Optimization design method for wind turbine fairing

Abstract

The invention discloses a wind turbine fairing optimization design method, which relates to the technical field of fluid machinery and engineering, and comprises the following steps: a basic parameter acquisition and working condition defining step: collecting wind turbine model parameters, and measuring the wind condition of an operation area; the flow guide cover is divided into a windward section, a middle section and a wake flow section; a multi-objective optimization system construction step: determining aerodynamic resistance, structural strength, weight and wake flow influence optimization objectives; an optimization algorithm iteration solving step, wherein an improved non-dominated sorting genetic algorithm is adopted; a simulation verification and parameter correction step: selecting a scheme to perform CFD and finite element simulation; and a physical prototype manufacturing and performance testing step: manufacturing a prototype, and carrying out wind tunnel and field testing. The design efficiency of the air guide sleeve is improved, systematicness is optimized, pneumatic, structure and weight multi-target balance is achieved, cross-model universality is improved, and the problems that traditional design is dependent on experience and unbalanced in performance are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of fluid machinery and engineering, and in particular to a method for optimizing the design of a wind turbine fairing. BACKGROUND

[0002] As a key aerodynamic component of a wind turbine, the fairing directly affects the airflow characteristics, blade aerodynamic load distribution and overall energy conversion efficiency, and is crucial to the stable operation of the wind turbine under complex wind conditions. Current fairing design still has many technical bottlenecks, making it difficult to meet the needs of high-power wind turbines for high efficiency, reliability and lightweight.

[0003] Traditional fairing design relies heavily on the experience of engineers to carry out a cycle of iteration of "profile drawing - performance testing - parameter modification", which is inefficient and lacks systematicness. In the design process, a single circular arc or conic curve is often used to construct the profile, without fully considering the differences in airflow characteristics in different regions (upwind section, middle section and wake section) of the fairing, which leads to airflow separation in the middle section and complex wake vortex structure, increasing aerodynamic resistance and exacerbating blade aerodynamic load fluctuations. At the same time, the selection of design variables is often based on subjective judgment, without screening key parameters through sensitivity analysis, making the optimization direction blind and difficult to obtain a global optimal solution that takes into account aerodynamic performance and structural characteristics. This is common with the traditional design of Darrieus water turbine fairings, which relies on experience and is inefficient.

[0004] The lack of multi-objective collaborative optimization mechanism and the insufficient accuracy of simulation models further restrict the design quality. Existing designs often focus on minimizing aerodynamic resistance, ignoring key indicators such as structural strength, weight and wake effects, leading to stress concentration and excessive deformation of the optimized fairing under extreme wind speed conditions, or increasing the hub load due to excessive weight. In simulation analysis, the turbulence model often uses empirical parameters without modification based on actual wind condition data, resulting in large prediction errors in airflow transition location and wake velocity distribution, and unable to accurately guide the adjustment of design parameters. In addition, the material selection and layer design lack dynamic matching with structural loads, such as the use of equal angle arrangement for carbon fiber composite material layers, which fails to optimize the layer angle according to the load gradient in different regions of the fairing, resulting in waste of material performance or insufficient local strength.

[0005] The poor adaptability and cross-model versatility exacerbate the practical application problems. In the design process, the single rated wind speed condition is usually taken as the reference, and the influence of complex conditions such as turbulence, yaw and extreme wind speed is not fully considered, which leads to large performance fluctuations of the fairing in actual operation, especially in the turbulence condition, the influence range of the wake is expanded, and the operation of the downstream unit is affected. At the same time, for wind turbines of different power levels and hub sizes, a complete design process needs to be carried out again, and a standardized parameter scaling and adaptation method has not been formed, which results in long design cycle and high cost, and it is difficult to meet the needs of the development of wind turbine series. The superposition of these problems makes the traditional wind turbine fairing design present the status of "low efficiency, poor performance, poor precision and weak adaptation", which restricts the improvement of the performance of the whole wind turbine. SUMMARY

[0006] The wind turbine fairing optimization design method provided by the present application solves the problems mentioned in the prior art.

[0007] In order to achieve the above purpose, the present application adopts the following technical scheme: a wind turbine fairing optimization design method, comprising:

[0008] The basic parameter acquisition and condition definition step collects the target wind turbine model parameters, hub size, blade number and rated power; five typical working conditions of rated wind speed, extreme wind speed, turbulence, yaw and start are divided, the corresponding wind speed range, duration ratio and load characteristics are recorded, and a working condition parameter database is established;

[0009] The fairing multi-section parameterized modeling step uses a three-dimensional modeling software to construct a parameterized model, which is divided into a windward section, a middle section and a wake section; the windward section is a hemispherical transition structure, the middle section is a variable curvature revolving body, and the wake section is a converging conical surface; key design variables are defined, and in view of the uneven vertical distribution characteristics of wind speed in the wind turbine operation area, the whole outline of the fairing is designed in a non-central symmetric form: the radial dimensions of the windward section and the middle section are designed differently along the vertical direction according to the wind speed gradient, the influence of the uneven vertical distribution of wind speed on airflow is offset through the asymmetric structure, and the power fluctuation of the fan is reduced;

[0010] The multi-objective optimization system construction step constructs a two-layer system, the core target is that the net power generation under the equal diameter is increased by more than 5% than the reference, and the starting wind speed is less than or equal to 3.5 m / s; the secondary target includes a resistance coefficient less than or equal to 0.25 and a structural stress less than or equal to 80% of the yield strength; the weights are allocated according to the condition ratio and sensitivity, the rated wind speed condition focuses on power, and the starting condition focuses on starting wind speed; the analytic hierarchy process is used to build a hierarchical structure, the consistency test (CR<0.1) is carried out, the comprehensive weight is obtained, and the optimization function is constructed by using the weighted normalization method (w i is the comprehensive weight, f i (x) is the target normalized function), and the structure deformation, power constraints and process constraints are clear;

[0011] The optimization algorithm iteration solving step adopts an improved non-dominated sorting genetic algorithm for iteration optimization; a model is constructed and a target value is simulated in each generation, and the optimal solution set is screened through fast non-dominated sorting, and the Pareto optimal solution set is output after iteration convergence;

[0012] The simulation verification and parameter correction step selects three schemes from the optimal solution set for verification, for example, aerodynamic simulation uses CFD software and a k-ω SST model, a structured grid, and calculation of a drag coefficient; structural simulation uses finite element software.

[0013] Further, the method further comprises a fairing profile curve bionic optimization step, a composite curve equation is used to construct a middle section variable-curvature profile, the curve equation is y=kx 3 +mx 2 +nx+p+qsin(rx); wherein y is a profile radial coordinate, x is an axial coordinate, k is a cubic term coefficient, m is a quadratic term coefficient, n is a linear term coefficient, p is a constant term, q is a sine correction coefficient, and r is a correction frequency; the profile curve fitted through the equation forms a nodular micro-convex structure in the middle section, and a rough band or a turbulent flow control band is arranged in the middle section airflow separation area; the rough band adopts a silicon carbide particle convex structure, and the turbulent flow control band adopts a metal lattice grid structure, the boundary layer airflow is disturbed through the rough band / turbulent flow control band, and the airflow separation starting point is further delayed to move backward; and the nodular micro-convex structure and the rough band / turbulent flow control band synergistically induce the formation of stable attached vortex flow around the fairing, enhance the convergence ability of the airflow, and improve the wind energy density per unit area; the airflow separation starting point is moved backward, the airflow transition in the middle section is smooth, the local pressure loss is reduced, the buffer capacity for turbulent flow is enhanced, the surface pressure fluctuation amplitude of the fairing is reduced under turbulent flow conditions, and the convergence ability of the fairing is improved through the stable attached vortex flow, so that the wind speed at the inlet area of the fairing is increased by 8%-12% compared with the conventional design.

[0014] Further, the method further comprises a composite material layering gradient optimization step, a gradient layering optimization equation is used to determine the layering angle of each layer ; wherein θ is the layering angle of the i-th layer, i is the layering sequence number, α is a basic angle gradient coefficient, β is an angle sine adjustment amplitude, γ is an angle sine change coefficient, and δ is an angle root adjustment coefficient; through the equation, the layering angle of the windward section is increased in a gradient manner as the sequence number increases, and the layering angle of the wake flow section is flexibly transitioned in combination with the root term; at the same time, a variable-thickness layering process is introduced in the layering process.

[0015] Further, in the simulation verification and parameter correction step, the aerodynamic performance simulation adopts CFD analysis coupled with a transition model, and the measured turbulent intensity data is used to correct the turbulent dissipation rate through a turbulent dissipation rate correction equation Dynamic adjustment model parameters; wherein, ω is the corrected turbulent dissipation rate, δ is the turbulent base correction factor, v is the turbulent fluctuation velocity, u is the main flow velocity, ε is the initial dissipation rate, ζ is the velocity gradient correction coefficient, is the wall surface normal velocity gradient; the correction makes the average error of the wake flow velocity distribution simulation result and the wind tunnel test decrease; in addition, the transition starting criterion of the transition model needs to be corrected, and the transition determination coefficient based on the local turbulent intensity and the pressure gradient is introduced.

[0016] Further, in the optimization algorithm iterative solving step, an improved non-dominated sorting genetic algorithm with multi-strategy fusion is adopted, specifically including: initialization stage; in the crossover operation, arithmetic crossover is mainly used in the first 40 generations, and an "elite reservation + neighborhood guidance" mechanism is introduced; single-point crossover and simulated binary crossover are combined in the last 40 generations; adaptive mutation rate is used in the mutation operation; when iteration reaches the 40th generation, local search based on the response surface method is started, a quadratic response surface model is constructed in the neighborhood of the design variable, and the local optimal solution is solved and replaces the original individual; through these strategies, the convergence accuracy of the final Pareto solution set is improved, and the distribution of the solution set on the three core objectives of aerodynamic drag, structural strength and weight is more uniform.

[0017] Further, in the physical prototype manufacturing and performance testing step, the fatigue performance test adopts a combined method of "graded loading + acoustic emission monitoring": the graded loading is based on the wind turbine load spectrum; an acoustic emission monitoring system is arranged with 8 sensors to capture acoustic emission signals generated by internal micro-cracks in real time; the damage position is located by ultrasonic C scanning, the damage area parameters are adjusted by returning to the parameterized modeling step, and re-iterative optimization is performed.

[0018] Further, in the basic parameter acquisition and working condition definition step, the wind speed distribution measurement adopts a combined method of "laser Doppler anemometer + unmanned aerial vehicle aerial survey": the laser Doppler anemometer is arranged at the hub height, and measurement points are arranged in multiple directions to obtain the near-ground wind speed profile; the unmanned aerial vehicle carries a weather radar to conduct aerial survey above the wind turbine to obtain high-altitude wind speed, turbulence intensity and wind shear index; the ground and high-altitude data are combined to fit the full-height wind speed distribution using the logarithmic law wind profile model.

[0019] Further, it further includes a guide vane aerodynamic noise active control step, in which noise prediction of a micro-perforated plate sound absorption structure is introduced in the aerodynamic performance simulation, and the aerodynamic noise sound pressure level is corrected by an equation The calculation is: wherein, L is the sound pressure level of aerodynamic noise, mu is the airflow noise coefficient, u is the airflow velocity, d is the fairing characteristic size, phi is the noise radiation coefficient, eta is the sound absorption structure correction coefficient, delta p is the pressure difference before and after the micro-perforated plate, f is the noise frequency, rho is the air density, and c is the sound velocity; when the simulation predicted noise sound pressure level exceeds 65dB, a micro-perforated plate sound absorption layer is additionally arranged on the inner wall of the fairing wake flow section, and the wake flow section contraction angle is optimized to reduce the airflow separation noise; in addition, the installation angle of the micro-perforated plate needs to be optimized to be inclined to the inner wall of the wake flow section.

[0020] Further, in the multi-section parameterized modeling step of the fairing, the connection of each section adopts double constraint design of "curvature continuity + stress guide": curvature continuity constraint is realized through B-spline curve interpolation; stress guide constraint is realized by pre-simulating the stress distribution in the three-dimensional model of the connection area, adjusting the connection arc radius and transition length, and making the stress flow lines smoothly transition along the connection surface; for the connection requirement of non-central symmetric contour, in the area where the vertical wind speed difference is significant (such as within the range of hub height ± 5m), the connection arc radius is additionally optimized (increased by 10%-15% compared with the symmetric design), so that the stress concentration coefficient at the connection under the non-symmetric structure is still controlled within 1.2; at the same time, a "ring stiffener + variable wall thickness" structure is designed around the blade mounting hole, the edge of the mounting hole is rounded, the wall thickness around the hole gradually increases, and the maintenance window adopts "outward flange + sealing rubber strip" design.

[0021] Further, it further includes a cross-model adaptive design verification step, which adopts a "parameter scaling + local adaptation" strategy for different types of wind turbines: the hub diameter ratio lambda of the target model to the reference model is the scaling coefficient, and the scaling formula is P target = lambda * P base *(1+xi*(lambda-1)); the local adaptation focuses on adjusting the blade mounting hole position, maintenance window size and layup scheme for the scaled parameters; cross-model test selects three different types of wind turbines for adaptive test, and the test indicators include aerodynamic drag coefficient deviation, structural strength margin, installation compatibility and power generation efficiency change.

[0022] Compared with the prior art, the beneficial effects of the present application are:

[0023] The design efficiency and optimization systematization are significantly improved. The basic parameter acquisition and working condition definition step constructs a comprehensive working condition database to provide accurate data input for design; the multi-section parameterized modeling divides the fairing into functional areas, clearly defines the key design variables and value range, and breaks away from experience dependence. The optimization algorithm iteration solving step adopts an improved non-dominated sorting genetic algorithm, combined with Latin hypercube sampling, multi-strategy crossover and variation mechanisms, to greatly improve the convergence accuracy and efficiency of the Pareto optimal solution set, avoid blind iteration, shorten the period of traditional experience design, and obtain the global optimal solution.

[0024] Multi-objective performance balance and structural reliability comprehensive optimization. The multi-objective optimization system construction step establishes multi-dimensional targets of aerodynamics, structure, weight and wake, and realizes precise balance of core performance under different working conditions by combining with dynamic weight distribution. The composite material layer gradient optimization realizes precise matching of layer angle and load gradient through individualized layer equation and variable thickness process, improves structural strength and fatigue life while realizing lightweight, and solves the problems of single performance and material waste in traditional design. The simulation verification and parameter correction step uses CFD analysis and turbulent flow parameter dynamic correction of coupling transition model to significantly improve the prediction accuracy of flow field, provide accurate guidance for parameter optimization, and ensure the stability of the design scheme in actual operation.

[0025] Working condition adaptability and running stability are greatly enhanced. The basic parameter acquisition adopts the joint method of "ground measurement + high-altitude aerial survey", which can accurately capture the wind characteristics at all altitudes, and can make the fairing maintain excellent performance under the conditions of rated wind speed, turbulence and extreme wind speed by combining with multi-working condition weight distribution. The aerodynamic noise active control step effectively reduces the operating noise and improves the environmental adaptability through noise prediction and sound absorption structure optimization. The fatigue performance test uses the "graded loading + acoustic emission monitoring" method to identify weak points and optimize them in advance, which significantly improves the reliability and life of the fairing in long-term operation.

[0026] Cross-model generality and design expandability are significantly improved. The cross-model adaptive design verification step establishes a standardized parameter scaling and local adaptation method, which realizes the rapid adaptation of fairings for different power level wind turbines based on the hub diameter ratio, avoids repeated design, and forms a series of solutions. At the same time, a cross-model design database is established to provide a reference for subsequent new model design, which greatly shortens the research and development cycle, reduces the design cost, and perfectly adapts to the series development needs of wind turbines.

[0027] Overall, the present application realizes the efficiency, accuracy and standardization of wind turbine fairing design through the integration of data-driven parameterized modeling, intelligent multi-objective optimization, accurate simulation verification and cross-model adaptation. It not only improves the aerodynamic performance, structural reliability and lightweight level of the fairing, but also enhances the working condition adaptability and cross-model generality, providing a key guarantee for the energy conversion efficiency improvement and stable operation of the whole wind turbine, and has significant technical value and application prospect. BRIEF DESCRIPTION OF DRAWINGS

[0028] Figure 1 A schematic block diagram of a wind turbine fairing optimization design method proposed by the present application is shown in the figure;

[0029] Figure 2 A relationship diagram of key design variables and aerodynamic drag coefficients for each section of the fairing is shown in the figure;

[0030] Figure 3 Pareto optimal solution set diagram of the combination of the fairwater design variable (x m , L D ) group;

[0031] Figure 4 The relationship diagram of the composite material layer number and the structure stress and the layer angle. DETAILED DESCRIPTION

[0032] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0033] In the description of the present application, it should be understood that the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the purpose of facilitating the description of the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation of the present application.

[0034] In addition, the terms "first", "second" are only for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "multiple" is two or more, unless otherwise explicitly and specifically limited. In addition, the terms "mounting", "connecting", "connection" should be broadly understood, for example, it can be fixedly connected, or it can be detachably connected, or integrally connected; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium, or it can be the communication between two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances, and the present application will be further described in detail with reference to the drawings.

[0035] With reference to Figures 1 to 4 : A fairwater optimization design method of a wind turbine, comprising:

[0036] The basic parameter acquisition and working condition definition step collects the model parameters, hub size, number of blades and rated power of the target wind turbine, determines the fairing installation space boundary conditions, measures the wind speed distribution, turbulence intensity and wind direction change frequency of the wind turbine operating area, divides five typical working conditions, including rated wind speed working condition, extreme wind speed working condition, turbulence working condition, yaw working condition and starting working condition, records the wind speed range, duration ratio and load characteristics of each working condition, establishes a working condition parameter database, the data acquisition period is not less than 30 days, and the sampling frequency is set to 1 Hz, ensuring the comprehensiveness of working condition coverage and the accuracy of data.

[0037] The fairing multi-section parameterized modeling step uses a three-dimensional modeling software to construct a fairing parameterized model, divides the fairing into three functional areas of a windward section, a middle section and a wake section, the windward section adopts a hemispherical transition structure, the middle section is a variable curvature body of revolution structure, and the wake section is a shrinkage conical surface structure, defines 12 key design variables including the curvature radius of the windward section, the maximum outer diameter of the middle section, the length of the middle section, the shrinkage angle of the wake section, the wall thickness distribution and the arc radius of the connection of each section, sets the variable value range, wherein the curvature radius of the windward section is 1.2-1.5 times the hub radius, the maximum outer diameter of the middle section is not more than 2.0 times the hub outer diameter, the shrinkage angle of the wake section is controlled between 15°-25°, and the arc radius of the connection of each section is not less than 50 mm, ensuring the adjustability and structural rationality of the model parameters, and in view of the vertical wind speed distribution unevenness of the wind turbine operating area (based on the full height wind speed distribution data obtained by the “laser Doppler anemometer + unmanned aerial vehicle aerial survey” in the basic parameter acquisition step), the non-central symmetric form is adopted to design the overall outline of the fairing: wherein the radial dimensions of the windward section and the middle section are designed along the vertical according to the wind speed gradient difference (the radial dimension in the area with high wind speed is 5%-8% larger than the reference value, and the radial dimension in the area with low wind speed is 3%-5% smaller than the reference value), the influence of the vertical wind speed distribution unevenness on the airflow flow is offset through the asymmetric structure, and the fan power fluctuation is reduced.

[0038] The multi-objective optimization system construction step constructs a two-layer system, the core target is to improve the net power generation by ≥5% under the same diameter, and the starting wind speed is ≤3.5 m / s; the secondary target includes resistance coefficient ≤0.25 and structural stress ≤80% of yield strength; the weights are allocated according to the working condition proportion and sensitivity, the rated wind speed working condition focuses on power, and the starting working condition focuses on starting wind speed; the analytic hierarchy process is used to build a hierarchical structure, the consistency test (CR<0.1) is passed, the comprehensive weight is obtained, and the optimization function is constructed by the weighted normalization method (is the comprehensive weight, and is the target normalized function), and the structure deformation, power and process constraints are clear.

[0039] The optimization algorithm iteration solving step adopts the improved non-dominated sorting genetic algorithm for iteration optimization, the population size is initialized to 120, the chromosome coding adopts the real number coding mode corresponding to 12 design variables, the crossover probability is set to 0.85, the mutation probability is set to 0.05, the iteration number is set to 80 generations, in the iteration process of each generation, first, the fairing model is constructed based on the current population parameters, the aerodynamic and structural performance simulation is carried out, each objective function value is calculated, then the optimal solution set is selected through the fast non-dominated sorting, the solution set diversity is maintained by using the crowding distance operator, the local search strategy is introduced when the iteration reaches the 40th generation, fine optimization is carried out on the neighborhood of the optimal individual, until the iteration converges, and the Pareto optimal solution set is output.

[0040] The simulation verification and parameter correction step selects three representative schemes from the Pareto optimal solution set for detailed simulation verification, the aerodynamic performance simulation adopts the CFD software, selects the k-omega SST turbulence model, divides the structured grid, the grid number is controlled between 2 million and 3 million, the boundary condition is set to the inlet wind speed according to the typical working condition distribution, the outlet is the pressure outlet, the wall surface adopts the no-slip condition, the drag coefficient, the flow field velocity distribution and the wake vortex structure of each scheme are calculated; the structural performance simulation adopts the finite element software, applies the wind load, the centrifugal load and the gravity load, analyzes the structural stress distribution, the deformation condition and the natural frequency, compares the deviation between the simulation results and the optimization target, and fine tunes and corrects the design variables, and the correction amplitude is not more than 5% of the initial value.

[0041] In the application, the fairing contour curve bionic optimization step is also included, in the parameterized modeling process, the control mechanism of the fin limb leading edge nodule structure of the beluga whale on the airflow is referred to, the composite curve equation is adopted to construct the middle segment variable curvature contour, the curve equation is y=kx 3 +mx 2+ nx + p + qsin(rx). Wherein, y is the profile radial coordinate, x is the axial coordinate, k is the cubic term coefficient, m is the quadratic term coefficient, n is the linear term coefficient, p is the constant term, q is the sine correction coefficient, and r is the correction frequency. The profile curve fitted by the equation forms a nodular micro-convex structure in the middle section, and a rough band or a turbulent flow control band is arranged in the middle section airflow separation area (1 / 3-1 / 2 middle section length away from the windward section vertex). The rough band adopts a silicon carbide particle protruding structure (protruding height 0.5-1.0mm, particle spacing 5-10mm), and the turbulent flow control band adopts a metal hollow grid structure (grid aperture 2-3mm, grid thickness 1-1.5mm). The rough band / turbulent flow control band disturbs the boundary layer airflow, further delays the airflow separation starting point, and the nodular micro-convex structure cooperates with the rough band / turbulent flow control band to induce a stable attached vortex flow around the fairing, enhances the convergence ability of the airflow, and improves the wind energy density per unit area. The airflow separation starting point is moved back by more than 30%, the middle section airflow transition is smooth by 15%, the local pressure loss is reduced by 22%, and the turbulent flow buffering capacity is enhanced, and the fairing surface pressure fluctuation amplitude is reduced by 18% under turbulent flow conditions. In addition, a sensitivity analysis is also needed for the size of the bionic structure, the nodal spacing is selected to be 1 / 8-1 / 6 of the middle section circumference of the fairing, and the nodal height is selected to be 1%-3% of the maximum outer diameter of the middle section, so that the airflow control effect can be achieved under different wind speeds without increasing additional aerodynamic resistance, and the convergence ability is improved by the stable attached vortex flow, so that the wind speed at the fairing inlet area is increased by 8%-12% compared with the traditional design.

[0042] In the present application, a composite material layer gradient optimization step is also included, before the prototype is manufactured, based on the structure simulation results and the material mechanics performance, a gradient layer optimization equation is used to determine the layer angle of each layer Wherein, θ is the i-th layer angle, i is the layer number, α is the basic angle gradient coefficient, β is the angle sine adjustment amplitude, γ is the angle sine change coefficient, and δ is the angle root adjustment coefficient. Through the equation, the windward section layer angle increases with the sequence number to cope with high impact load, and the wake section layer angle realizes flexible transition combined with the root term to adapt to the lightweight demand under low load, and the overall structure fatigue resistance is improved by 25%, and the weight is reduced by 12% compared with the equal angle layer scheme. At the same time, a variable thickness layer process is introduced in the layering process, the windward section layer thickness is linearly thickened from 2mm to 4mm, and the wake section layer thickness is linearly thinned from 3mm to 1.5mm, further optimizing the structure stiffness distribution, under the extreme wind speed condition, the maximum stress concentration coefficient is reduced from 1.8 to 1.3, avoiding local premature failure.

[0043] In the simulation verification and parameter correction step of the application, the CFD analysis of the coupling transition model is adopted for aerodynamic performance simulation, combined with the measured turbulent intensity data, and the turbulent dissipation rate correction equation is used to correct the turbulent dissipation rate Dynamic adjustment of model parameters, wherein ω is the corrected turbulent dissipation rate, δ is the turbulent basic correction factor, v is the turbulent fluctuation velocity, u is the main flow velocity, ε is the initial dissipation rate, ζ is the velocity gradient correction coefficient, The wall normal velocity gradient is the correction, which reduces the average error of the flow velocity distribution simulation result in the wake flow area from 12% to below 4.5%, especially in the yawing condition, the flow velocity recovery degree prediction accuracy at 5 times the hub diameter downstream of the fairing is improved to 92%. In addition, the transition start criterion of the transition model needs to be corrected, and the transition judgment coefficient based on the local turbulent intensity and the pressure gradient is introduced, so that the prediction error of the transition position is controlled within 5%, which provides support for accurate evaluation of the wake influence range.

[0044] In the application, in the optimization algorithm iteration solving step, the improved non-dominated sorting genetic algorithm (NSGA-III) with multi-strategy fusion is adopted, which specifically includes: in the initialization stage, 120 initial populations are generated based on Latin hypercube sampling to ensure that the design variables are uniformly distributed in the value range; in the crossover operation, the arithmetic crossover is mainly used in the first 40 generations, and the crossover probability is linearly increased from 0.85 to 0.95, and the "elite reservation + neighborhood guidance" mechanism is introduced, 5 new individuals are generated within 0.1 times the variable range around the top 10 individuals of each generation; in the last 40 generations, single-point crossover and simulated binary crossover are combined, and the crossover probability is stabilized at 0.85; the adaptive mutation rate is used in the mutation operation, when the population similarity exceeds 0.7, the mutation rate is increased from 0.05 to 0.15; when iteration reaches the 40th generation, the local search based on the response surface method is started, the quadratic response surface model is constructed in the neighborhood of the design variables of the current top 20 individuals, the local optimal solution is solved and the original individual is replaced. Through these strategies, the convergence accuracy of the final Pareto solution set is improved by 30%, the iteration efficiency is improved by 25%, and the distribution of the solution set on the three core objectives of aerodynamic resistance, structural strength and weight is more uniform, providing more high-quality options for scheme selection.

[0045] In the application, in the physical prototype manufacturing and performance testing step, the fatigue performance test adopts the combined method of "graded loading + acoustic emission monitoring": the graded loading divides the fatigue load into 5 levels according to the wind turbine load spectrum, each level is cycled for 10 5Secondly, the loading frequency is consistent with the operating frequency of the wind turbine during the cycle process; the acoustic emission monitoring system is arranged with 8 sensors located at the vertex of the windward section, the middle section on both sides and the junction of the wake section, the sampling frequency is 2MHz, and the acoustic emission signals generated by the micro-cracks in the material are captured in real time; if the vibration amplitude increases by 20% or the acoustic emission signal is continuously abnormal during the test, the test is immediately stopped, the damage position is located by ultrasonic C scanning, the parameterized modeling step is returned to adjust the damage area parameters, and the iteration optimization is restarted. The method can identify the fatigue weak point in advance, so that the fatigue life safety factor of the finally designed fairing under the equivalent 10-year operating load is ≥1.5.

[0046] In the application, in the step of obtaining basic parameters and defining working conditions, the wind speed distribution measurement adopts a combined mode of "laser Doppler anemometer + unmanned aerial vehicle aerial survey": the laser Doppler anemometer is arranged at the hub height, and measurement points are arranged in multiple directions, each point is continuously measured for 30 minutes, the sampling frequency is 1Hz, and the near-ground wind speed profile is obtained; the unmanned aerial vehicle carries a weather radar and performs aerial survey at a height of 100m above the wind turbine to obtain high-altitude wind speed, turbulence intensity and wind shear index; the wind speed distribution at all altitudes is fitted by using a logarithmic law wind profile model in combination with the ground and high-altitude data, and the calculation accuracy of the wind shear index is improved to ±0.02. The method provides more accurate airflow incidence angle basis for the design of the fairing under the yawing working condition, so that the aerodynamic drag prediction error of the fairing under the yawing working condition is reduced from 15% to within 7%. At the same time, the wind speed data of different seasons and different weather types need to be classified and counted, and a seasonal correction model of wind speed distribution is established to further improve the accuracy of the working condition definition.

[0047] In the application, the step of actively controlling the aerodynamic noise of the fairing further includes introducing noise prediction of a micro-perforated plate sound absorption structure in the aerodynamic performance simulation, and calculating the aerodynamic noise sound pressure level L by a correction equation . Wherein, L is the aerodynamic noise sound pressure level, μ is the airflow noise coefficient, u is the airflow velocity, d is the characteristic size of the fairing, φ is the noise radiation coefficient, η is the sound absorption structure correction coefficient, Δp is the pressure difference before and after the micro-perforated plate, f is the noise frequency, ρ is the air density, and c is the sound speed. When the simulation predicted noise sound pressure level exceeds 65dB, a micro-perforated plate sound absorption layer is additionally arranged on the inner wall of the wake section of the fairing, so that the broadband noise attenuation amount reaches 8-12dB, and the contraction angle of the wake section is optimized to reduce the airflow separation noise, and finally the operating noise of the fairing is controlled below 60dB (10m away from the unit).

[0048] In the step of multi-section parametric modeling of the invention, the connection of each section adopts the design of "curvature continuity + stress guide": the curvature continuity constraint ensures the continuity of the first and second derivatives of the profile curve of each section by B-spline curve interpolation, and the curvature change rate of the connection area is less than or equal to 0.01 mm -1 , so that the pressure gradient change rate of the airflow at the connection is less than or equal to 3 Pa / mm; the stress guide constraint pre-simulates the stress distribution in the three-dimensional model of the connection area, adjusts the radius of the connection arc and the transition length, makes the stress flow lines smoothly transition along the connection surface, and controls the stress concentration factor at the connection within 1.2; for the connection requirements of non-central symmetric profiles, in the area where the vertical wind speed difference is significant (such as within the range of ± 5 m from the hub height), the radius of the connection arc is additionally optimized (increased by 10%-15% compared to the symmetric design), to ensure that the stress concentration factor at the connection under the non-symmetric structure is still controlled within 1.2; at the same time, the "annular reinforcing rib + variable wall thickness" structure is designed around the blade mounting hole, the edge of the mounting hole is rounded, the wall thickness around the hole gradually increases, and the maintenance window adopts the "outward flange + sealing rubber strip" design, which not only ensures the convenience of installation and maintenance, but also makes the structural strength margin of the mounting hole and the maintenance window area ≥1.5 times, without permanent deformation under extreme wind speed conditions.

[0049] In the invention, the step of cross-model adaptive design verification is also included, which adopts the "parameter scaling + local adaptation" strategy for different types of wind turbines (hub diameter range 60-120 m): parameter scaling takes the hub diameter ratio λ of the target model and the reference model as the scaling coefficient, and linearly scales the key parameters of the fairing, the scaling formula is P target = λ × P base × (1 + ξ × (λ - 1)); local adaptation focuses on adjusting the blade mounting hole position, maintenance window size and layup scheme for the scaled parameters; cross-model testing selects three different types of wind turbines for adaptive testing, and the test indicators include aerodynamic drag coefficient deviation, structural strength margin, installation compatibility and power generation efficiency change, and all indicators meet the requirements to determine that the adaptation is qualified. This step forms a series of fairing design schemes covering 1.5-6 MW power levels, which improves the universality of the method to more than 90%, and shortens the development cycle of the fairing of new models by 40%. At the same time, a cross-model design database needs to be established to record the adaptive parameters and test results of different models, providing a reference for subsequent new model design.

[0050] The specific implementation of the system is further illustrated by two embodiments as follows:

[0051] Example 1: 1.5 MW onshore wind turbine fairing optimization design (Hebei Zhangbei wind farm application scenario)

[0052] This embodiment is aimed at the hub fairing upgrade demand of 1.5MW double-fed wind turbine (rotor diameter 77m, hub diameter 3m, blade number 3, rated power 1500kW) in Zhangbei wind farm of Hebei province. The annual average wind speed in this area is 6.2m / s, the turbulence intensity is A class, the yawing working condition accounts for 18%, and the traditional hub fairing has large aerodynamic resistance due to the single circular arc type line, which leads to airflow turbulence in the hub area and increases the energy loss of the whole machine. The scheme of the application realizes precise optimization, and the specific implementation process is as follows:

[0053] 1. Basic parameter acquisition and working condition definition

[0054] 1.1 Core parameter collection Clearly distinguish the rotor diameter and the hub diameter (avoid confusing the reference datum):

[0055] Wind turbine parameters: rotor diameter 77m (swept area 4657m 2 ), hub diameter 3m (radius 1.5m), blade root chord length 3.8m, rated wind speed 12m / s, limit wind speed 25m / s, hub installation height 80m;

[0056] Material parameters: the traditional fairing adopts glass fiber composite material (yield strength 220MPa), and the scheme selects T700 carbon fiber composite material (yield strength 300MPa).

[0057] 1.2 Wind condition measurement and working condition division

[0058] Joint measurement of "laser Doppler anemometer + unmanned aerial vehicle aerial survey": laser Doppler anemometer: arranged at the hub height (80m), 3 measurement points (spacing 1m, to avoid blind area) are arranged in the wind direction and the vertical wind direction, continuous measurement for 30 days, sampling frequency 1Hz, and the near-surface wind speed profile is obtained;

[0059] Unmanned aerial vehicle aerial survey: equipped with weather radar to measure at a height of 100m to obtain high-altitude wind speed, turbulence intensity (0.12 measured) and wind shear index (0.12);

[0060] Working condition database (5 typical working conditions):

[0061] Rated wind speed working condition: 10-14m / s, accounting for 35%, aerodynamic load is stable, and the core requirement is to reduce resistance;

[0062] Limit wind speed working condition: 20-25m / s, accounting for 2%, wind load peak value is high, and the core requirement is structural safety; Turbulence working condition:

[0063] Turbulence intensity >0.16, accounting for 20%, load fluctuation is large, and the core requirement is to suppress tail flow turbulence;

[0064] Yaw condition: Yaw angle 15°-30°, accounting for 18%, air inflow angle changes, the core requirement is to balance drag and structural stress;

[0065] Start-up conditions: 3-5m / s, accounting for 25%. At low wind speeds, it is necessary to reduce start-up resistance, and the core requirement is smooth airflow.

[0066] 2. Multi-segment parametric modeling

[0067] A parametric model was built using SolidWorks, clearly defined as a hub fairing (only covering the hub, not the rotor), divided into 3 functional segments, and 12 key design variables were defined (the value range conforms to the engineering conventions for small-diameter fairings):

[0068] 2.1 Structure and Design Variables of Each Section Windward Section: Hemispherical transition structure, with the design variable being the radius of curvature R1 (2.0-2.5m, 1.3-1.7 times the hub radius) to ensure close contact with the hub and easy airflow adhesion; Middle Section: Variable curvature rotating body, with the design variables being the maximum outer diameter D (4.0-4.5m, ≤1.5 times the hub diameter, to avoid exceeding the rotor sweep area) and the middle section length L2 (3.5-4.0m); Wake Section: Contracting conical surface, with the design variables being the contraction angle θ (15°-25°) and the connecting arc radius R (20-30mm, to avoid stress concentration).

[0069] 2.2 Mid-section biomimetic profile optimization: A composite curve equation is used to construct the mid-section variable curvature profile (to adapt to small diameter sizes and avoid excessive bulging that increases drag): y = kx 3 +mx 2 +nx+p+qsin(rx) Variable definition: y is the radial coordinate of the profile (m), x is the axial coordinate (m); Coefficient values: k = 0.005, m = -0.08, n = 0.6, p = 1.5 (hub radius, ensuring coaxiality with the hub), q = 0.05, r = 0.8; Bionic structure: forming nodular micro-convexity (spacing 1.2m, about 1 / 8 of the circumference of the middle section; height 0.1m, 2% of the maximum outer diameter of the middle section), delaying airflow separation without increasing additional drag.

[0070] 3. Multi-objective optimization and algorithm iteration 3.1 Optimization objectives and weights (aligned with actual engineering priorities) With "reducing energy loss, ensuring structural reliability, and achieving lightweighting" as the core, four main objectives are established (weights are allocated based on the frequency and impact of operating conditions):

[0071] Minimize the aerodynamic drag coefficient (Cd): Cd≤0.3 under rated wind speed conditions, with a weight of 0.4 (directly affecting power generation efficiency);

[0072] Structural maximum stress (σ) minimization: σ≤280 MPa (93% of T700 yield strength) at extreme wind speed, weight 0.5 (safety priority);

[0073] Weight (W) minimization: total weight ≤0.3 t (0.31 t for traditional solution), weight 0.08 (reduce hub load);

[0074] Wake influence radius (Rw) minimization: Rw≤30 m at turbulent flow, weight 0.02 (reduce flow disturbance to blade root).

[0075] 3.2 Algorithm iteration parameters Improved NSGA-III algorithm (ensure convergence and diversity of solution set) is adopted:

[0076] Population initialization: size 120, real number coding corresponding to 12 design variables;

[0077] Crossover operation: arithmetic crossover is mainly used in the first 40 generations (probability 0.85→0.95), and "elite preservation + neighborhood guidance" is introduced (10 optimal individuals are reserved each generation, and 5 new individuals are generated within 0.1 times the variable range); single-point crossover + simulated binary crossover (probability 0.85) is switched to in the last 40 generations;

[0078] Mutation operation: adaptive mutation rate (when the similarity of the population is greater than 0.7, the mutation rate is increased from 0.05 to 0.15);

[0079] Local search: when iteration is 40, a quadratic response surface model is constructed for the optimal 20 individuals, the local optimal solution is solved and the original individual is replaced, and finally 32 groups of Pareto optimal solutions are output.

[0080] 4. Simulation verification and parameter correction

[0081] Three schemes are selected from the Pareto optimal solution set, and "CFD aerodynamic simulation + finite element structural simulation" is used for verification:

[0082] 4.1 Aerodynamic simulation (ANSYS Fluent)

[0083] Turbulence model: k-ω SST coupling transition model, structured grid (1.8 million units, wall surface first layer grid height 0.01 mm, ensuring y + <1);

[0084] Boundary conditions: inlet wind speed 12 m / s (rated working condition), outlet pressure outlet, wall surface without slip;

[0085] Turbulence parameter correction: the turbulence dissipation rate correction equation is used where δ=0.85, ζ=0.03, and the wake flow speed prediction error is reduced from 12% to 4.5% compared with the traditional solution.

[0086] 4.2 Structure simulation (ANSYS APDL)

[0087] Load application: wind load 800 Pa (corresponding to limit wind speed 25 m / s), centrifugal load 300 N (generated by hub rotation), gravity load;

[0088] Constraint condition: fairing-hub connection surface is fixed and displacement is not allowed;

[0089] Key results: traditional scheme stress 260 MPa (close to glass fiber yield strength), optimal scheme stress 230 MPa (T700 safety margin sufficient).

[0090] 4.3 Parameter correction

[0091] Final determination of optimal parameters (correction amplitude ≤3%):

[0092] Upwind segment R1 = 2.3 m, middle segment D = 4.2 m (1.4 times the hub diameter, ≤1.5 times the engineering upper limit), L2 = 3.9 m, wake segment θ = 20°;

[0093] Performance index: Cd = 0.22, σ = 230 MPa, W = 0.27 t, Rw = 28 m.

[0094] 5. Prototype manufacturing and performance testing

[0095] 5.1 Prototype manufacturing (adapt to small-diameter fairing process)

[0096] Material: T700 carbon fiber composite material (weather resistance modified, adapted to Zhangbei-30°C low temperature);

[0097] Process: molding (pressure 12 MPa, temperature 120°C), layer angle determined by gradient equation

[0098] Determination (α = 0.4, β = 1.8, γ = 0.06, δ = 0.9);

[0099] Thickness design: upwind segment 1→2 mm linear thickening (impact resistance), wake segment 1.5→0.8 mm linear thinning (lightweight), surface roughness Ra 1.6 μm.

[0100] 5.2 Performance testing (stage verification)

[0101] Wind tunnel test: low-speed wind tunnel laboratory (wind speed 3-25 m / s), Cd = 0.22 under rated wind speed (error 3.6% compared with simulation), lift coefficient CL = 0.05 (no additional lift interference);

[0102] Field installation test: 60 days, record the whole machine net power generation efficiency (traditional 33% → optimized 37%), vibration amplitude ≤0.1mm / s (no resonance);

[0103] Fatigue test: "graded loading + acoustic emission monitoring", according to the wind turbine load spectrum, 5 levels of load (10 5 times per level), 8 sensors have no abnormal signal (no microcracks).

[0104] 6. Operation effect data characterization

[0105] Table 1: Comparison of performance of traditional and optimized fairing

[0106]

[0107]

[0108] Explanation: The data in Table 1 comes from 60 days of field testing, accurately reflecting the performance difference between the two schemes in actual operation. From the core indicators, the traditional hub fairing has a high aerodynamic drag coefficient of 0.35 due to the use of a single circular arc profile and glass fiber material, resulting in early airflow separation in the hub area and high energy loss. The maximum stress of the structure is 260MPa, which is close to the yield strength of glass fiber (220MPa), and the safety margin is insufficient. The weight of 0.31t increases the load on the hub, and the wake influence radius of 35m also easily interferes with the airflow at the blade root, resulting in a net power generation efficiency of only 33% for the whole machine.

[0109] The optimized scheme realizes performance leap through three major improvements: first, the bionic profile (composite curve + micro convex nodules) delays airflow separation, and the drag coefficient is reduced to 0.22, a reduction of 37%; second, T700 carbon fiber gradient layering + variable thickness process reduces weight by 13% (0.27t) while controlling stress at 230MPa (far below the material yield strength of 300MPa); third, wake optimization reduces the influence radius by 20% (28m), reducing airflow interference. The net power generation efficiency of the whole machine is increased by 4%, perfectly adapting to the turbulent flow and yawing conditions of the wind farm, solving the problems of "high resistance, high stress, and poor adaptation" of the traditional design.

[0110] Example 2: Optimization design of 3MW offshore wind turbine fairing (application scenario of Guangdong East Sea offshore wind farm)

[0111] This example is the design of a fairing for a 3MW direct-drive wind turbine (hub diameter 130m, number of blades 3, rated power 3000kW) in Guangdong East Sea offshore wind farm. The area has an average annual wind speed of 8.5m / s and a maximum wind speed of 32m / s, with severe salt spray corrosion. The traditional fairing has problems of excessive noise and difficulty in cross-model adaptation, and the use of the scheme of the present application realizes efficient adaptation, and the specific implementation process is as follows.

[0112] 1. Basic parameter acquisition and working condition definition

[0113] Collect wind turbine parameters: hub radius 65 m, blade root chord length 5.2 m, rated wind speed 13 m / s, limit wind speed 32 m / s. Use laser Doppler anemometer + unmanned aerial vehicle aerial survey, measurement point spacing 3 m, obtain hub height (100 m) wind shear exponent 0.10, salt spray concentration 50 mg / m 3 . Divide working conditions: rated wind speed working condition (12-15 m / s, 40% of the proportion), limit wind speed working condition (28-32 m / s, 3% of the proportion), turbulence working condition (turbulence intensity > 0.18, 22% of the proportion), yaw working condition (yaw angle 10°-25°, 15% of the proportion), start-up working condition (4-6 m / s, 20% of the proportion), and establish a working condition database containing corrosion factors.

[0114] 2. Multi-section parameterized modeling and bionic optimization

[0115] Modeling is performed using UGNX, the curvature radius of the windward segment is 85 m (1.3 times the hub radius), the maximum outer diameter of the middle segment is 120 m, and the middle segment is constructed by a composite curve equation y = kx 3 + mx 2 + nx + p + q sin (rx) (k = 0.0015, m = -0.2, n = 10.5, p = 65, q = 1.2, r = 0.1), the pitch distance of the micro-convex structure is 18 m, and the height is 2.2 m. The contraction angle of the wake segment is 22°, the connection arc radius is 80 mm, and the blade mounting hole (fillet radius 15 mm) and maintenance window (outward flange sealing) are reserved.

[0116] 3. Multi-objective optimization and layup design

[0117] Optimization objectives: Cd minimization, σ ≤ 300 MPa, W ≤ 3.5 t, noise L ≤ 60 dB. Improved NSGA-III algorithm is used, population 120, iteration 80 generations, local search started at the 40th generation. The layup equation is θ = αi + β sin (γi) + δ√i, α = 0.6, β = 2.5, γ = 0.09, δ = 1.3, the thickness of the windward segment gradually changes from 3 to 5 mm, and the thickness of the wake segment gradually changes from 4 to 2 mm. Salt spray resistant carbon fiber composite material is selected, and corrosion resistant coating is added between the layers.

[0118] 4. Simulation verification and noise optimization

[0119] CFD simulation grid 3 million, turbulence correction equation (δ = 0.9, ζ = 0.04), wake prediction error 3.8%. Noise calculation uses L = μ × u 3 / d × φ + η × Δp 2(f x p x c), m = 0.008, f = 0.75, n = 0.005, f = 500 Hz, p = 1.225 kg / m 3 , c = 340 m / s, predicted noise 68 dB, tail flow section added micro-perforated plate sound absorption layer (inclined 8°), noise reduced to 58 dB.

[0120] 5. Cross-model adaptation and testing

[0121] Based on the 1.5 MW benchmark model, adapted according to the scaling formula Ptarget = l x Pbase x (1 + xi x (l - 1)), l = 130 / 82 = 1.585, xi = 0.12, adjust the middle section length to 85 m, and the outer diameter to 120 m. Prototype testing: wind tunnel wind speed 4-32 m / s, field test for 60 days, fatigue test acoustic emission monitoring is normal, cross-model adaptation deviation is 4.2%.

[0122] 6. Operation effect data characterization

[0123] Table 2: Performance of optimized fairing under different working conditions

[0124] Operating condition type Aerodynamic drag coefficient Structural stress Noise level Wake radius Adaptation deviation Rated wind speed operating condition 0.07 230 MPa 58 dB 130m - Turbulent flow operating condition 0.09 265 MPa 60 dB 145m - Extreme wind speed operating condition 0.11 290 MPa 62 dB 160m - Cross-model adaptation (3 MW -> 2 MW) 0.075 240 MPa 59 dB 135m 4.2％ Conventional solution (rated operating condition) 0.15 330 MPa 72 dB 190m 12％

[0125] Explanation: Table 2 data comes from 60 days of offshore testing, the traditional scheme has an aerodynamic drag coefficient of 0.15 under rated working conditions, a noise of 72 dB, and a stress of 330 MPa under extreme wind speed, which is close to the limit. The cross-model adaptation deviation is 12%. The optimized scheme of the present application has a drag coefficient of 0.07 under rated working conditions, and a noise of 58 dB; the performance is stable under turbulent and extreme working conditions, and the stress is lower than 300 MPa; the cross-model adaptation deviation is only 4.2%. The bionic structure and noise control design adapt to complex wind conditions and environmental protection requirements, the layer optimization improves corrosion resistance, the cross-model method shortens the design cycle by 40%, and perfectly meets the series needs of offshore wind turbines.

[0126] Reference Figure 2 : Directly quantifies the influence of core variables on aerodynamic performance in multi-section parameterized modeling, and is a key basis for breaking away from traditional empirical design. When the curvature radius of the windward section increases from 50 m to 62 m, the semi-spherical transition structure makes the airflow flow more smoothly, and the drag coefficient decreases by 33.3% (0.12→0.08), which verifies the "curvature matching airflow" design logic; the maximum outer diameter of the middle section is 79 m, which has the optimal drag coefficient of 0.09, and too large outer diameter increases the windward area (when the outer diameter is 75 m→82 m, the drag coefficient first decreases and then increases), and too small outer diameter causes poor airflow contraction; the minimum resistance is 0.08 when the contraction angle of the tail flow section is 22°, and angles greater than 22° are prone to cause tail flow turbulence (the resistance rises to 0.09 when the angle is 25°), and angles less than 18° cause the tail flow diffusion range to expand. Through the chart, the optimal interval of each section variable can be quickly locked, providing quantitative input for multi-objective optimization.

[0127] Reference Figure 3 This figure highlights the core value of the multi-objective optimization of this invention. Traditional designs often optimize based on a single metric, which can easily lead to "high..." Accompanied by "high σ" or "low σ sacrifice" This invention is based on the Kriging response model of iSIGHT software, and generates multiple sets of x through Latin hypercube sampling. m With L D The Pareto solution set was obtained through CFD numerical calculations. The scatter plot shows that the optimal solution (1.689, 0.1139) is located at the vertex of the non-dominated front line, achieving... The balance between the maximum and minimum σ solves the pain point of traditional design that "it is difficult to balance energy output and stability", and provides a quantitative basis for scheme selection.

[0128] Reference Figure 4 This chart illustrates the core value of gradient optimization for composite material layups, addressing the issue of "strength redundancy or insufficiency" in traditional constant-angle layups. As the layup number increases (10→50), the layup angle increases gradient (15°→45°), precisely matching the load distribution of the fairing: layups 10-20 correspond to the wake section, with a low angle (15°-25°) achieving a flexible transition to accommodate low loads; layups 30-50 correspond to the windward section, with a high angle (35°-45°) enhancing wind resistance. Structural stress continuously decreases with angle optimization (280→240MPa), a reduction of 14.3%, and remains below the material's yield strength (280MPa) throughout, while traditional constant-angle layups (e.g., 30°) still reach a stress of 270MPa at i=50. The chart verifies the scientific validity of the gradient layup equation, achieving synergy between "load-structure-material," balancing strength and lightweight (12% lighter than traditional layups after optimization).

[0129] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for optimizing the design of a wind turbine fairing, characterized in that, The method comprises the following steps: The basic parameter acquisition and working condition definition step collects target wind turbine model parameters, hub size, blade number and rated power; five typical working conditions of rated wind speed, limit wind speed, turbulence, yaw and start are divided, corresponding wind speed range, duration ratio and load characteristics are recorded, and a working condition parameter database is established; The multi-section parameterized modeling step of the fairing adopts a three-dimensional modeling software to construct a parameterized model, which is divided into a windward section, a middle section and a wake section; the windward section is a hemispherical transition structure, the middle section is a variable curvature revolving body, and the wake section is a tapered conical surface; key design variables are defined; in view of the uneven vertical distribution characteristics of wind speed in the operation area of the wind turbine (the full-height wind speed distribution data obtained by the “laser Doppler anemometer + unmanned aerial vehicle aerial survey” in the basic parameter acquisition step), the non-central symmetric form is adopted to design the overall profile of the fairing: the radial size of the windward section and the middle section is designed differently along the vertical direction according to the wind speed gradient (the radial size in the area with high wind speed is 5%-8% larger than the reference value, and the radial size in the area with low wind speed is 3%-5% smaller than the reference value), the influence of the uneven vertical distribution of wind speed on the airflow flow is offset through the asymmetric structure, and the power fluctuation of the wind turbine is reduced; The multi-objective optimization system construction step constructs a two-layer system, the core target is to improve the net power generation by ≥5% under the condition of equal diameter, and the starting wind speed is ≤3.5 m / s; The secondary target includes a resistance coefficient ≤0.25 and a structural stress ≤80% of the yield strength; the weights are allocated according to the working condition ratio and sensitivity, the rated wind speed working condition focuses on power, and the starting working condition focuses on starting wind speed; The analytic hierarchy process is used to build a hierarchical structure, the comprehensive weight is obtained through consistency check, the optimization function is constructed by weighted normalization method, and the structure deformation, power constraint and process constraint are clearly defined; The optimization algorithm iteration solving step adopts an improved non-dominated sorting genetic algorithm for iteration optimization; a model is constructed and simulated to calculate the target value in each generation, the optimal solution set is selected through fast non-dominated sorting, and the Pareto optimal solution set is output after iteration convergence; The simulation verification and parameter correction step selects three schemes from the optimal solution set for verification, for example, the aerodynamic simulation uses CFD software and k-ω SST model, structured grid, and calculates the resistance coefficient; the structural simulation uses finite element software.

2. A method of optimizing the design of a wind turbine fairing according to claim 1, characterized in that, The step of bionic optimization of the fairing profile curve is also included, and a composite curve equation is used to construct the variable-curvature profile in the middle section, and the curve equation is y=kx 3 +mx 2 +nx+p+qsin(rx); wherein y is a profile radial coordinate, x is an axial coordinate, k is a cubic term coefficient, m is a quadratic term coefficient, n is a linear term coefficient, p is a constant term, q is a sine correction coefficient, and r is a correction frequency; the profile curve fitted by the equation forms a nodular micro-convex structure in the middle section, and a rough band or a turbulent flow control band is arranged in the middle section airflow separation prone area; the rough band adopts a silicon carbide particle convex structure, and the turbulent flow control band adopts a metal hollow grid structure, the boundary layer airflow is disturbed through the rough band / turbulent flow control band, and the airflow separation starting point is further delayed to move backward; and the nodular micro-convex structure and the rough band / turbulent flow control band synergistically induce the formation of stable attached turbulent flow around the fairing, the airflow separation starting point is moved backward, the airflow transition in the middle section is smooth, the local pressure loss is reduced, the buffer capacity to turbulent flow is enhanced, and the surface pressure fluctuation amplitude of the fairing is reduced under the turbulent flow condition.

3. A method of optimizing the design of a windmill fairing according to claim 1, characterized in that, Also included is a composite material ply gradient optimization step using a gradient ply optimization equation determining the ply angle of each layer; wherein θ is the ply angle of the i-th layer, i is the ply sequence number, α is the base angle gradient coefficient, β is the angle sine adjustment amplitude, γ is the angle sine variation coefficient, and δ is the angle root adjustment coefficient; through the equation, the windward segment ply angle increases in gradient with the increase of the sequence number, and the wake segment ply angle realizes flexible transition in combination with the root term; at the same time, the "variable thickness ply" process is introduced in the ply process.

4. A method of optimizing the design of a windmill fairing according to claim 1, characterized in that, In the simulation verification and parameter correction step, the CFD analysis of the coupling transition model is used for aerodynamic performance simulation, combined with the measured turbulence intensity data, and the turbulence dissipation rate correction equation is used Dynamic adjustment model parameters; wherein, ω is the corrected turbulence dissipation rate, δ is the turbulence basic correction factor, v is the turbulence fluctuation velocity, u is the main flow velocity, ε is the initial dissipation rate, ζ is the velocity gradient correction coefficient, The wall normal velocity gradient; the correction makes the average error of the simulation results of the wake flow velocity distribution and the wind tunnel test decrease; in addition, the transition start criterion of the transition model needs to be corrected, and the transition judgment coefficient based on the local turbulence intensity and the pressure gradient is introduced.

5. A method of optimizing the design of a windmill fairing according to claim 1, characterized in that, In the optimization algorithm iteration solving step, the improved non-dominated sorting genetic algorithm with multiple strategies is adopted, which specifically includes: an initialization stage; in the crossover operation, arithmetic crossover is mainly used in the first 40 generations, and an elite reservation + neighborhood guide mechanism is introduced at the same time; single-point crossover and simulated binary crossover are combined in the last 40 generations; the mutation operation adopts an adaptive mutation rate; when the iteration reaches the 40th generation, the local search based on the response surface method is started, a quadratic response surface model is constructed in the neighborhood of the design variable, the local optimal solution is solved and the original individual is replaced; through these strategies, the convergence accuracy of the final Pareto solution set is improved, and the distribution of the solution set on the three core targets of aerodynamic resistance, structural strength and weight is more uniform.

6. A method of optimizing the design of a windmill fairing according to claim 1, characterized in that, In the physical prototype manufacturing and performance testing step, the fatigue performance test adopts a combined method of "graded loading + acoustic emission monitoring": graded loading is based on the load spectrum of the wind turbine; the acoustic emission monitoring system is arranged with 8 sensors to capture the acoustic emission signals generated by the internal micro-cracks of the material in real time; the damage location is located by ultrasonic C scanning, and the damage area parameters are adjusted by returning to the parameterized modeling step.

7. A method of optimizing the design of a windmill fairing according to claim 1, characterized in that, In the basic parameter acquisition and working condition definition step, the wind speed distribution measurement adopts a combined method of "laser Doppler anemometer + unmanned aerial vehicle aerial survey": the laser Doppler anemometer is arranged at the hub height, and the measurement points are arranged in multiple directions to obtain the near-ground wind speed profile; the unmanned aerial vehicle carries a weather radar to conduct aerial survey above the wind turbine to obtain high-altitude wind speed, turbulence intensity and wind shear index; combined with the ground and high-altitude data, the logarithmic law wind profile model is used to fit the wind speed distribution at all altitudes.

8. A method of optimizing the design of a windmill fairing according to claim 1, characterized in that, Also included is a step of active control of the aerodynamic noise of the fairing, in which a noise prediction of the micro-perforated panel sound absorption structure is introduced in the aerodynamic performance simulation, and the sound pressure level of the aerodynamic noise is corrected by an equation L = μu2dφηΔp2f2ρc2; wherein L is the sound pressure level of the aerodynamic noise, μ is the airflow noise coefficient, u is the airflow velocity, d is the characteristic size of the fairing, φ is the noise radiation coefficient, η is the sound absorption structure correction coefficient, Δp is the pressure difference before and after the micro-perforated panel, f is the noise frequency, ρ is the air density, and c is the sound speed; when the simulated predicted noise sound pressure level exceeds 65 dB, a micro-perforated panel sound absorption layer is additionally arranged on the inner wall of the wake section of the fairing, and the contraction angle of the wake section is optimized to reduce the airflow separation noise; in addition, the installation angle of the micro-perforated panel needs to be optimized to be inclined to the inner wall of the wake section.

9. A method of optimizing the design of a windmill fairing according to claim 1, characterized in that, In the multi-section parameterized modeling step of the fairing, the connection of each section adopts a double-constraint design of "curvature continuity + stress guiding": the curvature continuity constraint is realized by B-spline curve interpolation; the stress guiding constraint is to pre-simulate the stress distribution in the three-dimensional model of the connection area, and by adjusting the connection arc radius and transition length, the stress flow line is smoothly transitioned along the connection surface; for the connection requirements of non-central symmetric contour, in the area where the vertical wind speed difference is significant, the connection arc radius is additionally optimized, and the stress concentration coefficient at the connection of the asymmetric structure is still controlled within 1.2; at the same time, ring stiffeners + variable wall thickness structure are designed around the blade mounting hole, the mounting hole edge is rounded, the hole wall thickness gradually increases, and the maintenance window adopts the design of "outward flange + sealing rubber strip".

10. A method of optimizing the design of a windmill fairing according to claim 1, characterized in that, Also included is a cross-model adaptive design verification step, which adopts a "parameter scaling + local adaptation" strategy for different types of wind turbines: the ratio of the hub diameter of the target model to the reference model λ is the scaling coefficient, and the scaling formula is P target = λ × P base × (1 + ξ × (λ - 1)); local adaptation focuses on adjusting the blade mounting hole position, maintenance window size, and layup scheme for the scaled parameters; cross-model testing selects three different types of wind turbines for adaptation testing, and the test indicators include aerodynamic drag coefficient deviation, structural strength margin, installation compatibility, and power generation efficiency change.

Images