Controller software adjustment method, model training method, device, and storage medium

Abstract

The application discloses a controller software adjustment method, a model training method, equipment and a computer readable storage medium, and the method comprises the following steps: obtaining target installation scale parameter values, target installation environment parameter values, a renewable energy power generation platform simulation model and software control logic of a to-be-installed renewable energy power generation platform; predicting simulation controller parameter values of the renewable energy power generation platform according to the target installation scale parameter values, the target installation environment parameter values and the software control logic through the renewable energy power generation platform simulation model; if the simulation controller parameter values do not meet preset power generation conditions, adjusting the software control logic and returning to predicting the simulation controller parameter values of the to-be-installed renewable energy power generation platform according to the target installation scale, the target installation environment and the software control logic through the renewable energy power generation platform simulation model. The application realizes a scheme for adjusting controller software through simulation, so that the software adjustment cost of the controller is reduced.

Description

Technical Field

[0001] This invention relates to the field of power generation technology, and in particular to a controller software adjustment method, a model training method, an apparatus, and a computer-readable storage medium. Background Technology

[0002] In the field of power generation technology, the controllers of renewable energy power generation platforms typically use fixed software control logic. However, to ensure the power generation efficiency of renewable energy power generation platforms, they may be installed in unconventional environments such as open spaces or remote areas. This can easily lead to a situation where the fixed software control logic is incompatible with the renewable energy power generation platform installed in an unconventional environment. Adjusting the software control logic after the renewable energy power generation platform has been installed would result in excessively high adjustment costs for the controller's software control logic. Summary of the Invention

[0003] The main objective of this invention is to provide a controller software adjustment method, a model training method, an apparatus, and a computer-readable storage medium. The aim is to provide a scheme for adjusting the software control logic of a controller by obtaining a simulation model of a renewable energy power generation platform through model training and by simulating the renewable energy power generation platform to be installed based on the simulation model to obtain the simulated controller parameter values. This reduces the software adjustment cost of the controller.

[0004] To achieve the above objectives, the present invention provides a controller software adjustment method, the controller software adjustment method comprising the following steps:

[0005] Obtain the target installation scale parameter values, the corresponding target installation environment parameter values, the simulation model of the renewable energy power generation platform, and the software control logic corresponding to the renewable energy power generation platform to be installed;

[0006] The simulation model of the renewable energy power generation platform predicts the simulated controller parameter values ​​of the renewable energy power generation platform to be installed based on the target installation scale parameter values, the target installation environment parameter values, and the software control logic. The simulated controller parameter values ​​are the controller parameter values ​​in the simulation of the controller running with the software control logic in the renewable energy power generation platform to be installed.

[0007] If the simulated controller parameter values ​​do not meet the preset power generation conditions, the software control logic is adjusted, and the process returns to the step of predicting the simulated controller parameter values ​​of the renewable energy power generation platform to be installed based on the target installation scale, the target installation environment, and the software control logic using the renewable energy power generation platform simulation model, until the simulated controller parameter values ​​meet the preset power generation conditions.

[0008] Optionally, the analog controller parameter values ​​include at least one of the analog controller's voltage, current, and power.

[0009] Prior to the step of adjusting the software control logic, the method further includes:

[0010] Based on the simulated controller parameter values, the power generation benefits of the renewable energy power generation platform to be installed are predicted, wherein the power generation benefits include at least one of power generation efficiency and power generation;

[0011] Determine whether the power generation benefit is greater than a preset benefit threshold;

[0012] If so, then the parameter values ​​of the analog controller are determined to meet the preset power generation conditions;

[0013] If not, it is determined that the parameter values ​​of the analog controller do not meet the preset power generation conditions.

[0014] Optionally, the analog controller parameter values ​​include harmonic content.

[0015] Prior to the step of adjusting the software control logic, the method further includes:

[0016] Determine whether the parameter value of the analog controller is less than a preset harmonic content threshold;

[0017] If so, then the parameter values ​​of the analog controller are determined to meet the preset power generation conditions;

[0018] If not, it is determined that the parameter values ​​of the analog controller do not meet the preset power generation conditions.

[0019] To achieve the above objectives, the present invention provides a model training method, which includes the following steps:

[0020] Multiple training samples are obtained, wherein each training sample includes an input feature data and a training label corresponding to the input feature data. The input feature data includes an installation environment parameter value, an installation scale parameter value, and a preset software control logic. The training label includes the actual controller parameter value when the controller in the renewable energy power generation platform is controlled by the preset software control logic under the installation environment defined by the installation environment parameter value and the installation scale defined by the installation scale parameter value.

[0021] Generate real eye maps corresponding to the multiple training samples, wherein the real eye maps are used to characterize the correspondence between the input feature data and the real controller parameter values ​​in each of the training samples;

[0022] Based on the input feature data, the simulation controller parameter values ​​are predicted by the simulation model of the renewable energy power generation platform to be trained, and a simulation eye diagram corresponding to the simulation model of the renewable energy power generation platform to be trained is generated based on the simulation controller parameter values ​​and the input feature data. The simulation eye diagram is used to characterize the correspondence between the input feature data and the simulation controller parameter values ​​in each training sample.

[0023] Based on the overlap between the real eye diagram and the simulated eye diagram, the simulation model of the renewable energy power generation platform to be trained is iteratively optimized to obtain the simulation model of the renewable energy power generation platform.

[0024] Optionally, the simulation model of the renewable energy power generation platform to be trained includes a fully virtual renewable energy power generation platform simulation model to be trained.

[0025] The step of predicting the simulation controller parameter values ​​based on the input feature data using the simulation model of the renewable energy power generation platform to be trained includes:

[0026] The input feature data is mapped to the simulation controller parameter values ​​through the simulation model of the fully virtual renewable energy power generation platform to be trained.

[0027] Optionally, the simulation model of the renewable energy power generation platform to be trained includes a simulation model of a semi-physical renewable energy power generation platform to be trained, and the simulation model of the semi-physical renewable energy power generation platform to be trained further includes a physical controller.

[0028] The step of predicting the simulation controller parameter values ​​based on the input feature data using the simulation model of the renewable energy power generation platform to be trained includes:

[0029] The physical controller converts the installation environment parameter values ​​into input installation environment electrical signal values ​​according to preset parameter conversion rules.

[0030] The input installation environment electrical signal value, the installation scale parameter value, and the preset software control logic are concatenated into semi-physical input feature data;

[0031] The semi-physical input feature data is mapped to the simulation controller parameter values ​​through the simulation model of the semi-physical renewable energy power generation platform to be trained.

[0032] Optionally, the physical controller includes environmental parameter sensors.

[0033] Before the step of converting the installation environment parameter values ​​into input installation environment electrical signal values ​​by the physical controller according to a preset parameter conversion rule, the method further includes:

[0034] According to the preset parameter conversion rules, the physical controller converts the installation environment parameter values ​​into regular environment parameter electrical signal values.

[0035] The analytical installation environment parameter values ​​are obtained by analyzing the electrical signal values ​​of the training installation environment parameters using the sensor.

[0036] The preset parameter conversion rules are adjusted and optimized based on the difference between the installation environment parameter values ​​and the parsed installation environment parameter values.

[0037] Optionally, the step of iteratively optimizing the simulation model of the renewable energy power generation platform to be trained based on the overlap between the real eye diagram and the simulated eye diagram to obtain the simulation model of the renewable energy power generation platform includes:

[0038] If the simulation model of the renewable energy power generation platform to be trained includes the simulation model of the fully virtual renewable energy power generation platform to be trained, then it is determined whether the overlap is greater than the first preset overlap threshold.

[0039] If the overlap degree is greater than the first preset overlap degree threshold, then the simulation model of the fully virtual renewable energy power generation platform to be trained is used as the simulation model of the renewable energy power generation platform; or...

[0040] If the simulation model of the renewable energy power generation platform to be trained includes the simulation model of the semi-physical renewable energy power generation platform to be trained, then it is determined whether the overlap is greater than the second preset overlap threshold, wherein the second preset overlap threshold is greater than the first preset overlap threshold.

[0041] If the overlap is greater than the second preset overlap threshold, then the semi-physical renewable energy power generation platform simulation model to be trained will be used as the renewable energy power generation platform simulation model.

[0042] Optionally, before the step of iteratively optimizing the simulation model of the renewable energy power generation platform to be trained based on the overlap between the real eye diagram and the simulated eye diagram to obtain the simulation model of the renewable energy power generation platform, the method further includes:

[0043] Identify and remove spiking signals from the real eye diagram.

[0044] To achieve the above objectives, the present invention also provides a controller software adjustment device, the controller software adjustment device comprising:

[0045] The first acquisition module is used to acquire the target installation scale parameter values, the corresponding target installation environment parameter values, the simulation model of the renewable energy power generation platform, and the software control logic corresponding to the renewable energy power generation platform to be installed.

[0046] The prediction module is used to predict the simulation controller parameter values ​​of the renewable energy power generation platform to be installed based on the target installation scale parameter values, the target installation environment parameter values, and the software control logic through the renewable energy power generation platform simulation model. The simulation controller parameter values ​​are the controller parameter values ​​in the simulation of the controller running with the software control logic in the renewable energy power generation platform to be installed.

[0047] An adjustment module is used to adjust the software control logic if the simulated controller parameter values ​​do not meet the preset power generation conditions, and then return to the step of predicting the simulated controller parameter values ​​of the renewable energy power generation platform to be installed based on the target installation scale, the target installation environment, and the software control logic through the renewable energy power generation platform simulation model, until the simulated controller parameter values ​​meet the preset power generation conditions.

[0048] To achieve the above objectives, the present invention also provides a model training apparatus, the model training apparatus comprising:

[0049] The second acquisition module is used to acquire multiple training samples, wherein each training sample includes an input feature data and a training label corresponding to the input feature data. The input feature data includes an installation environment parameter value, an installation scale parameter value, and a preset software control logic. The training label includes the actual controller parameter value when the controller in the renewable energy power generation platform is controlled by the preset software control logic under the installation environment limited by the installation environment parameter value and the installation scale limited by the installation scale parameter value.

[0050] The first generation module is used to generate real eye maps corresponding to the plurality of training samples, wherein the real eye maps are used to characterize the correspondence between the input feature data in each of the training samples and the real controller parameter values;

[0051] The second generation module is used to predict the simulation controller parameter values ​​through the simulation model of the renewable energy power generation platform to be trained based on the input feature data, and generate the simulation eye diagram corresponding to the simulation model of the renewable energy power generation platform to be trained based on the simulation controller parameter values ​​and the input feature data. The simulation eye diagram is used to characterize the correspondence between the input feature data and the simulation controller parameter values ​​in each training sample.

[0052] The iteration module is used to iteratively optimize the simulation model of the renewable energy power generation platform to be trained based on the overlap between the real eye diagram and the simulated eye diagram, so as to obtain the simulation model of the renewable energy power generation platform.

[0053] To achieve the above objectives, the present invention also provides a controller software adjustment device, the controller software adjustment device comprising: a memory, a processor, and a controller software adjustment program stored in the memory and executable on the processor, wherein the controller software adjustment program, when executed by the processor, implements the steps of the controller software adjustment method as described above.

[0054] To achieve the above objectives, the present invention also provides a model training device, the model training device comprising: a memory, a processor, and a model training program stored in the memory and executable on the processor, wherein the controller software adjustment program, when executed by the processor, implements the steps of the model training method as described above.

[0055] Furthermore, to achieve the above objectives, the present invention also proposes a computer-readable storage medium storing a controller software adjustment program and a model training program. When the controller software adjustment program is executed by a processor, it implements the steps of the controller software adjustment method as described above, and when the model training program is executed by a processor, it implements the steps of the model training method as described above.

[0056] In this invention, the target installation scale parameter value, the target installation environment parameter value, the renewable energy power generation platform simulation model, and the software control logic corresponding to the renewable energy power generation platform to be installed are obtained. Based on the target installation scale parameter value, the target installation environment parameter value, and the software control logic, the renewable energy power generation platform simulation model predicts the simulated controller parameter value of the renewable energy power generation platform to be installed. The simulated controller parameter value is the controller parameter value simulating the operation of the controller in the renewable energy power generation platform to be installed using the software control logic. If the simulated controller parameter value does not meet the preset power generation conditions, the software control logic is adjusted, and the process returns to the step of predicting the simulated controller parameter value of the renewable energy power generation platform to be installed based on the target installation scale, the target installation environment, and the software control logic, until the simulated controller parameter value meets the preset power generation conditions. By simulating the renewable energy power generation platform to be installed, the simulated controller parameter value obtained from the simulation can be judged. This allows for adjustment of the controller's software control logic before the renewable energy power generation platform is installed and when the simulated controller parameter value does not meet the preset power generation conditions, thereby reducing the cost of adjusting the controller's software control logic.

[0057] In this invention, multiple training samples are obtained. Each training sample includes one input feature data and a corresponding training label. The input feature data includes an installation environment parameter value, an installation scale parameter value, and a preset software control logic. The training label includes the actual controller parameter value when the controller in the renewable energy power generation platform is controlled by the preset software control logic under the installation environment defined by the installation environment parameter value and the installation scale defined by the installation scale parameter value. A true eye diagram corresponding to the multiple training samples is generated. The true eye diagram is used to characterize the correspondence between the input feature data and the actual controller parameter value in each training sample. Based on the input feature data, the simulated controller parameter value is predicted by the simulation model of the renewable energy power generation platform to be trained, and the simulated controller parameter value is then used to predict the simulated controller parameter value. The simulation eye diagram corresponding to the simulation model of the renewable energy power generation platform to be trained is generated by the input feature data and the input feature data. The simulation eye diagram is used to characterize the correspondence between the input feature data and the simulation controller parameter values ​​in each training sample. The simulation model of the renewable energy power generation platform to be trained is iteratively optimized according to the overlap between the real eye diagram and the simulation eye diagram to obtain the simulation model of the renewable energy power generation platform. The simulation model of the renewable energy power generation platform to be trained is trained according to various installation environment parameter values, installation scale parameter values, various preset software control logics and real controller parameter values ​​of the installed renewable energy power generation platform. This makes the simulation controller parameter values ​​output by the trained renewable energy power generation platform simulation model based on the input feature data more realistic, thereby improving the simulation accuracy of the renewable energy power generation platform simulation model. Attached Figure Description

[0058] Figure 1 This is a schematic diagram of the hardware operating environment involved in the embodiments of the present invention;

[0059] Figure 2 This is another schematic diagram of the hardware operating environment involved in the embodiments of the present invention;

[0060] Figure 3 This is a flowchart illustrating the first embodiment of the controller software adjustment method of the present invention;

[0061] Figure 4 This is a flowchart illustrating the first embodiment of the model training method of the present invention;

[0062] Figure 5 This is a schematic diagram of the functional modules of a preferred embodiment of the controller software adjustment device of the present invention;

[0063] Figure 6This is a schematic diagram of the functional modules of a preferred embodiment of the model training device of the present invention.

[0064] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0065] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0066] like Figure 1 As shown, Figure 1 This is a schematic diagram of a device structure in the hardware operating environment involved in the embodiments of the present invention.

[0067] It should be noted that the controller software adjustment device in this embodiment of the invention can be a smartphone, personal computer, server, or other device, and no specific limitation is made here.

[0068] like Figure 1 As shown, the controller software adjustment device may include: a processor 1001, such as a CPU; a network interface 1004; a user interface 1003; a memory 1005; and a communication bus 1002. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display screen or an input unit such as a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be high-speed RAM or non-volatile memory, such as a disk drive. Optionally, the memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0069] Those skilled in the art will understand that Figure 1 The device structure shown does not constitute a limitation on the controller software adjustment device, which may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0070] like Figure 1 As shown, the memory 1005, as a computer storage medium, may include an operating system, a network communication module, a user interface module, and a controller software adjustment program. The operating system is a program that manages and controls the hardware and software resources of the device, supporting the operation of the controller software adjustment program and other software or programs. Figure 1In the device shown, the user interface 1003 is mainly used for data communication with the client; the network interface 1004 is mainly used for establishing a communication connection with the server; and the processor 1001 can be used to call the controller software adjustment program stored in the memory 1005 and perform the following operations:

[0071] Obtain the target installation scale parameter values, the corresponding target installation environment parameter values, the simulation model of the renewable energy power generation platform, and the software control logic corresponding to the renewable energy power generation platform to be installed;

[0072] The simulation model of the renewable energy power generation platform predicts the simulated controller parameter values ​​of the renewable energy power generation platform to be installed based on the target installation scale parameter values, the target installation environment parameter values, and the software control logic. The simulated controller parameter values ​​are the controller parameter values ​​in the simulation of the controller running with the software control logic in the renewable energy power generation platform to be installed.

[0073] If the simulated controller parameter values ​​do not meet the preset power generation conditions, the software control logic is adjusted, and the process returns to the step of predicting the simulated controller parameter values ​​of the renewable energy power generation platform to be installed based on the target installation scale, the target installation environment, and the software control logic using the renewable energy power generation platform simulation model, until the simulated controller parameter values ​​meet the preset power generation conditions.

[0074] Furthermore, the analog controller parameter values ​​include at least one of the analog controller's voltage, current, and power. Before adjusting the software control logic, the processor 1001 can also call the controller software adjustment program stored in the memory 1005 to perform the following operations:

[0075] Based on the simulated controller parameter values, the power generation benefits of the renewable energy power generation platform to be installed are predicted, wherein the power generation benefits include at least one of power generation efficiency and power generation;

[0076] Determine whether the power generation benefit is greater than a preset benefit threshold;

[0077] If so, then the parameter values ​​of the analog controller are determined to meet the preset power generation conditions;

[0078] If not, it is determined that the parameter values ​​of the analog controller do not meet the preset power generation conditions.

[0079] Furthermore, the analog controller parameter values ​​include harmonic content. Before adjusting the software control logic, the processor 1001 can also call the controller software adjustment program stored in the memory 1005 to perform the following operations:

[0080] Determine whether the parameter value of the analog controller is less than a preset harmonic content threshold;

[0081] If so, then the parameter values ​​of the analog controller are determined to meet the preset power generation conditions;

[0082] If not, it is determined that the parameter values ​​of the analog controller do not meet the preset power generation conditions.

[0083] like Figure 2 As shown, Figure 2 This is a schematic diagram of another device structure in the hardware operating environment involved in the embodiments of the present invention.

[0084] It should be noted that the model training device in this embodiment of the invention can be a smartphone, a personal computer, a server, or other devices, and no specific limitations are imposed here.

[0085] like Figure 2 As shown, the model training device may include: a processor 1006, such as a CPU; a network interface 1009; a user interface 1008; a memory 1010; and a communication bus 1007. The communication bus 1007 is used to establish communication between these components. The user interface 1008 may include a display screen and an input unit such as a keyboard; optionally, the user interface 1008 may also include a standard wired interface or a wireless interface. The network interface 1009 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1010 may be high-speed RAM or stable non-volatile memory, such as a disk drive. Optionally, the memory 1010 may also be a storage device independent of the aforementioned processor 1006.

[0086] Those skilled in the art will understand that Figure 2 The device structure shown does not constitute a limitation on the model training device and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0087] like Figure 2 As shown, the memory 1010, as a computer storage medium, may include an operating system, a network communication module, a user interface module, and a model training program. The operating system is a program that manages and controls the device's hardware and software resources, supporting the operation of the model training program and other software or programs. Figure 2 In the device shown, the user interface 1008 is mainly used for data communication with the client; the network interface 1009 is mainly used for establishing a communication connection with the server; and the processor 1006 can be used to call the model training program stored in the memory 1010 and perform the following operations:

[0088] Multiple training samples are obtained, wherein each training sample includes an input feature data and a training label corresponding to the input feature data. The input feature data includes an installation environment parameter value, an installation scale parameter value, and a preset software control logic. The training label includes the actual controller parameter value when the controller in the renewable energy power generation platform is controlled by the preset software control logic under the installation environment defined by the installation environment parameter value and the installation scale defined by the installation scale parameter value.

[0089] Generate real eye maps corresponding to the multiple training samples, wherein the real eye maps are used to characterize the correspondence between the input feature data and the real controller parameter values ​​in each of the training samples;

[0090] Based on the input feature data, the simulation controller parameter values ​​are predicted by the simulation model of the renewable energy power generation platform to be trained, and a simulation eye diagram corresponding to the simulation model of the renewable energy power generation platform to be trained is generated based on the simulation controller parameter values ​​and the input feature data. The simulation eye diagram is used to characterize the correspondence between the input feature data and the simulation controller parameter values ​​in each training sample.

[0091] Based on the overlap between the real eye diagram and the simulated eye diagram, the simulation model of the renewable energy power generation platform to be trained is iteratively optimized to obtain the simulation model of the renewable energy power generation platform.

[0092] Furthermore, the simulation model of the renewable energy power generation platform to be trained includes a fully virtual renewable energy power generation platform simulation model to be trained, and the operation of predicting the simulation controller parameter values ​​through the simulation model of the renewable energy power generation platform to be trained based on the input feature data includes:

[0093] The input feature data is mapped to the simulation controller parameter values ​​through the simulation model of the fully virtual renewable energy power generation platform to be trained.

[0094] Furthermore, the simulation model of the renewable energy power generation platform to be trained includes a semi-physical renewable energy power generation platform simulation model to be trained, which further includes a physical controller. The operation of predicting the simulation controller parameter values ​​through the simulation model of the renewable energy power generation platform to be trained based on the input feature data includes:

[0095] The physical controller converts the installation environment parameter values ​​into input installation environment electrical signal values ​​according to preset parameter conversion rules.

[0096] The input installation environment electrical signal value, the installation scale parameter value, and the preset software control logic are concatenated into semi-physical input feature data;

[0097] The semi-physical input feature data is mapped to the simulation controller parameter values ​​through the simulation model of the semi-physical renewable energy power generation platform to be trained.

[0098] Furthermore, the physical controller includes an environmental parameter sensor. Before the operation of converting the installation environment parameter value into an input installation environment electrical signal value by the physical controller according to a preset parameter conversion rule, the processor 1006 can also be used to call the model training program stored in the memory 1010 and perform the following operations:

[0099] According to the preset parameter conversion rules, the physical controller converts the installation environment parameter values ​​into regular environment parameter electrical signal values.

[0100] The analytical installation environment parameter values ​​are obtained by analyzing the electrical signal values ​​of the training installation environment parameters using the sensor.

[0101] The preset parameter conversion rules are adjusted and optimized based on the difference between the installation environment parameter values ​​and the parsed installation environment parameter values.

[0102] Furthermore, the step of iteratively optimizing the simulation model of the renewable energy power generation platform to be trained based on the overlap between the real eye diagram and the simulated eye diagram to obtain the simulation model of the renewable energy power generation platform includes:

[0103] If the simulation model of the renewable energy power generation platform to be trained includes the simulation model of the fully virtual renewable energy power generation platform to be trained, then it is determined whether the overlap is greater than the first preset overlap threshold.

[0104] If the overlap degree is greater than the first preset overlap degree threshold, then the simulation model of the fully virtual renewable energy power generation platform to be trained is used as the simulation model of the renewable energy power generation platform; or...

[0105] If the simulation model of the renewable energy power generation platform to be trained includes the simulation model of the semi-physical renewable energy power generation platform to be trained, then it is determined whether the overlap is greater than the second preset overlap threshold, wherein the second preset overlap threshold is greater than the first preset overlap threshold.

[0106] If the overlap is greater than the second preset overlap threshold, then the semi-physical renewable energy power generation platform simulation model to be trained will be used as the renewable energy power generation platform simulation model.

[0107] Furthermore, before iteratively optimizing the simulation model of the renewable energy power generation platform to be trained based on the overlap between the real eye diagram and the simulated eye diagram to obtain the simulation model of the renewable energy power generation platform, the processor 1006 can also be used to call the model training program stored in the memory 1010 to perform the following operations:

[0108] Identify and remove spiking signals from the real eye diagram.

[0109] Based on the above structure, various embodiments of the controller software adjustment method are proposed.

[0110] Reference Figure 3 , Figure 3 This is a flowchart illustrating the first embodiment of the controller software adjustment method of the present invention.

[0111] This invention provides an embodiment of a controller software adjustment method. It should be noted that although the logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order. In this embodiment, the executing entity of the controller software adjustment method can be a personal computer, smartphone, server, or other device; no limitation is made in this embodiment. For ease of description, the execution entity is omitted from the description of each embodiment. In this embodiment, the controller software adjustment method includes:

[0112] Step S10: Obtain the target installation scale parameter value, the target installation environment parameter value, the simulation model of the renewable energy power generation platform, and the software control logic corresponding to the renewable energy power generation platform to be installed;

[0113] In this embodiment, it should be noted that the renewable energy power generation platform is a platform that generates electricity using renewable energy sources, such as wind power, solar power, hydropower, and other non-fossil energy sources. The target installation scale parameter value is the platform scale parameter value of the renewable energy power generation platform to be installed, and the target installation environment parameter value is the environmental parameter value of the installation environment corresponding to the renewable energy power generation platform to be installed.

[0114] In one feasible embodiment, when the renewable energy power generation platform is a photovoltaic (PV) power generation platform, the target installation scale parameter is the target PV installation scale parameter value, and the target installation environment parameter value is the target PV installation environment parameter value. The target PV installation scale parameter value includes at least one of PV panel parameter values, PV grid parameter values, and PV circuit topology parameter values. The PV panel parameter values ​​include at least one of parameters such as PV panel type, PV panel power specification, and PV panel material. The PV panel type can be a single-sided PV panel, or a double-sided PV panel, or other types. The PV panel power specification can be 250W, or... The power rating can be 270W, 310W, or other specifications. The photovoltaic panel material can be amorphous silicon, monocrystalline silicon, polycrystalline silicon, or other materials. The target photovoltaic panel installation environment parameters include one of the following: sunshine parameters, temperature, meteorological parameters, meteorological conditions, and cloud parameters. The sunshine parameters include at least one of solar irradiance and angle of incidence. The meteorological parameters include at least one of rainfall, snowfall, hail size, rainfall duration, snowfall duration, visible particulate matter information, and hail duration. The meteorological conditions include at least one of snowy days, rainy days, foggy days, sunny days, cloudy days, and hail days.

[0115] In another feasible embodiment, when the renewable energy power generation platform is a wind power generation platform, the target installation scale parameter is the target wind power installation scale parameter value, the target installation environment parameter value is the target wind power installation environment parameter value, the target wind power installation scale parameter value includes the wind power platform blade specifications, wind power grid parameter values, and wind power circuit topology parameter values, the wind power platform blade specifications include at least one of the following: number of blades, blade shape, blade size, and blade material, and the target wind power installation environment parameter value includes the blade rotation speed, wherein the blade rotation speed is the rotation speed of the blades in the wind power generation platform under the constraints of the installation environment.

[0116] In another feasible embodiment, when the renewable energy power generation platform is an energy storage power generation platform, the target installation scale parameter is the target energy storage installation scale parameter value, the target installation environment parameter value is the target energy storage installation environment parameter value, the target energy storage installation scale parameter value includes the energy storage platform battery capacity, energy storage grid parameter value and energy storage circuit topology parameter value, and the target energy storage installation environment parameter value includes at least one of battery temperature and battery output voltage.

[0117] Step S20: Based on the target installation scale parameter value, the target installation environment parameter value, and the software control logic, the simulation model of the renewable energy power generation platform predicts the simulation controller parameter value of the renewable energy power generation platform to be installed. The simulation controller parameter value is the controller parameter value when the controller in the simulated renewable energy power generation platform to be installed is running with the software control logic.

[0118] In this embodiment, it should be noted that the controller is a device that converts the direct current of the renewable energy power generation platform into alternating current. When the renewable energy platform is a photovoltaic power generation platform, the controller is an inverter.

[0119] For example, the target installation scale parameter value, the target installation environment parameter value, and the software control logic are concatenated into target input feature data, and the target input feature data is mapped to the simulation controller parameter value of the renewable energy power generation platform to be installed through the renewable energy power generation platform simulation model.

[0120] In one feasible embodiment, the renewable energy power generation platform simulation model includes a fully virtual renewable energy power generation platform simulation model trained by the following fully virtual renewable energy power generation platform simulation model to be trained, through which the target input feature data is mapped to the simulation controller parameter values ​​of the renewable energy power generation platform to be installed.

[0121] In another feasible embodiment, the renewable energy power generation platform simulation model includes a semi-physical renewable energy power generation platform simulation model trained by the semi-physical renewable energy power generation platform simulation model to be trained, through which the target input feature data is mapped to the simulation controller parameter values ​​of the renewable energy power generation platform to be installed.

[0122] Step S30: If the simulated controller parameter value does not meet the preset power generation condition, the software control logic is adjusted, and the process returns to the step of predicting the simulated controller parameter value of the renewable energy power generation platform to be installed based on the target installation scale, the target installation environment, and the software control logic using the renewable energy power generation platform simulation model, until the simulated controller parameter value meets the preset power generation condition.

[0123] For example, it is determined whether the simulation controller parameter value meets the preset power generation condition. If the simulation controller parameter value does not meet the preset power generation condition, the software control logic is adjusted according to the simulation controller parameter value, and the process returns to the step of predicting the simulation controller parameter value of the renewable energy power generation platform to be installed based on the target installation scale, the target installation environment, and the software control logic through the renewable energy power generation platform simulation model, until the simulation controller parameter value meets the preset power generation condition.

[0124] In this invention, the target installation scale parameter value, the target installation environment parameter value, the renewable energy power generation platform simulation model, and the software control logic corresponding to the renewable energy power generation platform to be installed are obtained. Based on the target installation scale parameter value, the target installation environment parameter value, and the software control logic, the renewable energy power generation platform simulation model predicts the simulated controller parameter value of the renewable energy power generation platform to be installed. The simulated controller parameter value is the controller parameter value simulating the operation of the controller in the renewable energy power generation platform to be installed using the software control logic. If the simulated controller parameter value does not meet the preset power generation conditions, the software control logic is adjusted, and the process returns to the step of predicting the simulated controller parameter value of the renewable energy power generation platform to be installed based on the target installation scale, the target installation environment, and the software control logic, until the simulated controller parameter value meets the preset power generation conditions. By simulating the renewable energy power generation platform to be installed, the simulated controller parameter value obtained from the simulation can be judged. This allows for adjustment of the controller's software control logic before the renewable energy power generation platform is installed and when the simulated controller parameter value does not meet the preset power generation conditions, thereby reducing the cost of adjusting the controller's software control logic.

[0125] Furthermore, based on the first embodiment of the controller software adjustment method described above, a second embodiment of the controller software adjustment method of the present invention is proposed. In this embodiment, the analog controller parameter values ​​include at least one of the voltage, current, and power of the analog controller. Before step S30, the method further includes:

[0126] Step A10: Based on the parameter values ​​of the simulation controller, predict the power generation benefits of the renewable energy power generation platform to be installed, wherein the power generation benefits include at least one of power generation efficiency and power generation.

[0127] In this embodiment, it should be noted that the voltage includes at least one of DC-side voltage and AC-side voltage, the current includes at least one of DC-side current and AC-side current, and the power includes at least one of DC-side input power and AC-side output power.

[0128] For example, a preset benefit calculation formula is obtained, and the power generation benefit of the renewable energy power generation platform to be installed is calculated based on the preset benefit calculation formula and the parameter values ​​of the simulation controller.

[0129] In one feasible embodiment, the preset benefit calculation formula includes a power generation efficiency calculation formula and a power generation calculation formula. The power generation efficiency of the renewable energy power generation platform to be installed is calculated based on the power generation efficiency calculation formula and the simulation controller parameter values, and / or the power generation of the renewable energy power generation platform to be installed is calculated based on the power generation calculation formula and the simulation controller parameter values.

[0130] Step A20: Determine whether the power generation benefit is greater than a preset benefit threshold;

[0131] In this embodiment, it should be noted that the preset benefit threshold is a pre-set critical value for determining whether the power generation benefit of the renewable energy power generation platform meets the requirements.

[0132] In one feasible embodiment, it is determined whether the power generation efficiency is greater than a preset efficiency threshold, wherein the preset efficiency threshold is a pre-set critical value for determining that the power generation efficiency of the renewable energy power generation platform meets the requirements, and / or, it is determined whether the power generation is greater than a preset power generation threshold, wherein the preset power generation threshold is a pre-set critical value for determining that the power generation of the renewable energy power generation platform meets the requirements.

[0133] Step A30: If yes, then determine that the parameter values ​​of the analog controller meet the preset power generation conditions;

[0134] In one feasible implementation, if the power generation efficiency is greater than a preset efficiency threshold, and / or if the power generation is greater than a preset power generation threshold, then the analog controller parameter value is determined to meet the preset power generation condition.

[0135] Step A40: If not, then it is determined that the parameter values ​​of the analog controller do not meet the preset power generation conditions.

[0136] In one feasible implementation, if the power generation efficiency is less than or equal to a preset efficiency threshold, and / or if the power generation is less than or equal to a preset power generation threshold, then it is determined that the analog controller parameter value does not meet the preset power generation condition.

[0137] By assessing the power generation benefits of the renewable energy power generation platform, namely its power generation efficiency and output, the software control logic on the renewable energy power generation platform to be installed can be designed to ensure the platform's power generation benefits.

[0138] The analog controller parameter values ​​include harmonic content, and prior to step S30, the following is also included:

[0139] Step B10: Determine whether the parameter value of the analog controller is less than the preset harmonic content threshold;

[0140] In this embodiment, it should be noted that the preset harmonic content threshold is a pre-set harmonic content critical value for determining that the renewable energy power generation platform has no risk of equipment damage. The harmonic content can be the harmonic content on the AC side of the controller.

[0141] For example, it is determined whether the harmonic content is less than a preset harmonic content threshold.

[0142] Step B20: If yes, then determine that the parameter values ​​of the analog controller meet the preset power generation conditions;

[0143] Step B30: If not, then it is determined that the parameter values ​​of the simulation controller do not meet the preset power generation conditions.

[0144] For example, steps B20 to B30 include: if the harmonic content is less than a preset harmonic content threshold, then the analog controller parameter value is determined to meet the preset power generation condition; if the harmonic content is greater than or equal to the preset harmonic content threshold, then the analog controller parameter value is determined to not meet the preset power generation condition.

[0145] It is understandable that when the harmonic content on the AC side of the controller is high, it can easily damage the renewable energy power generation platform. Therefore, by judging whether the parameter value of the analog controller is less than the preset harmonic content threshold, if so, it is determined that the parameter value of the analog controller meets the preset power generation condition; if not, it is determined that the parameter value of the analog controller does not meet the preset power generation condition. This allows for timely adjustment of the controller's software control logic when the harmonic content is high, thus avoiding equipment damage that can easily occur when the installed renewable energy power generation platform uses fixed software control logic, thereby ensuring the health of the renewable energy power generation platform after installation.

[0146] Based on the above structure, various embodiments of the model training method are proposed.

[0147] Reference Figure 4 , Figure 4 This is a flowchart illustrating the first embodiment of the model training method of the present invention.

[0148] This invention provides an embodiment of a model training method. It should be noted that although the logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order. In this embodiment, the executing entity of the model training method can be a personal computer, smartphone, server, or other device; no limitation is made in this embodiment. For ease of description, the execution entity is omitted from the description of each embodiment. In this embodiment, the model training method includes:

[0149] Step C10: Obtain multiple training samples, wherein each training sample includes an input feature data and a training label corresponding to the input feature data. The input feature data includes an installation environment parameter value, an installation scale parameter value, and a preset software control logic. The training label includes the actual controller parameter value when the controller in the renewable energy power generation platform is controlled by the preset software control logic under the installation environment limited by the installation environment parameter value and the installation scale limited by the installation scale parameter value.

[0150] In this embodiment, it should be noted that the renewable energy platform is an already installed platform.

[0151] Step C20: Generate the real eye diagrams corresponding to the multiple training samples, wherein the real eye diagrams are used to characterize the correspondence between the input feature data and the real controller parameter values ​​in each of the training samples;

[0152] In one feasible implementation, the input feature data in each of the training samples is used as the x-axis value and the real controller parameter value is used as the y-axis value; or, the real controller parameter value is used as the x-axis value and the input feature data in each of the training samples is used as the y-axis value. At least one straight line is fitted to generate the real eye diagram corresponding to the multiple training samples. For example, input feature data 1 corresponds to real controller parameter 1, and input feature data 2 corresponds to real controller parameter 2.

[0153] Step C30: Based on the input feature data, the simulation controller parameter values ​​are predicted by the simulation model of the renewable energy power generation platform to be trained, and a simulation eye diagram corresponding to the simulation model of the renewable energy power generation platform to be trained is generated based on the simulation controller parameter values ​​and the input feature data. The simulation eye diagram is used to characterize the correspondence between the input feature data and the simulation controller parameter values ​​in each training sample.

[0154] For example, the input feature data is mapped to analog controller parameter values ​​through a simulation model of the renewable energy power generation platform to be trained.

[0155] In one feasible embodiment, the input feature data in each of the training samples is used as the abscissa value and the simulation controller parameter value is used as the ordinate value; or, the simulation controller parameter value is used as the abscissa value and the input feature data in each of the training samples is used as the ordinate value. At least one straight line is fitted to generate the simulation eye diagram corresponding to the simulation model of the renewable energy power generation platform to be trained. For example, input feature data 1 corresponds to simulation controller parameter 1, and input feature data 2 corresponds to simulation controller parameter 2.

[0156] Step C40: Based on the overlap between the real eye diagram and the simulated eye diagram, iteratively optimize the simulation model of the renewable energy power generation platform to be trained to obtain the simulation model of the renewable energy power generation platform.

[0157] For example, the overlap between the real eye diagram and the simulated eye diagram is obtained. Based on the overlap, it is determined whether the simulation model of the renewable energy power generation platform to be trained meets the preset model accuracy. If the simulation model of the renewable energy power generation platform to be trained meets the preset model accuracy, then the simulation model of the renewable energy power generation platform to be trained is used as the renewable energy power generation platform simulation model. If the simulation model of the renewable energy power generation platform to be trained does not meet the preset model accuracy, then the adjustable support control model is updated using the gradient of the loss calculation of the simulation model of the renewable energy power generation platform to be trained, which is calculated based on the overlap, through a preset model update method. The preset model update method includes gradient descent and gradient ascent methods, etc.

[0158] Prior to step C40, the procedure further includes:

[0159] Step C01: Identify the spiking signals in the real eye diagram and remove the spiking signals from the real eye diagram.

[0160] It is understandable that when obtaining the actual controller parameter values ​​of a renewable energy power generation platform, there may be noise signals in the environment. However, the simulated controller parameter values ​​predicted by the simulation model of the renewable energy power generation platform to be trained are simulated in an ideal environment or in a laboratory environment, so there are no noise signals. If the actual eye diagram and the simulated eye diagram are directly compared, the actual controller parameter values ​​in the actual eye diagram may be subject to noise interference, resulting in the inaccuracy of the trained simulation model of the renewable energy power generation platform.

[0161] By identifying and removing glitch signals in the real eye diagram, the noise interference of the real controller parameter values ​​in the real eye diagram is avoided, thereby improving the accuracy of the simulation model of the renewable energy power generation platform obtained through training.

[0162] In this invention, multiple training samples are obtained. Each training sample includes one input feature data and a corresponding training label. The input feature data includes an installation environment parameter value, an installation scale parameter value, and a preset software control logic. The training label includes the actual controller parameter value when the controller in the renewable energy power generation platform is controlled by the preset software control logic under the installation environment defined by the installation environment parameter value and the installation scale defined by the installation scale parameter value. A true eye diagram corresponding to the multiple training samples is generated. The true eye diagram is used to characterize the correspondence between the input feature data and the actual controller parameter value in each training sample. Based on the input feature data, the simulated controller parameter value is predicted by the simulation model of the renewable energy power generation platform to be trained, and the simulated controller parameter value is then used to predict the simulated controller parameter value. The simulation eye diagram corresponding to the simulation model of the renewable energy power generation platform to be trained is generated by the input feature data and the input feature data. The simulation eye diagram is used to characterize the correspondence between the input feature data and the simulation controller parameter values ​​in each training sample. The simulation model of the renewable energy power generation platform to be trained is iteratively optimized according to the overlap between the real eye diagram and the simulation eye diagram to obtain the simulation model of the renewable energy power generation platform. The simulation model of the renewable energy power generation platform to be trained is trained according to various installation environment parameter values, installation scale parameter values, various preset software control logics and real controller parameter values ​​of the installed renewable energy power generation platform. This makes the simulation controller parameter values ​​output by the trained renewable energy power generation platform simulation model based on the input feature data more realistic, thereby improving the simulation accuracy of the renewable energy power generation platform simulation model.

[0163] Furthermore, based on the first embodiment of the above model training method, a second embodiment of the controller software adjustment method of the present invention is proposed. In this embodiment, the simulation model of the renewable energy power generation platform to be trained includes a fully virtual renewable energy power generation platform simulation model to be trained, and step C30 includes:

[0164] Step C31: The input feature data is mapped to the simulation controller parameter values ​​through the simulation model of the fully virtual renewable energy power generation platform to be trained.

[0165] The simulation model of the renewable energy power generation platform to be trained includes a simulation model of the semi-physical renewable energy power generation platform to be trained, which further includes a physical controller. Step C30 includes:

[0166] Step D10: The physical controller converts the installation environment parameter values ​​into input installation environment electrical signal values ​​according to preset parameter conversion rules;

[0167] For example, the installation environment parameter values ​​are mapped to input installation environment electrical signal values ​​through the preset parameter conversion rules.

[0168] The physical controller includes environmental parameter sensors, and prior to step D10, the following steps are also included:

[0169] Step D01: According to the preset parameter conversion rules, the installation environment parameter values ​​are converted into regular environment parameter electrical signal values ​​by the physical controller;

[0170] For example, the installation environment parameter values ​​are mapped to rule environment parameter electrical signal values ​​through the preset parameter conversion rules.

[0171] Step D02: The analytical installation environment parameter values ​​are obtained by analyzing the electrical signal values ​​of the training installation environment parameters through the sensor;

[0172] Step D03: Adjust and optimize the preset parameter conversion rule based on the difference between the installation environment parameter value and the parsed installation environment parameter value.

[0173] For example, the difference between the installation environment parameter value and the parsed installation environment parameter value is obtained, and it is determined whether the difference is less than a preset difference threshold. If the difference is greater than or equal to the preset difference threshold, the preset parameter conversion rule is adjusted according to the difference, and the process returns to the step of converting the installation environment parameter value into a regular environment parameter electrical signal value through the physical controller according to the preset parameter conversion rule, until the difference is less than the preset difference threshold.

[0174] Step D20: The input installation environment electrical signal value, the installation scale parameter value, and the preset software control logic are concatenated into semi-physical input feature data;

[0175] Step D30: The semi-physical input feature data is mapped to the simulation controller parameter values ​​through the simulation model of the semi-physical renewable energy power generation platform to be trained.

[0176] It is understandable that when using a fully virtual renewable energy power generation platform simulation model for model prediction, directly simulating environmental parameter values ​​can easily lead to discrepancies between the simulated and actual environmental parameter values, resulting in low prediction accuracy of the renewable energy power generation platform simulation model. By using a physical controller to convert the installation environmental parameter values ​​into input installation environmental electrical signal values ​​according to preset parameter conversion rules; concatenating the input installation environmental electrical signal values, the installation scale parameter values, and the preset software control logic into semi-physical input feature data; and mapping this semi-physical input feature data to the simulated controller parameter values ​​through the semi-physical renewable energy power generation platform simulation model to be trained, inaccurate environmental parameter values ​​are converted into accurate environmental electrical signal values, thereby improving the prediction accuracy of the renewable energy power generation platform simulation model.

[0177] Furthermore, based on the first and / or second embodiments of the above model training method, a third embodiment of the model training method of the present invention is proposed. In this embodiment, step C40 includes:

[0178] Step C41: If the simulation model of the renewable energy power generation platform to be trained includes the simulation model of the fully virtual renewable energy power generation platform to be trained, then determine whether the overlap is greater than the first preset overlap threshold.

[0179] Step C42: If the overlap degree is greater than the first preset overlap degree threshold, then the simulation model of the fully virtual renewable energy power generation platform to be trained is used as the simulation model of the renewable energy power generation platform; or,

[0180] For example, if the overlap is greater than the first preset overlap threshold, the simulation model of the fully virtual renewable energy power generation platform to be trained is used as the simulation model of the renewable energy power generation platform; if the overlap is less than or equal to the first preset overlap threshold, the process returns to the step of obtaining multiple training samples until the overlap is greater than the first preset overlap threshold.

[0181] Step C43: If the simulation model of the renewable energy power generation platform to be trained includes the simulation model of the semi-physical renewable energy power generation platform to be trained, then determine whether the overlap is greater than the second preset overlap threshold, wherein the second preset overlap threshold is greater than the first preset overlap threshold.

[0182] Step C44: If the overlap is greater than the second preset overlap threshold, then the simulation model of the semi-physical renewable energy power generation platform to be trained is used as the simulation model of the renewable energy power generation platform.

[0183] For example, if the overlap is greater than the second preset overlap threshold, the semi-physical renewable energy power generation platform simulation model to be trained is used as the renewable energy power generation platform simulation model; if the overlap is less than or equal to the second preset overlap threshold, the process returns to the step of obtaining multiple training samples until the overlap is greater than the second preset overlap threshold.

[0184] Understandably, since the semi-physical renewable energy power generation platform simulation model to be trained contains a physical controller that converts inaccurate environmental parameter values ​​into correct environmental electrical signal values, the accuracy requirement for the semi-physical renewable energy power generation platform simulation model to be trained is higher than that for the fully virtual renewable energy power generation platform simulation model to be trained. However, it requires the construction of a physical controller, thus incurring more time costs. Therefore, when efficiency is prioritized, the fully virtual renewable energy power generation platform simulation model to be trained can be used to improve the efficiency of model training, while when accuracy is prioritized, the semi-physical renewable energy power generation platform simulation model to be trained can be used to improve the accuracy of model prediction.

[0185] Furthermore, embodiments of the present invention also propose a controller software adjustment device, referring to... Figure 5 The controller software adjustment device includes:

[0186] The first acquisition module 10 is used to acquire the target installation scale parameter value, the target installation environment parameter value, the simulation model of the renewable energy power generation platform and the software control logic corresponding to the renewable energy power generation platform to be installed;

[0187] The prediction module 20 is used to predict the simulation controller parameter values ​​of the renewable energy power generation platform to be installed based on the target installation scale parameter values, the target installation environment parameter values, and the software control logic through the renewable energy power generation platform simulation model. The simulation controller parameter values ​​are the controller parameter values ​​when the controller in the simulated renewable energy power generation platform to be installed is running with the software control logic.

[0188] The adjustment module 30 is used to adjust the software control logic if the simulation controller parameter value does not meet the preset power generation condition, and return to the step of predicting the simulation controller parameter value of the renewable energy power generation platform to be installed based on the target installation scale, the target installation environment and the software control logic through the renewable energy power generation platform simulation model, until the simulation controller parameter value meets the preset power generation condition.

[0189] Furthermore, the analog controller parameter values ​​include at least one of the analog controller's voltage, current, and power. Prior to the step of adjusting the software control logic, the controller software adjustment device is further configured to:

[0190] Based on the simulated controller parameter values, the power generation benefits of the renewable energy power generation platform to be installed are predicted, wherein the power generation benefits include at least one of power generation efficiency and power generation;

[0191] Determine whether the power generation benefit is greater than a preset benefit threshold;

[0192] If so, then the parameter values ​​of the analog controller are determined to meet the preset power generation conditions;

[0193] If not, it is determined that the parameter values ​​of the analog controller do not meet the preset power generation conditions.

[0194] Furthermore, the analog controller parameter values ​​include harmonic content, and prior to the step of adjusting the software control logic, the controller software adjustment device is further configured to:

[0195] Determine whether the parameter value of the analog controller is less than a preset harmonic content threshold;

[0196] If so, then the parameter values ​​of the analog controller are determined to meet the preset power generation conditions;

[0197] If not, it is determined that the parameter values ​​of the analog controller do not meet the preset power generation conditions.

[0198] All embodiments of the controller software adjustment device of the present invention can refer to the various embodiments of the controller software adjustment method of the present invention, and will not be described again here.

[0199] Furthermore, embodiments of the present invention also propose a model training device, referring to... Figure 6 The model training device includes:

[0200] The second acquisition module 40 is used to acquire multiple training samples, wherein each training sample includes an input feature data and a training label corresponding to the input feature data. The input feature data includes an installation environment parameter value, an installation scale parameter value, and a preset software control logic. The training label includes the actual controller parameter value when the controller in the renewable energy power generation platform is controlled by the preset software control logic under the installation environment limited by the installation environment parameter value and the installation scale limited by the installation scale parameter value.

[0201] The first generation module 50 is used to generate real eye maps corresponding to the plurality of training samples, wherein the real eye maps are used to characterize the correspondence between the input feature data in each of the training samples and the real controller parameter values;

[0202] The second generation module 60 is used to predict the simulation controller parameter values ​​through the simulation model of the renewable energy power generation platform to be trained based on the input feature data, and generate a simulation eye diagram corresponding to the simulation model of the renewable energy power generation platform to be trained based on the simulation controller parameter values ​​and the input feature data. The simulation eye diagram is used to characterize the correspondence between the input feature data and the simulation controller parameter values ​​in each training sample.

[0203] The iteration module 70 is used to iteratively optimize the simulation model of the renewable energy power generation platform to be trained based on the overlap between the real eye diagram and the simulated eye diagram, so as to obtain the simulation model of the renewable energy power generation platform.

[0204] Furthermore, the simulation model of the renewable energy power generation platform to be trained includes a fully virtual renewable energy power generation platform simulation model to be trained, and the second generation module 60 is also used for:

[0205] The input feature data is mapped to the simulation controller parameter values ​​through the simulation model of the fully virtual renewable energy power generation platform to be trained.

[0206] Furthermore, the simulation model of the renewable energy power generation platform to be trained includes a simulation model of the semi-physical renewable energy power generation platform to be trained, and the simulation model of the semi-physical renewable energy power generation platform to be trained also includes a physical controller. The second generation module 60 is further used for:

[0207] The physical controller converts the installation environment parameter values ​​into input installation environment electrical signal values ​​according to preset parameter conversion rules.

[0208] The input installation environment electrical signal value, the installation scale parameter value, and the preset software control logic are concatenated into semi-physical input feature data;

[0209] The semi-physical input feature data is mapped to the simulation controller parameter values ​​through the simulation model of the semi-physical renewable energy power generation platform to be trained.

[0210] Furthermore, the physical controller includes an environmental parameter sensor. Before the step of converting the installation environment parameter values ​​into input installation environment electrical signal values ​​by the physical controller according to a preset parameter conversion rule, the model training device is also used for:

[0211] According to the preset parameter conversion rules, the physical controller converts the installation environment parameter values ​​into regular environment parameter electrical signal values.

[0212] The analytical installation environment parameter value is obtained by analyzing the electrical signal value of the rule installation environment parameter by the environmental parameter sensor;

[0213] The preset parameter conversion rules are adjusted and optimized based on the difference between the installation environment parameter values ​​and the parsed installation environment parameter values.

[0214] Furthermore, the iteration module 70 is also used for:

[0215] If the simulation model of the renewable energy power generation platform to be trained includes the simulation model of the fully virtual renewable energy power generation platform to be trained, then it is determined whether the overlap is greater than the first preset overlap threshold.

[0216] If the overlap degree is greater than the first preset overlap degree threshold, then the simulation model of the fully virtual renewable energy power generation platform to be trained is used as the simulation model of the renewable energy power generation platform; or...

[0217] If the simulation model of the renewable energy power generation platform to be trained includes the simulation model of the semi-physical renewable energy power generation platform to be trained, then it is determined whether the overlap is greater than the second preset overlap threshold, wherein the second preset overlap threshold is greater than the first preset overlap threshold.

[0218] If the overlap is greater than the second preset overlap threshold, then the semi-physical renewable energy power generation platform simulation model to be trained will be used as the renewable energy power generation platform simulation model.

[0219] Furthermore, before the step of iteratively optimizing the simulation model of the renewable energy power generation platform to be trained based on the overlap between the real eye diagram and the simulated eye diagram to obtain the simulation model of the renewable energy power generation platform, the model training device is also used for:

[0220] Identify and remove spiking signals from the real eye diagram.

[0221] All embodiments of the model training device of the present invention can refer to the various embodiments of the model training method of the present invention, and will not be described again here.

[0222] Furthermore, embodiments of the present invention also propose a computer-readable storage medium storing a controller software adjustment program, which, when executed by a processor, implements the steps of the controller software adjustment method described below.

[0223] The various embodiments of the controller software adjustment device and computer-readable storage medium of the present invention can be referred to the various embodiments of the controller software adjustment method of the present invention, and will not be repeated here.

[0224] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0225] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0226] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0227] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A controller software adjustment method, characterized in that, The controller software adjustment method includes the following steps: Obtain the target installation scale parameter values, the corresponding target installation environment parameter values, the renewable energy power generation platform simulation model, and the software control logic for the renewable energy power generation platform to be installed; The simulation model of the renewable energy power generation platform predicts the simulated controller parameter values ​​of the renewable energy power generation platform to be installed based on the target installation scale parameter values, the target installation environment parameter values, and the software control logic. The simulated controller parameter values ​​are the controller parameter values ​​in the simulation of the controller running with the software control logic in the renewable energy power generation platform to be installed. If the simulated controller parameter values ​​do not meet the preset power generation conditions, the software control logic is adjusted, and the process returns to the step of predicting the simulated controller parameter values ​​of the renewable energy power generation platform to be installed based on the target installation scale, the target installation environment, and the software control logic using the renewable energy power generation platform simulation model, until the simulated controller parameter values ​​meet the preset power generation conditions. The steps for constructing the simulation model of the renewable energy power generation platform include: Multiple training samples are obtained, wherein each training sample includes an input feature data and a training label corresponding to the input feature data. The input feature data includes an installation environment parameter value, an installation scale parameter value, and a preset software control logic. The training label includes the actual controller parameter value when the controller in the renewable energy power generation platform is controlled by the preset software control logic under the installation environment defined by the installation environment parameter value and the installation scale defined by the installation scale parameter value. Generate real eye maps corresponding to the multiple training samples, wherein the real eye maps are used to characterize the correspondence between the input feature data and the real controller parameter values ​​in each of the training samples; Based on the input feature data, the simulation controller parameter values ​​are predicted by the simulation model of the renewable energy power generation platform to be trained, and a simulation eye diagram corresponding to the simulation model of the renewable energy power generation platform to be trained is generated based on the simulation controller parameter values ​​and the input feature data. The simulation eye diagram is used to characterize the correspondence between the input feature data and the simulation controller parameter values ​​in each training sample. Based on the overlap between the real eye diagram and the simulated eye diagram, the simulation model of the renewable energy power generation platform to be trained is iteratively optimized to obtain the simulation model of the renewable energy power generation platform.

2. The controller software adjustment method as described in claim 1, characterized in that, The analog controller parameter values ​​include at least one of the analog controller's voltage, current, and power. Prior to the step of adjusting the software control logic, the method further includes: Based on the simulated controller parameter values, the power generation benefits of the renewable energy power generation platform to be installed are predicted, wherein the power generation benefits include at least one of power generation efficiency and power generation; Determine whether the power generation benefit is greater than a preset benefit threshold; If so, then the parameter values ​​of the analog controller are determined to meet the preset power generation conditions; If not, it is determined that the parameter values ​​of the analog controller do not meet the preset power generation conditions.

3. The controller software adjustment method as described in claim 1, characterized in that, The analog controller parameter values ​​include harmonic content. Prior to the step of adjusting the software control logic, the method further includes: Determine whether the parameter value of the analog controller is less than a preset harmonic content threshold; If so, then the parameter values ​​of the analog controller are determined to meet the preset power generation conditions; If not, it is determined that the parameter values ​​of the analog controller do not meet the preset power generation conditions.

4. A model training method, characterized in that, The model training method includes the following steps: Multiple training samples are obtained, wherein each training sample includes an input feature data and a training label corresponding to the input feature data. The input feature data includes an installation environment parameter value, an installation scale parameter value, and a preset software control logic. The training label includes the actual controller parameter value when the controller in the renewable energy power generation platform is controlled by the preset software control logic under the installation environment defined by the installation environment parameter value and the installation scale defined by the installation scale parameter value. Generate real eye maps corresponding to the multiple training samples, wherein the real eye maps are used to characterize the correspondence between the input feature data and the real controller parameter values ​​in each of the training samples; Based on the input feature data, the simulation controller parameter values ​​are predicted by the simulation model of the renewable energy power generation platform to be trained, and a simulation eye diagram corresponding to the simulation model of the renewable energy power generation platform to be trained is generated based on the simulation controller parameter values ​​and the input feature data. The simulation eye diagram is used to characterize the correspondence between the input feature data and the simulation controller parameter values ​​in each training sample. Based on the overlap between the real eye diagram and the simulated eye diagram, the simulation model of the renewable energy power generation platform to be trained is iteratively optimized to obtain the simulation model of the renewable energy power generation platform.

5. The model training method as described in claim 4, characterized in that, The simulation model of the renewable energy power generation platform to be trained includes a fully virtual renewable energy power generation platform simulation model to be trained. The step of predicting the simulation controller parameter values ​​based on the input feature data using the simulation model of the renewable energy power generation platform to be trained includes: The input feature data is mapped to the simulation controller parameter values ​​through the simulation model of the fully virtual renewable energy power generation platform to be trained.

6. The model training method as described in claim 4, characterized in that, The simulation model of the renewable energy power generation platform to be trained includes a semi-physical renewable energy power generation platform simulation model to be trained, which also includes a physical controller. The step of predicting the simulation controller parameter values ​​based on the input feature data using the simulation model of the renewable energy power generation platform to be trained includes: The physical controller converts the installation environment parameter values ​​into input installation environment electrical signal values ​​according to preset parameter conversion rules. The input installation environment electrical signal value, the installation scale parameter value, and the preset software control logic are concatenated into semi-physical input feature data; The semi-physical input feature data is mapped to the simulation controller parameter values ​​through the simulation model of the semi-physical renewable energy power generation platform to be trained.

7. The model training method as described in claim 6, characterized in that, The physical controller includes environmental parameter sensors. Before the step of converting the installation environment parameter values ​​into input installation environment electrical signal values ​​by the physical controller according to a preset parameter conversion rule, the method further includes: According to the preset parameter conversion rules, the physical controller converts the installation environment parameter values ​​into regular environment parameter electrical signal values. The analytical installation environment parameter value is obtained by analyzing the electrical signal value of the regular environmental parameter through the environmental parameter sensor; The preset parameter conversion rules are adjusted and optimized based on the difference between the installation environment parameter values ​​and the parsed installation environment parameter values.

8. The model training method as described in claim 4, characterized in that, The step of iteratively optimizing the simulation model of the renewable energy power generation platform to be trained based on the overlap between the real eye diagram and the simulated eye diagram to obtain the simulation model of the renewable energy power generation platform includes: If the simulation model of the renewable energy power generation platform to be trained includes the simulation model of the fully virtual renewable energy power generation platform to be trained, then it is determined whether the overlap is greater than the first preset overlap threshold. If the overlap degree is greater than the first preset overlap degree threshold, then the simulation model of the fully virtual renewable energy power generation platform to be trained is used as the simulation model of the renewable energy power generation platform; or... If the simulation model of the renewable energy power generation platform to be trained includes the simulation model of the semi-physical renewable energy power generation platform to be trained, then it is determined whether the overlap is greater than the second preset overlap threshold, wherein the second preset overlap threshold is greater than the first preset overlap threshold. If the overlap is greater than the second preset overlap threshold, then the semi-physical renewable energy power generation platform simulation model to be trained will be used as the renewable energy power generation platform simulation model.

9. The model training method according to any one of claims 4 to 8, characterized in that, Before the step of iteratively optimizing the simulation model of the renewable energy power generation platform to be trained based on the overlap between the real eye diagram and the simulated eye diagram to obtain the simulation model of the renewable energy power generation platform, the method further includes: Identify and remove spiking signals from the real eye diagram.

10. A controller software adjustment device, characterized in that, The controller software adjustment device includes: a memory, a processor, and a controller software adjustment program stored in the memory and executable on the processor, wherein when the controller software adjustment program is executed by the processor, it implements the steps of the controller software adjustment method as described in any one of claims 1 to 3.

11. A model training device, characterized in that, The model training device includes: a memory, a processor, and a model training program stored in the memory and executable on the processor, wherein the model training program, when executed by the processor, implements the steps of the model training method as described in any one of claims 4 to 9.

12. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a controller software adjustment program or a model training program, wherein when the controller software adjustment program is executed by a processor, it implements the steps of the controller software adjustment method as described in any one of claims 1 to 3, or the model training program implements the steps of the model training method as described in any one of claims 4 to 9.

Images