individual represents a virtual impedance coefficient, and N is a positive integer; the virtual impedance coefficient optimization model calculates the fitness value of each neuron individual in the initial neuron population based on the objective function, wherein the objective function is constructed based on overcurrent, frequency deviation and voltage harmonic distortion rate; the virtual impedance coefficient optimization model iteratively updates the initial neuron population based on the fitness value until the iteration number threshold is reached and the iteration is stopped; the virtual impedance coefficient optimization model selects the neuron individual with the largest fitness value as the target virtual impedance coefficient based on the updated neuron population.

[0057] Specifically, in order to achieve dynamic optimization of the virtual impedance coefficient, this application defines the optimization goal of the model, which is to minimize the overcurrent, frequency deviation, and voltage harmonic distortion rate of the photovoltaic storage grid-connected system, thereby obtaining the objective function, which can be expressed as:

[0058] f(x i )=α·max(I output -1.5,0)+β·|Δf|+γTHD s (Formula 2)

[0059] Among them, f(x i ) represents the objective function value, α, β, γ are weight coefficients, which penalize overcurrent, frequency deviation and voltage harmonic distortion rate respectively, max(I output -1.5,0) indicates system overcurrent, I output =(V inverter -V pcc ) / (Z L +Z V (t)), represents the current flowing through the inverter control circuit, V inverter is the virtual impedance voltage amplitude, V pcc is the voltage amplitude of the common connection point of the switching load, Z L is the load impedance value, Z V (t) is the virtual resistance value, |Δf| represents the frequency deviation of the photovoltaic and energy storage grid-connected system, and THD s is the voltage harmonic distortion rate of the photovoltaic energy storage grid-connected system.

[0060] Secondly, when the target virtual impedance value is output through the virtual impedance coefficient optimization model, the following steps are specifically included: Generation of initial neuron population: the virtual impedance coefficient optimization model first generates a set of initial neuron populations P = {x1, x2, x3, ..., x NThis population consists of N individual neurons, where N is a positive integer and each individual neuron represents a possible value of the virtual impedance coefficient Kr(t).

[0061] Specifically, each neuron can be calculated based on the following formula:

[0062] x i =l b +rand(0,1)×(u b -l b )(Formula 3)

[0063] Among them, x i represents the i-th neuron individual, l b Represents the lower bound, u b Represents the upper bound, rand(0, 1) is a random number between 0 and 1.

[0064] After constructing the initial neuron population, the fitness value of each individual neuron in the initial population is calculated based on the constructed objective function. This objective function comprehensively considers factors such as overcurrent, frequency deviation, and voltage harmonic distortion. This setting ensures that the optimization process not only focuses on overcurrent suppression but also considers system frequency stability and voltage quality, ensuring the overall excellent performance of the control strategy.

[0065] The virtual impedance coefficient optimization model then iteratively updates the initial neuron population based on the calculated fitness value (i.e., the objective function value). This process follows an attractor trend strategy, a coupled interference strategy, and an information projection strategy, dynamically adjusting and optimizing the states of individual neurons to gradually approach the optimal virtual impedance coefficient value. Iterative updates continue until a pre-set iteration threshold is reached, at which point the optimization process is considered to have converged, or if the fitness value no longer improves significantly after several consecutive iterations, the model automatically stops iterating.

[0066] After iterations cease, the virtual impedance coefficient optimization model selects the neuron with the highest fitness value from the final updated neuron population and determines the virtual impedance coefficient represented by it as the target virtual impedance coefficient. This coefficient is the optimal solution in the optimization process and performs best in suppressing overcurrent, maintaining system frequency stability, and optimizing voltage quality. It directly guides the virtual impedance control strategy during black start, achieving the most efficient and stable power system recovery.

[0067] Furthermore, the virtual impedance coefficient optimization model iteratively updates the initial neuron population based on the fitness value, including the following steps: step 1, the virtual impedance coefficient optimization model optimizes the initial neuron population based on the fitness value of the individual neuron and the attractor trend strategy to obtain an attractor-optimized neuron population; step 2, the virtual impedance coefficient optimization model optimizes the initial neuron population based on the coupling interference strategy to obtain a coupling-interference-optimized neuron population; step 3, the virtual impedance coefficient optimization model optimizes the neuron population based on the information projection strategy and in combination with the attractor-based neuron population and the coupling-interference-based neuron population to obtain an information-projection-optimized neuron population; step 4, the virtual impedance coefficient optimization model calculates the fitness value of each individual neuron in the information-projection-optimized neuron population through a pre-established objective function, and screens the information-projection-optimized neuron population according to the fitness value to obtain a screened neuron population, and mutates the screened neuron population through a mutation strategy to generate a new neuron population, and uses the new neuron population as the initial neuron population for the next round of iteration; repeating the above steps 1 to 4 to iteratively update the neuron population.

[0068] Specifically, the iterative update process is divided into the following steps: driving the neural population towards the optimal decision direction based on the attractor trend strategy, thereby optimizing the neuron population; secondly, using the coupled interference strategy to make the neural population deviate from the attractor, thereby optimizing the neuron population; and finally, controlling the communication between neural populations through the information projection strategy. The matrix representing the adjacency and communication strength is used to represent the communication subspace of the information projection between neural groups. Here, multiple neuron populations are optimized and updated to obtain new neuron populations. For the updated neuron population, the fitness is evaluated here through the objective function, and neuron individuals with larger fitness values ​​are selected to construct a new neuron population. Mutations are introduced into it to maintain the diversity of the population, thereby obtaining the initial neuron population for the next iteration and entering the next iteration.

[0069] By integrating attractor trend strategies, coupled interference strategies, and information projection strategies, supplemented by fitness-based screening and mutation mechanisms, a comprehensive and efficient dynamic optimization framework has been formed. This framework, capable of operating simultaneously across multiple dimensions, not only accelerates the optimization process and ensures the algorithm quickly converges to the optimal solution, but also effectively avoids the trap of local optimality by introducing randomness and diversity, significantly improving the model's ability to find the optimal virtual impedance coefficient.

[0070] Furthermore, the virtual impedance coefficient optimization model optimizes the initial neuron population based on the attractor trend strategy, and the steps of obtaining the attractor-based neuron population include: the virtual impedance coefficient optimization model sorts the neuron individuals according to the fitness values ​​of the neuron individuals to obtain a sorted list; calculates the number of attractors to be selected based on the attractor ratio and the total number of neuron individuals in the initial neuron population; selects attractors from the sorted list based on the number of attractors to obtain an attractor set; optimizes the initial neuron population based on the attractors in the attractor set to obtain a neuron population optimized based on the attractors, wherein, when optimizing the initial neuron population based on the attractors in the attractor set, the neuron individuals in the initial neuron population except the attractors are controlled to randomly converge to any attractor.

[0071] In some embodiments, in a photovoltaic grid-connected system, by applying the concept of neural population dynamics to the optimization process of the virtual impedance coefficient, the characteristics of multiple attractors can be used to improve the performance of the optimization algorithm. The attractor trend strategy ensures that the algorithm can develop in the direction of the optimal solution, and the characteristics of multiple attractors introduce a competition mechanism so that the algorithm will not converge to the local optimal solution too early during the exploration process, but can explore the solution space more comprehensively, thereby increasing the probability of finding the global optimal solution. In the process of optimization based on multiple attractors, the fitness values ​​of individual neurons are first sorted according to the numerical size, and then the number of attractors to be selected is determined. The number of attractors is determined based on the attractor ratio of the model parameters and the total number of individual neurons in the neuron population. The attractors with larger fitness values ​​are selected according to the number of attractors to obtain an attractor set, and then the other individual neurons are optimized based on the random attractors in the attractor set, so that the other individual neurons are close to the attractor (the current optimal solution) and converge to one of these attractors, thereby obtaining a neuron population optimized based on the attractor.

[0072] The algorithm for updating individual neurons based on the attractor trend strategy can be expressed as:

[0073] x attract,i =l·r1 2 ·(att k -x i )+w i ,w i ~N(0,UL) (Formula 4)

[0074] Among them, l represents the attractor ratio, r1 represents a random number between 0 and 1, and att k is the kth attractor (representing the first a*n neuron individuals with the best fitness in the current population), w i To simulate noise.

[0075] Furthermore, the virtual impedance coefficient optimization model optimizes the initial neuron population based on the coupling interference strategy, and the steps of obtaining the neuron population based on coupling interference include: calculating the average value of other neuron populations except the initial neuron population, and optimizing the initial neuron population according to the average value to obtain a first optimized neuron population; calculating the average value of the difference between the initial neuron population and other neuron populations, and optimizing the initial neuron population according to the average value of the difference to obtain a second optimized neuron population; generating a neuron population based on coupling interference optimization based on the first optimized neuron population, the second optimized neuron population and the global scaling factor.

[0076] In some embodiments, the process of optimizing the initial neuron population based on a coupled perturbation strategy aims to enhance the algorithm's global search capabilities and avoid premature convergence to local optima by introducing interactions between populations. Specifically, the virtual impedance coefficient optimization model first calculates the average neural state of all neuron populations other than the initial neuron population. This average represents the dynamic trend of the entire population and can be considered a global information indicator. The model then uses this average to perturb the state of the initial neuron population, prompting it to go beyond its current local environment and explore a wider solution space. After this perturbation, the state of the resulting first optimized neuron population is updated to reflect the influence of the population average dynamics, which helps improve the algorithm's global search capabilities. Next, the model calculates the average difference between the initial neuron population and the other neuron populations mentioned above and uses this difference to optimize the initial population. The average difference reveals the relative differences between individuals and the population. Using this difference in optimization can encourage individual neurons in the initial population to not only follow the population trend but also understand and adapt to differences between populations, increasing the diversity of solutions and reducing the likelihood of premature convergence, thereby obtaining the second optimized neuron population. Building on the previous two steps, the model further combines the first and second optimized neuron populations with a pre-set global scaling factor to generate the final neuron population optimized based on coupled interference. The global scaling factor regulates the coupling strength, affecting the amplitude and direction of the perturbation, ensuring that the perturbation is neither too intense to cause algorithm instability nor too weak to affect exploration. In this way, the model leverages coupled interference between populations to effectively balance exploration and exploitation, guiding the population's evolution towards a more optimal solution.

[0077] Specifically, the algorithm for optimizing the initial neuron population based on the coupled interference strategy can be expressed as:

[0078]

[0079] x couple,i =d(x add,i +x dif,i )(Formula 7)

[0080] Among them, r2 and r3 represent random numbers between 0 and 0.5, which control the additive coupling and diffusion coupling strength respectively, and x j is the individual neuron in other neuron populations, n is the number of individual neurons in the neuron population, d is the global scaling factor of the coupling interference, x add,i That is, the individual neurons in the first optimized neuron population, x dif,i Optimize individual neurons in the second neuron population.

[0081] Furthermore, the virtual impedance coefficient optimization model is based on the information projection strategy, and combines the attractor-based neuron population and the coupled interference-based neuron population to optimize the initial neuron population. The steps of obtaining the information projection-based neuron population include: generating an adjacency matrix and a communication intensity matrix, and performing secondary optimization on the attractor-optimized neuron population according to the adjacency matrix and the communication intensity matrix to obtain a third optimized neuron population; performing secondary optimization on the coupled interference-optimized neuron population according to the adjacency matrix and the communication intensity matrix and the ratio of the number of iterations to obtain a fourth optimized neuron population; generating an information projection-based neuron population based on the initial neuron population, the third optimized neuron population and the fourth optimized neuron population.

[0082] In some embodiments, an information projection strategy is used to effectively manage and optimize communications between populations, implement more sophisticated and effective control strategies, protect power electronic equipment, and ensure the stable operation of the power system. Specifically, first, the virtual impedance coefficient optimization model automatically generates an adjacency matrix and a communication intensity matrix. The adjacency matrix is ​​used to define which individuals between populations can exchange information, that is, which individuals' states can affect each other. The communication intensity matrix quantifies the degree of information exchange and determines the update speed and direction of individual states between populations. Based on these two matrices, the model can finely control the flow of information between neuron populations based on attractor optimization, ensuring that the information of individuals with higher fitness (attractors) can be effectively transmitted to other individuals, thereby obtaining a third optimized neuron population. This population not only inherits the excellent characteristics of the attractor, but also enhances the synergy and diversity within the population through the information projection strategy.

[0083] The model then performs a secondary optimization on the coupled-interference optimization-based neuron population. The adjacency matrix and communication strength matrix are still used to control information exchange, but the model also considers the ratio of the current iteration count to the maximum iteration count to dynamically adjust the intensity and method of the information projection strategy. In the early stages of the optimization process, the model may prioritize exploratory communication between populations to discover more potential high-quality solutions. In the later stages, the model tends to enhance synergy between populations to promote convergence and optimization. This process ultimately generates a fourth optimized neuron population, whose state reflects the combined effects of the coupled-interference and information projection strategies, as well as the dynamic adjustments of the optimization process. Finally, based on the combined application of the information projection strategy, the model fuses the information from the initial, third, and fourth optimized neuron populations to form a neuron population based on information projection. This combined optimized population combines the convergence advantages of the attractor trend strategy, the diversity advantages of the coupled-interference strategy, and the efficient information sharing of the information projection strategy, achieving a good balance between the algorithm's global search capability and rapid convergence.

[0084] The algorithm for optimizing the neuron population based on the information projection strategy can be expressed as:

[0085] x C_attract,i =C attract,i R attract,i x attract,i (Formula 8)

[0086]

[0087] x new,i =x i +x C_attract,i +x C_couple,i (Formula 10)

[0088] Among them, C attract,i is the adjacency matrix, which determines whether to transmit information of a specific dimension; R attract,i is the communication strength matrix, which controls the information transmission weight; FE is the number of current function evaluations (iterations); FE max Represents the iteration threshold (termination condition); r4 is a random number between 0 and 1.

[0089] Step S103 , substituting the power data and the target virtual impedance coefficient into the virtual impedance calculation formula, calculating the virtual impedance value, and adjusting the reference signal of the inverter control loop based on the virtual impedance value to suppress the overcurrent caused by the black start.

[0090] In the above step S103, the acquired power data (such as real-time monitoring data such as voltage and current) is integrated with the target virtual impedance coefficient obtained by optimizing the virtual impedance coefficient optimization model as the input parameter of the virtual impedance calculation formula. These power data reflect the real-time status of the photovoltaic grid-connected system during the black start process, and the target virtual impedance coefficient is optimized according to the real-time requirements and constraints of the system, and is used to guide the subsequent virtual impedance value calculation. The virtual impedance value that can suppress overcurrent under specific working conditions is calculated by the virtual impedance calculation formula. The calculation formula comprehensively considers the power system parameters, the target virtual impedance coefficient and the real-time working conditions of the system to ensure that the calculated virtual impedance value can accurately reflect the system requirements, thereby achieving effective suppression of overcurrent. Based on the calculated virtual impedance value, the reference signal in the inverter control loop is dynamically adjusted. This is achieved by modifying the current or voltage reference point of the inverter to ensure that the inverter can adjust its output according to the optimized virtual impedance parameters, thereby limiting overcurrent, protecting power electronic equipment, and maintaining stable operation of the system.

[0091] Furthermore, the virtual impedance calculation formula is expressed as: Among them, Z v (t) is the virtual impedance, E is the output voltage amplitude of the converter, V PCC is the voltage amplitude of the common connection point of the switching load, I max is the maximum allowable current of the converter, k r (t) is the virtual impedance coefficient, I(t) is the current value of the overcurrent control system, I ref It is the current reference value of the overcurrent control system.

[0092] In some embodiments, the fundamental principle behind adding virtual impedance is to simulate the physical inductive behavior of a synchronous generator by introducing a specific virtual impedance loop. When the system experiences a sudden change in current, the virtual impedance generates a response inversely proportional to the change in current. This response generates a voltage within the virtual impedance, thereby suppressing sudden current fluctuations. This method of adding virtual impedance can be used not only to suppress inrush currents during the startup of transformers and induction motors, but also to mitigate short-term overcurrents generated when connected to the power grid.

[0093] Due to the limitations of power electronic components, the injected current should be limited to within 1.5 pu of the rated current (the overcurrent capacity of a grid-type energy storage circulator can reach 3 pu) to prevent converter failure. When the reference voltage and reference current are inconsistent, using only the current limiter to limit the current reference value may cause system instability during and after load application. When a current limiting strategy is adopted when the load is applied, the terminal voltage of the converter cannot track the voltage reference, resulting in a larger current reference value. However, after the black start is completed, it must be ensured that the converter can operate in normal mode. Therefore, it is necessary to coordinate voltage and current control to ensure system stability.

[0094] The virtual impedance calculation formula combines the virtual impedance coefficient, the converter output voltage, the voltage at the common connection point of the switching load, the maximum allowable current of the converter, the current reference value of the overcurrent control system, and the current value of the overcurrent control system to calculate the virtual impedance to achieve adaptive adjustment of the virtual impedance. It can be adjusted in real time according to the actual operating status of the grid-connected system to suppress overcurrent phenomena.

[0095] Through the above steps, the power data of the photovoltaic storage grid-connected system is collected, and the power data is preprocessed to extract the power characteristic vector, wherein the power data includes at least voltage data, current data and frequency data. Then the power characteristic vector is input into the virtual impedance coefficient optimization model, and the target virtual impedance coefficient is output. The virtual impedance coefficient optimization model integrates the attractor trend strategy, the coupling interference strategy and the information projection strategy, and performs iterative calculation in a dynamic optimization manner to obtain the target virtual impedance coefficient. Finally, the power data and the target virtual impedance coefficient are substituted into the virtual impedance calculation formula to calculate the virtual impedance value, and the reference signal of the inverter control loop is adjusted based on the virtual impedance value to suppress the overcurrent caused by the black start.

[0096] In this embodiment, the virtual impedance coefficient is optimally calculated based on the actual situation of grid-connected operation through a pre-constructed virtual impedance coefficient optimization model, so as to provide the optimal virtual impedance parameters in the current state according to the real-time operating state of the power grid system, so as to suppress the overcurrent generated by the black start, thereby achieving the purpose of dynamically optimizing the virtual resistance parameters and obtaining the technical effect of improving the stability of overcurrent suppression, thereby solving the technical problem in the related art of low overcurrent suppression stability when suppressing overcurrent based on a fixed virtual impedance control strategy.

[0097] The following describes in detail another optional specific implementation.

[0098] Iterative optimization of virtual impedance is the core of the embodiment of the present invention. Figure 2 is a schematic diagram of an optional virtual impedance coefficient optimization process according to an embodiment of the present invention, such as Figure 2 As shown in Figure 2, the virtual impedance coefficient optimization process includes:

[0099] Step 1, start;

[0100] Step 2: Determine the objective function;

[0101] Step 3: Construct the initial neuron population;

[0102] Step 4: Determine whether the current number of iterations is less than the maximum number of iterations. If so, proceed to step 5; if not, proceed to step 16.

[0103] Step 5: Perform fitness evaluation and calculate the fitness value of individual neurons in the neuron population through the objective function;

[0104] Step 6: sort individual neurons;

[0105] Step 7: Select an attractor and set i = a*n+1;

[0106] Step 8: Determine whether i is less than n. If yes, go to step 9; if not, go to step 10.

[0107] Step 9: Randomly select an attractor, substitute it into Formula 4 to calculate, optimize the neuron population, and set i = i + 1, and repeat steps 8 to 9;

[0108] Step 10, let i=1;

[0109] Step 11, determine whether i is less than n, if so, go to step 12, if not, go to step 13;

[0110] Step 12: Optimize the neuron population by performing calculations using formulas 5, 6, and 7, and set i = i + 1, and repeat steps 11 to 12.

[0111] Step 13, let i=1;

[0112] Step 14: determine whether i is less than n. If so, proceed to step 15. If not, return to step 4 and repeat steps 4 to 14.

[0113] Step 15: Optimize the neuron population by performing calculations using formulas 8, 9, and 10, and set i = i + 1, and repeat steps 14 to 15.

[0114] Step 16: Obtain the optimal neuron individual and determine the optimal virtual impedance coefficient;

[0115] Step 17, end.

[0116] In an embodiment of the present invention, a black start overcurrent suppression strategy based on an adaptive virtual resistor and a neural population dynamic optimization algorithm is proposed. By introducing an adaptive virtual resistor, the overcurrent phenomenon generated when special loads such as transformers are connected during the black start process is suppressed, thereby extending the service life of the converter and enhancing the intelligent characteristics of the photovoltaic storage system.

[0117] The following describes it in detail with reference to another embodiment.

[0118] Example 2

[0119] A black start overcurrent suppression device provided in this embodiment includes multiple implementation units, each implementation unit corresponds to each implementation step in the above-mentioned embodiment 1. Its specific implementation methods and beneficial effects can be referred to the above-mentioned method embodiments and will not be repeated here.

[0120] Figure 3 FIG. 1 is a schematic diagram of an optional black start overcurrent suppression device according to an embodiment of the present invention. Figure 3 As shown, the black start overcurrent suppression device may include: a collection unit 31, an output unit 32, and a suppression unit 33, wherein:

[0121] The acquisition unit 31 is used to collect power data of the photovoltaic energy storage grid-connected system, pre-process the power data, and extract the power feature vector, wherein the power data includes at least voltage data, current data, and frequency data;

[0122] An output unit 32 is configured to input the power characteristic vector into a virtual impedance coefficient optimization model and output a target virtual impedance coefficient, wherein the virtual impedance coefficient optimization model integrates an attractor trend strategy, a coupled interference strategy, and an information projection strategy, and performs iterative calculations in a dynamic optimization manner to obtain the target virtual impedance coefficient;

[0123] The suppression unit 33 is used to substitute the power data and the target virtual impedance coefficient into the virtual impedance calculation formula, calculate the virtual impedance value, and adjust the reference signal of the inverter control loop based on the virtual impedance value to suppress the overcurrent caused by the black start.

[0124] The above-mentioned overcurrent suppression device for black start collects power data of the photovoltaic storage grid-connected system through the collection unit 31, preprocesses the power data, and extracts the power characteristic vector, wherein the power data at least includes voltage data, current data and frequency data; inputs the power characteristic vector into the virtual impedance coefficient optimization model through the output unit 32, and outputs the target virtual impedance coefficient, wherein the virtual impedance coefficient optimization model integrates the attractor trend strategy, the coupling interference strategy and the information projection strategy, and performs iterative calculation in a dynamic optimization manner to obtain the target virtual impedance coefficient; substitutes the power data and the target virtual impedance coefficient into the virtual impedance calculation formula through the suppression unit 33, calculates the virtual impedance value, and adjusts the reference signal of the inverter control loop based on the virtual impedance value to suppress the overcurrent caused by the black start.

[0125] In this embodiment, the virtual impedance coefficient is optimally calculated based on the actual situation of grid-connected operation through a pre-constructed virtual impedance coefficient optimization model, so as to provide the optimal virtual impedance parameters in the current state according to the real-time operating state of the power grid system, so as to suppress the overcurrent generated by the black start, thereby achieving the purpose of dynamically optimizing the virtual resistance parameters and obtaining the technical effect of improving the stability of overcurrent suppression, thereby solving the technical problem in the related art of low overcurrent suppression stability when suppressing overcurrent based on a fixed virtual impedance control strategy.

[0126] Furthermore, the output unit includes: a first generating subunit, which is used for the virtual impedance coefficient optimization model to generate a group of initial neuron populations based on the power characteristic vector, wherein the initial neuron population contains N neuron individuals, each neuron individual represents a virtual impedance coefficient, and N is a positive integer; a first calculating subunit, which is used for the virtual impedance coefficient optimization model to calculate the fitness value of each neuron individual in the initial neuron population based on the objective function, wherein the objective function is constructed based on overcurrent, frequency deviation and voltage harmonic distortion rate; a first iterative subunit, which is used for the virtual impedance coefficient optimization model to iteratively update the initial neuron population based on the fitness value until the iteration number threshold is reached and the iteration is stopped; a first selecting subunit, which is used for the virtual impedance coefficient optimization model to select the neuron individual with the largest fitness value as the target virtual impedance coefficient based on the updated neuron population.

[0127] Furthermore, the first iterative subunit includes: a first optimization module, which is used for step one, the virtual impedance coefficient optimization model optimizes the initial neuron population based on the fitness value of the neuron individual and the attractor trend strategy to obtain the neuron population based on the attractor optimization; a second optimization module, which is used for step two, the virtual impedance coefficient optimization model optimizes the initial neuron population based on the coupling interference strategy to obtain the neuron population based on the coupling interference optimization; a third optimization module, which is used for step three, the virtual impedance coefficient optimization model optimizes the neuron population based on the information projection strategy and combines the neuron population based on the attractor and the neuron population based on the coupling interference Optimize and obtain a neuron population optimized based on information projection; the first generation module is used for step four, the virtual impedance coefficient optimization model calculates the fitness value of each individual neuron in the neuron population optimized based on information projection through a pre-established objective function, and screens the neuron population optimized based on information projection according to the fitness value to obtain a screened neuron population, and mutates the screened neuron population through a mutation strategy to generate a new neuron population, and uses the new neuron population as the initial neuron population for the next round of iteration; the first repetition module is used to repeat steps one to four above to iteratively update the neuron population.

[0128] Furthermore, the first optimization module includes: a first sorting submodule, which is used to sort the neuron individuals according to the fitness values ​​of the neuron individuals in the virtual impedance coefficient optimization model to obtain a sorted list; a first calculation submodule, which is used to calculate the number of attractors to be selected based on the attractor ratio and the total number of neuron individuals in the initial neuron population; a first selection submodule, which is used to select attractors from the sorted list based on the number of attractors to obtain an attractor set; a first optimization submodule, which is used to optimize the initial neuron population based on the attractors in the attractor set to obtain a neuron population optimized based on the attractors, wherein, when optimizing the initial neuron population based on the attractors in the attractor set, the neuron individuals other than the attractors in the initial neuron population are controlled to randomly converge to any attractor.

[0129] Furthermore, the second optimization module includes: a second calculation submodule, used to calculate the average value of other neuron populations except the initial neuron population, and optimize the initial neuron population based on the average value to obtain a first optimized neuron population; a third calculation submodule, used to calculate the difference average value between the initial neuron population and other neuron populations, and optimize the initial neuron population based on the difference average value to obtain a second optimized neuron population; a first generation submodule, used to generate a neuron population based on coupled interference optimization based on the first optimized neuron population, the second optimized neuron population and the global scaling factor.

[0130] Furthermore, the third optimization module includes: a second generation submodule, used to generate an adjacency matrix and a communication intensity matrix, and perform secondary optimization on the neuron population based on attractor optimization according to the adjacency matrix and the communication intensity matrix to obtain a third optimized neuron population; a second optimization submodule, used to perform secondary optimization on the neuron population based on coupling interference optimization according to the adjacency matrix and the communication intensity matrix and the ratio of the number of iterations to obtain a fourth optimized neuron population; a third generation submodule, used to generate a neuron population based on information projection based on the initial neuron population, the third optimized neuron population and the fourth optimized neuron population.

[0131] Furthermore, the virtual impedance calculation formula is expressed as: Among them, Z v (t) is the virtual impedance, E is the output voltage amplitude of the converter, V PCC is the voltage amplitude of the common connection point of the switching load, I max is the maximum allowable current of the converter, k r (t) is the virtual impedance coefficient, I(t) is the current value of the overcurrent control system, I ref It is the current reference value of the overcurrent control system.

[0132] The above-mentioned black start overcurrent suppression device may further include a processor and a memory. The above-mentioned receiving unit 20, search unit 21, comparison unit 22, sending unit 23, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to implement corresponding functions.

[0133] The processor includes a core, which retrieves the corresponding program unit from the memory. One or more cores can be provided, and the overcurrent generated by the black start can be suppressed by adjusting the core parameters.

[0134] The above-mentioned memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0135] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is further provided, the computer-readable storage medium including a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any of the above-mentioned black start overcurrent suppression methods.

[0136] According to another aspect of an embodiment of the present invention, an electronic device is also provided, including one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by one or more processors, the one or more processors implement any of the above-mentioned black start overcurrent suppression methods.

[0137] According to another aspect of an embodiment of the present invention, a computer program product is provided. The computer program product includes a computer program, wherein when the computer program is executed by a processor, any of the above-mentioned black start overcurrent suppression methods is implemented.

[0138] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program initialized with the following method steps: collecting power data of a photovoltaic energy storage grid-connected system, preprocessing the power data, and extracting a power characteristic vector, wherein the power data includes at least voltage data, current data, and frequency data; inputting the power characteristic vector into a virtual impedance coefficient optimization model, and outputting a target virtual impedance coefficient, wherein the virtual impedance coefficient optimization model integrates an attractor trend strategy, a coupling interference strategy, and an information projection strategy, and performs iterative calculations in a dynamic optimization manner to obtain a target virtual impedance coefficient; substituting the power data and the target virtual impedance coefficient into a virtual impedance calculation formula, calculating the virtual impedance value, and adjusting the reference signal of the inverter control loop based on the virtual impedance value to suppress overcurrent caused by black start.

[0139] The present application also provides a computer program product, which, when executed on a data processing device, is also suitable for executing a program initialized with the following method steps: inputting the power characteristic vector into the virtual impedance coefficient optimization model, and outputting the target virtual impedance coefficient includes: the virtual impedance coefficient optimization model generates a set of initial neuron populations based on the power characteristic vector, wherein the initial neuron population contains N neuron individuals, each neuron individual represents a virtual impedance coefficient, and N is a positive integer; the virtual impedance coefficient optimization model calculates the fitness value of each neuron individual in the initial neuron population based on the objective function, wherein the objective function is constructed based on overcurrent, frequency deviation and voltage harmonic distortion rate; the virtual impedance coefficient optimization model iteratively updates the initial neuron population based on the fitness value until the iteration number threshold is reached and the iteration is stopped; the virtual impedance coefficient optimization model selects the neuron individual with the largest fitness value as the target virtual impedance coefficient based on the updated neuron population.

[0140] The present application also provides a computer program product, which, when executed on a data processing device, is also suitable for executing a program for initializing the following method steps: the steps of iteratively updating the initial neuron population based on the fitness value of the virtual impedance coefficient optimization model include: step 1, the virtual impedance coefficient optimization model optimizes the initial neuron population based on the fitness value of the individual neurons and the attractor trend strategy to obtain a neuron population based on attractor optimization; step 2, the virtual impedance coefficient optimization model optimizes the initial neuron population based on the coupling interference strategy to obtain a neuron population based on coupling interference optimization; step 3, the virtual impedance coefficient optimization model optimizes the initial neuron population based on the information projection strategy and combines the basic The neuron population based on the attractor and the neuron population based on coupled interference are optimized to obtain the neuron population optimized based on information projection; in step four, the virtual impedance coefficient optimization model calculates the fitness value of each individual neuron in the neuron population based on information projection through a pre-established objective function, and screens the neuron population optimized based on information projection according to the fitness value to obtain the screened neuron population, and mutates the screened neuron population through a mutation strategy to generate a new neuron population, and uses the new neuron population as the initial neuron population for the next round of iteration; repeat steps one to four to iteratively update the neuron population.

[0141] The present application also provides a computer program product, which, when executed on a data processing device, is also suitable for executing a program initialized with the following method steps: a virtual impedance coefficient optimization model optimizes an initial neuron population based on an attractor trend strategy, and the steps of obtaining an attractor-based neuron population include: the virtual impedance coefficient optimization model sorts the neuron individuals according to their fitness values ​​to obtain a sorted list; the number of attractors to be selected is calculated based on the attractor ratio and the total number of neuron individuals in the initial neuron population; attractors are selected from the sorted list based on the number of attractors to obtain an attractor set; the initial neuron population is optimized based on the attractors in the attractor set to obtain an attractor-optimized neuron population, wherein, when the initial neuron population is optimized based on the attractors in the attractor set, the neuron individuals in the initial neuron population other than the attractors are controlled to randomly converge to any attractor.

[0142] The present application also provides a computer program product, which, when executed on a data processing device, is also suitable for executing a program initialized with the following method steps: a virtual impedance coefficient optimization model optimizes the initial neuron population based on a coupling interference strategy, and the steps of obtaining a neuron population based on coupling interference include: calculating the average value of other neuron populations except the initial neuron population, and optimizing the initial neuron population based on the average value to obtain a first optimized neuron population; calculating the average value of the difference between the initial neuron population and other neuron populations, and optimizing the initial neuron population based on the average value of the difference to obtain a second optimized neuron population; generating a neuron population based on coupling interference optimization based on the first optimized neuron population, the second optimized neuron population and the global scaling factor.

[0143] The present application also provides a computer program product, which, when executed on a data processing device, is also suitable for executing a program initialized with the following method steps: the virtual impedance coefficient optimization model is based on the information projection strategy, and the initial neuron population is optimized in combination with the attractor-based neuron population and the coupled interference-based neuron population, and the steps of obtaining the neuron population based on information projection include: generating an adjacency matrix and a communication intensity matrix, and performing secondary optimization on the attractor-optimized neuron population according to the adjacency matrix and the communication intensity matrix to obtain a third optimized neuron population; performing secondary optimization on the coupled interference-optimized neuron population according to the adjacency matrix and the communication intensity matrix and the ratio of the number of iterations to obtain a fourth optimized neuron population; generating a neuron population based on information projection based on the initial neuron population, the third optimized neuron population and the fourth optimized neuron population.

[0144] The present application also provides a computer program product, which, when executed on a data processing device, is further adapted to execute a program for initializing the following method steps: the virtual impedance calculation formula is expressed as: Among them, Z v (t) is the virtual impedance, E is the output voltage amplitude of the converter, V PCC is the voltage amplitude of the common connection point of the switching load, I max is the maximum allowable current of the converter, k r (t) is the virtual impedance coefficient, I(t) is the current value of the overcurrent control system, I ref It is the current reference value of the overcurrent control system.

[0145] Figure 4 FIG. 1 is a hardware structure block diagram of an electronic device (or mobile device) that performs an overcurrent suppression method for a black start according to an embodiment of the present invention. Figure 4 As shown, the electronic device may include one or more processors ( Figure 4402a, 402b, ..., 402n are used to illustrate that the processor may include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 404 for storing data. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, a keyboard, a power supply and / or a camera. It will be understood by those skilled in the art that Figure 4 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 4 More or fewer components than shown, or with Figure 4 Different configurations shown.

[0146] The serial numbers of the above embodiments of the present invention are for description only and do not represent the advantages or disadvantages of the embodiments.

[0147] In the above embodiments of the present invention, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0148] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only exemplary. For example, the division of the units can be a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0149] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0150] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0151] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, etc. Various media that can store program codes.

[0152] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A black start overcurrent suppression method, characterized in that: include: Collecting power data of the photovoltaic energy storage grid-connected system, preprocessing the power data, and extracting power feature vectors, wherein the power data includes at least voltage data, current data, and frequency data; Inputting the power characteristic vector into a virtual impedance coefficient optimization model and outputting a target virtual impedance coefficient, wherein the virtual impedance coefficient optimization model integrates an attractor trend strategy, a coupling interference strategy, and an information projection strategy, and performs iterative calculations in a dynamic optimization manner to obtain the target virtual impedance coefficient; The power data and the target virtual impedance coefficient are substituted into a virtual impedance calculation formula to calculate a virtual impedance value, and a reference signal of an inverter control loop is adjusted based on the virtual impedance value to suppress overcurrent caused by black start.

2. The method according to claim 1, characterized in that The step of inputting the power characteristic vector into a virtual impedance coefficient optimization model and outputting a target virtual impedance coefficient includes: The virtual impedance coefficient optimization model generates a set of initial neuron populations based on the power characteristic vector, wherein the initial neuron population includes N neuron individuals, each neuron individual represents a virtual impedance coefficient, and N is a positive integer; The virtual impedance coefficient optimization model calculates the fitness value of each individual neuron in the initial neuron population based on an objective function, wherein the objective function is constructed based on overcurrent, frequency deviation and voltage harmonic distortion rate; The virtual impedance coefficient optimization model iteratively updates the initial neuron population based on the fitness value until a threshold number of iterations is reached and the iteration is stopped; The virtual impedance coefficient optimization model selects the neuron individual with the largest fitness value as the target virtual impedance coefficient based on the updated neuron population.

3. The method according to claim 2, characterized in that The step of iteratively updating the initial neuron population based on the fitness value by the virtual impedance coefficient optimization model includes: Step 1: The virtual impedance coefficient optimization model optimizes the initial neuron population based on the fitness value of the individual neurons and the attractor trend strategy to obtain a neuron population optimized based on the attractor; Step 2: The virtual impedance coefficient optimization model optimizes the initial neuron population based on a coupling interference strategy to obtain a neuron population optimized based on coupling interference; Step 3: The virtual impedance coefficient optimization model is based on the information projection strategy and is combined with the attractor-based neuron population and the coupled interference-based neuron population to optimize the neuron population to obtain a neuron population optimized based on information projection; Step 4: The virtual impedance coefficient optimization model calculates the fitness value of each individual neuron in the information projection-based neuron population through a pre-established objective function, and screens the neuron population optimized based on information projection according to the fitness value to obtain the screened neuron population, and performs a mutation operation on the screened neuron population through a mutation strategy to generate a new neuron population, and uses the new neuron population as the initial neuron population for the next round of iteration; Repeat steps 1 to 4 above to iteratively update the neuron population.

4. The method according to claim 3, characterized in that The virtual impedance coefficient optimization model optimizes the initial neuron population based on the attractor trend strategy, and the step of obtaining the attractor-based neuron population includes: The virtual impedance coefficient optimization model sorts the individual neurons according to the fitness values ​​of the individual neurons to obtain a sorted list; Calculating the number of attractors to be selected based on the attractor ratio and the total number of individual neurons in the initial neuron population; Selecting attractors from the sorted list based on the number of attractors to obtain an attractor set; The initial neuron population is optimized based on the attractors in the attractor set to obtain a neuron population optimized based on the attractors, wherein when the initial neuron population is optimized based on the attractors in the attractor set, the neuron individuals in the initial neuron population except the attractors are controlled to randomly converge to any of the attractors.

5. The method according to claim 3, characterized in that The virtual impedance coefficient optimization model optimizes the initial neuron population based on the coupling interference strategy to obtain the neuron population based on coupling interference, including the following steps: Calculating an average value of other neuron populations except the initial neuron population, and optimizing the initial neuron population according to the average value to obtain a first optimized neuron population; Calculating an average difference between the initial neuron population and the other neuron populations, and optimizing the initial neuron population according to the average difference to obtain a second optimized neuron population; A neuron population based on coupled interference optimization is generated based on the first optimized neuron population, the second optimized neuron population and a global scaling factor.

6. The method according to claim 3, characterized in that The virtual impedance coefficient optimization model is based on an information projection strategy and optimizes the initial neuron population by combining an attractor-based neuron population and a coupled interference-based neuron population. The steps of obtaining the information projection-based neuron population include: generating an adjacency matrix and a communication intensity matrix, and performing secondary optimization on the attractor-optimized neuron population according to the adjacency matrix and the communication intensity matrix to obtain a third optimized neuron population; performing secondary optimization on the neuron population based on coupled interference optimization according to the adjacency matrix, the communication strength matrix, and the ratio of the number of iterations to obtain a fourth optimized neuron population; A neuron population based on information projection is generated based on the initial neuron population, the third optimized neuron population, and the fourth optimized neuron population.

7. The method according to claim 1, characterized in that The virtual impedance calculation formula is expressed as: Among them, Z v (t) is the virtual impedance, E is the output voltage amplitude of the converter, V PCC is the voltage amplitude of the common connection point of the switching load, I max is the maximum allowable current of the converter, k r (t) is the virtual impedance coefficient, I(t) is the current value of the overcurrent control system, I ref It is the current reference value of the overcurrent control system.

8. A black start overcurrent suppression device, characterized in that: include: A collection unit, configured to collect power data of the photovoltaic energy storage grid-connected system, pre-process the power data, and extract power feature vectors, wherein the power data includes at least voltage data, current data, and frequency data; an output unit, configured to input the power characteristic vector into a virtual impedance coefficient optimization model and output a target virtual impedance coefficient, wherein the virtual impedance coefficient optimization model integrates an attractor trend strategy, a coupling interference strategy, and an information projection strategy, and performs iterative calculation in a dynamic optimization manner to obtain the target virtual impedance coefficient; A suppression unit is used to substitute the power data and the target virtual impedance coefficient into a virtual impedance calculation formula, calculate a virtual impedance value, and adjust a reference signal of an inverter control loop based on the virtual impedance value to suppress overcurrent caused by black start.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the black start overcurrent suppression method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: The device comprises one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the black start overcurrent suppression method according to any one of claims 1 to 7.

Images