Photovoltaic power station technical transformation scheme generation method and system based on multi-dimensional data

Abstract

This invention discloses a method and system for generating photovoltaic power plant technical upgrade schemes based on multi-dimensional data, belonging to the field of photovoltaic intelligent operation and maintenance. The method collects data such as electrical parameters, hot spot temperature gradients, and dust accumulation / shading loss rates from three types of intelligent agents: monitoring, inspection, and cleaning. It constructs a string-level health model and calculates the overall health of the power plant. A dual-threshold coupling trigger mechanism is formed by combining technical thresholds and incremental cost per kilowatt-hour, generating technical upgrade instructions when conditions are met. Candidate schemes such as component replacement, inverter upgrades, and digital transformation are matched from a strategy library. Four hard constraints—MPPT voltage window, capacity ratio, grid-type short-circuit ratio, and support load—are verified using digital twin electrical topology, eliminating incompatible schemes. Investment costs are dynamically adjusted by accessing real-time prices in the industry chain via API, calculating incremental internal rate of return (IRR) and sorting by return, outputting a JSON-structured executable technical upgrade sequence containing equipment model, engineering quantity, budget, and construction period. This invention achieves end-to-end closed-loop technical upgrade decision-making, improving decision accuracy and scheme feasibility.

Description

Technical Field

[0001] This invention belongs to the field of intelligent operation and maintenance of photovoltaic power plants, specifically involving a method for generating technical upgrade schemes for photovoltaic power plants based on multi-dimensional data. Background Technology

[0002] As of the first quarter of 2025, my country's cumulative installed photovoltaic (PV) capacity had exceeded 945 GW, of which 174 GW, or nearly 20%, were older power plants connected to the grid in 2018 or earlier. With increasing operating years, PV power plants generally face problems such as module power degradation, hot spot effects, dust accumulation and shading, and decreased inverter efficiency, leading to a year-on-year decrease in system efficiency (PR) and a continuous decline in power generation revenue. Upgrading existing power plants has become a key means to enhance asset value and ensure investment returns.

[0003] Currently, the decision-making process for upgrading photovoltaic power plants mainly relies on the following two methods: One approach is to replace equipment based on its lifespan in the equipment ledger: technical upgrade plans are formulated according to the design lifespan of equipment such as components and inverters (usually 10-25 years). However, there is a lack of detailed assessment of the actual health status of the equipment, which leads to "replacing before it breaks down" resulting in capital waste, or "not replacing when it breaks down" resulting in power generation loss.

[0004] The second method involves issuing a technical improvement report after a manual on-site inspection: maintenance personnel or third-party testing agencies conduct sampling inspections of the power station and propose technical improvement suggestions based on their experience. This method relies on expert experience, has a decision-making cycle of up to several weeks, and different personnel may have significantly different judgments on the same power station, lacking a unified quantitative evaluation standard.

[0005] The monitoring intelligent agent described in this invention is a SCADA system and IV curve monitoring device deployed in a power plant; the inspection intelligent agent is a drone or robot equipped with a dual-spectrum infrared camera; and the cleaning intelligent agent is a cleaning robot with power generation gain feedback function.

[0006] A search revealed the following technical defects in the currently disclosed patented technologies: (1) Technological upgrades rely on a single threshold or human experience, lacking precise criteria. Some patents (such as CN121638068A) only output the probability distribution of the remaining lifespan of the components, without coupling the overall health of the power plant with dynamic economic boundary conditions, and cannot automatically determine "when to start technological upgrades".

[0007] (2) There is a technical gap between the diagnostic results and the technical improvement plan. Existing patents (such as CN120806735B) only output evaluation values ​​such as net present value (NPV) and internal rate of return (IRR) to answer whether it is worth repairing, but do not automatically generate specific technical improvement plans (such as which type of components to replace, whether to upgrade the inverter, and what kind of digital transformation to adopt). Maintenance personnel still need to manually convert the diagnostic report into a construction plan, resulting in a broken decision-making chain.

[0008] (3) The candidate schemes lack electrical topology compatibility verification. The compatibility of technical upgrade schemes (such as replacing high-power components and upgrading grid-type inverters) with the original electrical equipment of the power plant (inverter MPPT voltage window, capacity ratio, short-circuit ratio, support load) relies on manual verification, which often results in the schemes being unfeasible or the power generation gain not meeting expectations after implementation.

[0009] (4) The economic assessment uses static costs, which leads to inaccurate calculations. The existing method uses fixed equipment unit prices and electricity prices, and does not incorporate real-time price fluctuation data of the photovoltaic industry chain (the price per watt of the module has dropped from 38 yuan to about 0.7 yuan), resulting in a serious deviation from the actual incremental internal rate of return (ΔIRR) calculation.

[0010] (5) The output results are disconnected from the project execution. The output of existing technologies is mostly numerical reports or evaluation rankings, which cannot be directly converted into executable construction instructions. They require secondary manual processing, which is inefficient and prone to errors.

[0011] In summary, there is an urgent need in this field for a photovoltaic power plant technical upgrade decision-making method that can automatically complete the end-to-end closed loop of "health quantification → dual threshold triggering → scheme generation → topology verification → economic deduction → executable sequence output" to solve the technical problems in the existing technology, such as inaccurate technical upgrade triggering, disconnect between diagnosis and scheme, lack of electrical compatibility verification, distorted economic calculation, and output that cannot be directly executed. Summary of the Invention

[0012] To address the aforementioned problems, the purpose of this invention is to provide a method and system for generating photovoltaic power plant technical upgrade schemes based on multi-dimensional data.

[0013] To solve the above-mentioned technical problems, the first technical solution adopted by the present invention is as follows: a method for generating a photovoltaic power station technical upgrade scheme based on multi-dimensional data, comprising the following steps: S1: Multi-source data acquisition and power plant health measurement, acquire string current, voltage and fill factor collected by monitoring agent, hot spot temperature gradient and structural coupling degree collected by inspection agent, and dust accumulation and shading loss rate collected by cleaning agent, construct string-level health model and calculate the overall health of power plant. S2: Technical and economic dual threshold coupling judgment, compare the overall health of the power station with the preset technical threshold, and calculate the incremental cost per kilowatt-hour based on the incremental investment in technical upgrades, changes in annual operation and maintenance costs, annual increase in power generation, equipment life cycle and electricity price fluctuations. When the overall health of the power station is lower than the technical threshold and the incremental cost per kilowatt-hour is not greater than 0, trigger the technical upgrade scheme generation instruction. S3: Candidate technical upgrade scheme generation: In response to the technical upgrade scheme generation instruction, a set of candidate technical upgrade schemes containing at least one of component replacement, inverter upgrade, and digital transformation is obtained from the preset strategy library; S4: Electrical topology hard constraint verification. The candidate technical modification scheme set is mapped to the digital twin electrical topology model of the power plant. MPPT voltage window matching verification, capacity ratio compliance verification, network short-circuit ratio verification and support load margin verification are performed in sequence. Schemes that do not meet the constraints are eliminated to obtain a set of compatible technical modification schemes. S5: Dynamic economic simulation and executable sequence output. It obtains equipment prices in the photovoltaic industry chain in real time through API interface, dynamically adjusts the investment cost of each compatible technical transformation scheme, calculates the corresponding incremental internal rate of return, sorts them from high to low incremental internal rate of return, and generates and outputs executable technical transformation sequences.

[0014] In the above technical solution, the string-level health model mentioned in step S1 is: , in: , , These are the measured current, measured voltage, and measured fill factor of the i-th string, respectively. , , These are the current reference value, voltage reference value, and fill factor reference value for the same type of healthy string, respectively; This is the amount of fill factor decay. , , , The weights are dynamically determined using the entropy weight method and satisfy the following conditions: + + + =1; The overall health status of the power station is: Where N is the total number of strings in the entire site. .

[0015] In the above technical solution, the formula for calculating the incremental electricity cost in step S2 is as follows: Among them: For incremental investment in technological upgrading, For annual changes in operation and maintenance costs, To increase annual power generation, For the life cycle of technical upgrade equipment, This refers to changes in electricity prices.

[0016] In the above technical solution, step S2 further includes constructing a dual-threshold coupling trigger function: ; in: The preset trigger threshold ranges from 0.5 to 0.8, with a default value of 0.65. It can be dynamically adjusted based on the confidence level of historical power plant data. When... < and When ≤ 0, > Trigger the technical modification plan generation command. Coefficient , Based on the power station's historical PR deviation With electricity price volatility Adaptive adjustment using the least squares method satisfies + =1.

[0017] In the above technical solution, the four hard constraint verification rules in step S4 are as follows: (1) MPPT voltage window matching: The maximum power point voltage of the new module string falls within the inverter MPPT voltage range, and the open circuit voltage of the new module string does not exceed the inverter's maximum input voltage; (2) Capacity ratio compliance: The ratio of the total capacity of the DC side to the rated capacity of the AC side is between 1.1 and 1.5; (3) Grid-type short-circuit ratio verification: The ratio of the short-circuit capacity at the grid connection point to the rated power of the inverter is not less than 1.5; (4) Support load margin: The total weight of the new components does not exceed 85% of the original support design load.

[0018] In the above technical solution, the steps for generating candidate technical modification schemes and verifying electrical topology hard constraints are as follows: S'1: Scenario generation, the engine matches the following types from the strategy library: Component replacement (replacing low-power components with high-power components, and replacing over-degrading components with components of equal power). Inverter upgrade (centralized to string inverter, with added grid-connected functionality) Digital transformation (deployment of intelligent IV scanning system, optimizer / shutdown device).

[0019] Let the generation The candidate solution, the first Each scheme is denoted as ; S'2: For each candidate solution Map it to the digital twin electrical topology model of the power plant, and perform the following four hard constraint checks in sequence: ①MPPT voltage window matching: in For the inverter's MPPT voltage tracking range, For the open-circuit voltage of the new component series, This is the voltage at the maximum power point; ② Compliance with capacity ratio: in This represents the total capacity on the DC side. Rated capacity for the AC side.

[0020] ③ Grid-type short-circuit ratio verification (for inverter upgrades): in The short-circuit capacity at the grid connection point (MVA). The inverter's rated power (MW); ④ Support load margin: in This represents the total weight of the newly added components. The original support was designed to withstand the load. S'3: If candidate solutions If all four constraints are satisfied, the candidate solution is retained in the compatible solution set; if the candidate solution... If any of the constraints is not satisfied, it is marked as a "topological conflict" and removed. All solutions that satisfy all four constraints constitute a set of compatible solutions. .

[0021] In the above technical solution, the dynamic investment cost correction formula in step S5 is: Where: CAPEX0 is the benchmark investment cost; , , , These are the benchmark weekly average prices of components and inverters, respectively. , These refer to the required quantities of components and inverters, respectively. In the above technical solution, the incremental internal rate of return mentioned in step S5... Solve using the following formula: ,in .

[0022] In the above technical solution, the executable technical modification sequence output in step S5 is JSON structured data, which includes: technical modification type, equipment model, bill of quantities, dynamic investment budget, incremental internal rate of return, investment payback period, electrical compatibility verification results, construction priority, expected start time and construction period.

[0023] To achieve the above objectives, the second technical solution of the present invention is: a photovoltaic power plant technical upgrade scheme generation system based on multi-dimensional data, characterized in that it includes: a data acquisition module for acquiring multi-dimensional operation and maintenance data of three types of intelligent agents: monitoring, inspection, and cleaning; a health quantification module for constructing a string health model and calculating the overall health of the power plant; a dual threshold triggering module for completing technical threshold judgment and economic boundary constraint calculation, and outputting technical upgrade triggering instructions; a scheme generation module for matching and generating a set of candidate technical upgrade schemes according to the triggering instructions; a topology verification module for performing digital twin electrical hard constraint verification on candidate technical upgrade schemes and outputting a set of compatible schemes; and an economic deduction module for accessing real-time industry chain prices for dynamic economic calculation and outputting an executable technical upgrade sequence sorted by incremental internal rate of return.

[0024] In summary, the advantages of using the technical solution of this invention compared to traditional technical means are as follows: I. This invention combines technical and economic factors, employing a dual-threshold nonlinear coupling triggering mechanism. A dual-threshold coupling triggering function is established, adaptively adjusted based on the power plant's historical PR deviation and electricity price volatility, triggering a preset technical upgrade plan and generating instructions. This avoids wasting capital on premature upgrades and prevents revenue loss due to delayed upgrades.

[0025] This enables proactive and precise triggering of technological upgrading decisions, avoiding the subjectivity of human experience and misjudgments based on single thresholds.

[0026] II. This invention, based on digital twin electrical topology hard constraint verification, maps candidate technical modification schemes to a digital twin electrical topology model, and enforces four physical constraint verifications: It has higher compatibility with the existing electrical equipment in the power station, avoiding problems such as MPPT window mismatch, capacity ratio violation, insufficient short-circuit ratio or excessive support load during on-site implementation.

[0027] The first-time approval rate of technological upgrading plans increased by ≥40%, and the capital misallocation rate decreased by ≥35%.

[0028] III. The present invention utilizes a dynamic supply chain price-driven incremental IRR rolling evaluation and executable sequence output, which dynamically adjusts investment costs by accessing the average price of components and inverters at weekly frequencies via API. ), and calculate the incremental internal rate of return (ΔIRR) for each option on a rolling basis. kThe final output is an executable technical modification sequence sorted in descending order of ΔIRR. The sequence includes: equipment model, engineering quantity, dynamic budget, ΔIRR, investment payback period, electrical compatibility verification report, and construction priority, which can be directly issued to the work order system / ERP to trigger resource allocation.

[0029] The IRR calculation error is reduced by ≥15%, and the decision-making cycle is compressed from the weekly level to the hourly level, realizing a technical closed loop from "assessment value" to "engineering execution".

[0030] Fourth, this invention integrates multi-dimensional operation and maintenance data of photovoltaic power plants into a continuous quantitative health model. It integrates string current, voltage, and fill factor of the monitoring agent, hot spot gradient direction and structural coupling degree of the inspection agent, and dust accumulation and shading loss rate of the cleaning agent. The weights are dynamically determined by the entropy weight method to construct a power plant health quantification model with a continuous score of 0-100.

[0031] It provides a refined state input for dual-threshold triggering, and the health assessment accuracy is significantly higher than that of existing single degradation model methods. Attached Figure Description

[0032] The foregoing and other objects, features and advantages of the present invention will become apparent from the following detailed description taken in conjunction with the accompanying drawings.

[0033] Figure 1 This is a schematic diagram of the principle framework of the present invention; Detailed Implementation

[0034] Based on the preferred embodiments of the present invention, and through the following description, those skilled in the art can make various changes and modifications without departing from the technical concept of the present invention. The technical scope of the present invention is not limited to the contents of the specification, but must be determined according to the scope of the claims.

[0035] The invention will be further described with reference to the following figures: Example 1 I. Data Collection In this embodiment, data acquisition is based on a deployed intelligent operation and maintenance system, which includes the following three intelligent agents: The monitoring agent collects the current of 186 strings every 5 minutes through the SCADA system. ,Voltage Perform cascade-level IV curve scanning once a month to extract fill factors. and string dispersion rate.

[0036] Inspection AI Agent: A full-site inspection is conducted quarterly using a drone equipped with a dual-spectrum infrared camera to obtain the hotspot temperature gradient direction consistency index. and structural coupling (Calculated using the "Intelligent Diagnosis Method for Hot Spots in Photovoltaic Modules Based on Multi-Dimensional Images" submitted by the applicant), outputting the hot spot type and severity.

[0037] Cleaning AI: The monitoring system records the robot's work path and power generation gain before and after each cleaning session, and calculates the dust accumulation and occlusion loss rate in each area. .

[0038] The digital twin orchestrator integrates the above data to construct a digital twin model of all elements of the power plant (based on the IEC 61850SCL file and the power plant single-line diagram, including the topology relationships of inverter nodes, string branches, transformer substation nodes and grid connection points).

[0039] II. Measurement of Health We will use a typical string (numbered S-0123) from a power plant as an example. This string consists of 22 280Wp modules connected in series. Current measured data: String current = 7.85, the reference current for healthy string circuits of the same model. = 8.30 String voltage = 620V, reference voltage = 660V Fill factor = 0.68, baseline fill factor Attenuation = 0.08 Dust accumulation and shading loss rate = 0.09 The weights, dynamically determined based on the historical data of the power station using the entropy weight method, are as follows: = 0.25, = 0.20, =0.40, =0.15.

[0040] Substitute into formula (1) to calculate the health of the string: This means the health score for this string is 91.9 (out of 100). After calculating the health score for each of the 186 strings at the entire station, the overall health score of the power station is... .

[0041] III. Dual Threshold Trigger Judgment Technical threshold: setting = 72 (corresponding to PR decay ≥15%). Current = 68.3 < 72, which meets the technical requirements.

[0042] Economic boundary conditions: Calculating the cost of power plant upgrades .

[0043] The preliminary technical upgrade plan is as follows: replace the 280Wp modules with 550Wp monocrystalline silicon modules (replacing all severely degraded strings and hot spot fault strings, accounting for about 35% of the total), and replace 200 500kW centralized inverters with 175kW string inverters (8 units per 1.25MW).

[0044] Technological upgrading incremental investment The following costs were incurred: 125,000 module purchases, with an average weekly price of 0.71 yuan / W for 550Wp modules (obtained via API), resulting in a module cost of approximately 125,000 × 550 × 0.71 = 48.81 million yuan; 800 175kW string inverters were purchased at an average price of 0.12 yuan / W, resulting in a cost of approximately 800 × 175,000 × 0.12 = 16.8 million yuan; installation and other expenses amounted to approximately 8 million yuan; totaling approximately 73.61 million yuan.

[0045] Changes in annual maintenance costs The new equipment reduces maintenance costs by approximately 20% compared to the old equipment, resulting in annual savings of approximately 600,000 yuan. = -600,000 yuan / year.

[0046] Annual increase in power generation After the technological upgrade, due to improved component efficiency, elimination of hot spots, and improved inverter efficiency, the annual power generation is expected to increase from 138 million kWh to 155 million kWh, an increase of 17 million kWh per year.

[0047] Equipment life cycle of technical upgrade = 10 years (based on a 25-year warranty for components, but the economic calculation for technological upgrades uses a 10-year discount period).

[0048] Electricity price changes = 0 (Electricity price remains unchanged at 0.32 yuan / kWh).

[0049] Benchmark cost per kilowatt-hour before technological upgrade = 0.28 yuan / kWh (including operation and maintenance and depreciation).

[0050] Substitute into formula (3): Calculated = 0.398 yuan / kWh, far greater than 0, meaning the increase in cost per kilowatt-hour after the upgrade is too high and economically unsustainable. However, analysis revealed that the reason was that the solution replaced 35% of the modules, resulting in excessive investment. The system's automatic adjustment strategy: only replace the 12% of modules with severe hot spot failures (approximately 42,857 units), and strings with a dispersion rate >20% (approximately 8%), totaling approximately 20% module replacements. Recalculation: Component cost: 42,857 × 550 × 0.71 = 16.73 million yuan The cost of the inverter remains at 16.8 million yuan, while the installation and construction cost has been reduced to 4 million yuan. total = 37.53 million yuan Annual increase in power generation = 9 million kWh (due to only replacing some components) Substitute into formula (3): Yuan / kWh, still greater than 0.

[0051] Further optimization strategy: Retain the original centralized inverter, only replace the modules and add a smart optimizer (to resolve MPPT mismatch). The optimizer cost is approximately 0.05 yuan / W, replacing 20% ​​of the modules costs 16.73 million yuan, optimizer cost: 100MW × 0.05 = 5 million yuan, installation cost 2 million yuan, totaling... = 23.73 million yuan, annual increase in power generation = 8 million kWh, Not much has changed. Substitute: It remains positive. Ultimately, the system is connected to the local power grid company's announced renewable energy pricing policy: starting in 2025, existing power plants will participate in market-based transactions, with the average settlement price expected to rise to 0.38 yuan / kWh. = 0.38 - 0.32 = 0.06 yuan / kWh. Recalculate: at this time The value is >0, indicating poor economic viability. Based on current industry chain prices and electricity rates, simply replacing components is not cost-effective. Therefore, the system recommends an "inverter upgrade + digital transformation" solution: replacing centralized inverters with string inverters (16.8 million yuan) and deploying a smart IV scanning system (500,000 yuan), without replacing the components. This solution involves an investment of 17.3 million yuan, resulting in an annual increase in power generation of approximately 4 million kWh due to improved inverter efficiency and MPPT optimization. Changes ignored =0.06, calculated as follows: It still does not meet the economic boundary. The system ultimately determines that, under the current conditions of a component price of 0.71 yuan / W and an electricity price of 0.38 yuan / kWh, any technological upgrade plan... All are greater than 0, therefore not satisfied. The economic triggering condition. Therefore, the dual threshold triggering function. Not exceeding This does not trigger the technical modification plan generation command.

[0052] This embodiment demonstrates the practical operation of the dual-threshold mechanism: even if the health level is below the technical threshold, if the economic boundary is not satisfied (i.e., ... (If the cost per kilowatt-hour increases after the technical upgrade, the system will not blindly recommend technical upgrades, thus avoiding ineffective investment.)

[0053] IV. Triggering Scenario Demonstration (Another Power Plant) To fully demonstrate the subsequent steps of this invention, a 50MW power plant in East China is used as an example. This power plant has been operating for 6 years, and its modules are PERC monocrystalline 390Wp. The current health status is... = 65 (PR degradation 12%), the average price of components in the industry chain drops to 0.62 yuan / W, and the local market-based electricity price is 0.45 yuan / kWh. Calculations show that adopting the "equal power replacement of over-degraded components + inverter upgrade" solution... = 9.5 million yuan, with an annual increase in power generation of 6 million kWh. = -200,000 yuan / year = 0, substituting gives Yuan / kWh, because The value of >0 is still not met. However, when considering the factor of rising electricity prices (expected to rise to 0.48 yuan / kWh), = 0.03, = 0.125 - 0.03 = 0.095 > 0. The system ultimately recommends the "replace only severely degraded components (10%) + intelligent optimizer" solution, with an investment of 4.5 million yuan and an annual increase in power generation of 3 million kWh. = (450) / 3000 = 0.15, still greater than 0. In reality, due to higher subsidies, the owner of the power station still decided to upgrade the technology, but this system strictly adheres to economic boundaries and will not trigger [the upgrade]. Only when a scheme is calculated... It will only trigger automatically at certain times.

[0054] To demonstrate the complete process, we assume a scenario that satisfies two thresholds: the health status of a power plant. = 60, the average price of components in the industrial chain drops to 0.50 yuan / W, the electricity price is 0.55 yuan / kWh, 20% of the components are replaced with high-power components (550W replacing the original 335W), the investment is 7 million yuan, and the annual power generation increases by 8 million kWh. = 700 / 8000 = 0.0875, =0.05, then = 0.0875 -0.05 = 0.0375 > 0. Still a positive value. If the component price further decreases to 0.45 yuan / W, and the investment is 6.3 million yuan, then... = 630 / 8000 = 0.07875, subtracting 0.05 gives 0.02875 > 0. To achieve ≤ 0, the component price needs to be extremely low or the electricity price extremely high, which is rare in reality. Therefore, the dual threshold mechanism of this invention is relatively conservative and conforms to the engineering principle that "technical upgrades should have clear economic viability".

[0055] We assume an ideal triggering scenario: power plant health. = 55, module price 0.45 yuan / W, electricity price 0.65 yuan / kWh, replacing 20% ​​of the modules requires an investment of 6.3 million yuan, and will increase annual power generation by 10 million kWh. = 630 / 10000 = 0.063, = 0.65 - 0.55 = 0.10, then = 0.063 - 0.10 = -0.037 ≤ 0, satisfying the economic boundary condition. At this point... < Both thresholds are satisfied. > Trigger the technical modification plan generation command.

[0056] V. Generation of Candidate Technological Upgrade Schemes Upon responding to the trigger command, the solution generation engine matches the following candidate solutions from the preset strategy library: Option A: Replace the existing 335Wp modules with 550Wp monocrystalline silicon modules, with a replacement ratio of 20% (approximately 30,000 units), while retaining the original centralized inverter.

[0057] - Option B: Replace 20% of the components (same as Option A) and upgrade the centralized inverter to a string inverter (175kW×80 units).

[0058] Option C: Replace only the severely over-degraded components (10%) and deploy the Smart Optimizer (one per component).

[0059] Option D: Upgrade the centralized inverter to a string inverter without replacing the components.

[0060] VI. Verification of Hard Constraints in Electrical Topology The four schemes mentioned above are mapped to the digital twin electrical topology model of the power plant, and four hard constraint checks are performed.

[0061] Validation of Solution A: - MPPT voltage window matching: The original centralized inverter's MPPT range was 450V~850V. 22 550Wp modules (operating voltage approximately 42V) were connected in series to achieve this. = 924V, exceeding the 850V upper limit. Not satisfied.

[0062] - Capacity Ratio: Due to the increase in module power, the total DC capacity increases after a 20% replacement. Calculations show... = 1.58>1.5). Not satisfied.

[0063] Option A was eliminated.

[0064] Validation of Solution B: - The MPPT range of the string inverter is 200V~1000V, with 22 550Wp modules connected in series (V_{mpp,string}} = 924V), which is within the range; the open circuit voltage is approximately 52V×22=1144V, and the maximum input voltage of the inverter is 1500V, which meets the requirements.

[0065] - Short-circuit ratio verification: Short-circuit capacity at grid connection point = 250 MVA, total inverter capacity 50MW, = 250 / 50 = 5.0 ≥ 1.5, which satisfies the condition.

[0066] - Support Load: The new component weighs 22kg / piece, while the original component weighs 19kg / piece, an increase of approximately 16%. The original support design load needs to be verified. Referring to the original design documents, the support design load is 25kg / piece, with a safety margin of approximately 31.6%. Calculations... = 0.85 = 21.25kg The actual weight of the new component is 22kg, which is greater than 21.25kg, failing to meet the load margin requirement (exceeding it by approximately 3.5%). Therefore, Option B is also eliminated.

[0067] Validation of Scheme C: - Only 10% of the components were replaced, and string inverters were still used. However, the voltage of the replaced component strings was still within the MPPT range, the capacity ratio was 1.32 (satisfied), the SCR was satisfied, and in terms of load, only 10% of the brackets could bear 22kg, while the rest could still bear 19kg. After overall calculation, the average load increased by 1.5%, which was far less than the 15% margin, so it was satisfied.

[0068] - All validations passed, scheme C is retained.

[0069] Validation of Scheme D: - Upgrade the inverter only, without replacing the modules. The module string voltage remains the same as the original 22 335Wp modules connected in series. =32V×22=704V, within the MPPT range; capacity ratio 1.2; SCR satisfied; support load unchanged. All passed. Scheme D retained.

[0070] The final set of compatible solutions is {Solution C, Solution D}.

[0071] VII. Dynamic Economic Deduction The system connects to the PVinfoLink platform via API to obtain the average component price for the current week (week 15 of 2025). =0.45 yuan / W, average price of string inverters = 0.11 yuan / W. The benchmark weekly prices were 0.48 yuan / W and 0.12 yuan / W, respectively.

[0072] Option C (Replace 10% of components + optimizer): - Number of modules: 30,000 units × 550Wp = 16.5MWp, module cost = 16.5e6 × 0.45 = 7.425 million yuan.

[0073] - Optimizer: 30,000 × 50 yuan / each = 1.5 million yuan.

[0074] - Installation and construction cost: 1 million yuan.

[0075] - = 742.5 + 150 + 100 = 9.925 million yuan.

[0076] - Annual increase in electricity generation = 3.5 million kWh.

[0077] - Changes in annual maintenance costs = -150,000 yuan / year (the optimizer can reduce manual inspection).

[0078] Electricity Price = 0.65 yuan / kWh (fixed).

[0079] - Life cycle = 10 years.

[0080] - calculate Substitute into equation (10): Solving using Newton's iterative method yields the following results: 19.2%, with an investment payback period of approximately 4.3 years.

[0081] Option D (replace only the inverter): - Inverter procurement: 80 units × 175kW × 0.11 yuan / W = 80 × 175,000 × 0.11 = 1.54 million yuan.

[0082] - Installation and commissioning fee: 500,000 yuan.

[0083] - = 2.04 million yuan.

[0084] - Annual increase in electricity generation = 1.2 million kWh (due to improved inverter efficiency and MPPT optimization).

[0085] - = -100,000 yuan / year (new inverter operation and maintenance costs are reduced).

[0086] - Substitute into the equation: Solving 42.6, with an investment payback period of approximately 2.1 years.

[0087] 8. Output executable technical modification sequence The system generates an executable technical modification sequence in JSON format, as shown in the example below: json { "station_id": "ES_50MW_001", "timestamp": "2025-04-13T10:30:00Z", "tech_seq": [ { "rank": 1, "type": "inverter_upgrade", "description": "Replace the existing 500kW centralized inverter with a 175kW string inverter", "equipment_model": "SUN2000-175KTL-H0", "quantity": 80, "unit_price": 0.11, "total_capex": 2040000, "delta_irr": 0.426, "payback_years": 2.1, "electrical_check": {"mppt": "pass", "dc_ac_ratio": "pass", "scr":"pass", "load": "pass"}, "priority": 1, "estimated_start_date": "2025-05-01", "estimated_duration_days": 30 }, { "rank": 2, "type": "module_replacement_optimizer", "description": "Replace 10% of severely over-degraded components and install a smart optimizer", "equipment_model": "550W mono PERC + Optimizer", "quantity": 30000, "unit_price": 0.45, "total_capex": 9925000, "delta_irr": 0.192, "payback_years": 4.3, "electrical_check": {"mppt": "pass", "dc_ac_ratio": "pass", "scr":"pass", "load": "pass"}, "priority": 2, "estimated_start_date": "2025-06-15", "estimated_duration_days": 45 } ] } This sequence can be directly sent to the power plant work order system to trigger material procurement and construction scheduling.

[0088] The above embodiments fully demonstrate the technical solution and beneficial effects of the present invention. Those skilled in the art can reproduce all the technical content of the present invention without creative effort based on the teachings of these embodiments.

Claims

1. A method for generating photovoltaic power plant technical upgrade schemes based on multi-dimensional data, characterized in that, Includes the following steps: S1: Multi-source data acquisition and power plant health measurement, acquire string current, voltage and fill factor collected by monitoring agent, hot spot temperature gradient and structural coupling degree collected by inspection agent, and dust accumulation and shading loss rate collected by cleaning agent, construct string-level health model and calculate the overall health of power plant. S2: Technical and economic dual threshold coupling judgment, compare the overall health of the power station with the preset technical threshold, and calculate the incremental cost per kilowatt-hour based on the incremental investment in technical upgrades, changes in annual operation and maintenance costs, annual increase in power generation, equipment life cycle and electricity price fluctuations. When the overall health of the power station is lower than the technical threshold and the incremental cost per kilowatt-hour is not greater than 0, trigger the technical upgrade scheme generation instruction. S3: Candidate technical upgrade scheme generation: In response to the technical upgrade scheme generation instruction, a set of candidate technical upgrade schemes containing at least one of component replacement, inverter upgrade, and digital transformation is obtained from the preset strategy library; S4: Electrical topology hard constraint verification. The candidate technical modification scheme set is mapped to the digital twin electrical topology model of the power plant. MPPT voltage window matching verification, capacity ratio compliance verification, network short-circuit ratio verification and support load margin verification are performed in sequence. Schemes that do not meet the constraints are eliminated to obtain a set of compatible technical modification schemes. S5: Dynamic economic simulation and executable sequence output. It obtains equipment prices in the photovoltaic industry chain in real time through API interface, dynamically adjusts the investment cost of each compatible technical transformation scheme, calculates the corresponding incremental internal rate of return, sorts them from high to low incremental internal rate of return, and generates and outputs executable technical transformation sequences.

2. The method for generating photovoltaic power plant technical upgrade schemes based on multi-dimensional data according to claim 1, characterized in that, The string-level health model mentioned in step S1 is as follows: , in: , , These are the measured current, measured voltage, and measured fill factor of the i-th string, respectively. , , These are the current reference value, voltage reference value, and fill factor reference value for the same type of healthy string, respectively; This is the amount of fill factor decay. , , , The weights are dynamically determined using the entropy weight method and satisfy the following conditions: + + + =1; The overall health status of the power station is: Where N is the total number of strings in the entire site. .

3. The method for generating photovoltaic power plant technical upgrade schemes based on multi-dimensional data according to claim 1, characterized in that, The formula for calculating the incremental electricity cost in step S2 is as follows: Among them: For incremental investment in technological upgrading, For annual changes in operation and maintenance costs, To increase annual power generation, For the life cycle of technical upgrade equipment, This refers to changes in electricity prices.

4. The method for generating photovoltaic power plant technical upgrade schemes based on multi-dimensional data according to claim 3, characterized in that, Step S2 also includes constructing a dual-threshold coupled trigger function: in: The preset trigger threshold ranges from 0.5 to 0.8, with a default value of 0.

65. It can be dynamically adjusted based on the confidence level of historical power plant data. When... < and When ≤ 0, > Trigger the technical modification plan generation command. Coefficient , Based on the power station's historical PR deviation With electricity price volatility Adaptive adjustment using the least squares method satisfies + =1.

5. The method for generating photovoltaic power plant technical upgrade schemes based on multi-dimensional data according to claim 1, characterized in that, The four hard constraint verification rules mentioned in step S4 are as follows: (1) MPPT voltage window matching: The maximum power point voltage of the new module string falls within the inverter MPPT voltage range, and the open circuit voltage of the new module string does not exceed the inverter's maximum input voltage; (2) Capacity ratio compliance: The ratio of the total capacity of the DC side to the rated capacity of the AC side is between 1.1 and 1.5; (3) Grid-type short-circuit ratio verification: The ratio of the short-circuit capacity at the grid connection point to the rated power of the inverter is not less than 1.5; (4) Support load margin: The total weight of the new components does not exceed 85% of the original support design load.

6. The method for generating photovoltaic power plant technical upgrade schemes based on multi-dimensional data according to claim 5, characterized in that, The steps for generating candidate technical modification schemes and verifying hard constraints on electrical topology are as follows: S'1: Scenario generation, the engine matches the following types from the strategy library: Component replacement (replacing low-power components with high-power components, and replacing over-degrading components with components of equal power). Inverter upgrade (centralized to string inverter, with added grid-connected functionality) Digital transformation (deployment of intelligent IV scanning system, optimizer / shutdown device). Let the generation The candidate solution, the first Each scheme is denoted as ; S'2: For each candidate solution Map it to the digital twin electrical topology model of the power plant, and perform the following four hard constraint checks in sequence: ①MPPT voltage window matching: in For the inverter's MPPT voltage tracking range, For the open-circuit voltage of the new component series, This is the voltage at the maximum power point; ② Compliance with capacity ratio: in This represents the total capacity on the DC side. Rated capacity for the AC side. ③ Grid-type short-circuit ratio verification (for inverter upgrades): in The short-circuit capacity at the grid connection point (MVA). The inverter's rated power (MW); ④ Support load margin: in This represents the total weight of the newly added components. The original support was designed to withstand the load. S'3: If candidate solutions If all four constraints are satisfied, the candidate solution is retained in the compatible solution set; if the candidate solution... If any of the constraints is not satisfied, it is marked as a "topological conflict" and removed. All solutions that satisfy all four constraints constitute a set of compatible solutions. .

7. The method for generating photovoltaic power plant technical upgrade schemes based on multi-dimensional data according to claim 1, characterized in that, The dynamic investment cost correction formula mentioned in step S5 is as follows: Where: CAPEX0 is the benchmark investment cost; , , , These are the benchmark weekly average prices of components and inverters, respectively. , These are the required quantities of components and inverters, respectively.

8. The method for generating photovoltaic power plant technical upgrade schemes based on multi-dimensional data according to claim 1, characterized in that, The incremental internal rate of return mentioned in step S5 Solve using the following formula: ,in .

9. The method for generating photovoltaic power plant technical upgrade schemes based on multi-dimensional data according to claim 1, characterized in that, The executable technical modification sequence output in step S5 is JSON structured data, which includes: technical modification type, equipment model, bill of quantities, dynamic investment budget, incremental internal rate of return, investment payback period, electrical compatibility verification results, construction priority, expected start time and construction period.

10. A photovoltaic power plant technical upgrade scheme generation system based on multi-dimensional data, characterized in that, The photovoltaic power plant technical upgrade scheme generation method based on multi-dimensional data according to any one of claims 1 to 9 includes: a data acquisition module for acquiring multi-dimensional operation and maintenance data of three types of intelligent agents: monitoring, inspection, and cleaning; a health quantification module for constructing a string health model and calculating the overall health of the power plant; a dual threshold triggering module for completing technical threshold judgment and economic boundary constraint calculation, and outputting technical upgrade triggering instructions; a scheme generation module for matching and generating a set of candidate technical upgrade schemes according to the triggering instructions; a topology verification module for performing digital twin electrical hard constraint verification on the candidate technical upgrade schemes and outputting a set of compatible schemes; and an economic deduction module for accessing real-time industry chain prices for dynamic economic calculation and outputting an executable technical upgrade sequence sorted by incremental internal rate of return.

Images