Photovoltaic power station design method based on digital twin model

Abstract

The invention discloses a photovoltaic power station design method based on a digital twin model, and belongs to the technical field of photovoltaic power station design. The problems that an existing photovoltaic power station design process is low in transparency and poor in man-machine interaction, and a design conclusion is difficult to trace are solved. Receiving and analyzing a natural language design instruction input by a user; generating a preliminary multi-dimensional design parameter set based on an analysis result, and constructing a parameter negotiation strategy to perform interactive negotiation with a user so as to determine a final multi-dimensional design parameter set; and based on the simulation data, generating a design scheme, and between a conclusion sentence in the scheme and the simulation data supporting the conclusion. By introducing an interactive negotiation link in which a user deeply participates and a sentence-level traceability mechanism for ensuring the credibility of a conclusion, a traditional closed design process is converted into an open, transparent and verifiable collaborative process, and the transparency and the credibility of a design scheme are remarkably improved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of photovoltaic power station design, and particularly relates to a photovoltaic power station design method based on a digital twin model. BACKGROUND

[0002] As an important part of renewable energy, the pros and cons of the early design of a photovoltaic power station directly determine the power generation benefit and the return on investment of the whole life cycle of the project. The design of a photovoltaic power station is a complex system engineering, and needs to comprehensively consider the geographical and meteorological conditions, the topography and geomorphology, the power grid access conditions and the performance parameters of key equipment such as photovoltaic components, inverters and supports of the project site. In order to cope with this complexity, professional computer simulation software is generally used for auxiliary design in the industry.

[0003] In the prior art practice, the design process is usually dominated by experienced engineers. The engineers first pre-decompose and convert a high-level design goal into a series of accurate and complete bottom-level technical parameters according to the project requirements, and then manually input these parameters into the design software. The software runs the built-in physical and electrical models according to the input, performs 8760-hour hourly power generation performance simulation, and finally outputs a report on key performance indicators including annual power generation, performance ratio (PR) and levelized cost of electricity (LCOE).

[0004] However, this traditional design mode relying on professional software has inherent limitations in the depth of human-computer interaction and the transparency of the design process. The existing process often presents a one-way and non-interactive feature, that is, the parameters are "injected" from the human to the machine, and there is a lack of an effective mechanism to intelligently guide, suggest or negotiate design boundaries with the designer based on high-level goals. This not only puts high requirements on the professional experience of the designers, but also makes the human-computer interaction process appear rigid.

[0005] More importantly, there is a logical gap between the massive and discrete raw data output from the simulation software and the comprehensive design scheme presented to the decision maker in the form of natural language. Each conclusive statement in the final scheme, such as "selecting XX type component can make the power generation increase by 3%", the specific simulation data and boundary conditions behind it are often missing or vague in its final text form. This conversion process from data to conclusion lacks a clear and machine-readable attribution mechanism, making the entire design argument chain opaque or in a "black box" state to some extent. When the design scheme needs to be audited, reviewed or compared with multiple schemes, it is often very difficult to quickly and accurately trace back to the original input conditions and simulation data fragments relied on by a specific conclusion, thereby affecting the overall credibility of the design scheme and the decision-making efficiency of the later operation and optimization. SUMMARY

[0006] The technical problem to be solved by the present application is that the existing model-based photovoltaic power station design method is usually a "black box" process, and users are difficult to participate in the decision-making of key parameters, and the generation logic of the design scheme is not transparent, which leads to the final output scheme being difficult to be fully trusted and verified.

[0007] To solve the above technical problems, the present application provides a photovoltaic power station design method based on a digital twin model, a device, an electronic equipment and a storage medium.

[0008] The first aspect of the present application provides a photovoltaic power station design method based on a digital twin model, which introduces a user interactive negotiation mechanism to determine the core input parameters of the design, and uses a digital twin model for simulation, and finally generates a design scheme with traceable conclusions and verifiable processes, thereby significantly improving the transparency of the design process and the credibility of the final results.

[0009] In one specific embodiment, the method comprises the following steps: First, receive the design instruction of the photovoltaic power station. The instruction can be a natural language description input by the user, which contains the preliminary requirements and expected goals of the power station.

[0010] Then, based on the design instruction, a preliminary multi-dimensional design parameter set is generated. This step aims to convert unstructured user instructions into structured parameter sets that can be used for model simulation. A feasible implementation is to first perform semantic analysis on the design instruction to identify the core design entities (such as geographic location, installed capacity) and constraint conditions (such as budget limit, land area); then, based on the identified information, through synonym expansion, theme decomposition or parameter combination, etc. Strategy, build a preliminary multi-dimensional design parameter set with rich content and multiple dimensions, lay a foundation for subsequent comprehensive simulation.

[0011] Subsequently, the parameter negotiation strategy corresponding to the preliminary multi-dimensional design parameter set is constructed, and the parameter negotiation strategy is presented to the user through a human-computer interaction interface for interactive negotiation by the user. This step is one of the core innovations of the present application, which changes the hidden parameter selection process in traditional design methods into a transparent link that users can observe and intervene.

[0012] After the interactive consultation of the user, a final multi-dimensional design parameter set is determined according to the consultation feedback of the user. The consultation feedback of the user reflects the deep consideration and final confirmation of the user on the design details, and specifically can include at least one operation of adding, deleting or modifying a parameter entry in the preliminary multi-dimensional design parameter set. For example, the user can add a parameter of focusing on the influence of shadow blocking on power generation efficiency, or delete an economic evaluation parameter which the user considers unimportant. In order to accurately reflect the intention of the user, the process of determining the final multi-dimensional design parameter set can be formally described. Specifically, the consultation feedback of the user can be parsed into a parameter set to be added and a parameter set to be removed , and the final multi-dimensional design parameter set is calculated by the following formula

[0013] ; wherein, is the preliminary multi-dimensional design parameter set. The formula ensures that the parameter set for final simulation is the product of the combination of system intelligence and user expertise.

[0014] Then, the final multi-dimensional design parameter set is input into a preset digital twin model for running to obtain simulation data. The digital twin model is a high-fidelity virtual model which can integrate geographic information data of the location of the photovoltaic power station, meteorological historical data, and physical performance model of the photovoltaic component, so as to accurately simulate the running situation in the real world. The obtained simulation data is an objective basis for subsequent generation of design scheme, and specifically can include performance indicators, environmental impact factors and economic benefit evaluation results corresponding to each parameter entry in the final multi-dimensional design parameter set.

[0015] Finally, based on the simulation data, a design scheme containing sentence-level provenance mapping is generated. This step aims to solve the problem of "unexplainable" design results. The specific implementation manner is: first, generating a preliminary text of the design scheme according to the simulation data; then, for each sentence containing a design conclusion in the preliminary text, locating and associating one or more simulation data segments serving as the basis for the determination of the sentence in the simulation data, so as to build a sentence-level provenance mapping from the conclusion to the evidence.

[0016] In order to enable the user to intuitively understand the source of the conclusion, the final presentation form of the design scheme can include: at the end of each sentence containing a design conclusion, an index pointing to the associated simulation data segment is attached; at the same time, at the end of the design scheme, a complete list containing all cited simulation data segments and their source parameters is provided.

[0017] The establishment process of the sentence-level traceability mapping can be accurately described by a function relationship. In a specific embodiment, the sentence-level traceability mapping is established by the following function relationship:

[0018] wherein, is the index of the Chinese text sentence of the design scheme, is a set composed of the simulation data, is a simulation data segment in the set, is the index of the data segment, and the supports relationship indicates that the simulation data segment is a sentence provides direct and verifiable judgment basis.

[0019] In addition, in a preferred embodiment, the method can further include collecting and analyzing the negotiation feedback data of the user, for continuously optimizing the generation strategy of the preliminary multi-dimensional design parameter set, so that the system can continuously learn and propose an initial strategy closer to the user's demand in subsequent design tasks.

[0020] The second aspect of the present application provides a photovoltaic power station design device based on a digital twin model, which comprises: an instruction receiving module configured to receive a design instruction of a photovoltaic power station; a parameter set generating module configured to generate a preliminary multi-dimensional design parameter set based on the design instruction; a negotiation strategy constructing module configured to construct a parameter negotiation strategy corresponding to the preliminary multi-dimensional design parameter set, and present the parameter negotiation strategy to a user through a human-computer interaction interface for interactive negotiation by the user; a parameter set determining module configured to determine a final multi-dimensional design parameter set according to negotiation feedback of the user; a simulation running module configured to input the final multi-dimensional design parameter set into a preset digital twin model for running to obtain simulation data; a scheme generating module configured to generate a design scheme comprising a sentence-level traceability mapping based on the simulation data.

[0021] The third aspect of the present application provides an electronic device comprising a processor and a memory, wherein the memory stores a computer program, and the processor is configured to implement the method according to the first aspect of the present application when executing the computer program.

[0022] The fourth aspect of the present application provides a computer readable storage medium, which stores a computer program, and the program is executed by a processor to implement the method according to the first aspect of the present application. ​

[0023] The application provides a photovoltaic power station design method based on a digital twin model.

[0024] 1. Significantly improves the transparency and credibility of the design scheme. The application directly associates each key conclusion sentence with one or more specific simulation data segments behind it through the unique sentence-level traceability mapping mechanism in the finally generated design scheme text. This makes any conclusion of the design scheme traceable and verifiable, completely solves the "black box" problem of the opaque demonstration process in the traditional design report, and greatly enhances the credibility of the scheme in the review and audit process.

[0025] 2. Enhances the flexibility and human-machine collaboration depth of the design process. The application introduces an interactive parameter negotiation link, changes the traditional one-way "man-machine" parameter input mode to a two-way, collaborative "man-machine" dialogue mode. Users are no longer passive inputters, but can actively modify, add or delete the preliminary parameter strategy generated by the system, and observe and adjust in real time, which gives the design process great flexibility and enables better combination of human experience and wisdom and machine computing power.

[0026] 3. Reduces the threshold for using professional design software. By introducing a natural language instruction analysis module, the application allows users to start the design task with high-level, general language that is more in line with human thinking habits, rather than requiring users to master all underlying and accurate technical parameters in advance. The system can automatically decompose high-level instructions into executable parameter sets, which enables non-senior engineers or even project managers to conveniently use the system for preliminary scheme conception and evaluation.

[0027] 4. Improves the overall efficiency and consistency of photovoltaic power station design. The application automates and integrates the complete process from receiving instructions, negotiating parameters, driving simulation to generating traceable reports. This not only reduces the large amount of repetitive labor required for manual conversion and data transfer between different software and documents, but also effectively avoids data inconsistencies or errors caused by human negligence, thereby improving design efficiency while ensuring the standardization and accuracy of the final output results.

[0028] 5. Ensures that the design conclusion is based on high-fidelity physical reality. The application uses a digital twin model as the core simulation engine and ensures that all parameters input into the model are confirmed by the user through interactive negotiation. This means that each simulation operation is an accurate simulation of a carefully considered design scheme in a virtual environment that closely approximates physical reality, resulting in higher quality simulation data and more reliable design conclusions, providing solid data support for project investment decisions. BRIEF DESCRIPTION OF DRAWINGS

[0029] Fig. 1 Flow chart of the method of the present application; Fig. 2 Schematic diagram of digital twin model data interaction of an embodiment of the present application; Fig. 3 Schematic diagram of the hardware structure of an electronic device of an embodiment of the present application. DETAILED DESCRIPTION

[0030] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and do not constitute a limitation on the present application.

[0031] Reference Figs. 1 to 3 A photovoltaic power station design method based on a digital twin model, which aims to provide a transparent, interactive and result-traceable design environment, can be deployed on a single server, a personal computer or a distributed computing platform. The system can include an instruction analysis module, a parameter negotiation module, a digital twin simulation module, a scheme generation and traceability module, a human-computer interaction interface and a data storage unit.

[0032] The instruction analysis module functions as a bridge between the system and the user's natural language input. When the user inputs a descriptive design instruction through the human-computer interaction interface, the module is responsible for receiving the instruction and performing deep semantic analysis on it using natural language processing technology to accurately identify the core design entities, performance constraints and user's potential intentions implied in the instruction, and finally output structured instruction information that can be processed by subsequent modules.

[0033] The parameter negotiation module, connected with the instruction analysis module, is the core functional unit that realizes the human-computer collaborative design concept of the present application. After receiving the structured instruction information, the module first automatically constructs a preliminary multi-dimensional design parameter set according to the preset generation strategy. Then, the module encapsulates the parameter set into a clear parameter negotiation strategy and presents it to the user through the human-computer interaction interface. During the user's interactive negotiation process, the parameter negotiation module is responsible for receiving the user's negotiation feedback in real time, such as adding, deleting or modifying operations on the parameters, and dynamically updating the parameter set according to the feedback until the user finally confirms.

[0034] The digital twin simulation module is connected with the parameter negotiation module, and is a calculation engine for high-fidelity performance prediction of the application. After receiving the final multi-dimensional design parameter set determined by the parameter negotiation module, the module takes it as an accurate input to drive the internally integrated digital twin model to perform simulation operation. The digital twin model is pre-constructed and stored in the data storage unit, and integrates geographic information data of the target site, years of high-precision weather data, and physical performance models of various photovoltaic components. After the simulation operation is completed, the module outputs a series of structured simulation data, which can be stored in the data storage unit.

[0035] The scheme generation and tracing module is connected with the digital twin simulation module, and is responsible for converting machine-readable simulation data into design schemes that can be understood and verified by human users. The module first integrates and analyzes the simulation data to generate a preliminary text of the design scheme. The key function is to accurately construct the sentence-level trace mapping between each independent sentence containing design conclusions in the scheme and the corresponding simulation data segment while generating the text. This mapping relationship ensures that any conclusion in the scheme has an objective and verifiable data source support.

[0036] The human-computer interaction interface, as the only window for information exchange between users and the system, interacts with the parameter negotiation module and the scheme generation and tracing module. On the one hand, it is responsible for visualizing the parameter negotiation strategy constructed by the parameter negotiation module to the user and collecting all the negotiation operations of the user; on the other hand, it is responsible for presenting the final design scheme generated by the scheme generation and tracing module, which contains sentence-level trace mapping, to the user in a friendly manner, such as through inline index and interactive reference list.

[0037] The data storage unit provides data support for the operation of the entire system, and exchanges data with the parameter negotiation module, the digital twin simulation module, and the scheme generation and tracing module. The unit is responsible for persistently storing the digital twin model itself, the basic geographic and weather data, the user input design instructions, the intermediate version parameter set of the negotiation process, the final determined parameter set, and all simulation data generated by the simulation operation.

[0038] In the embodiment, the flow of the above-mentioned modules working together is as follows: the user submits a design instruction through the human-computer interaction interface, and the instruction analysis module analyzes and transmits the result to the parameter negotiation module. The parameter negotiation module generates a preliminary strategy and carries out multi-round negotiation with the user through the human-computer interaction interface until the user confirms and forms a final multi-dimensional design parameter set. The final parameter set is sent to the digital twin simulation module for simulation, and original simulation data is output. Finally, the scheme generation and traceability module generates a design scheme with complete sentence-level traceability mapping using the simulation data, and finally presents it to the user through the human-computer interaction interface. This architecture changes the traditional closed design process into an open, transparent and user-involved collaborative process.

[0039] In the embodiment, the specific flow of the method of the present application executed by each module in the system architecture will be described in detail. The flow starts with the user inputting a high-level design goal and ends with the generation of a completely traceable detailed design scheme driven by a digital twin model.

[0040] The execution of the method starts with the reception of a design instruction and the generation of a preliminary multi-dimensional design parameter set. Specifically, when the user inputs a design instruction in the form of natural language through the human-computer interaction interface , the instruction analysis module first processes it. The processing process includes capturing key parameters such as geographic location "Nevada" and installed capacity "50MW" using named entity recognition technology; at the same time, it uses intent classification technology to understand the user's core goal, such as "maximizing annual power generation" or "optimizing initial investment return rate".

[0041] Based on the above-mentioned structured information parsed, the parameter negotiation module immediately starts the generation process of the preliminary multi-dimensional design parameter set . This process is not simply a list of identified entities, but is expanded through a series of strategies to ensure the comprehensiveness of subsequent simulation. For example, the identified geographic location "Nevada" will trigger the system to automatically associate and include parameter entries related to the high solar radiation and high temperature environment in this region, such as "component temperature attenuation coefficient", "inverter cooling scheme", etc. If the user's instruction is complex, the system will also use a topic decomposition strategy to decompose it into multiple sub-parameter sets for combination, and finally form an initial set of parameter entries . .

[0042] Next, the system enters the core interactive parameter negotiation link. The parameter negotiation module encapsulates the generated preliminary multi-dimensional design parameter set into a user-friendly parameter negotiation strategy and presents it through the human-computer interaction interface. The interface displays each parameter entry and provides the user with explicit interaction controls, allowing the user to perform operations of adding, deleting or modifying parameters. Each operation of the user is considered as a negotiation feedback by the system . The negotiation process is a deterministic, user-led converging process. The system updates the parameter set accurately according to the negotiation feedback of the user. Specifically, the operation of the user is parsed into two sets: one is the set of parameters decided to be added by the user , and the other is the set of parameters decided to be removed by the user o the multi-dimensional design parameter set for final simulation is obtained by the following set operations:

[0043]

[0044] wherein, is the preliminary multi-dimensional design parameter set; is the set of parameters to be removed; is the set of parameters to be added; represents the difference set operation of sets; represents the union set operation of sets. This process will continue until the user issues a final confirmation instruction through the interface, at which time the obtained is the only input for the next stage.

[0045] After obtaining the final multi-dimensional design parameter set confirmed by the user, the digital twin simulation module is activated. The module inputs each parameter entry in as accurate boundary conditions and configuration information into the internal digital twin model. The model is a high-fidelity virtual power station environment that has internalized topographic data at a specific geographic coordinate, hourly weather data (including irradiance, temperature, wind speed, etc.) over the past decades, and detailed physical performance models of photovoltaic components, inverters, and other equipment from different manufacturers under different working conditions. The core of the simulation run is to simulate the hourly performance of the photovoltaic power station in one or more typical operating years in the future. After the operation is completed, the digital twin simulation module outputs a structured, machine-readable simulation data set . The data set is composed of independent simulation data segments, each of which contains specific numerical values, units, and their corresponding input parameter sources, such as "under the parameter "component type A", the average power generation efficiency in the first year is 21.5%" or "corresponding to "inverter scheme B", the system availability is 99.8%".

[0046] Finally, the system performs a design scheme generation step containing sentence-level traceability mapping. The scheme generation and traceability module receives the simulation data set and generate a preliminary design text in natural language form based on the dataset The core task of this generation process is to ensure that the text content is completely faithful to the simulation data.

[0047] Most importantly, while generating each sentence , the module performs a rigorous attribution operation to build a sentence-level provenance map for the sentence. This means that for any conclusive statement in the design, such as "Based on the above analysis, the use of component type A can increase annual power generation by 3%", the system accurately locates the data segment or segments that support the conclusion , and establishes an unalterable link between the sentence and the data segment.

[0048] This provenance mapping relationship can be formally described by the following function:

[0049] In this function, is the unique index of the sentence in the design; is the unique index of the data segment in the simulation dataset; is the complete set of simulation data segments; and the supports relationship represents a certain logical support relationship, i.e., the establishment of the sentence is a direct calculation or logical derivation result of the simulation data segment . Finally, the mapping relationship is attached to the end of each conclusive sentence in the form of an index subscript, and combined with the complete reference list at the end of the design, the entire logical chain of the design is presented to the user.

[0050] In order to further illustrate the technical solutions provided by the present application, the complete execution process of the method will be described below through a specific application scenario example. This example aims to demonstrate how the method combines the user's professional judgment with the system's precise simulation capability in a real-world photovoltaic power station design task.

[0051] In this example, it is assumed that a design engineer receives a design task to design a 100MW photovoltaic power station for a specific geographic coordinate in the Gobi land of Qinghai Province, China. The engineer inputs the following initial design instructions through the human-computer interaction interface: "Design a 100MW photovoltaic power station for the land at coordinates (E95.XX, N37.XX) in Qinghai Province, focusing on evaluating the combined benefits of double-sided components and fixed supports, flat single-axis tracking supports, and considering the impact of extreme sand and dust weather on power generation."

[0052] Upon receiving the instruction, the instruction parsing module immediately parses it. The module identifies the core entities "Qinghai Province Coordinate (E95.XX, N37.XX)", "100 MW", "Bifacial Module", "Fixed Mount", "Flat Single-axis Tracking Mount", and a key constraint condition "Consider Extreme Dust Weather Impact". Based on this, the parameter negotiation module generates a preliminary multi-dimensional design parameter set, which not only contains the parameters explicitly mentioned by the user, but also automatically extends associated parameters based on the built-in knowledge base, such as "Module Height from Ground", "Tracking Axis Rotation Angle Range", "Local Annual Average Dusty Days", "Transmittance Attenuation Model under Dust Weather", etc.

[0053] Subsequently, the human-computer interaction interface displays this parameter negotiation strategy to the engineer in the form of a list. When reviewing, the engineer uses his professional experience to conduct interactive negotiation. He thinks that the "Transmittance Attenuation Model under Dust Weather" generated by the system is too general, so he manually modifies the parameter entry and specifies a more specific attenuation model recognized by the industry that better fits the local actual situation. In addition, he adds a new parameter entry: "Evaluate the impact of different cleaning cycles (e.g. monthly vs quarterly) on annual power generation" to quantify the relationship between operation and maintenance costs and power generation gains. After the engineer completes these operations and clicks Confirm, the system generates the final multi-dimensional design parameter set based on the aforementioned set operation rules.

[0054] Upon receiving this parameter set deeply customized by the engineer, the digital twin simulation module immediately starts the simulation. The module retrieves detailed elevation data for the corresponding coordinate plot from the data storage unit, as well as past 20 years of hourly solar irradiance, wind speed, temperature, and historical dust weather records. The simulation process parallelly calculates the annual 8760-hour power generation performance of "Bifacial Module + Fixed Mount" and "Bifacial Module + Flat Single-axis Tracking Mount" under different cleaning cycles, different heights from the ground, etc. After the simulation is completed, a simulation data set containing hundreds of independent data segments is generated, each clearly recording its input parameters and output results, such as: "Input: {Module Type: Bifacial, Mount: Flat Single-axis, Cleaning Cycle: Monthly}, Output: {First Year Power Generation: XXXX MWh, KPR: 85.1%}".

[0055] Finally, the scheme generation and provenance module starts to generate the final design scheme based on this detailed simulation dataset. The module explicitly states in the generated text: “Simulation results show that, compared to the fixed support, the annual power generation can be increased by about 18.5% by adopting the flat single-axis tracking support.” The index “” at the end of the sentence directly links to the corresponding data segment in the reference list, which accurately displays the original simulation values of the power generation of the two schemes. The scheme continues to elaborate: “Further analysis shows that, after considering the local high-sand-dust impact, increasing the component cleaning period from once every quarter to once every month can bring an additional 2.1% power generation gain.” Here, “” links to two simulation data segments, which display the power generation simulation results under two different cleaning periods, providing double data support for the conclusion.

[0056] Finally, the engineer obtains a complete design scheme. The scheme not only gives clear technical selection recommendations, but also each key conclusion in the scheme is corresponded to the simulation data driven by the specific input parameters confirmed by him through the index subscript at the end of the sentence. By clicking on these indexes, he can instantly check the original data source of the conclusion, making the entire design process completely transparent, credible and traceable.

[0057] Correspondingly, the present application also provides a photovoltaic power station design device based on a digital twin model. The device is the physical carrier for implementing the foregoing method. In a specific embodiment, the device can be a standalone server, a workstation, or a distributed system composed of multiple computing nodes. The device logically includes an instruction parsing module, a parameter negotiation module, a digital twin simulation module, a scheme generation and provenance module, a human-computer interaction interface, and a data storage unit. The functions of these modules have been described in detail in the foregoing system architecture section, and will not be repeated here.

[0058] At the hardware level, the device can include but is not limited to one or more processors, a memory, a communication interface, and a human-computer interaction interface. The processor, as the operation core of the device, is responsible for executing the computer program instructions stored in the memory to realize the functions of the above-mentioned logical modules. The memory can be any type of volatile or non-volatile storage medium, such as random access memory (RAM), read-only memory (ROM), hard disk drive (HDD), or solid state drive (SSD), for storing program codes required to execute the method of the present application and various data during processing. The communication interface is used for data exchange with external networks or databases, such as obtaining real-time weather data. The human-computer interaction interface connects devices such as displays, keyboards, and mice to build the aforementioned human-computer interaction interface.

[0059] The present application also provides an electronic device, which can be a personal computer, a server, a mobile terminal or any device with computing and storage capability. The electronic device internally comprises a processor and a memory, and the memory is solidified or installed with a computer program. When the electronic device is running, the processor reads and executes the computer program in the memory, so as to complete all steps of the photovoltaic power station design method based on digital twin model described in any of the foregoing embodiments.

[0060] The present application also provides a computer readable storage medium. The storage medium can be non-transient, such as an optical disc, a U disk, a mobile hard disk or a storage array on a server. A series of computer program instructions are stored on the computer readable storage medium, and when the instructions are loaded and executed by one or more processors, the electronic device or computing system executing the instructions can complete all steps of the photovoltaic power station design method based on digital twin model described in any of the foregoing embodiments.

[0061] In the embodiments of the present application, it should be noted that the term "module" can refer to a software program segment, a hardware circuit, or a combination of the two. For software implementation, these modules are program units stored in the memory and executed by the processor; for hardware implementation, these modules can be independent integrated circuits or different functional areas integrated on the same chip.

[0062] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit it. Although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A photovoltaic power plant design method based on a digital twin model, characterized in that, Includes the following steps: Receive design instructions for photovoltaic power plants; Based on the design instructions, a preliminary multidimensional design parameter set is generated; Construct a parameter negotiation strategy corresponding to the preliminary multidimensional design parameter set, and present the parameter negotiation strategy to the user through a human-computer interaction interface for interactive negotiation. Based on user feedback, the final multi-dimensional design parameter set was determined. The final multidimensional design parameter set is input into a preset digital twin model for execution to obtain simulation data; Based on the simulation data, a design scheme containing sentence-level source mapping is generated.

2. The method as described in claim 1, characterized in that, The steps for generating the preliminary multidimensional design parameter set include: The design instructions are semantically parsed to identify core design entities and constraints; Based on the core design entities and constraints, the preliminary multidimensional design parameter set is constructed through strategies such as synonym expansion, topic decomposition, or parameter combination.

3. The method as described in claim 1, characterized in that, The user's negotiation feedback includes at least one operation: adding, deleting, or modifying parameter entries in the preliminary multidimensional design parameter set.

4. The method as described in claim 3, characterized in that, The specific steps for determining the final multidimensional design parameter set are as follows: parsing the user's negotiation feedback into a set of parameters to be added. and the set of parameters to be removed ; The final multidimensional design parameter set is calculated using the following formula. : ; in, This is the preliminary multidimensional design parameter set.

5. The method as described in claim 1, characterized in that, The simulation data includes performance indicators, environmental impact factors, and economic benefit assessment results that correspond one-to-one with each parameter item in the final multidimensional design parameter set.

6. The method as described in claim 1, characterized in that, The steps for generating a design scheme containing sentence-level source mapping include: A preliminary design scheme text is generated based on the simulation data; For each sentence in the preliminary text that contains a design conclusion, locate and associate one or more simulation data segments in the simulation data that serve as the basis for its determination, thereby constructing the sentence-level source mapping.

7. The method as described in claim 6, characterized in that, The final presentation of the design scheme includes: At the end of each sentence containing the design conclusion, an index pointing to the associated simulation data segment is appended; And at the end of the design scheme, a complete list containing all referenced simulation data fragments and their source parameters is provided.

8. The method as described in claim 6, characterized in that, The sentence-level source mapping This is established through the following functional relationship: in, For the indexing of text sentences in the design scheme, The set of simulation data, For the simulation data fragments in the set, The index is for the data segment, and supports represents the simulation data segment. For sentences Provide the basis for judgment.

9. The method as described in claim 1, characterized in that, The digital twin model integrates geographic information data of the photovoltaic power station location, historical meteorological data, and physical performance models of photovoltaic modules.

10. The method as described in claim 1, characterized in that, The method further includes: Collect and analyze the user's negotiation feedback data to optimize the generation strategy of the preliminary multidimensional design parameter set.

Images