Simulation system and simulation method of photovoltaic station

Abstract

The embodiment of the invention discloses a simulation system and simulation method for a photovoltaic station, and the method comprises the steps: employing a filter capacitor at an output side of a grid-connected unit as a decoupling point of a large inertia link, and deploying a detailed model corresponding to each decoupled grid-connected unit to each first FPGA in a multi-FPGA structure in master-slave connection; and deploying the convergence model corresponding to the decoupled transformer to a second FPGA in the plurality of FPGA structures in master-slave connection. A plurality of FPGAs are adopted to respectively undertake simulation of each power electronic device in the photovoltaic station, so that small-step simulation of the photovoltaic station can be realized. The plurality of FPGAs adopt a master-slave connection structure, so that signal interaction among the FPGAs can be reduced, the problem that the number of IOs in a simulation system is large is solved, and expansion of the simulation system is facilitated. When the capacity of the photovoltaic station changes, a corresponding number of FPGAs can be connected into or removed from the simulation system, so that the simulation system has good expandability.

Description

Technical Field

[0001] This invention relates to the field of new energy technology, and in particular to a simulation system and simulation method for photovoltaic power plants. Background Technology

[0002] As the capacity of new energy equipment increases and new energy systems become more complex, the requirements of the power grid for new energy equipment become higher and higher. From product development, testing, certification to actual project implementation, power conversion equipment applied to new energy systems cannot do without real-time simulation technology.

[0003] Most existing real-time simulation devices possess hardware-in-the-loop (HIL) simulation capabilities for single devices, but their system-level simulation capabilities for large-scale photovoltaic (PV) power plants are relatively insufficient. The numerous parallel devices within PV power plants place higher demands on the I / O resources for external information interaction of the simulation system. Furthermore, PV power plants contain a large number of DC-DC converters, inverters, and other power electronic devices. Due to the high switching frequencies of these devices, real-time simulation requires simulation steps at the microsecond level or lower, necessitating that HIL simulations for PV power plants simultaneously meet the demands of large scale and small step size. Existing real-time simulation systems have significant limitations in their simulation capabilities, exhibiting scalability differences and poor system adaptability for PV power plants of varying sizes. This is particularly true for distributed PV power plant simulations, where the large number of deployed devices results in large-scale simulations, making it difficult for existing simulation systems to be expanded using independent computing units. Therefore, how to utilize distributed computing hardware to support large systems and achieve synchronous collaborative simulation, while fully and efficiently coordinating distributed computing power, is a key challenge in PV power plant simulation. Summary of the Invention

[0004] This application provides a simulation system and method for photovoltaic power plants, which enables small-step simulation and simulation scalability.

[0005] Firstly, this application provides a simulation system for a photovoltaic (PV) power plant. The PV power plant includes multiple grid-connected units and a transformer. Each grid-connected unit includes PV modules, a power converter, and a filter. The output terminal of the PV module is connected to the input terminal of the power converter, and the output terminal of the power converter is connected to the input terminal of the filter. The output terminals of multiple filters are connected in parallel and then connected to the transformer. The filter includes a filter capacitor. The transformer is used for connection to the power grid. The simulation system includes multiple first field-programmable gate arrays (FPGAs) and second FPGAs. The number of first FPGAs is consistent with the number of grid-connected units. Each first FPGA is used to simulate the detailed model corresponding to each deployed grid-connected unit in parallel. The second FPGA is used to simulate the convergence model corresponding to the deployed transformer. The detailed model corresponding to each grid-connected unit and the convergence model corresponding to the transformer are decoupled at the filter capacitors. Each filter capacitor is equivalent to an AC voltage source connected to the convergence model corresponding to the transformer, and each filter capacitor is equivalent to a first current source connected to the detailed model corresponding to the grid-connected unit.

[0006] In this application, the filter capacitor on the output side of the grid-connected unit is used as a decoupling point for large inertia links. The detailed models corresponding to each decoupled grid-connected unit are deployed to the first FPGA (slave FPGA) in a master-slave connected multi-FPGA structure. The convergence model corresponding to the decoupled transformer is deployed to the second FPGA (master FPGA) in a master-slave connected multi-FPGA structure to construct the simulation model corresponding to the photovoltaic power station. Multiple FPGAs are used to simulate each power electronic device in the photovoltaic power station, and small-step simulation of each power electronic device in the photovoltaic power station can be realized in each FPGA. Furthermore, the multiple FPGAs adopt a master-slave connection structure, and the first FPGAs are set in parallel without communication. Only the first FPGAs and the second FPGAs communicate with each other, which can reduce the signal interaction between FPGAs and solve the problem of a large number of I / O in the simulation system. The simulation system architecture using multiple FPGAs connected in a master-slave manner results in less data interaction between the second FPGA (master FPGA) and the multiple first FPGAs (slave FPGAs), making it easy to expand the simulation system. When the capacity of a photovoltaic power station changes, i.e., when the number of grid-connected units in the photovoltaic power station is increased or decreased, the simulation system can be adapted to the photovoltaic power station by adding or removing the corresponding number of first FPGAs (i.e., slave FPGAs) in the simulation system, thus enabling the simulation system to have good scalability.

[0007] In some embodiments of this application, each first FPGA is configured to: simulate a detailed model corresponding to the grid-connected unit based on the current signals of the AC voltage sources received from the second FPGA, simulate and obtain voltage signals of the first current sources, and then send the generated voltage signals of the first current sources to the second FPGA. Correspondingly, the second FPGA is configured to: simulate a convergence model corresponding to the transformer based on the voltage signals of the first current sources received from each first FPGA, simulate and obtain current signals of each AC voltage source, and then send the corresponding current signals of each AC voltage source to each first FPGA.

[0008] In some embodiments of this application, each first FPGA is specifically used to: simulate the detailed model corresponding to the grid-connected unit based on the current signals of the AC voltage sources sent by the second FPGA in the previous simulation cycle, and send the voltage signals of the first current sources generated in the current simulation cycle to the second FPGA. Correspondingly, the second FPGA is specifically used to: simulate the convergence model corresponding to the transformer based on the voltage signals of the first current sources sent by each first FPGA in the previous simulation cycle, and send the current signals of each AC voltage source generated in the current simulation cycle to each first FPGA.

[0009] The first FPGA uses the current signal from the previous simulation cycle to calculate the voltage signal for the current simulation cycle, and the second FPGA uses the voltage signal from the previous simulation cycle to calculate the current signal. This allows the first and second FPGAs to perform simulation calculations independently and in parallel, thereby significantly reducing the computational complexity of the simulation system, reducing the simulation step size, and implementing simulation calculations in a simpler way.

[0010] In some embodiments of this application, the power converter specifically includes a DC-DC converter and an inverter; the output terminal of the photovoltaic module is connected to the input terminal of the DC-DC converter, the output terminal of the DC-DC converter is connected to the input terminal of the inverter via a DC capacitor, and the output terminal of the inverter is connected to the input terminal of a filter. Decoupling is achieved at the DC capacitor, dividing the equivalent model corresponding to the power converter into an equivalent model corresponding to the DC-DC converter and an equivalent model corresponding to the inverter. The DC capacitor is equivalent to a second current source connected to the equivalent model corresponding to the DC-DC converter, and also equivalent to a DC voltage source connected to the equivalent model corresponding to the inverter. Each first FPGA simulates the deployed equivalent models corresponding to the DC-DC converter and the inverter, respectively. Each first FPGA is used to simulate the deployed equivalent models corresponding to the DC-DC converter and the inverter, respectively. The equivalent models corresponding to the DC-DC converter and the inverter are decoupled at the DC capacitor, where the DC capacitor is equivalent to a second current source connected to the equivalent model corresponding to the DC-DC converter, and also equivalent to a DC voltage source connected to the equivalent model corresponding to the inverter.

[0011] Utilizing the gradual voltage variation and inertial characteristics of the DC capacitor, the DC capacitor is selected as the decoupling point for the detailed model corresponding to the grid-connected unit, and a decoupling method is used to segment the detailed model. By decomposing the simulation model at the module level, the equivalent models corresponding to the segmented DC-DC converter and the equivalent models corresponding to the inverter can be simulated and calculated in parallel, thereby reducing the simulation step size.

[0012] In some embodiments of this application, each first FPGA is used to: simulate the equivalent model corresponding to the DC converter based on the current signal of the DC voltage source generated by the equivalent model corresponding to the inverter, and generate the voltage signal of the second current source. Accordingly, each first FPGA simulates the equivalent model corresponding to the inverter based on the voltage signal of the second current source generated by the equivalent model corresponding to the DC converter and the current signal of the AC voltage source sent by the second FPGA, and respectively generates the current signal of the DC voltage source and the voltage signal of the first current source sent to the second FPGA.

[0013] In some embodiments of this application, each first FPGA is specifically used to simulate the equivalent model corresponding to the DC converter based on the current signal of the DC voltage source generated by the equivalent model corresponding to the inverter in the previous simulation cycle, and generate the voltage signal of the second current source in the current simulation cycle. Correspondingly, each first FPGA is specifically used to simulate the equivalent model corresponding to the inverter based on the voltage signal of the second current source generated by the equivalent model corresponding to the DC converter in the previous simulation cycle and the current signal of the AC voltage source sent by the second FPGA in the previous simulation cycle, respectively generating the current signal of the DC voltage source in the current simulation cycle and sending the voltage signal of the first current source generated in the current simulation cycle to the second FPGA.

[0014] In the first FPGA, the equivalent model corresponding to the DC-DC converter uses the current signal from the previous simulation cycle to calculate the voltage signal for the current simulation cycle, and the equivalent model corresponding to the inverter uses the current and voltage signals from the previous simulation cycle to calculate the current and voltage signals for the current simulation cycle. This allows the two equivalent models to perform simulation calculations independently and in parallel, thereby significantly reducing the computational complexity of the simulation system, reducing the simulation step size, and implementing a simpler simulation calculation method.

[0015] In some embodiments of this application, each first FPGA is further configured to: simulate the equivalent model corresponding to the deployed photovoltaic module based on the terminal voltage signal generated by the equivalent model corresponding to the DC-DC converter, and send the generated current source signal to the equivalent model corresponding to the DC-DC converter. Accordingly, each first FPGA simulates the equivalent model corresponding to the DC-DC converter based on the current source signal generated by the equivalent model corresponding to the photovoltaic module and the current signal of the DC voltage source generated by the equivalent model corresponding to the inverter, and generates the terminal voltage signal and the voltage signal of the second current source.

[0016] In some embodiments of this application, each first FPGA is specifically used to simulate the equivalent model corresponding to the deployed photovoltaic module based on the terminal voltage signal generated by the equivalent model corresponding to the DC-DC converter in the previous simulation cycle, and to send the current source signal generated in the current simulation cycle to the equivalent model corresponding to the DC-DC converter. Accordingly, each first FPGA is specifically used to simulate the equivalent model corresponding to the DC-DC converter based on the current source signal generated by the equivalent model corresponding to the photovoltaic module in the previous simulation cycle and the current signal of the DC voltage source generated by the equivalent model corresponding to the inverter in the previous simulation cycle, and to generate the terminal voltage signal and the voltage signal of the second current source for the current simulation cycle.

[0017] In the first FPGA, the equivalent model corresponding to the photovoltaic module uses the terminal voltage signal of the previous simulation cycle to calculate the current source signal of the current simulation cycle, and the equivalent model corresponding to the DC converter uses the current source signal and current signal of the previous simulation cycle to calculate the terminal voltage signal and voltage signal of the current simulation cycle. This allows the two equivalent models to perform simulation calculations independently and in parallel, thereby significantly reducing the computational complexity of the simulation system, reducing the simulation step size, and implementing a simpler simulation calculation method.

[0018] In some embodiments of this application, the filter in the grid-connected unit can specifically be an LCL filter circuit, that is, in addition to a filter capacitor, the filter may also include a first filter inductor and a second filter inductor; wherein, the first filter inductor is connected between the power converter and the filter capacitor, and the second filter inductor is connected between the filter capacitor and the transformer. Correspondingly, the convergence model corresponding to the transformer deployed in the second FPGA is connected to the equivalent model corresponding to the second filter inductor through a connection between the model and each AC voltage source.

[0019] Secondly, this application provides a simulation method for photovoltaic power plants, including:

[0020] A photovoltaic power station model is established, which includes multiple grid-connected units and transformers. Each grid-connected unit includes photovoltaic modules, a power converter, and a filter. The output terminal of the photovoltaic module is connected to the input terminal of the power converter, and the output terminal of the power converter is connected to the input terminal of the filter. The output terminals of multiple filters are connected in parallel and then connected to the transformer. The filter includes a filter capacitor, and the transformer is used to connect to the power grid.

[0021] Decoupling is achieved at the filter capacitors, and the photovoltaic power station model is divided into detailed models corresponding to each grid-connected unit and convergence models corresponding to the transformers. Each filter capacitor is equivalent to each AC voltage source connected to the convergence model corresponding to the transformer, and each filter capacitor is equivalent to each first current source connected to the detailed model corresponding to the grid-connected unit.

[0022] Each first FPGA simulates the detailed model corresponding to each grid-connected unit deployed, and the second FPGA simulates the convergence model corresponding to the deployed transformer.

[0023] In the simulation method provided in this application embodiment, the filter capacitor on the output side of the grid-connected unit is used as the decoupling point for the large inertia link. The detailed models corresponding to each grid-connected unit after decoupling are deployed to the first FPGA (i.e., slave FPGA) in the master-slave connected multi-FPGA structure. The convergence model corresponding to the decoupling transformer is deployed to the second FPGA (i.e., master FPGA) in the master-slave connected multi-FPGA structure to construct the simulation model corresponding to the photovoltaic power station. Multiple FPGAs are used to simulate each power electronic device in the photovoltaic power station, and small-step simulation of each power electronic device in the photovoltaic power station can be realized in each FPGA. Furthermore, the multiple FPGAs adopt a master-slave connection structure. The first FPGAs are set in parallel and do not need to communicate. Only the first FPGAs and the second FPGAs communicate and interact, which can reduce the signal interaction between FPGAs and solve the problem of a large number of I / O in the simulation system. The simulation system architecture composed of multiple FPGAs connected in a master-slave manner reduces the amount of data interaction between the second FPGA (i.e., master FPGA) and the multiple first FPGAs (i.e. slave FPGAs), making it easy to expand the simulation system. When the capacity of a photovoltaic power station changes, i.e., when the number of grid-connected units in the photovoltaic power station is increased or decreased, the simulation system can be adapted to the photovoltaic power station by adding or removing the corresponding number of first FPGAs (i.e., slave FPGAs) in the simulation system, thus enabling the simulation system to have good scalability.

[0024] In some embodiments of this application, each first FPGA can simulate the detailed model corresponding to the grid-connected unit based on the current signal of the AC voltage source received from the second FPGA, and simulate the voltage signal of the first current source, then send the generated voltage signal of the first current source to the second FPGA. Correspondingly, the second FPGA can simulate the convergence model corresponding to the transformer based on the voltage signal of the first current source received from each first FPGA, and simulate the current signal of each AC voltage source, then send the corresponding current signal of each AC voltage source to each first FPGA.

[0025] In some embodiments of this application, the simulation method includes: each first FPGA simulating a detailed model corresponding to the grid-connected unit based on the current signals of the AC voltage sources received from the second FPGA in the previous simulation cycle, and sending the voltage signals of the first current sources generated in the current simulation cycle to the second FPGA. Correspondingly, the second FPGA simulating a convergence model corresponding to the transformer based on the voltage signals of the first current sources received from each first FPGA in the previous simulation cycle, and sending the current signals of each AC voltage source generated in the current simulation cycle to each first FPGA.

[0026] The first FPGA uses the current signal from the previous simulation cycle to calculate the voltage signal for the current simulation cycle, and the second FPGA uses the voltage signal from the previous simulation cycle to calculate the current signal. This allows the first and second FPGAs to perform simulation calculations independently and in parallel, thereby significantly reducing the computational complexity of the simulation system, reducing the simulation step size, and implementing simulation calculations in a simpler way.

[0027] In some embodiments of this application, the power converter specifically includes a DC-DC converter and an inverter; the output terminal of the photovoltaic module is connected to the input terminal of the DC-DC converter, the output terminal of the DC-DC converter is connected to the input terminal of the inverter via a DC capacitor, and the output terminal of the inverter is connected to the input terminal of a filter. The simulation method further includes:

[0028] Decoupling is achieved at the DC capacitor, dividing the equivalent model of the power converter into an equivalent model for the DC converter and an equivalent model for the inverter. The DC capacitor is equivalent to a second current source connected to the equivalent model for the DC converter, and also equivalent to a DC voltage source connected to the equivalent model for the inverter. Each first FPGA simulates the deployed equivalent models for the DC converter and the inverter, respectively.

[0029] Utilizing the gradual voltage variation and inertial characteristics of the DC capacitor, the DC capacitor is selected as the decoupling point for the detailed model corresponding to the grid-connected unit, and a decoupling method is used to segment the detailed model. By decomposing the simulation model at the module level, the equivalent models corresponding to the segmented DC-DC converter and the equivalent models corresponding to the inverter can be simulated and calculated in parallel, thereby reducing the simulation step size.

[0030] In some embodiments of this application, the simulation method further includes:

[0031] Each first FPGA simulates the equivalent model of the DC-DC converter based on the current signal of the DC voltage source generated by the equivalent model of the inverter, and generates the voltage signal of the second current source. Correspondingly, each first FPGA simulates the equivalent model of the inverter based on the voltage signal of the second current source generated by the equivalent model of the DC-DC converter and the current signal of the AC voltage source sent by the second FPGA, and generates the current signal of the DC voltage source and the voltage signal of the first current source sent to the second FPGA, respectively.

[0032] In some embodiments of this application, the simulation method specifically includes:

[0033] Each first FPGA simulates the equivalent model of the DC-DC converter based on the current signal of the DC voltage source generated by the equivalent model of the inverter in the previous simulation cycle, generating the voltage signal of the second current source for the current simulation cycle. Correspondingly, each first FPGA simulates the equivalent model of the inverter based on the voltage signal of the second current source generated by the equivalent model of the DC-DC converter in the previous simulation cycle and the current signal of the AC voltage source sent by the second FPGA in the previous simulation cycle, generating the current signal of the DC voltage source for the current simulation cycle and sending the voltage signal of the first current source generated in the current simulation cycle to the second FPGA.

[0034] In the first FPGA, the equivalent model corresponding to the DC-DC converter uses the current signal from the previous simulation cycle to calculate the voltage signal for the current simulation cycle, and the equivalent model corresponding to the inverter uses the current and voltage signals from the previous simulation cycle to calculate the current and voltage signals for the current simulation cycle. This allows the two equivalent models to perform simulation calculations independently and in parallel, thereby significantly reducing the computational complexity of the simulation system, reducing the simulation step size, and implementing a simpler simulation calculation method.

[0035] In some embodiments of this application, the simulation method further includes:

[0036] Each first FPGA simulates the equivalent model of the deployed photovoltaic module based on the terminal voltage signal generated by the equivalent model of the DC-DC converter, and sends the generated current source signal to the equivalent model of the DC-DC converter. Correspondingly, each first FPGA simulates the equivalent model of the DC-DC converter based on the current source signal generated by the equivalent model of the photovoltaic module and the current signal of the DC voltage source generated by the equivalent model of the inverter, and generates the terminal voltage signal and the voltage signal of the second current source.

[0037] In some embodiments of this application, the simulation method specifically includes:

[0038] Each first FPGA simulates the equivalent model corresponding to the deployed photovoltaic module based on the terminal voltage signal generated by the equivalent model corresponding to the DC-DC converter in the previous simulation cycle, and sends the current source signal generated in the current simulation cycle to the equivalent model corresponding to the DC-DC converter. Correspondingly, each first FPGA simulates the equivalent model corresponding to the DC-DC converter based on the current source signal generated by the equivalent model corresponding to the photovoltaic module in the previous simulation cycle and the current signal of the DC voltage source generated by the equivalent model corresponding to the inverter in the previous simulation cycle, generating the terminal voltage signal and the voltage signal of the second current source for the current simulation cycle.

[0039] In the first FPGA, the equivalent model corresponding to the photovoltaic module uses the terminal voltage signal of the previous simulation cycle to calculate the current source signal of the current simulation cycle, and the equivalent model corresponding to the DC converter uses the current source signal and current signal of the previous simulation cycle to calculate the terminal voltage signal and voltage signal of the current simulation cycle. This allows the two equivalent models to perform simulation calculations independently and in parallel, thereby significantly reducing the computational complexity of the simulation system, reducing the simulation step size, and implementing a simpler simulation calculation method.

[0040] The technical effects of the corresponding solutions in the second aspect can be referenced from the technical effects that can be obtained by the corresponding solutions in the first aspect; the repetitions will not be detailed. Attached Figure Description

[0041] Figure 1 This is a schematic diagram of the structure of a distributed photovoltaic power station provided in an embodiment of this application;

[0042] Figure 2 This is a schematic diagram of the simulation system provided in the embodiments of this application;

[0043] Figure 3 This is a schematic diagram of the signal interaction between each FPGA in the simulation system provided in the embodiments of this application;

[0044] Figure 4 A schematic diagram of each simulation model in the simulation system provided in the embodiments of this application;

[0045] Figure 5a A schematic diagram illustrating the signal interaction between the equivalent model of the photovoltaic module and the equivalent model of the DC-DC converter in the simulation system provided in this application embodiment;

[0046] Figure 5b A schematic diagram illustrating the signal interaction between the equivalent model of the DC converter and the equivalent model of the inverter in the simulation system provided in this application embodiment;

[0047] Figure 5c A schematic diagram illustrating the signal interaction between the equivalent model of a single inverter and the convergence model of a transformer in the simulation system provided in this application embodiment;

[0048] Figure 5d A schematic diagram illustrating the signal interaction between the equivalent models of multiple inverters and the convergence model of the transformer in the simulation system provided in this application embodiment;

[0049] Figure 6a This is a schematic diagram of the simulation results of the simulation system provided in the embodiments of this application;

[0050] Figure 6b This is a schematic diagram of the offline simulation results for commercial use. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of this application clearer, the application will be described in further detail below with reference to the accompanying drawings. It should be noted that in the description of this application, "multiple" can be understood as "at least two". Furthermore, it should be understood that in the description of this application, terms such as "first" and "second" are used only to distinguish the purpose of the description and should not be construed as indicating or implying relative importance, nor as indicating or implying order.

[0052] It should be noted that the same reference numerals in the accompanying drawings of this application denote the same or similar structures, and therefore repeated descriptions of them will be omitted. Terms expressing position and direction described in this application are illustrative based on the accompanying drawings, but may be modified as needed, and all modifications are included within the scope of protection of this application. The accompanying drawings of this application are for illustrating relative positional relationships only and do not represent actual scale.

[0053] To facilitate understanding of the simulation system and simulation method for photovoltaic power stations in the embodiments of this application, some terms will be explained below.

[0054] Real-time simulation: Real-time simulation is a type of simulation mode, which is the opposite of non-real-time offline simulation. Real-time simulation is a simulation mode in which the simulation model runs strictly in sync with the real physical world.

[0055] Hardware-in-the-loop (HIL) simulation based on hardware is a subset of real-time simulation. A hardware-in-the-loop simulation system can be constructed based on a real-time simulation system. The latter connects the real-time running simulation model to the actual hardware device through a board with I / O interface to form a hardware-in-the-loop and exchange data in real time.

[0056] Currently, the existing technologies for collaborative simulation with distributed computing devices mainly include: (1) Real-time simulation system based on multi-core central processing unit (CPU): The simulation model is deployed in the CPU, and the CPU multi-core undertakes the simulation calculation task. Although the architecture modeling of the CPU multi-core real-time simulation system is relatively flexible and has a certain degree of versatility, the simulation step size is limited. This real-time simulation is mostly applied to the large-scale new energy power station scenario of modular multilevel converter (MMC). The power electronic model used in its multi-core simulation is an average value model, which cannot realize small step size simulation. The simulation method is to simulate the detailed behavior of power electronic equipment. (2) Real-time simulation system based on CPU + multiple field-programmable gate array (FPGA): The simulation model is deployed in a mixed manner in the CPU and FPGA. The CPU and multiple FPGAs have a one-to-many connection relationship. The CPU undertakes low-speed model calculation, and the FPGA undertakes high-speed model calculation. The architecture of the CPU + multiple FPGA real-time simulation system has a certain degree of versatility while retaining the ability of small step size simulation. However, due to the CPU undertaking the convergence and interaction of data from multiple FPGAs, the communication latency is greatly increased, which affects the simulation efficiency of the CPU. (3) Real-time simulation system based on multiple FPGAs: The simulation model is deployed on multiple FPGAs, each FPGA independently undertakes the simulation task, and the multiple FPGAs communicate with each other. The real-time simulation system architecture based on multiple FPGAs has the simulation capability of small steps and high simulation efficiency, but its versatility is low, and the communication delay problem of FPGAs on the same board is unavoidable.

[0057] Distributed photovoltaic (PV) power plants are characterized by a large number of grid-connected units, which are arranged in a structured parallel configuration. Because each grid-connected unit contains power electronic equipment, a small simulation step size is required during hardware-in-the-loop (HIL) simulation. Furthermore, the interaction between grid-connected units and external devices consumes significant I / O resources, hindering simulation scalability and expansion.

[0058] To address the above issues, this application provides a simulation system for photovoltaic power plants, which is designed for simulating distributed photovoltaic power plants.

[0059] Specifically, refer to Figure 1 The photovoltaic power station includes multiple grid-connected units 100 and a transformer 200. Each grid-connected unit 100 includes a photovoltaic module 101, a power converter 102, and a filter 103. The output terminal of the photovoltaic module 101 is connected to the input terminal of the power converter 102, and the output terminal of the power converter 102 is connected to the input terminal of the filter 103. The output terminals of multiple filters 103 are connected in parallel and then connected to the transformer 200, which is used to connect to the power grid 300. Exemplarily, the output terminals of the filters 103 of the multiple grid-connected units 100 are converged to the PCC node via an AC bus and then connected to the transformer 200. The filter 103 includes a filter capacitor C1, which, exemplarily, can be connected in series between the output terminal of the power converter 102 and the grounding port.

[0060] Reference Figure 1 The simulation system provided in this application, specifically for the structure of a distributed photovoltaic (PV) power station where multiple grid-connected units 100 are connected in parallel to a transformer 200, can utilize the inertial characteristic of the gradually changing voltage of the filter capacitor C1 in each grid-connected unit 100. The filter capacitor C1 of the filter 103 in each grid-connected unit 100 of the PV power station can be selected as the decoupling point for each simulation model. The decoupling point is shown as the dashed line in the figure. The simulation model corresponding to the PV power station is divided using a decoupling method. This decomposition of the simulation model at the system level addresses the problem that the simulation system cannot accommodate the expansion of PV power stations of different scales due to the large number of parallel units in the PV power station.

[0061] Reference Figure 2 The simulation system provided in this application specifically includes: multiple first FPGAs and second FPGAs. The number of first FPGAs is consistent with the number of multiple grid-connected units 100 in the photovoltaic power station. All first FPGAs are communicatively connected to the second FPGAs, forming a many-to-one master-slave connection structure. Specifically, the convergence model 200' corresponding to the transformer 200 after the photovoltaic power station simulation model is segmented can be deployed to the second FPGA (master FPGA) for simulation calculation. The detailed models 100' corresponding to each grid-connected unit 100 after the photovoltaic power station simulation model is segmented can be deployed to each first FPGA (slave FPGA) for simulation calculation. In other words, each first FPGA is used to simulate the detailed models 100' corresponding to each deployed grid-connected unit 100 in parallel; the second FPGA is used to simulate the convergence model 200' corresponding to the deployed transformer 200.

[0062] In the simulation system provided in this application embodiment, the filter capacitor C1 on the output side of the grid-connected unit 100 is used as the decoupling point for the large inertia link. The detailed models 100' corresponding to each decoupled grid-connected unit 100 are deployed to the first FPGA (i.e., slave FPGA) in the master-slave connected multi-FPGA structure, and the convergence model 200' corresponding to the decoupled transformer 200 is deployed to the second FPGA (i.e., master FPGA) in the master-slave connected multi-FPGA structure to construct the simulation model corresponding to the photovoltaic power station. Multiple FPGAs are used to simulate each power electronic device in the photovoltaic power station, and small-step simulation of each power electronic device in the photovoltaic power station can be realized in each FPGA. Furthermore, the multiple FPGAs adopt a master-slave connection structure, and the first FPGAs are set in parallel without communication. Only the first FPGAs and the second FPGAs communicate and interact, which can reduce the signal interaction between FPGAs and solve the problem of a large number of I / O in the simulation system. The simulation system architecture is constructed by using multiple FPGAs connected in a master-slave manner, which reduces the amount of data interaction between the second FPGA (i.e., master FPGA) and the multiple first FPGAs (i.e. slave FPGAs), making it easy to expand the simulation system. When the capacity of a photovoltaic power station changes, i.e., when the number of grid-connected units 100 in the photovoltaic power station is increased or decreased, the simulation system can be adapted to the photovoltaic power station by connecting or removing the corresponding number of first FPGAs (i.e., slave FPGAs) in the simulation system, thus enabling the simulation system to have good scalability.

[0063] Reference Figure 2 In this application, the detailed models corresponding to each grid-connected unit 100 and the convergence model 200' corresponding to the transformer 200 are decoupled at the filter capacitor C1. Each filter capacitor C1 can be equivalent to each AC voltage source Uc connected to the convergence model 200' corresponding to the transformer 200, and each filter capacitor C1 is equivalent to each first current source Ic connected to the detailed model 100' corresponding to the grid-connected unit 100. Furthermore, an equivalent model 300' corresponding to the power grid 300 connected to the transformer 200 will also be deployed in the second FPGA. When the power grid 300 is a three-phase power grid, correspondingly, the transformer 200 and the filter capacitor C1 are also a three-phase transformer and a three-phase filter capacitor, respectively. Therefore, the AC voltage sources Uc equivalent to each filter capacitor C1 deployed in the second FPGA can specifically be three-phase AC voltage sources Uca, Ucb, and Ucc, and the first current sources Ic equivalent to each filter capacitor C1 deployed in the first FPGA can specifically be three-phase first current sources Ica, Icb, and Icc. Figure 2The diagram only shows the first and last FPGAs among the multiple first FPGAs, namely first FPGA1 and first FPGAn. The first current sources Ic deployed in first FPGA1 can be three-phase first current sources Ica1, Icb1 and Icc1. The first current sources Ic deployed in first FPGAn can be three-phase first current sources Ican, Icbn and Iccn. Correspondingly, the AC voltage sources Uc deployed in the second FPGA can be three-phase AC voltage sources Uca1, Ucb1 and Ucc1, Ucan, Ucbn and Uccn.

[0064] Reference Figure 3 In some embodiments of this application, each first FPGA can receive the current signal i from the AC voltage source Uc sent by the second FPGA. c A detailed model corresponding to the grid-connected unit 100 was simulated, and the voltage signal v of the first current source Ic was obtained from the simulation. c Then, the voltage signal v of the generated first current source Ic is sent to the second FPGA. c Accordingly, the second FPGA can receive the voltage signal v of the first current source Ic sent by each of the first FPGAs. c Simulations were performed on the convergence model corresponding to transformer 200, and the current signals i of each AC voltage source Uc were obtained from the simulation. c Then, the current signal i corresponding to each AC voltage source Uc is sent to each of the first FPGAs. c .exist Figure 3 Only four of the multiple first FPGAs are shown in the diagram: first FPGA1, first FPGA2, first FPGA3, and first FPGAn. Specifically, each first current source Ic deployed within first FPGA1 can receive the three-phase current signal i sent by the second FPGA. ca1 i cb1 and i cc1 And send a three-phase voltage signal v to the corresponding AC voltage source Uc deployed in the second FPGA. ca1 v cb1 and v cc1 The first current sources Ic deployed within the first FPGA2 can specifically receive the three-phase current signals i sent by the second FPGA. ca2 i cb2 and i cc2 And send a three-phase voltage signal v to the corresponding AC voltage source Uc deployed in the second FPGA. ca2 v cb2 and v cc2 The first current sources Ic deployed within the first FPGA3 can specifically receive the three-phase current signals i sent by the second FPGA. ca3i cb3 and i cc3 And send a three-phase voltage signal v to the corresponding AC voltage source Uc deployed in the second FPGA. ca3 v cb3 and v cc3 The first current sources Ic deployed within the first FPGA n can specifically receive the three-phase current signal i sent by the second FPGA. can i cbn and i ccn And send a three-phase voltage signal v to the corresponding AC voltage source Uc deployed in the second FPGA. can v cbn and v ccn .

[0065] Reference Figure 1 In some embodiments of this application, the filter 103 in the grid-connected unit 100 can specifically be an LCL filter circuit, that is, in addition to the filter capacitor C1, the filter 103 may also include a first filter inductor L1 and a second filter inductor L2; wherein, the first filter inductor L1 is connected between the power converter 102 and the filter capacitor C1, and the second filter inductor L2 is connected between the filter capacitor C1 and the transformer 200. Accordingly, referring to Figure 2 The convergence model 200' corresponding to the transformer 200 deployed in the second FPGA is connected to the three-phase AC voltage sources Uca1, Ucb1 and Ucc1, Ucan, Ucbn and Uccn by the equivalent model L2' corresponding to the second filter inductor L2.

[0066] Reference Figure 1 In some embodiments of this application, the power converter 102 of each grid-connected unit 100 in the photovoltaic power station may specifically include a DC / DC converter and an DC / AC inverter; the output terminal of the photovoltaic module 101 is connected to the input terminal of the DC / DC converter, the output terminal of the DC / DC converter is connected to the input terminal of the DC / AC inverter through a DC capacitor C0, and the output terminal of the DC / AC inverter is connected to the input terminal of the filter 103.

[0067] Reference Figure 2In the simulation system provided in this application embodiment, in order to achieve small-step simulation, the inertial characteristic of the gradually changing voltage of DC capacitor C0 can be utilized to select DC capacitor C0 as the decoupling point of the detailed model 100' corresponding to the grid-connected unit 100, and the detailed model 100' is divided using a decoupling method. The simulation model is decomposed at the module level, and the equivalent models corresponding to the divided DC converter and inverter are simulated and calculated in parallel to reduce the simulation step size. Specifically, each first FPGA is used to simulate the equivalent model DC / DC' corresponding to the deployed DC converter DC / DC and the equivalent model DC / AC' corresponding to the inverter DC / AC, respectively. The equivalent model DC / DC' corresponding to the DC converter DC / DC and the equivalent model DC / AC' corresponding to the inverter DC / AC are decoupled at DC capacitor C0. DC capacitor C0 is equivalent to the second current source Is connected to the equivalent model DC / DC' corresponding to the DC converter DC / DC, and DC capacitor C0 is equivalent to the DC voltage source Us connected to the equivalent model DC / AC' corresponding to the inverter DC / AC.

[0068] Reference Figure 2 In some embodiments of this application, each first FPGA can generate a current signal i from the DC voltage source Us based on the equivalent DC / AC model corresponding to the inverter DC / AC. s Simulations were performed on the equivalent model DC / DC' of the DC / DC converter to generate the voltage signal v of the second current source Is. s ; and based on the voltage signal v of the second current source Is generated according to the equivalent model DC / DC' corresponding to the DC / DC converter. s And the current signal i from the AC voltage source Uc sent by the second FPGA. c Simulations were performed on the equivalent model DC / AC' corresponding to the inverter's DC / AC, generating the current signal i of the DC voltage source Us. s and the voltage signal v of the first current source Ic sent to the second FPGA c .

[0069] Reference Figure 2 In some embodiments of this application, each first FPGA is further configured to deploy an equivalent model 101' corresponding to the photovoltaic module 101. Each first FPGA can simulate the equivalent model 101' corresponding to the deployed photovoltaic module 101 based on the terminal voltage signal Upv generated by the equivalent model DC / DC' corresponding to the DC / DC converter, and send the generated current source signal Ipv to the equivalent model DC / DC' corresponding to the DC / DC converter; and based on the current source signal Ipv generated by the equivalent model 101' corresponding to the photovoltaic module 101 and the current signal i of the DC voltage source Us generated by the equivalent model DC / AC' corresponding to the inverter DC / AC.s Simulations were performed on the equivalent model DC / DC' of the DC / DC converter to generate the terminal voltage signal Upv and the voltage signal v of the second current source Is. s .

[0070] The simulation system provided in the embodiments of this application is described below with specific examples.

[0071] In this embodiment, at the system level, the transformer 200 and each grid-connected unit 100 in the photovoltaic power station are separated and placed in different computing cores, i.e., different FPGAs; at the module level, the circuits inside the grid-connected unit 100 are separated and placed in the same computing core, i.e., the same FPGA. The grid-connected unit 100 includes a photovoltaic module 101, a DC / DC converter, and a DC / AC inverter, wherein the filter capacitor C1 connected to the inverter DC / AC will generate a voltage signal v c Transmit the signal to transformer 200, and simultaneously receive the current signal i transmitted from transformer 200. c .

[0072] Specifically, the photovoltaic modules 101 included in the grid-connected unit 100 can be composed of Np photovoltaic strings connected in parallel, and each photovoltaic string contains Ns photovoltaic cells connected in series. The DC / DC converter included in the grid-connected unit 100 can be a three-level boost circuit, and the DC / AC inverter can be a three-level NPC inverter circuit. A DC capacitor C0 is provided between the DC / AC inverter and the DC / DC converter, and the filter 103 connected to the output terminal of the DC / AC inverter can be an LCL filter circuit.

[0073] Reference Figure 4 Taking a single grid-connected unit 100 as an example, the corresponding detailed model segmentation includes the equivalent model 101' corresponding to the photovoltaic module 101, the equivalent model DC / DC' corresponding to the DC / DC converter, and the equivalent model DC / AC' corresponding to the DC / AC inverter.

[0074] Reference Figure 5a The signal interaction between the equivalent model 101' corresponding to the photovoltaic module 101 and the equivalent model DC / DC' corresponding to the DC / DC converter includes: the equivalent model 101' corresponding to the photovoltaic module 101 receives the terminal voltage signal Upv generated by the low voltage terminal of the equivalent model DC / DC' corresponding to the DC / DC converter, and generates the corresponding current source signal Ipv according to its UI characteristic curve and returns it to the equivalent model DC / DC' corresponding to the DC / DC converter.

[0075] Reference Figure 5bThe signal interaction between the equivalent model DC / DC' of the DC / DC converter and the equivalent model DC / AC' of the DC / AC inverter includes: the high-voltage side of the equivalent model DC / DC' receives the current signal i generated on the DC side of the equivalent model DC / AC' of the DC / AC inverter. s and the voltage signal v s Return the equivalent model DC / AC' to the inverter DC / AC.

[0076] Reference Figure 5c The signal interaction between the equivalent model DC / AC' of the inverter DC / AC and the convergence model 200' of the transformer 200 includes: the equivalent model DC / AC' of the inverter DC / AC receives the current signal i generated by the convergence model 200' of the transformer 200. c and the voltage signal v c Return to the convergence model 200' corresponding to transformer 200.

[0077] Reference Figure 5d The signal interaction between the equivalent model DC / AC' of each inverter DC / AC in the multiple grid-connected units 100 and the aggregation model 200' of the transformer 200 includes: the aggregation model 200' of the transformer 200 receiving the voltage signal v generated by the equivalent model DC / AC' of each inverter DC / AC. c and the current signal i c The equivalent DC / AC' of each inverter DC / AC is returned. For each additional grid-connected unit 100, the corresponding parallel nodes are expanded in the aggregation model 200' corresponding to the transformer 200.

[0078] In the simulation system of this application, in order to reduce the coupling between simulation models, the signal transmission between simulation systems can be separated by at least one simulation cycle. That is, each simulation model uses the signal sent by other simulation models in the previous simulation cycle that it receives and interacts with to perform simulation calculations. This decouples the originally tightly coupled simulation models and allows them to operate independently in parallel, thereby significantly reducing the computational complexity of each simulation model and reducing the simulation step size.

[0079] The following is the calculation derivation related to connecting transformer 200 after multiple grid-connected units (100 units) are connected in parallel:

[0080] If there are N grid-connected units 100 corresponding to detailed models 100' connected to the convergence model 200' corresponding to transformer 200, then based on the voltage signals transmitted by the N grid-connected units 100' to the convergence model 200' corresponding to transformer 200, and the current signals transmitted by the convergence model 200' corresponding to transformer 200 to the N grid-connected units 100', their respective model matrices can be listed:

[0081]

[0082] Among them, G1, G2…G N G A These are the node admittance matrices listed based on parameters such as capacitance and inductance of grid-connected unit 100 and transformer 200; V1, V2…V N V A These are phasors composed of the voltage signals transmitted from each grid-connected unit 100 to the transformer 200, b1, b2…b N b A b'1, b'2, ..., b' are phasors composed of the current signals transmitted from transformer 200 to each grid-connected unit 100, respectively. N b' A S represents the phasor composed of the historical current signals transmitted from transformer 200 to each grid-connected unit 100. A1 S A2 …S AN S is the associated quantity of the current signal transmitted from the reaction transformer 200 to each grid-connected unit 100. 1A S 2A …S NA S is a correlated quantity that reflects the voltage signal transmitted from the grid-connected unit 100 to the transformer 200. A1 S A2 …S AN S 1A S 2A …S NA Only the elements in the row corresponding to the V phase and the column corresponding to the b phase are 1, and all other elements are 0.

[0083] Combining the above equations, we obtain the augmented matrix equation:

[0084]

[0085] As can be seen, the augmented matrix has a large order and strong coupling, making the calculation complex. Therefore, a simulation delay of one cycle can be added to the simulation system. That is, the voltage signal transmitted from the grid-connected unit 100 to the transformer 200 is the voltage signal from the previous moment, i.e., the previous simulation cycle. The current signal fed back from the transformer 200 to the grid-connected unit 100 is also the current signal from the previous moment, i.e., the previous simulation cycle. The historical current signals of each simulation model also adopt the historical current signals calculated from the previous moment. The matrices of each simulation model then become:

[0086]

[0087] In this context, the parameter with the superscript (t) represents the parameter quantity at the current time t, and the parameter with the superscript (t-1) represents the parameter quantity at the previous time t.

[0088] Combining the above equations, we can obtain the augmented matrix:

[0089]

[0090] As can be seen, the b parameters on the right-hand side all depend on the variables of the previous time step, and the main matrix has a diagonal structure. Therefore, the simulation models of the above N grid-connected units 100 and transformer 200 can be solved independently and in parallel. The calculation steps are as follows:

[0091] 1. First, analyze the voltage signal. and historical current signals Perform initialization processing.

[0092] 2. Using time (k-1) and The current time (k) is calculated.

[0093] 3. Calculate the current time (k). The calculation formula is: And the historical current signal was calculated.

[0094] 4. Then proceed to the next iteration calculation.

[0095] The above analysis shows that using data from the previous simulation cycle to calculate the parameters of the current simulation cycle with a one-cycle delay can significantly reduce the coupling between simulation models, decouple the originally tightly coupled simulation models, and allow each simulation model to operate independently and in parallel. This greatly reduces the computational complexity of the simulation system, reduces the simulation step size, and makes the simulation calculation simpler to implement.

[0096] Based on this, in some embodiments of this application, each first FPGA is specifically used to: simulate the detailed model corresponding to the grid-connected unit based on the current signal of the AC voltage source sent by the second FPGA in the previous simulation cycle, and send the voltage signal of the first current source generated in the current simulation cycle to the second FPGA. Correspondingly, the second FPGA is specifically used to: simulate the convergence model corresponding to the transformer based on the voltage signal of the first current source sent by each first FPGA in the previous simulation cycle, and send the current signals of each AC voltage source generated in the current simulation cycle to each first FPGA. The first FPGA uses the current signal of the previous simulation cycle to calculate the voltage signal of the current simulation cycle, and the second FPGA uses the voltage signal of the previous simulation cycle to calculate the current signal. This allows the first FPGA and the second FPGA to perform simulation calculations independently and in parallel, thereby significantly reducing the computational complexity of the simulation system, reducing the simulation step size, and implementing a simpler simulation calculation method.

[0097] In some embodiments of this application, each first FPGA is specifically used to: simulate the equivalent model corresponding to the DC converter based on the current signal of the DC voltage source generated by the equivalent model corresponding to the inverter in the previous simulation cycle, and generate the voltage signal of the second current source in the current simulation cycle; and simulate the equivalent model corresponding to the inverter based on the voltage signal of the second current source generated by the equivalent model corresponding to the DC converter in the previous simulation cycle and the current signal of the AC voltage source sent by the second FPGA in the previous simulation cycle, respectively generating the current signal of the DC voltage source in the current simulation cycle and sending the voltage signal of the first current source generated in the current simulation cycle to the second FPGA. In the first FPGA, the equivalent model corresponding to the DC converter uses the current signal of the previous simulation cycle to calculate the voltage signal of the current simulation cycle, and the equivalent model corresponding to the inverter uses the current signal and voltage signal of the previous simulation cycle to calculate the current signal and voltage signal of the current simulation cycle. This allows the two equivalent models to perform simulation calculations independently and in parallel, thereby significantly reducing the computational complexity of the simulation system, reducing the simulation step size, and implementing a simpler simulation calculation method.

[0098] In some embodiments of this application, each first FPGA is specifically used for: simulating the equivalent model corresponding to the deployed photovoltaic module based on the terminal voltage signal generated by the equivalent model corresponding to the DC-DC converter in the previous simulation cycle; sending the current source signal generated in the current simulation cycle to the equivalent model corresponding to the DC-DC converter; and simulating the equivalent model corresponding to the DC-DC converter based on the current source signal generated by the equivalent model corresponding to the photovoltaic module in the previous simulation cycle and the current signal of the DC voltage source generated by the equivalent model corresponding to the inverter in the previous simulation cycle, generating the terminal voltage signal and the voltage signal of the second current source in the current simulation cycle. In the first FPGA, the equivalent model corresponding to the photovoltaic module uses the terminal voltage signal of the previous simulation cycle to calculate the current source signal of the current simulation cycle, and the equivalent model corresponding to the DC-DC converter uses the current source signal and the current signal of the previous simulation cycle to calculate the terminal voltage signal and the voltage signal of the current simulation cycle. This allows the two equivalent models to perform simulation calculations independently and in parallel, thereby significantly reducing the computational complexity of the simulation system, reducing the simulation step size, and implementing a simpler simulation calculation method.

[0099] Reference Figure 6a Using the simulation system provided in the embodiments of this application, a real-time simulation of a photovoltaic power station is performed with a simulation step size of 1µs. (Refer to...) Figure 6b By comparing the results of offline simulation of photovoltaic power plants using commercial offline simulation systems, and by comparing the three-phase voltage signals output from the photovoltaic power plant to the power grid obtained by the simulation system, it can be seen that the simulation accuracy of the simulation system in this application can reach the accuracy of offline simulation.

[0100] Based on the same inventive concept, embodiments of this application also provide a simulation method for photovoltaic power stations, including:

[0101] A photovoltaic power station model is established, which includes multiple grid-connected units and transformers. Each grid-connected unit includes photovoltaic modules, a power converter, and a filter. The output terminal of the photovoltaic module is connected to the input terminal of the power converter, and the output terminal of the power converter is connected to the input terminal of the filter. The output terminals of multiple filters are connected in parallel and then connected to the transformer. The filter includes a filter capacitor, and the transformer is used to connect to the power grid.

[0102] Decoupling is achieved at the filter capacitors, and the photovoltaic power station model is divided into detailed models corresponding to each grid-connected unit and convergence models corresponding to the transformers. Each filter capacitor is equivalent to each AC voltage source connected to the convergence model corresponding to the transformer, and each filter capacitor is equivalent to each first current source connected to the detailed model corresponding to the grid-connected unit.

[0103] Each first FPGA simulates the detailed model corresponding to each grid-connected unit deployed, and the second FPGA simulates the convergence model corresponding to the deployed transformer.

[0104] In the simulation method provided in this application embodiment, the filter capacitor on the output side of the grid-connected unit is used as the decoupling point for the large inertia link. The detailed models corresponding to each grid-connected unit after decoupling are deployed to the first FPGA (i.e., slave FPGA) in the master-slave connected multi-FPGA structure. The convergence model corresponding to the decoupling transformer is deployed to the second FPGA (i.e., master FPGA) in the master-slave connected multi-FPGA structure to construct the simulation model corresponding to the photovoltaic power station. Multiple FPGAs are used to simulate each power electronic device in the photovoltaic power station, and small-step simulation of each power electronic device in the photovoltaic power station can be realized in each FPGA. Furthermore, the multiple FPGAs adopt a master-slave connection structure. The first FPGAs are set in parallel and do not need to communicate. Only the first FPGAs and the second FPGAs communicate and interact, which can reduce the signal interaction between FPGAs and solve the problem of a large number of I / O in the simulation system. The simulation system architecture composed of multiple FPGAs connected in a master-slave manner reduces the amount of data interaction between the second FPGA (i.e., master FPGA) and the multiple first FPGAs (i.e. slave FPGAs), making it easy to expand the simulation system. When the capacity of a photovoltaic power station changes, i.e., when the number of grid-connected units in the photovoltaic power station is increased or decreased, the simulation system can be adapted to the photovoltaic power station by adding or removing the corresponding number of first FPGAs (i.e., slave FPGAs) in the simulation system, thus enabling the simulation system to have good scalability.

[0105] In some embodiments of this application, each first FPGA can simulate the detailed model corresponding to the grid-connected unit based on the current signal of the AC voltage source received from the second FPGA, and simulate the voltage signal of the first current source, then send the generated voltage signal of the first current source to the second FPGA. Correspondingly, the second FPGA can simulate the convergence model corresponding to the transformer based on the voltage signal of the first current source received from each first FPGA, and simulate the current signal of each AC voltage source, then send the corresponding current signal of each AC voltage source to each first FPGA.

[0106] In some embodiments of this application, the simulation method includes: each first FPGA simulating a detailed model corresponding to the grid-connected unit based on the current signals of the AC voltage sources received from the second FPGA in the previous simulation cycle, and sending the voltage signals of the first current sources generated in the current simulation cycle to the second FPGA. Correspondingly, the second FPGA simulating a convergence model corresponding to the transformer based on the voltage signals of the first current sources received from each first FPGA in the previous simulation cycle, and sending the current signals of each AC voltage source generated in the current simulation cycle to each first FPGA.

[0107] The first FPGA uses the current signal from the previous simulation cycle to calculate the voltage signal for the current simulation cycle, and the second FPGA uses the voltage signal from the previous simulation cycle to calculate the current signal. This allows the first and second FPGAs to perform simulation calculations independently and in parallel, thereby significantly reducing the computational complexity of the simulation system, reducing the simulation step size, and implementing simulation calculations in a simpler way.

[0108] In some embodiments of this application, the power converter specifically includes a DC-DC converter and an inverter; the output terminal of the photovoltaic module is connected to the input terminal of the DC-DC converter, the output terminal of the DC-DC converter is connected to the input terminal of the inverter via a DC capacitor, and the output terminal of the inverter is connected to the input terminal of a filter. The simulation method further includes:

[0109] Decoupling is achieved at the DC capacitor, dividing the equivalent model of the power converter into an equivalent model for the DC converter and an equivalent model for the inverter. The DC capacitor is equivalent to a second current source connected to the equivalent model for the DC converter, and also equivalent to a DC voltage source connected to the equivalent model for the inverter. Each first FPGA simulates the deployed equivalent models for the DC converter and the inverter, respectively.

[0110] Utilizing the gradual voltage variation and inertial characteristics of the DC capacitor, the DC capacitor is selected as the decoupling point for the detailed model corresponding to the grid-connected unit, and a decoupling method is used to segment the detailed model. By decomposing the simulation model at the module level, the equivalent models corresponding to the segmented DC-DC converter and the equivalent models corresponding to the inverter can be simulated and calculated in parallel, thereby reducing the simulation step size.

[0111] In some embodiments of this application, the simulation method further includes:

[0112] Each first FPGA simulates the equivalent model of the DC-DC converter based on the current signal of the DC voltage source generated by the equivalent model of the inverter, and generates the voltage signal of the second current source. Correspondingly, each first FPGA simulates the equivalent model of the inverter based on the voltage signal of the second current source generated by the equivalent model of the DC-DC converter and the current signal of the AC voltage source sent by the second FPGA, and generates the current signal of the DC voltage source and the voltage signal of the first current source sent to the second FPGA, respectively.

[0113] In some embodiments of this application, the simulation method specifically includes:

[0114] Each first FPGA simulates the equivalent model of the DC-DC converter based on the current signal of the DC voltage source generated by the equivalent model of the inverter in the previous simulation cycle, generating the voltage signal of the second current source for the current simulation cycle. Correspondingly, each first FPGA simulates the equivalent model of the inverter based on the voltage signal of the second current source generated by the equivalent model of the DC-DC converter in the previous simulation cycle and the current signal of the AC voltage source sent by the second FPGA in the previous simulation cycle, generating the current signal of the DC voltage source for the current simulation cycle and sending the voltage signal of the first current source generated in the current simulation cycle to the second FPGA.

[0115] In the first FPGA, the equivalent model corresponding to the DC-DC converter uses the current signal from the previous simulation cycle to calculate the voltage signal for the current simulation cycle, and the equivalent model corresponding to the inverter uses the current and voltage signals from the previous simulation cycle to calculate the current and voltage signals for the current simulation cycle. This allows the two equivalent models to perform simulation calculations independently and in parallel, thereby significantly reducing the computational complexity of the simulation system, reducing the simulation step size, and implementing a simpler simulation calculation method.

[0116] In some embodiments of this application, the simulation method further includes:

[0117] Each first FPGA simulates the equivalent model of the deployed photovoltaic module based on the terminal voltage signal generated by the equivalent model of the DC-DC converter, and sends the generated current source signal to the equivalent model of the DC-DC converter. Correspondingly, each first FPGA simulates the equivalent model of the DC-DC converter based on the current source signal generated by the equivalent model of the photovoltaic module and the current signal of the DC voltage source generated by the equivalent model of the inverter, and generates the terminal voltage signal and the voltage signal of the second current source.

[0118] In some embodiments of this application, the simulation method specifically includes:

[0119] Each first FPGA simulates the equivalent model corresponding to the deployed photovoltaic module based on the terminal voltage signal generated by the equivalent model corresponding to the DC-DC converter in the previous simulation cycle, and sends the current source signal generated in the current simulation cycle to the equivalent model corresponding to the DC-DC converter. Correspondingly, each first FPGA simulates the equivalent model corresponding to the DC-DC converter based on the current source signal generated by the equivalent model corresponding to the photovoltaic module in the previous simulation cycle and the current signal of the DC voltage source generated by the equivalent model corresponding to the inverter in the previous simulation cycle, generating the terminal voltage signal and the voltage signal of the second current source for the current simulation cycle.

[0120] In the first FPGA, the equivalent model corresponding to the photovoltaic module uses the terminal voltage signal of the previous simulation cycle to calculate the current source signal of the current simulation cycle, and the equivalent model corresponding to the DC converter uses the current source signal and current signal of the previous simulation cycle to calculate the terminal voltage signal and voltage signal of the current simulation cycle. This allows the two equivalent models to perform simulation calculations independently and in parallel, thereby significantly reducing the computational complexity of the simulation system, reducing the simulation step size, and implementing a simpler simulation calculation method.

[0121] The simulation system and method for photovoltaic power plants provided in this application utilize the filter capacitor on the output side of the grid-connected unit as a decoupling point for large inertia links. The detailed models corresponding to each decoupled grid-connected unit are deployed to the first FPGA (slave FPGA) in a master-slave connected multi-FPGA structure. The convergence model corresponding to the decoupled transformer is deployed to the second FPGA (master FPGA) in the master-slave connected multi-FPGA structure to construct the simulation model corresponding to the photovoltaic power plant. Multiple FPGAs are used to simulate each power electronic device in the photovoltaic power plant, allowing for small-step simulation of each power electronic device within each FPGA. Furthermore, the multiple FPGAs adopt a master-slave connection structure, with the first FPGAs connected in parallel and requiring no communication; only communication occurs between the first and second FPGAs. This reduces signal interaction between FPGAs, addressing the issue of a large number of I / O operations in the simulation system. The master-slave connection of multiple FPGAs in the simulation system architecture minimizes data interaction between the second FPGA (master FPGA) and the multiple first FPGAs (slave FPGAs), facilitating system expansion. When the capacity of a photovoltaic power station changes, i.e., when the number of grid-connected units in the photovoltaic power station is increased or decreased, the simulation system can be adapted to the photovoltaic power station by adding or removing the corresponding number of first FPGAs (i.e., slave FPGAs) in the simulation system, thus enabling the simulation system to have good scalability.

[0122] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A simulation system for a photovoltaic power station, characterized in that, The photovoltaic power station includes multiple grid-connected units and a transformer. Each grid-connected unit includes a photovoltaic module, a power converter, and a filter. The output terminal of the photovoltaic module is connected to the input terminal of the power converter, the output terminal of the power converter is connected to the input terminal of the filter, and the output terminals of multiple filters are connected in parallel and then connected to the transformer. The filter includes a filter capacitor, and the transformer is used to connect to the power grid. The simulation system includes: multiple first field-programmable gate arrays (FPGAs) and second FPGAs, wherein the number of first FPGAs is the same as the number of grid-connected units; Each of the first FPGAs is used to simulate the detailed models corresponding to each of the deployed grid-connected units in parallel, and the second FPGA is used to simulate the convergence model corresponding to the deployed transformer. The detailed models corresponding to each of the grid-connected units and the convergence models corresponding to the transformer are decoupled at the filter capacitors. Each filter capacitor is equivalent to each AC voltage source connected to the convergence model corresponding to the transformer, and each filter capacitor is equivalent to each first current source connected to the detailed model corresponding to the grid-connected unit.

2. The simulation system as described in claim 1, characterized in that, Each of the first FPGAs is configured to: simulate the detailed model corresponding to the grid-connected unit based on the current signal of the AC voltage source received from the second FPGA, and send the generated voltage signal of the first current source to the second FPGA; The second FPGA is used to: simulate the convergence model corresponding to the transformer based on the voltage signals of the first current sources sent by each of the first FPGAs, and send the current signals of each of the corresponding AC voltage sources to each of the first FPGAs.

3. The simulation system as described in claim 2, characterized in that, Each of the first FPGAs is specifically used to: simulate the detailed model corresponding to the grid-connected unit based on the current signal of the AC voltage source sent by the second FPGA in the previous simulation cycle, and send the voltage signal of the first current source generated in the current simulation cycle to the second FPGA; The second FPGA is specifically used to: simulate the convergence model corresponding to the transformer based on the voltage signals of the first current sources sent by each of the first FPGAs in the previous simulation cycle, and send the current signals of each of the AC voltage sources generated in the current simulation cycle to each of the first FPGAs.

4. The simulation system according to any one of claims 1-3, characterized in that, The power converter includes a DC-DC converter and an inverter; the output terminal of the photovoltaic module is connected to the input terminal of the DC-DC converter, the output terminal of the DC-DC converter is connected to the input terminal of the inverter via a DC capacitor, and the output terminal of the inverter is connected to the input terminal of the filter. Each of the first FPGAs is used to simulate the equivalent model corresponding to the deployed DC-DC converter and the equivalent model corresponding to the inverter, respectively. The equivalent model corresponding to the DC-DC converter and the equivalent model corresponding to the inverter are decoupled at the DC capacitor. The DC capacitor is equivalent to a second current source connected to the equivalent model corresponding to the DC-DC converter, and the DC capacitor is equivalent to a DC voltage source connected to the equivalent model corresponding to the inverter.

5. The simulation system as described in claim 4, characterized in that, Each of the first FPGAs is configured to: simulate the equivalent model corresponding to the DC converter based on the current signal of the DC voltage source generated according to the equivalent model corresponding to the inverter, and generate the voltage signal of the second current source; and simulate the equivalent model corresponding to the inverter based on the voltage signal of the second current source generated according to the equivalent model corresponding to the DC converter and the current signal of the AC voltage source sent by the second FPGA, and generate the current signal of the DC voltage source and the voltage signal of the first current source sent to the second FPGA, respectively.

6. The simulation system as described in claim 5, characterized in that, Each of the first FPGAs is specifically used to: simulate the equivalent model corresponding to the DC converter based on the current signal of the DC voltage source generated by the equivalent model corresponding to the inverter in the previous simulation cycle, and generate the voltage signal of the second current source in the current simulation cycle; and simulate the equivalent model corresponding to the inverter based on the voltage signal of the second current source generated by the equivalent model corresponding to the DC converter in the previous simulation cycle and the current signal of the AC voltage source sent by the second FPGA in the previous simulation cycle, respectively generating the current signal of the DC voltage source in the current simulation cycle and sending the voltage signal of the first current source generated in the current simulation cycle to the second FPGA.

7. The simulation system according to any one of claims 4-6, characterized in that, Each of the first FPGAs is further configured to: simulate the equivalent model corresponding to the deployed photovoltaic module based on the terminal voltage signal generated by the equivalent model corresponding to the DC converter, and send the generated current source signal to the equivalent model corresponding to the DC converter; and simulate the equivalent model corresponding to the DC converter based on the current source signal generated by the equivalent model corresponding to the photovoltaic module and the current signal of the DC voltage source generated by the equivalent model corresponding to the inverter, and generate the terminal voltage signal and the voltage signal of the second current source.

8. The simulation system as described in claim 7, characterized in that, Each of the first FPGAs is specifically used to: simulate the equivalent model corresponding to the deployed photovoltaic module based on the terminal voltage signal generated by the equivalent model corresponding to the DC converter in the previous simulation cycle; send the current source signal generated in the current simulation cycle to the equivalent model corresponding to the DC converter; and simulate the equivalent model corresponding to the DC converter based on the current source signal generated by the equivalent model corresponding to the photovoltaic module in the previous simulation cycle and the current signal of the DC voltage source generated by the equivalent model corresponding to the inverter in the previous simulation cycle, thereby generating the terminal voltage signal and the voltage signal of the second current source in the current simulation cycle.

9. The simulation system according to any one of claims 1-8, characterized in that, The filter further includes a first filter inductor and a second filter inductor, wherein the first filter inductor is connected between the power converter and the filter capacitor, and the second filter inductor is connected between the filter capacitor and the transformer; The convergence model corresponding to the transformer deployed in the second FPGA is connected to the equivalent model corresponding to the second filter inductor between each of the AC voltage sources.

10. A simulation method for a photovoltaic power station, characterized in that, include: A photovoltaic power station model is established, which includes multiple grid-connected units and a transformer. Each grid-connected unit includes a photovoltaic module, a power converter, and a filter. The output terminal of the photovoltaic module is connected to the input terminal of the power converter, the output terminal of the power converter is connected to the input terminal of the filter, and the output terminals of multiple filters are connected in parallel and then connected to the transformer. The filter includes a filter capacitor, and the transformer is used to connect to the power grid. Decoupling is performed at the filter capacitor, and the photovoltaic power station model is divided into a detailed model corresponding to each grid-connected unit and a convergence model corresponding to the transformer. Each filter capacitor is equivalent to each AC voltage source connected to the convergence model corresponding to the transformer, and each filter capacitor is equivalent to each first current source connected to the detailed model corresponding to the grid-connected unit. Each first FPGA simulates the detailed model corresponding to each of the deployed grid-connected units, and the second FPGA simulates the convergence model corresponding to the deployed transformer.

11. The simulation method as described in claim 10, characterized in that, include: Each of the first FPGAs simulates the detailed model corresponding to the grid-connected unit based on the current signal of the AC voltage source received from the second FPGA, and sends the generated voltage signal of the first current source to the second FPGA. The second FPGA simulates the convergence model corresponding to the transformer based on the voltage signals of the first current sources received from each of the first FPGAs, and sends the current signals of each of the corresponding AC voltage sources to each of the first FPGAs.

12. The simulation method as described in claim 11, characterized in that, include: Each of the first FPGAs simulates the detailed model corresponding to the grid-connected unit based on the current signal of the AC voltage source sent by the second FPGA in the previous simulation cycle, and sends the voltage signal of the first current source generated in the current simulation cycle to the second FPGA. The second FPGA simulates the convergence model corresponding to the transformer based on the voltage signals of the first current sources sent by each of the first FPGAs in the previous simulation cycle, and sends the current signals of each AC voltage source generated in the current simulation cycle to each of the first FPGAs.

13. The simulation method according to any one of claims 10-12, characterized in that, The power converter includes a DC-DC converter and an inverter; the output terminal of the photovoltaic module is connected to the input terminal of the DC-DC converter, the output terminal of the DC-DC converter is connected to the input terminal of the inverter via a DC capacitor, and the output terminal of the inverter is connected to the input terminal of the filter. The simulation method further includes: Decoupling is performed at the DC capacitor, and the equivalent model corresponding to the power converter is divided into the equivalent model corresponding to the DC converter and the equivalent model corresponding to the inverter. The DC capacitor is equivalent to the second current source connected to the equivalent model corresponding to the DC converter, and the DC capacitor is equivalent to the DC voltage source connected to the equivalent model corresponding to the inverter. Each of the first FPGAs simulates the equivalent model corresponding to the deployed DC-DC converter and the equivalent model corresponding to the inverter.

14. The simulation method as described in claim 13, characterized in that, Also includes: Each of the first FPGAs simulates the equivalent model corresponding to the DC converter based on the current signal of the DC voltage source generated by the equivalent model corresponding to the inverter, and generates the voltage signal of the second current source. Each of the first FPGAs simulates the equivalent model corresponding to the inverter based on the voltage signal of the second current source generated by the equivalent model corresponding to the DC converter and the current signal of the AC voltage source sent by the second FPGA, thereby generating the current signal of the DC voltage source and the voltage signal of the first current source sent to the second FPGA.

15. The simulation method as described in claim 14, characterized in that, Specifically, it includes: Each of the first FPGAs simulates the equivalent model corresponding to the DC converter based on the current signal of the DC voltage source generated in the previous simulation cycle according to the equivalent model corresponding to the inverter, and generates the voltage signal of the second current source in the current simulation cycle. Each of the first FPGAs simulates the equivalent model corresponding to the inverter based on the voltage signal of the second current source generated by the equivalent model corresponding to the DC converter in the previous simulation cycle and the current signal of the AC voltage source sent by the second FPGA in the previous simulation cycle, thereby generating the current signal of the DC voltage source in the current simulation cycle and sending the voltage signal of the first current source generated in the current simulation cycle to the second FPGA.

16. The simulation method according to any one of claims 13-15, characterized in that, Also includes: Each of the first FPGAs simulates the equivalent model corresponding to the deployed photovoltaic module based on the terminal voltage signal generated by the equivalent model corresponding to the DC converter, and sends the generated current source signal to the equivalent model corresponding to the DC converter. Each of the first FPGAs simulates the equivalent model corresponding to the DC converter based on the current source signal generated by the equivalent model corresponding to the photovoltaic module and the current signal of the DC voltage source generated by the equivalent model corresponding to the inverter, thereby generating the terminal voltage signal and the voltage signal of the second current source.

17. The simulation method as described in claim 16, characterized in that, Specifically, it includes: Each of the first FPGAs simulates the equivalent model corresponding to the deployed photovoltaic module based on the terminal voltage signal generated by the equivalent model corresponding to the DC converter in the previous simulation cycle, and sends the current source signal generated in the current simulation cycle to the equivalent model corresponding to the DC converter. Each of the first FPGAs simulates the equivalent model corresponding to the DC converter based on the current source signal generated by the equivalent model corresponding to the photovoltaic module in the previous simulation cycle and the current signal of the DC voltage source generated by the equivalent model corresponding to the inverter in the previous simulation cycle, thereby generating the terminal voltage signal and the voltage signal of the second current source for the current simulation cycle.

Images