A wind turbine unit modification scheme evaluation method and device based on operation data

Abstract

This invention discloses a method and apparatus for evaluating wind turbine retrofitting schemes based on operational data, relating to the field of wind power engineering technology. The method includes: collecting at least five years of operational data from the target wind farm; preprocessing the data using the DBSCAN-segmented quartile combination algorithm; restoring the free-flow wind speed through primary and secondary correction functions to separate the impact of wind resource fluctuations and obtain the net power generation revenue increment from the retrofit; and establishing a multi-dimensional comprehensive evaluation system to assess the degree to which the technological upgrade improves the overall benefits of the wind farm. The apparatus includes a data acquisition module, a correction module, and an evaluation module, which execute the steps described above. This invention solves the problems of insufficient representative wind measurement data and single evaluation indicators for retrofitting schemes. By utilizing existing operational data from wind farms, it can effectively isolate the impact of wind resource fluctuations, explore the net power generation revenue increment brought about by the technological upgrade of wind turbines, improve the accuracy and comprehensiveness of the evaluation, and provide a scientific basis for the selection of retrofitting schemes.

Description

Technical Field

[0001] This invention relates to the field of wind turbine retrofitting technology, specifically to a method and apparatus for evaluating wind turbine retrofitting schemes based on operational data. Background Technology

[0002] As wind farms age, older wind farms face problems such as immature early technologies, outdated control strategies, and aging equipment, leading to decreased wind turbine operating efficiency and frequent failures. Various wind turbine retrofitting solutions (such as control strategy optimization, blade lengthening, tower raising, and relocation / construction) have become key means to solve these problems. However, evaluating the optimization effect of wind turbine retrofitting solutions has always been an important issue for the industry. Especially as wind farms age, data from free-flow anemometers used in the early planning stage are prone to loss or insufficient representativeness, seriously affecting the accuracy of post-retrofit power generation assessment and making it difficult to scientifically judge the rationality and feasibility of the retrofitting solution.

[0003] In existing technologies, some assessment methods rely solely on a single data source or a simplified calculation model, failing to fully utilize existing wind farm operational data and exhibiting the following drawbacks: First, they cannot effectively distinguish between annual wind resource fluctuations and the benefits brought by the retrofit itself, leading to economic assessments relying on single-cycle data and easily misjudging the actual benefits of retrofits due to interannual differences in wind resources; second, the assessment process over-relies on economic parameters, while the uncertainty of economic factors is significant, resulting in poor reproducibility of the technical solutions; third, the existing evaluation index system is inadequate, focusing excessively on power output indicators while ignoring key factors such as power curve changes and reliability details, lacking independence between indicators, and failing to cover important evaluation dimensions such as generator group efficiency and cost per kilowatt-hour, resulting in assessment results that are not comprehensive or objective.

[0004] Therefore, there is an urgent need for a method and device that can fully utilize existing wind farm operation data, effectively isolate the impact of wind resource fluctuations, explore the incremental net power generation revenue brought about by the wind turbine technical transformation process, and comprehensively evaluate wind turbine transformation schemes based on multiple dimensions, so as to provide a scientific basis for the comparison and decision-making of transformation schemes. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a method and device for evaluating wind turbine retrofit schemes based on operational data, which solves the problems of lack of representative free-flow wind measurement data, inability to distinguish between wind resource fluctuations and retrofit benefits, and single evaluation indicators in the prior art, and realizes accurate and comprehensive evaluation of wind turbine retrofit schemes.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A method for evaluating wind turbine retrofitting schemes based on operational data includes the following steps: Step 1: Determine the initial wind turbine technical upgrade plan, collect turbine operation data for at least 5 full years since the target wind farm was connected to the grid. The turbine operation data includes SCADA system data, concurrent wind power prediction tower data and financial system data, and use the DBSCAN-segmented quartile combination algorithm to preprocess and clean the turbine operation data. Step 2: Select a candidate set of corrected wind turbines. Based on the SCADA system data and the concurrent wind power prediction tower data, restore the free flow wind speed of the turbines by using primary and secondary correction functions to obtain the corrected wind resource data of the turbines. Step 3: Determine the wind resource base year based on the corrected unit wind resource data, calculate the wind speed fluctuation coefficient for each year, and establish a wind resource fluctuation coefficient model. Step 4: Collect environmental data of the target wind farm, and construct a CFD fluid dynamics model by combining the corrected unit wind resource data. Based on the CFD fluid dynamics model and wind speed fluctuation coefficient, separate the impact of wind resource fluctuation to obtain the incremental net power generation revenue from the retrofit. Step 5: Based on the financial system data and the incremental net power generation revenue from the renovation, the net revenue volatility coefficient is introduced to correct the investment payback period calculation, and the benchmark investment payback period and the dynamically corrected investment payback period are output. Step 6: Establish a multi-dimensional comprehensive evaluation index system, which includes indicators such as power curve, availability, reliability, wind resources, power loss, and economic efficiency. Combine the net power generation revenue increase from the renovation, the investment payback period, and the scores of each evaluation index to assess the feasibility of the renovation plan.

[0007] In Step 1 above, the SCADA system data includes 10-minute average wind speed, wind direction, power data, fault record data, and operation and maintenance cost data; the financial system data includes planned renovation costs, operating costs, other costs, historical electricity data, and electricity price data; the processing flow of the DBSCAN-segmented quartile combination algorithm is as follows: first, the wind speed-power data is clustered using the DBSCAN algorithm to remove abnormal data clusters, then the intervals are divided according to wind speed at equal intervals, and the quartile method is used to further remove discrete abnormal data in each interval.

[0008] The process of selecting candidate modified wind turbine sets in Step 2 above is as follows: taking into account the different turbine models in the target wind farm, the relative positional relationship of each unit, the terrain level assessment results at the unit location, and the terrain conditions of the target wind farm, at least one wind turbine with more than 5 complete years of operating data is selected as the candidate modified wind turbine set. The candidate modified wind turbine set includes unit types with different single unit capacities and different hub heights.

[0009] The calculation formula for the first correction in Step 2 above is: ; in, The first correction factor is... The SCADA unit wind speed data before correction. This is the unit wind speed data after one correction. The average turbine wind speed of the target wind farm is given. The theoretical average wind speed of the target wind farm turbines; The theoretical average wind speed is obtained by dividing the wind speed into zones with a step size of 0.5 m / s, obtaining the average power curve over the entire annual operating cycle based on the SCADA system data of the target wind farm units, and interpolating the average power value of the units onto the average power curve to obtain the theoretical average wind speed.

[0010] The second correction process in Step 2 above is as follows: a. Divide the corrected unit wind speed sequence into sectors according to wind direction, and select the sectors with the highest frequency of wind direction as the effective sectors; b. Perform a correlation test on the wind resource sequence of the unit after one correction and the wind resource sequence of the wind power prediction tower during the same period to ensure that the correlation reaches above 0.85; c. The formula for calculating the second correction is: ; in, This represents the unit wind speed data after secondary correction for the i-th sector. This refers to the unit wind speed data after one correction for the i-th sector. and These are the slope and intercept of the linear fit between the two sets of data in the i-th sector, respectively.

[0011] The process of determining the wind resource base year in Step 3 above is as follows: Wind resource characteristic values ​​are calculated based on the corrected unit wind resource data. The wind resource characteristic values ​​include annual average wind speed, Weibull distribution parameters, and effective wind time rate. The year in the past 5 years in which the wind resource characteristic values ​​are close to the long-term average level is selected as the wind resource benchmark year. The formula for calculating the wind speed fluctuation coefficient is: ; in, Let be the wind speed fluctuation coefficient in year j. Let j be the average wind speed in year j. The base annual average wind speed.

[0012] The environmental data in Step 4 above includes topographic data, roughness data, and local air density of the wind field area. When constructing the CFD fluid dynamics model, WT, Openwind, Windfarmer, or WAsP software is used to simulate the wind field, and the model parameters are adjusted to ensure the convergence of the flow field simulation. The process for obtaining the incremental net power generation revenue from the renovation is as follows: a. Calculate the difference in benchmark power generation revenue before and after the retrofit using a CFD fluid dynamics model, based on the benchmark year's wind resources. By analyzing the difference in power curves before and after the modification, the additional benefits brought by the modification are calculated separately. ; b. Calculate the difference in actual power generation revenue by combining the wind speed fluctuation coefficients for each year. ; c. Based on the actual power generation revenue difference over at least five full years, calculate the net revenue standard deviation σ, and separate the impact of wind resource fluctuations through deviation analysis to obtain the incremental net power generation revenue from the retrofit. .

[0013] The formula for dynamically adjusting the payback period in Step 5 above is as follows: ; in, To dynamically adjust the investment payback period, Let σ be the benchmark payback period and σ be the standard deviation of net income. This represents the average net income over several years. The formula for calculating the benchmark payback period is as follows: ; in, The net cash flow increment for period j consists of the increase in power generation efficiency before and after the renovation, the increase in operating costs, the increase in other expenses, and the planned renovation costs.

[0014] The multi-dimensional comprehensive evaluation index system in Step 6 above includes: Power curve indicators: aging power curve deviation, power curve reachability; Availability metrics: Time availability, capacity availability; Reliability metrics: Mean time between repairs, maintenance performance factor, mean time to repair; Wind resource indicators: average wind speed, effective wind time rate; Power loss indicators: overall power consumption rate, transmission line loss rate, and power loss; Economic indicators: unit capacity operation and maintenance cost, cost per kilowatt-hour, and generator group generation efficiency.

[0015] The calculation formulas for each indicator in the above multi-dimensional comprehensive evaluation index system are as follows: Aging power curve offset: ; in This represents the actual power output of the unit when it was first put into operation at a wind speed of 1 m / s. The actual power of the unit within the statistical period when the wind speed is in m / s; Power curve reachability: ; in This represents the actual power generation of the generating unit within the statistical period. This represents the theoretical power generation of the generating unit within the statistical period. Time availability: ; in To calculate the duration of unit downtime due to faults during a statistical period, This represents the total duration of the statistical period. Capacity availability: ; in Unit capacity; Mean time between maintenance (MTBG): ; in This refers to the number of maintenance operations within the statistical period. Maintenance performance factor: ; in To calculate the total cost of unit failure repairs within the statistical period, This represents the total cost of unit operation and maintenance within the statistical period. Mean time to repair: ; in Let N be the failure duration of the i-th unit within the statistical period, and N be the number of units in the field. Effective wind time rate: ; in The duration of wind speed between the cut-in wind speed and the cut-out wind speed; Overall site power consumption rate: ; in This represents the actual power generation of the wind farm during the statistical period. To calculate the electricity purchased from the network within the statistical period, For the amount of electricity used in the internet during the statistical period; Outgoing line loss rate: ; in The power supply at the beginning of the transmission line. This refers to the power output at the end of the transmission line. Power loss: ; in Power loss due to fault To compensate for the power loss due to power rationing, To maintain the power loss, For other lost electricity; Unit capacity maintenance cost: ; in To calculate the total cost of wind farm operation and maintenance during the statistical period, This represents the total installed capacity within the facility. Cost per kilowatt-hour: ; in Total cost of power generation; Group power generation efficiency: ; in This is the unit's performance evaluation score.

[0016] The process of evaluating the feasibility of the renovation plan in Step 6 above is as follows: the fuzzy comprehensive evaluation method is used to weight and score each indicator, and the results of net power generation revenue increase, benchmark investment payback period, and dynamically corrected investment payback period are combined. If the investment payback period does not exceed the remaining lifespan of the unit and the comprehensive score reaches the preset threshold, the renovation plan is deemed feasible.

[0017] The apparatus for evaluating a wind turbine retrofit scheme based on operational data, as described above, includes a data acquisition module, a correction module, and an evaluation module. The data acquisition module is used to collect unit operation data and environmental data for at least five full years since the target wind farm was connected to the grid. The unit operation data includes SCADA system data, concurrent wind power prediction tower data and financial system data. The environmental data includes wind farm area topography data, roughness data and local air density. The DBSCAN-segmented quartile combination algorithm is used to preprocess and clean the unit operation data. The correction module is used to select a candidate set of wind turbines for correction, restore the free flow wind speed of the turbines through primary and secondary correction functions, obtain the corrected wind resource data of the turbines, determine the wind resource base year and calculate the wind speed fluctuation coefficient for each year, and establish a wind resource fluctuation coefficient model. The evaluation module is used to construct a CFD fluid dynamics model, separate the impact of wind resource fluctuations to obtain the net power generation revenue increment of the renovation; introduce a net revenue fluctuation coefficient to correct the investment payback period calculation, and output the benchmark investment payback period and the dynamically corrected investment payback period; establish a multi-dimensional comprehensive evaluation index system to comprehensively evaluate the feasibility of the renovation plan.

[0018] The preprocessing and cleaning process of the above-mentioned acquisition module includes: firstly, performing cluster analysis on the wind speed-power data using the DBSCAN algorithm to remove abnormal data clusters; then, dividing the data into intervals according to wind speed; and further removing discrete abnormal data within each interval using the quartile method.

[0019] The aforementioned correction module includes a primary correction unit, a secondary correction unit, and a reference year calibration unit; The primary correction unit is used to calculate the unit wind speed data after primary correction based on SCADA system data; The secondary correction unit is used to perform secondary sector-based correction on the wind speed data after the first correction, in conjunction with the data from the concurrent wind power prediction tower. The base year calibration unit is used to calculate wind resource characteristic values, determine the wind resource base year, and calculate the wind speed fluctuation coefficient for each year.

[0020] The aforementioned evaluation modules include a model building unit, a revenue calculation unit, a payback period correction unit, and a comprehensive evaluation unit; The model building unit is used to combine the corrected unit wind resource data and environmental data to build a CFD fluid dynamics model; The revenue calculation unit is used to separate the impact of wind resource fluctuations and calculate the incremental net power generation revenue from the renovation. The payback period correction unit is used to introduce the net income volatility coefficient and calculate the benchmark payback period and the dynamically corrected payback period. The comprehensive evaluation unit is used to establish a multi-dimensional comprehensive evaluation system that includes indicators such as power curve, availability, reliability, wind resources, power loss, and economic efficiency. It adopts the fuzzy comprehensive evaluation method to quantify the scoring and combines revenue data and payback period data to evaluate the feasibility of the renovation plan.

[0021] The generator group power generation efficiency index in the above-mentioned multi-dimensional comprehensive evaluation index system is calculated based on the weighted average of the capacity of each unit and its power generation efficiency score, ensuring the fairness of cross-regional wind farm evaluation; the cost per kilowatt-hour and unit capacity operation and maintenance cost indexes are used to quantitatively evaluate the economic sustainability of the renovation plan.

[0022] The wind turbine retrofitting scheme evaluation method and apparatus mentioned in this invention, based on operational data, has the following beneficial effects: 1. A new module for wind resource benchmark year calibration and net income splitting has been added. By combining wind resource characteristic indicators, the interannual fluctuation of wind resources and the income brought by the transformation itself have been successfully separated. This solves the pain point of existing technologies that cannot objectively quantify the real benefits of transformation and forms a unique technical path.

[0023] 2. A dynamic investment payback period correction method is introduced, which combines wind resource statistical characteristics and net income fluctuation coefficient to break through the limitations of traditional static calculation. At the same time, the fairness of cross-regional wind farm evaluation is ensured by the generator group power generation efficiency index, providing a more scientific quantitative method for the economic assessment of retrofitting.

[0024] 3. A multi-dimensional comprehensive evaluation index system covering six categories, including power curve, availability, reliability, wind resources, power loss, and economic efficiency, has been constructed. Key indicators such as aging power curve deviation, generator group generation efficiency, and cost per kilowatt-hour have been added to make up for the shortcomings of the existing evaluation system, which is characterized by single indicators and one-sided focus, and to achieve a comprehensive quantitative evaluation of the renovation plan.

[0025] 4. The entire evaluation process relies on existing operational data from the wind farm, has a high degree of process standardization, can improve the efficiency of comparing and selecting technical upgrade schemes, weaken the impact of uncertain economic factors, enhance the reproducibility of technical solutions, and provide comprehensive and reliable support for wind farm upgrade decisions. Attached Figure Description

[0026] The present invention will be further described below with reference to the accompanying drawings and embodiments: Figure 1 This is a flowchart of the wind turbine retrofitting scheme evaluation method based on operational data according to the present invention; Figure 2 This is a flowchart of the data preprocessing and wind speed correction process of the present invention; Figure 3 This is a diagram illustrating the system structure of the multi-dimensional comprehensive evaluation index of this invention. Figure 4 This is a schematic diagram of the wind turbine retrofitting scheme evaluation device based on operational data according to the present invention. Figure 5 This is a wind speed map at a height of 80m for the target wind field area in this embodiment of the invention; Figure 6 This is a diagram showing the fan layout of the modification scheme in an embodiment of the present invention. Detailed Implementation

[0027] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0028] like Figure 1-4As shown, a method for evaluating wind turbine retrofitting schemes based on operational data includes the following steps: Step 1: Determine the initial wind turbine technical upgrade plan, collect turbine operation data for at least 5 full years since the target wind farm was connected to the grid, including SCADA system data, concurrent wind power prediction tower data and financial system data, and use the DBSCAN-segmented quartile combination algorithm to preprocess and clean the turbine operation data.

[0029] Determine the information on the retained, dismantled, and newly built units for the target wind farm in the initial renovation plan. Collect 10-minute average wind speed, wind direction, power data, fault record data, and operation and maintenance cost data from the SCADA system, wind speed and wind direction data from surrounding concurrent wind power prediction towers, and planned renovation costs, operating costs, other expenses, historical electricity data, and electricity price data from the financial system.

[0030] Data preprocessing and cleaning employed the DBSCAN-segmented quartile combination algorithm. The specific process was as follows: First, the wind speed-power data was clustered using the DBSCAN algorithm, dividing the data into clusters based on the density distribution of data points and removing outlier clusters. Then, intervals were divided according to equal wind speed intervals, and the upper and lower edge values ​​within each interval were calculated using the quartile method to further remove discrete outlier data. This ensured that surrounding wind power prediction towers and operating wind turbines had concurrent wind speed and direction data, and that wind frequencies with a frequency ratio exceeding 60% were not within the influence range of obstacles.

[0031] Step 2: Select a candidate set of corrected wind turbines. Based on the SCADA system data and the concurrent wind power prediction tower data, restore the free flow wind speed of the turbines through primary and secondary correction functions to obtain the corrected wind resource data of the turbines.

[0032] Taking into account factors such as different turbine models, turbine locations, and terrain type within the target wind farm, at least one turbine with more than five full years of operational data was selected to form a candidate turbine set for correction. For turbines of the same type, a primary correction formula was used. The wind speed was initially corrected, with the theoretical average wind speed obtained through power curve interpolation. Subsequently, the first-corrected wind speed sequence was divided into sectors by wind direction, valid sectors were selected, and a correlation test was performed (correlation ≥ 0.85). A second-correction formula was then applied. Complete the restoration of free-flow wind speed.

[0033] Step 3: Determine the wind resource base year based on the corrected unit wind resource data, calculate the wind speed fluctuation coefficient for each year, and establish a wind resource fluctuation coefficient model.

[0034] Calculate the annual average wind speed, Weibull distribution parameters, effective wind time rate, and other characteristic values ​​of the corrected wind resource data. Select the year in the past 5 years whose wind resource characteristic values ​​are close to the long-term average level as the base year, and then apply the formula... The wind speed fluctuation coefficient for each year is calculated to provide a basis for subsequent revenue breakdown. The effective wind time rate reflects the proportion of time during which wind resources are available, providing more comprehensive support for wind resource assessment.

[0035] Step 4: Collect environmental data of the target wind farm, construct a CFD fluid dynamics model by combining the corrected unit wind resource data, and separate the impact of wind resource fluctuations based on the CFD fluid dynamics model and wind speed fluctuation coefficient to obtain the net power generation revenue increment of the retrofit.

[0036] Environmental data such as wind field topography, roughness, and air density were collected to construct a CFD fluid dynamics model and verify its rationality. The baseline revenue difference before and after the retrofit was calculated based on the base year's wind resources. The actual revenue difference was calculated by combining the annual wind speed fluctuation coefficient and the additional revenue from the retrofit (calculated through the difference in power curves before and after the retrofit). Deviation analysis was used to separate the impact of wind resource fluctuations, yielding the net increase in power generation revenue from the retrofit. .

[0037] Step 5: Based on the financial system data and the increase in net power generation revenue from the renovation, the net revenue volatility coefficient is introduced to adjust the investment payback period calculation, and the benchmark investment payback period and the dynamically adjusted investment payback period are output.

[0038] According to the formula The benchmark payback period is calculated using the standard deviation σ, which is calculated based on net income data for at least five full years, using the formula... The investment payback period is dynamically adjusted, which more objectively reflects the stability of the renovation benefits.

[0039] Step 6: Establish a multi-dimensional comprehensive evaluation index system, and evaluate the feasibility of the renovation plan by combining the net power generation revenue increase, investment payback period and scores of each evaluation index.

[0040] The multi-dimensional comprehensive evaluation index system covers six categories of indicators: power curve (aging power curve deviation, power curve reachability), availability (time availability, capacity availability), reliability (average maintenance interval, maintenance performance factor, average repair time), wind resources (average wind speed, effective wind hourly rate), power loss (comprehensive site power consumption rate, transmission line loss rate, power loss), and economic efficiency (unit capacity operation and maintenance cost, cost per kilowatt-hour, generator group generation efficiency). A fuzzy comprehensive evaluation method is used to weight and score each indicator, and the feasibility of the retrofit plan is comprehensively determined by considering whether the investment payback period exceeds the remaining lifespan of the generator unit.

[0041] Accordingly, the present invention also provides a wind turbine retrofitting scheme evaluation device based on operational data, including a data acquisition module, a correction module, and an evaluation module: The data acquisition module is used to collect unit operation data and environmental data for at least five full years since the target wind farm was connected to the grid. The unit operation data includes SCADA system data, concurrent wind power prediction tower data, and financial system data. The environmental data includes wind farm area topography data, roughness data, and local air density. The DBSCAN-segmented quartile combination algorithm is used to preprocess and clean the unit operation data.

[0042] The correction module is used to select a candidate set of wind turbines for correction, restore the free-flow wind speed of the turbines through primary and secondary correction functions, and obtain the corrected wind resource data of the turbines; determine the wind resource base year and calculate the wind speed fluctuation coefficient for each year, and establish a wind resource fluctuation coefficient model. The correction module includes a primary correction unit, a secondary correction unit, and a base year calibration unit, which respectively perform the primary correction, secondary correction, base year calibration, and fluctuation coefficient calculation functions.

[0043] The evaluation module is used to construct a CFD fluid dynamics model, separate the impact of wind resource fluctuations to obtain the net power generation revenue increment from the retrofit; introduce a net revenue fluctuation coefficient to correct the payback period calculation, and output the baseline payback period and the dynamically corrected payback period; establish a multi-dimensional comprehensive evaluation index system to comprehensively assess the feasibility of the retrofit scheme. The evaluation module includes a model construction unit, a revenue calculation unit, a payback period correction unit, and a comprehensive evaluation unit, which respectively perform the functions of model construction, revenue calculation, payback period correction, and comprehensive evaluation.

[0044] Example 1: This embodiment is applied to an onshore wind power technology renovation project in Hubei Province. The project has constructed 13 wind turbine units with a single unit capacity of 2.625MW and 58 wind turbine units with a single unit capacity of 0.85MW, with a total installed capacity of 83.425MW. The turbines have been in operation for more than 10 years. The plan is to remove some of the turbine locations and implement a technology renovation scheme of replacing small turbines with larger ones at the original locations.

[0045] Step 1: Determine the initial renovation plan, identify the units to be retained, demolished, and newly built, and collect unit operation data from 2019 to 2023 (5 full years). This includes 10-minute average wind speed, wind direction, power data, fault record data, and operation and maintenance cost data from the SCADA system at the hub height of each unit, wind speed and wind direction data from surrounding wind power prediction towers, and planned renovation costs, operating costs, other costs, historical electricity data, and electricity price data from the financial system. Data preprocessing is performed using the DBSCAN-segmented quartile combination algorithm: first, the DBSCAN algorithm is used to remove abnormal data clusters caused by fault shutdowns and communication anomalies; then, the data is divided into 0.5 m / s wind speed intervals, and discrete outliers are removed within each interval using the quartile method to verify that the wind power prediction tower data meets the wind direction frequency requirements.

[0046] Step 2: Taking into account the wind farm turbine type, unit location, and terrain conditions, select turbine positions G125, G132, and G135 to form a candidate corrected turbine set. For turbines of the same type, calculate the corrected wind speed data using the first correction formula, where the theoretical average wind speed is obtained through power curve interpolation. Then, divide the wind into sectors according to wind direction, select the top 60% of effective sectors by wind direction frequency, and after correlation verification, restore the wind speed using the second correction formula to obtain the corrected turbine wind resource data.

[0047] Step 3: Based on the corrected wind resource data, calculate the annual average wind speed, Weibull distribution parameters of wind speed, and effective wind hour rate for 2019-2023. Select 2022, the year with wind resources closest to the long-term average level, as the base year, and apply the formula... Calculate the wind speed fluctuation coefficients for 2021 and 2023. The effective wind time rate for 2022 was 78%, reflecting stable wind resource conditions and sufficient available time in that year.

[0048] Step 4: Collect wind field topographic data, roughness data, and local air density. Construct a CFD fluid dynamics model in WT software, adjust parameters to ensure flow field simulation convergence, and obtain the wind speed map at 80m height for the wind field area as shown below. Figure 5 As shown, the difference in benchmark revenue before and after the renovation is calculated based on wind resources in 2022 (base year). Combining the wind speed fluctuation coefficient and additional revenue from the renovation in each year, the actual revenue difference for each year is calculated. The net increase in power generation revenue from the renovation is obtained through deviation analysis.

[0049] Step 5: Calculate the benchmark payback period based on financial system data. The standard deviation and multi-year average net income are calculated based on net income data from five complete years, using the formula... Get dynamically adjusted investment payback period Year.

[0050] Step 6: Establish a multi-dimensional comprehensive evaluation index system and calculate the scores of each index: After the aging power curve deviation is modified, it decreases by 35%, power curve accessibility is improved to 92%, time availability is improved to 96%, the average maintenance interval is extended to 2180 hours, the maintenance performance factor decreases to 28%, the average repair time is shortened to 8.5 hours, the comprehensive field power consumption rate decreases to 3.2%, the transmission line loss rate decreases to 2.8%, the unit capacity operation and maintenance cost decreases to 0.08 yuan / W, the cost per kilowatt-hour decreases to 0.32 yuan / kWh, and the group generation efficiency is improved to 89.6%. A fuzzy comprehensive evaluation method is used to weight the score, resulting in a comprehensive score of 8.7 points (out of 10). Comprehensive evaluation results: The dynamic correction investment payback period of 6.98 years does not exceed the unit's lifespan; the comprehensive score of the technical indicators is ≥8 points, indicating that the modification plan is feasible. The final modified plan's wind turbine layout diagram is as follows. Figure 6 As shown.

[0051] This embodiment successfully achieved a precise and comprehensive evaluation of the renovation plan through the evaluation method of the present invention, providing a scientific basis for project plan design and verifying the practicality and effectiveness of the present invention.

Claims

1. A method for evaluating wind turbine retrofitting schemes based on operational data, characterized in that, Includes the following steps: Step 1: Determine the initial wind turbine technical upgrade plan, collect turbine operation data for at least 5 full years since the target wind farm was connected to the grid. The turbine operation data includes SCADA system data, concurrent wind power prediction tower data and financial system data, and use the DBSCAN-segmented quartile combination algorithm to preprocess and clean the turbine operation data. Step 2: Select a candidate set of corrected wind turbines. Based on the SCADA system data and the concurrent wind power prediction tower data, restore the free flow wind speed of the turbines by using primary and secondary correction functions to obtain the corrected wind resource data of the turbines. Step 3: Determine the wind resource base year based on the corrected unit wind resource data, calculate the wind speed fluctuation coefficient for each year, and establish a wind resource fluctuation coefficient model. Step 4: Collect environmental data of the target wind farm, and construct a CFD fluid dynamics model by combining the corrected unit wind resource data. Based on the CFD fluid dynamics model and wind speed fluctuation coefficient, separate the impact of wind resource fluctuation to obtain the incremental net power generation revenue from the retrofit. Step 5: Based on the financial system data and the incremental net power generation revenue from the renovation, the net revenue volatility coefficient is introduced to correct the investment payback period calculation, and the benchmark investment payback period and the dynamically corrected investment payback period are output. Step 6: Establish a multi-dimensional comprehensive evaluation index system, which includes indicators such as power curve, availability, reliability, wind resources, power loss, and economic efficiency. Combine the net power generation revenue increase from the renovation, the investment payback period, and the scores of each evaluation index to assess the feasibility of the renovation plan.

2. The method for evaluating wind turbine retrofitting schemes based on operational data according to claim 1, characterized in that, The SCADA system data in Step 1 includes 10-minute average wind speed, wind direction, power data, fault record data, and operation and maintenance cost data; the financial system data includes planned renovation costs, operating costs, other costs, historical electricity data, and electricity price data.

3. The method for evaluating wind turbine retrofitting schemes based on operational data according to claim 1, characterized in that, The process of selecting candidate modified wind turbine sets in Step 2 is as follows: taking into account the different turbine models in the target wind farm, the relative positional relationship of each unit, the terrain level assessment results at the location of the unit, and the terrain conditions of the target wind farm, at least one wind turbine with more than 5 complete years of operating data is selected as the candidate modified wind turbine set. The candidate modified wind turbines include unit types with different single unit capacities and different hub heights.

4. The method for evaluating wind turbine retrofitting schemes based on operational data according to claim 1, characterized in that, The calculation formula for the first correction in Step 2 is as follows: ； in, The first correction factor is... The SCADA unit wind speed data before correction. This is the unit wind speed data after one correction. The average turbine wind speed of the target wind farm is given. The theoretical average wind speed of the target wind farm turbines; The theoretical average wind speed is obtained by dividing the wind speed into zones with a step size of 0.5 m / s, obtaining the average power curve over the entire annual operating cycle based on the SCADA system data of the target wind farm units, and interpolating the average power value of the units onto the average power curve to obtain the theoretical average wind speed.

5. The method for evaluating wind turbine retrofitting schemes based on operational data according to claim 1, characterized in that, The secondary correction process in Step 2 is as follows: a. Divide the corrected unit wind speed sequence into sectors according to wind direction, and select the sectors with the highest frequency of wind direction as the effective sectors; b. Perform a correlation test on the wind resource sequence of the unit after one correction and the wind resource sequence of the wind power prediction tower during the same period to ensure that the correlation reaches above 0.85; c. The formula for calculating the second correction is: ; in, This represents the unit wind speed data after secondary correction for the i-th sector. This refers to the unit wind speed data after one correction for the i-th sector. and These are the slope and intercept of the linear fit between the two sets of data in the i-th sector, respectively.

6. The method for evaluating wind turbine retrofitting schemes based on operational data according to claim 1, characterized in that, The process of determining the wind resource base year in Step 3 is as follows: Wind resource characteristic values ​​are calculated based on the corrected unit wind resource data. The wind resource characteristic values ​​include annual average wind speed, Weibull distribution parameters, and effective wind time rate. The year in the past 5 years in which the wind resource characteristic values ​​are close to the long-term average level is selected as the wind resource benchmark year. The formula for calculating the wind speed fluctuation coefficient is: ; in, Let be the wind speed fluctuation coefficient in year j. Let j be the average wind speed in year j. The base annual average wind speed.

7. The method for evaluating wind turbine retrofitting schemes based on operational data according to claim 1, characterized in that, In Step 4, the environmental data includes topographic data, roughness data, and local air density of the wind field area. When constructing the CFD fluid dynamics model, WT, Openwind, Windfarmer, or WASP software is used for wind field simulation, and the model parameters are adjusted to ensure the convergence of the flow field simulation. The process for obtaining the incremental net power generation revenue from the renovation is as follows: a. Calculate the difference in benchmark power generation revenue before and after the retrofit using a CFD fluid dynamics model, based on the benchmark year's wind resources. By analyzing the difference in power curves before and after the modification, the additional benefits brought by the modification are calculated separately. ; b. Calculate the difference in actual power generation revenue by combining the wind speed fluctuation coefficients for each year. ; c. Based on the actual power generation revenue difference over at least five full years, calculate the net revenue standard deviation σ, and separate the impact of wind resource fluctuations through deviation analysis to obtain the incremental net power generation revenue from the retrofit. .

8. The method for evaluating wind turbine retrofitting schemes based on operational data according to claim 7, characterized in that, The formula for calculating the dynamically adjusted payback period in Step 5 is as follows: ； in, To dynamically adjust the investment payback period, Let σ be the benchmark payback period and σ be the standard deviation of net income. This represents the average net income over several years. The formula for calculating the benchmark payback period is as follows: ； in, The net cash flow increment for period j consists of the increase in power generation efficiency before and after the renovation, the increase in operating costs, the increase in other expenses, and the planned renovation costs.

9. The method for evaluating wind turbine retrofitting schemes based on operational data according to claim 1, characterized in that, The multi-dimensional comprehensive evaluation index system in Step 6 includes: Power curve indicators: aging power curve deviation, power curve reachability; Availability metrics: Time availability, capacity availability; Reliability metrics: Mean time between repairs, maintenance performance factor, mean time to repair; Wind resource indicators: average wind speed, effective wind time rate; Power loss indicators: overall power consumption rate, transmission line loss rate, and power loss; Economic indicators: unit capacity operation and maintenance cost, cost per kilowatt-hour, and generator group generation efficiency.

10. The method for evaluating wind turbine retrofitting schemes based on operational data according to claim 9, characterized in that, The calculation formulas for each indicator in the aforementioned multi-dimensional comprehensive evaluation index system are as follows: Aging power curve offset: ； in This represents the actual power output of the unit when it was first put into operation at a wind speed of 1 m / s. The actual power of the unit within the statistical period when the wind speed is in m / s; Power curve reachability: ； in This represents the actual power generation of the generating unit within the statistical period. This represents the theoretical power generation of the generating unit within the statistical period. Time availability: ； in To calculate the duration of unit downtime due to faults during a statistical period, This represents the total duration of the statistical period. Capacity availability: ； in Unit capacity; Mean time between maintenance (MTBG): ； in This refers to the number of maintenance operations within the statistical period. Maintenance performance factor: ； in To calculate the total cost of unit failure repairs within the statistical period, This represents the total cost of unit operation and maintenance within the statistical period. Mean time to repair: ； in Let N be the failure duration of the i-th unit within the statistical period, and N be the number of units in the field. Effective wind time rate: ； in The duration of wind speed between the cut-in wind speed and the cut-out wind speed; Overall site power consumption rate: ； in This represents the actual power generation of the wind farm during the statistical period. To calculate the electricity purchased from the network within the statistical period, For the amount of electricity used in the internet during the statistical period; Outgoing line loss rate: ； in The power supply at the beginning of the transmission line. This refers to the power output at the end of the transmission line. Power loss: ； in Power loss due to fault To compensate for the power loss due to power rationing, To maintain the power loss, For other lost electricity; Unit capacity maintenance cost: ； in To calculate the total cost of wind farm operation and maintenance during the statistical period, This represents the total installed capacity within the facility. Cost per kilowatt-hour: ； in Total cost of power generation; Group power generation efficiency: ； in This is the unit's performance evaluation score.

11. The method for evaluating wind turbine retrofitting schemes based on operational data according to claim 1, characterized in that, The process of evaluating the feasibility of the renovation plan in Step 6 is as follows: the fuzzy comprehensive evaluation method is used to weight and score each indicator, and the results of net power generation revenue increase, benchmark investment payback period, and dynamically corrected investment payback period are combined. If the investment payback period does not exceed the remaining lifespan of the unit and the comprehensive score reaches the preset threshold, the renovation plan is deemed feasible.

12. An apparatus for evaluating wind turbine retrofit schemes based on operational data as described in any one of claims 1-11, characterized in that, It includes a data acquisition module, a correction module, and an evaluation module; The data acquisition module is used to collect unit operation data and environmental data for at least five full years since the target wind farm was connected to the grid. The unit operation data includes SCADA system data, concurrent wind power prediction tower data and financial system data. The environmental data includes wind farm area topography data, roughness data and local air density. The DBSCAN-segmented quartile combination algorithm is used to preprocess and clean the unit operation data. The correction module is used to select a candidate set of wind turbines for correction, and restore the free flow wind speed of the unit through primary and secondary correction functions to obtain the corrected unit wind resource data. Determine the base year for wind resources and calculate the wind speed fluctuation coefficient for each year, and establish a wind resource fluctuation coefficient model; The evaluation module is used to construct a CFD fluid dynamics model, separate the impact of wind resource fluctuations to obtain the net power generation revenue increment of the retrofit; introduce a net revenue fluctuation coefficient to correct the investment payback period calculation, and output the benchmark investment payback period and the dynamically corrected investment payback period; establish a multi-dimensional comprehensive evaluation index system to comprehensively evaluate the feasibility of the retrofit scheme.

13. The wind turbine retrofitting scheme evaluation device based on operational data according to claim 12, characterized in that, The preprocessing and cleaning process of the acquisition module includes: first, performing cluster analysis on the wind speed-power data using the DBSCAN algorithm to remove abnormal data clusters; then, dividing the data into intervals at equal intervals according to wind speed; and further removing discrete abnormal data within each interval using the quartile method.

14. The wind turbine retrofitting scheme evaluation device based on operational data according to claim 12, characterized in that, The correction module includes a primary correction unit, a secondary correction unit, and a reference year calibration unit; The primary correction unit is used to calculate the unit wind speed data after primary correction based on SCADA system data; The secondary correction unit is used to perform secondary sector-based correction on the wind speed data after the first correction, in conjunction with the data from the concurrent wind power prediction tower. The base year calibration unit is used to calculate wind resource characteristic values, determine the wind resource base year, and calculate the wind speed fluctuation coefficient for each year.

15. The wind turbine retrofitting scheme evaluation device based on operational data according to claim 12, characterized in that, The evaluation module includes a model building unit, a revenue calculation unit, a payback period correction unit, and a comprehensive evaluation unit; The model building unit is used to combine the corrected unit wind resource data and environmental data to build a CFD fluid dynamics model; The revenue calculation unit is used to separate the impact of wind resource fluctuations and calculate the incremental net power generation revenue from the renovation. The payback period correction unit is used to introduce the net income volatility coefficient and calculate the benchmark payback period and the dynamically corrected payback period. The comprehensive evaluation unit is used to establish a multi-dimensional comprehensive evaluation system that includes indicators such as power curve, availability, reliability, wind resources, power loss, and economic efficiency. It adopts the fuzzy comprehensive evaluation method to quantify the scoring and combines revenue data and payback period data to evaluate the feasibility of the renovation plan.

16. The wind turbine retrofitting scheme evaluation device based on operational data according to claim 12, characterized in that, The generator group power generation efficiency index in the multi-dimensional comprehensive evaluation index system is calculated based on the weighted average of the capacity of each unit and its power generation efficiency score, ensuring the fairness of cross-regional wind farm evaluation; the cost per kilowatt-hour and unit capacity operation and maintenance cost indexes are used to quantitatively evaluate the economic sustainability of the renovation plan.

Images