Intelligent anti-icing system for wind turbine blades

Abstract

The present application relates to wind power blade anti-icing technical field, disclose a kind of wind turbine blade intelligent anti-icing system, system obtains external meteorological and unit mechanical state data, solves the relative wind speed amplitude of blade discrete leaf element node and airflow angle of attack;Airflow angle of attack is mapped as leading edge surface stagnation point coordinate, demarcate the physical impact boundary interval of liquid water droplet;Establish mass conservation equation calculation and update the water film thickness value of blade surface;Calculate chordwise aerodynamic shear stress and spanwise rotation centrifugal force, synthesis two-dimensional migration direction vector;According to physical impact boundary interval and two-dimensional migration direction vector generate dynamic power distribution matrix, output electric power instruction in combination with convection heat transfer coefficient, drive electric heating grid array heating.The present application concentrates distribution to real water droplet impact area and carries out compensation heating along water film migration path, effectively reduces the system anti-icing energy consumption, prevents overflow water from secondary icing.

Description

Technical Field

[0001] This invention relates to the field of wind turbine blade anti-icing technology, specifically to an intelligent anti-icing system for wind turbine blades. Background Technology

[0002] When wind turbines operate in low-temperature and high-humidity environments, the blade surface can capture supercooled water droplets from the air and cause icing. Blade icing alters the original aerodynamic shape, reduces wind energy capture efficiency, and increases the mechanical vibration load on the unit. Currently, the industry uses anti-icing systems that lay electric heating elements on or inside the blade surface to address this problem.

[0003] Existing blade electric heating anti-icing systems mostly adopt a heating control strategy with fixed area and constant power. In actual operation, the wind speed, rotational speed and pitch angle of the unit are constantly changing, which causes the flow field and airflow angle of attack of each section of the blade to change accordingly. The physical boundary area where liquid water droplets actually hit the leading edge of the blade exhibits dynamic fluctuation characteristics. Traditional fixed area heating methods cannot adapt to this dynamic change and often continuously output heat in areas where there is no water droplet impact, resulting in high overall energy consumption of the anti-icing system.

[0004] Furthermore, limited by the maximum output power of the heating element, the liquid water captured at the leading edge of the blade cannot evaporate completely in a short time. Under the combined action of surface aerodynamic shear stress and the centrifugal force of the rotating impeller, this residual liquid water will flow towards the trailing edge and tip of the blade. When the overflow water film flows out of the preset fixed heating area, it will undergo a phase change and form ice ridges due to the loss of heat replenishment and the influence of external cold air convection cooling, thus causing secondary icing. The existing anti-icing control mechanism lacks the ability to observe and track the flow trajectory of residual liquid water and cannot dynamically allocate compensation power along the migration path of the water film, making it difficult to eliminate the secondary icing problem and continuously affecting the safe and stable operation of the wind turbine. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an intelligent anti-icing system for wind turbine blades, which solves the problems of high overall energy consumption caused by fixed-area heating in existing anti-icing systems and the inability to cope with secondary icing caused by overflow water film flow.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] This invention provides an intelligent anti-icing system for wind turbine blades, comprising:

[0008] The system comprises the following modules: a data acquisition module, which acquires external environmental meteorological data and unit operating mechanical status data; a parameter calculation module, which calculates the relative wind speed amplitude and airflow angle of attack of discrete blade element nodes based on the mechanical status data; a boundary mapping module, which maps the airflow angle of attack to the physical chordal coordinates of the leading edge surface stagnation points, defining the physical impact boundary interval of liquid water droplets; a mass-heat observation module, which establishes a mass conservation equation based on the environmental meteorological data, calculates the current liquid water mass flux, and iteratively updates the water film thickness value on the blade surface; a vector synthesis module, which calculates the chordal aerodynamic shear stress and spanwise rotational centrifugal force based on the relative wind speed amplitude and the water film thickness value to synthesize a two-dimensional migration direction vector; a command generation module, which generates a dynamic power allocation matrix based on the physical impact boundary interval and the two-dimensional migration direction vector, and outputs an electric power command based on the convective heat transfer coefficient; and a heating execution module, which receives the electric power command and drives the electric heating grid array set on the blade surface to generate heat.

[0009] Furthermore, when calculating the relative wind speed amplitude and the airflow angle of attack, the parameter calculation module extracts the axial and circumferential induction factors when the airflow passes through the rotating blade plane. Combining the hub wind speed and the rotor speed, it solves for the axial and tangential relative velocities of the discrete blade element node. Based on the axial and tangential relative velocities, it performs spatial vector synthesis to obtain the inflow angle and the relative wind speed amplitude of the blade element node. The airflow angle of attack is then derived and calculated by combining the cross-sectional twist angle of the discrete blade element node and the pitch angle of the pitch system.

[0010] Furthermore, when defining the physical impact boundary interval, the boundary mapping module uses the airflow angle of attack as an addressing index to input the physical chord coordinates of the stagnation point on the leading edge surface into the airfoil aerodynamic database; using the stagnation point on the leading edge surface as a spatial reference, it introduces a supercooled water droplet impact trajectory calculation model to obtain the spatial motion trajectory of the water droplet in the flow field, establishes the physical coordinate limits of the farthest upper surface and the farthest lower surface of the liquid water droplet attached to the leading edge of the blade, and defines the chord length interval covered by the physical coordinate limits of the farthest upper surface and the farthest lower surface as the physical impact boundary interval.

[0011] Furthermore, the mass-heat observation module calculates the liquid water content of the incoming airflow, and calculates the current liquid water mass flux by combining the relative wind speed amplitude with the water droplet collection efficiency of the corresponding leaf element node; by superimposing the latent heat released by water freezing and the latent heat absorbed by liquid water vaporization, the phase change latent heat flux density required to completely evaporate the captured supercooled water droplets is calculated.

[0012] Preferably, when iteratively updating the water film thickness value, the mass-thermal observation module extracts the recorded value of the electric power command sent to the electrically heated grid array in the previous time step, retrieves the convective heat transfer coefficient and ambient temperature of the corresponding grid in the previous time step, calculates the remaining heat flux density surplus after offsetting surface convective heat dissipation, and converts the heat flux density surplus into the water mass flux consumed by evaporation in the previous time step; combining the current liquid water mass flux and the water mass flux consumed by evaporation in the previous time step, the module iteratively updates the water film thickness value in the discrete time domain based on the mass conservation equation.

[0013] Furthermore, the vector synthesis module combines the external air density and surface friction coefficient to calculate the chordal aerodynamic shear stress driving the water film migration based on the relative wind speed amplitude; it introduces the liquid water density constant and the wind turbine speed to calculate the spanwise rotational centrifugal force based on the water film thickness; and it uses the parallelogram rule to perform vector superposition operation on the chordal aerodynamic shear stress and the spanwise rotational centrifugal force to synthesize the two-dimensional migration direction vector characterizing the macroscopic flow trend of the residual liquid water.

[0014] Furthermore, when generating the dynamic power allocation matrix, the instruction generation module marks all the electric heating grid arrays covered by the physical impact boundary interval as the main evaporation zone in the matrix and assigns them a basic evaporation target temperature; using the spatial coordinates of the main evaporation zone as the origin reference, it locates the downstream adjacent electric heating grid arrays along the direction of the two-dimensional migration direction vector and marks them as dynamic flow guiding antifreeze layers in the matrix; for the electric heating grid arrays marked as dynamic flow guiding antifreeze layers, its target temperature is distributed in an exponentially decreasing manner along the direction of the two-dimensional migration direction vector.

[0015] Preferably, the instruction generation module extracts the physical geometric center coordinates of the grid in the dynamic flow-guiding antifreeze layer as the grid centroid, and constructs a spatial exponential decay function based on the physical spatial distance between the grid centroid and the stationary point on the leading edge surface to perform asymmetric temperature assignment, so as to form a spatial thermal gradient distribution.

[0016] Furthermore, the instruction generation module calculates the Reynolds number based on the relative wind speed amplitude and characteristic chord length, and introduces the Prandtl number to derive the Nusselt number to solve for the convective heat transfer coefficient of the corresponding grid; it calculates the basic sensible heat flux density based on the target temperature, external ambient temperature, and convective heat transfer coefficient assigned in the dynamic power allocation matrix, superimposes the latent heat flux density of phase change calculated by the mass-heating observation module, and integrates the heat transfer efficiency to solve for the electric power instruction required to be output by each electric heating grid array.

[0017] Furthermore, the heating execution module includes a solid-state relay group connected via an electrical control bus. The heating execution module converts the electrical power command into a pulse width modulation signal, calculates the duty cycle of the pulse width modulation signal based on the inherent maximum physical output power constant of the electric heating grid array under rated voltage, drives the solid-state relay group to perform high-frequency closing and opening actions, and adjusts the effective value of the transient current input to the multiple electric heating grid arrays.

[0018] This invention provides an intelligent anti-icing system for wind turbine blades. It has the following beneficial effects:

[0019] 1. This invention establishes the physical impact boundary range where liquid water droplets actually adhere by calculating the airflow angle of attack and mapping the chord coordinates of the stagnation points on the leading edge surface, and concentrates the electrical power to the main evaporation zone formed by this range. This method uses microscopic aerodynamic parameters to limit the heating range and avoids outputting ineffective heat to redundant areas without water droplet impact. Compared with the constant heating strategy in the whole area, it reduces the power consumption of the system.

[0020] 2. This invention establishes a mass conservation equation to observe the thickness of the residual water film, and combines the chordal aerodynamic shear stress and the spanwise rotational centrifugal force to synthesize a two-dimensional migration direction vector to locate the flow path of the water film toward the trailing edge of the blade. The system sets up a dynamic flow guiding and antifreeze layer downstream of the main evaporation zone along this vector direction and performs compensating heating to maintain the liquid characteristics of the overflow water during the migration process, prevent the water film from freezing and forming ice ridges due to leaving the heating zone, and ensure the integrity of the blade's aerodynamic shape.

[0021] 3. This invention introduces a spatial exponential decay function into the dynamic flow-guiding antifreeze layer to construct a decreasing thermal gradient distribution along the flow direction. Combined with the real-time convective heat transfer coefficient and phase change latent heat flux density, the target electric power command of each electric heating grid array is calculated. This mechanism matches the output power of the underlying hardware with the convective heat dissipation and latent heat consumption requirements of the corresponding node, preventing overheating of the grid surface while blocking icing, and improving the conversion efficiency of electrical energy into effective anti-icing thermal energy. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the system architecture of an intelligent anti-icing system for wind turbine blades according to the present invention;

[0023] Figure 2 This is a schematic diagram of the main flow of the control method for an intelligent anti-icing system for wind turbine blades according to the present invention;

[0024] Figure 3 This is a schematic diagram illustrating the specific operation process of data acquisition, parameter calculation, and boundary mapping in this invention.

[0025] Figure 4This is a schematic diagram illustrating the specific operation process of the mass-thermal observation and water film thickness numerical iteration of the present invention.

[0026] Figure 5 This is a schematic diagram illustrating the specific operation process of two-dimensional migration direction vector synthesis according to the present invention;

[0027] Figure 6 This is a schematic diagram illustrating the specific operation flow of the power command calculation and pulse width modulation signal generation of the present invention.

[0028] Figure 7 This is a schematic diagram comparing the energy consumption of the traditional uniform flow heating and dynamic anti-icing systems of the present invention;

[0029] Figure 8 This is a schematic diagram comparing the secondary icing thickness of the conventional uniform flow heating and dynamic anti-icing systems of the present invention.

[0030] Among them, 100 is the data acquisition module; 200 is the parameter calculation module; 300 is the boundary mapping module; 400 is the mass and heat observation module; 500 is the vector synthesis module; 600 is the instruction generation module; and 700 is the heating execution module. Detailed Implementation

[0031] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] See attached document Figure 1 This invention provides an intelligent anti-icing system for wind turbine blades, comprising: a data acquisition module 100, a parameter calculation module 200, a boundary mapping module 300, a mass and heat observation module 400, a vector synthesis module 500, an instruction generation module 600, and a heating execution module 700.

[0033] The data acquisition module 100 is connected to the cabin meteorological station and the data acquisition and monitoring control system in hardware, and is configured to read the external environmental meteorological data of the wind turbine and the mechanical status data of the unit in real time.

[0034] The parameter calculation module 200 is deployed in the main control computing unit and is communicatively connected to the data acquisition module 100. Based on the read mechanical state data, it calculates the relative airflow velocity and airflow angle of attack of each discrete node of the blade in the plane of rotation.

[0035] The boundary mapping module 300, connected to the parameter calculation module 200, maps the airflow angle of attack to the physical chord coordinates of the stagnation point on the leading edge surface of the blade by retrieving the airfoil aerodynamic database in the memory, thus defining the physical impact boundary range of the liquid water droplets.

[0036] The mass-thermal observation module 400 extracts heat output data from historical control cycles, combines it with external environmental meteorological data to establish a discrete-time step mass conservation equation, calculates the current liquid water mass flux, and iteratively updates the liquid water film thickness value on the blade surface.

[0037] The vector synthesis module 500 connects the parameter calculation module 200 and the mass and heat observation module 400. It introduces the liquid water film thickness value into the fluid dynamics calculation, solves the chordal aerodynamic shear stress and spanwise rotational centrifugal force, and synthesizes a two-dimensional migration direction vector characterizing the liquid water flow.

[0038] The instruction generation module 600 generates a dynamic power allocation matrix with spatial asymmetry characteristics based on the physical impact boundary interval defined by the boundary mapping module 300 and the two-dimensional migration direction vector output by the vector synthesis module 500, and outputs an electric power instruction in combination with the convective heat transfer coefficient.

[0039] The heating execution module 700 includes a multi-channel independent electric heating grid array and an associated solid-state relay group built into the blade surface. It receives the electric power command output by the command generation module 600 and drives the electric heating grid array to generate heat through pulse width modulation signal.

[0040] See attached document Figure 2 This invention provides a control method for an intelligent anti-icing system for wind turbine blades, comprising the following steps:

[0041] S100, the data acquisition module 100 acquires environmental meteorological data and unit operating status parameters. When the environmental meteorological data reaches the set icing threshold, the intelligent anti-icing system is activated to control the calculation process.

[0042] S200, the parameter calculation module 200 divides the discrete blade element nodes along the blade span, extracts the current wind speed, rotational speed and pitch angle from the unit's operating status parameters, and calculates the relative wind speed amplitude and airflow angle of attack for each discrete blade element node.

[0043] S300, the boundary mapping module 300 inputs the airflow angle of attack into the airfoil aerodynamic database for interpolation mapping, obtains the physical chord coordinates of the blade leading edge stagnation point at the current time step, and establishes the impact boundary range of the supercooled water droplet based on the preset impact model.

[0044] The S400 mass and heat observation module 400 calculates the liquid water content and water droplet collection efficiency of the incoming airflow based on environmental meteorological data and relative wind speed amplitude, and calculates the liquid water mass flux and latent heat flux density per unit surface area; it also estimates the liquid water content based on ambient temperature and relative humidity using a built-in atmospheric empirical model.

[0045] S500, the mass-heat observation module 400 extracts the recorded value of the electrical power command output by the command generation module 600 in the previous time step as heat feedback data, calculates the water mass flux consumed by evaporation, and combines it with the liquid water mass flux in the current time step to iteratively update the water film thickness value on the blade surface according to the mass conservation equation.

[0046] S600, the vector synthesis module 500 calculates the chordal aerodynamic shear stress based on the relative wind speed amplitude and the surface friction coefficient of the corresponding blade element node, introduces the updated water film thickness value to calculate the spanwise rotational centrifugal force, and performs vector synthesis on the chordal aerodynamic shear stress and the spanwise rotational centrifugal force to obtain a two-dimensional migration direction vector characterizing the overflow water flow.

[0047] S700, the instruction generation module 600 marks the electric heating grid array covered by the impact boundary interval as the main evaporation zone, addresses the downstream adjacent electric heating grid array along the direction of the two-dimensional migration vector and marks it as a dynamic flow guiding antifreeze layer, and constructs a dynamic power distribution matrix that has the liquid characteristics of maintaining the unevaporated overflow water and the target temperature decays asymmetrically along space.

[0048] S800, the instruction generation module 600 calculates the electric power instruction based on the convective heat transfer coefficient and latent heat flux density of the corresponding grid for the activated electric heating grid array in the dynamic power distribution matrix; and sets the input electric power required for the electric power instruction to be equal to the sum of convective heat transfer power, latent heat of vaporization power, and sensible heat heating power minus the pneumatic heating power according to the anti-icing thermal balance equation.

[0049] The heating execution module 700 converts the electrical power command into a pulse width modulation signal, which drives the associated solid-state relay group to close. When the data acquisition module 100 continuously monitors the external environmental meteorological data and the icing threshold condition is exceeded for a set time, the command generation module 600 outputs a zero power command, and the system stops the control calculation process and returns to a low-power data polling state.

[0050] See attached document Figure 3 In specific implementations, step S100 provided by the present invention may include the following steps:

[0051] S110, the data acquisition module 100 establishes a periodic data polling connection with the weather station in the nacelle of the wind turbine and the data acquisition and monitoring control system through the industrial communication bus to obtain real-time external environmental meteorological data and mechanical status data of the unit operation; the external environmental meteorological data specifically includes the ambient temperature and relative humidity outside the nacelle; the mechanical status data specifically includes the hub wind speed, rotor speed and pitch angle of the wind turbine at the current moment.

[0052] S120, the data acquisition module 100 performs threshold logic comparison on the acquired ambient temperature and relative humidity; the intelligent anti-icing system presets physical boundary conditions characterizing the probability of atmospheric icing, namely the icing temperature threshold and the icing humidity threshold. In this embodiment, the icing temperature threshold ranges from 0°C to 3°C, and the icing humidity threshold ranges from 80% to 95%; when the ambient temperature is lower than the icing temperature threshold and the relative humidity is higher than the icing humidity threshold, the system kernel generates a hardware interrupt signal, wakes up the anti-icing main control program, and activates the subsequent control calculation process; when the meteorological triggering conditions are not met, the system maintains a low-power data communication polling state and does not output control commands.

[0053] In specific implementations, step S200 provided by the present invention may include the following steps:

[0054] S210, the parameter calculation module 200 extracts the geometric three-dimensional model data of the physical blade and divides the blade into multiple independent discrete leaf element nodes along the spanwise dimension from the leaf root to the leaf tip. In this embodiment, the number of discrete leaf element nodes ranges from 20 to 50.

[0055] The geometric intrinsic properties of each discrete leaf element node are extracted, including the spanwise rotation radius and the cross-sectional twist angle of each node. Furthermore, each discrete leaf element node is mapped on the chord surface to a multi-path independent electric heating grid array built into the blade surface, establishing a spatial correspondence between one-dimensional spanwise nodes and two-dimensional heating grids.

[0056] S220, the parameter calculation module 200 acquires the mechanical state data of the unit operation and introduces the blade element momentum theory model; for each discrete blade element node, it extracts the axial induction factor and circumferential induction factor generated when air flows through the rotating blade plane; combined with the hub wind speed and the rotor speed, it calculates the axial relative velocity and tangential relative velocity of each discrete blade element node respectively. The specific fluid dynamics calculation formulas are as follows:

[0057] V ax =V·(1-a);

[0058] U tan =Ω·r·(1+a′);

[0059] In the formula, V a x V is the axial relative velocity, a is the hub wind speed, and a is the axial induction factor. tan Ω represents the tangential relative velocity, r represents the rotor speed, a′ represents the spanwise rotation radius of the discrete blade element node, and a′ represents the circumferential induction factor.

[0060] S230, the parameter calculation module 200, based on the axial and tangential relative velocities output, performs spatial vector synthesis to solve for the inflow angle and relative wind speed amplitude of the actual airflow entering the discrete blade element node; combining the inherent cross-sectional twist angle of each node's geometry and the pitch angle currently executed by the pitch system, it derives the airflow angle of attack of the blade in the three-dimensional plane of rotation. The specific geometric derivation formulas are as follows:

[0061]

[0062] α = φ - (β + θ);

[0063] In the formula, φ is the inflow angle; V ax U is the axial relative velocity. tan W represents the tangential relative velocity. r α is the relative wind speed amplitude; β is the airflow angle of attack; β is the blade pitch angle; θ is the cross-sectional structural twist angle of the discrete blade element node. The operation sign of θ depends on the reference twist direction defined during blade design, thereby completing the spatial state dimensionality reduction and transforming the macroscopic operating state parameters of the wind turbine into the microscopic aerodynamic inflow state of the blade cross section; arctan represents the inverse function of the tangent function in trigonometric functions.

[0064] In specific implementations, step S300 provided by the present invention may include the following steps:

[0065] S310, the boundary mapping module 300 calls the airfoil aerodynamic database preset in the main control computing unit memory; the airfoil aerodynamic database records the flow field distribution tensor of the target blade under different Reynolds numbers and different inflow conditions; the boundary mapping module 300 uses the calculated airflow angle of attack as the addressing index, and uses a multidimensional linear interpolation algorithm to match the spatial position of the stagnation point on the leading edge surface of the blade under the current flow field state in the airfoil aerodynamic database, and obtains the chordal arc length coordinate of the stagnation point on the physical profile of the blade cross section; the stagnation point represents the physical boundary origin where the airflow stagnates on the blade surface and the relative flow velocity drops to 0.

[0066] S320, the boundary mapping module 300 uses the chordal arc length coordinates of the stagnation point as the spatial reference and introduces a calculation model for the impact trajectory of supercooled water droplets; it performs flow field calculations for the corresponding cross section based on the Eulerian two-phase flow model or the Lagrange particle tracking method to obtain the spatial motion trajectory of supercooled water droplets in the airfoil flow field and the numerical value of collection efficiency.

[0067] S330, the boundary mapping module 300 establishes the farthest upper surface physical coordinate limit and the farthest lower surface physical coordinate limit of the liquid water droplet attached to the leading edge of the blade at the current time step based on the solution results of the supercooled water droplet impact trajectory calculation model; the chord length interval covered by the upper surface physical coordinate limit and the lower surface physical coordinate limit is defined as the physical impact boundary interval of the supercooled water droplet; the physical impact boundary interval of the supercooled water droplet represents the geometric surface mapping region within the current discrete time slice where there is only one external liquid water source capture.

[0068] See attached document Figure 4 In specific implementations, step S400 provided by the present invention may include the following steps:

[0069] S410, the mass and thermal observation module 400 receives ambient temperature and relative humidity parameters collected by meteorological sensors, and calculates the liquid water content of the incoming air under the current environmental conditions by combining it with a preset empirical model of saturated water vapor pressure.

[0070] S420, the mass and heat observation module 400 extracts the relative wind speed amplitude calculated by the parameter solving module and calculates the Stokes number, which characterizes the degree of deviation of the inertia of the supercooled water droplets from the streamline. Then, based on the Stokes number and combined with the geometric characteristics of the leading edge of the blade, the water droplet collection efficiency of each discrete leaf element node in the leading edge region is derived.

[0071] The S430 mass-thermal observation module 400 combines liquid water content, relative wind speed amplitude, and water droplet collection efficiency to calculate the liquid water mass flux impacting a unit surface area of ​​the leaf element at the current time step through multiplication operations. The corresponding calculation formula is as follows:

[0072]

[0073] In the formula, LWC represents the mass flux of liquid water captured during the current control cycle; W represents the liquid water content in the air; r η represents the relative wind speed amplitude; η is the water droplet collection efficiency; the water droplet collection efficiency is dimensionless and its value ranges between 0 and 1.

[0074] The S440 mass-heat observation module 400, based on a determined liquid water mass flux, is configured with energy redundancy control logic to cope with extreme operating conditions. It performs a conservative upper limit thermodynamic physical calibration by superimposing the latent heat released during water freezing and the latent heat absorbed during liquid water vaporization. The maximum theoretical latent heat flux density of phase change required for the complete evaporation of the captured supercooled water droplets on the blade surface is calculated, and the corresponding thermodynamic formula is as follows:

[0075]

[0076] In the formula, q lat L is the latent heat flux density of phase change;f L is the latent heat constant of water freezing. v The latent heat of vaporization of water is denoted as . The mass flux of liquid water captured during the current control cycle; the latent heat constant of water freezing is taken as 334 kJ / kg under standard atmospheric pressure, and the latent heat constant of water vaporization is taken as 2257 kJ / kg.

[0077] In specific implementations, step S500 provided by the present invention may include the following steps:

[0078] S510, the mass-thermal observation module 400 constructs a time-domain state observer based on discrete time steps in the control kernel, extracts the electric power command actually issued by the heating system to the corresponding physical heating grid in the previous time step, and retrieves the convective heat transfer coefficient and ambient temperature data of the corresponding grid in the previous time step.

[0079] S520, the mass-thermal observation module 400, based on the law of conservation of energy, calculates the excess heat flux density remaining after offsetting surface convective heat dissipation by the electrical power allocated in the previous time step, and converts the excess heat flux density into the actual water mass flux consumed by evaporation in the previous time step. The corresponding conversion formula is as follows:

[0080]

[0081] In the formula, This represents the water mass flux consumed by evaporation in the previous time step. The power command output at the previous time step; A zone This represents the physical area of ​​the electrically heated grid array at that location; T represents the convective heat transfer coefficient of the grid corresponding to the previous time step. tgt To set the target surface temperature; The ambient temperature at the previous time step; L v The latent heat of water vaporization is denoted by ; the invalid evaporation state before the system reaches the target temperature is forcibly set to 0 by taking the maximum value function.

[0082] The S530 mass-thermal observation module 400, based on the mass conservation equation, integrates the liquid water mass flux captured in the current time step and the water mass flux consumed by evaporation in the previous time step. It iteratively updates the residual water film thickness, representing the thickness of the microscopic liquid water layer that accumulates on the blade surface due to unevaporation, within the discrete time domain. The corresponding thickness iterative update formula is as follows:

[0083]

[0084] In the formula, This is the updated water film thickness value at the current time step; This is the water film thickness value left over from the previous time step; the water film thickness value is set to 0 at the initial running time of the system. This represents the liquid water mass flux at the current time step. The water mass flux consumed by evaporation in the previous time step; Δt is the time step of the system's discrete cycle, and the physical lower limit of the water film thickness is constrained to 0 in the formula by using a maximum value function; ρ w This is the density of liquid water.

[0085] See attached document Figure 5 In specific implementations, step S600 provided by the present invention may include the following steps:

[0086] The S610 vector synthesis module 500 extracts the relative wind speed amplitude calculated by the parameter solving module to calculate the chordal aerodynamic shear stress acting on the surface of the liquid water film. Combined with the external air density and surface friction coefficient, a fluid boundary layer tribological model is constructed to calculate the aerodynamic force components along the physical chord of the airfoil and pointing towards the trailing edge of the blade. The aerodynamic shear stress is represented by the physical force driving the water film on the blade surface to migrate towards the trailing edge. The corresponding formula for calculating the chordal aerodynamic shear stress is as follows:

[0087]

[0088] In the formula, For chordal aerodynamic shear stress; ρ a W represents the external air density. r C represents the relative wind speed amplitude. f The surface friction coefficient for the corresponding node is determined based on wind tunnel experimental data or empirical formulas for the Reynolds number of the target blade airfoil. It is the physical unit vector pointing along the chord towards the trailing edge of the blade.

[0089] S620, the vector synthesis module 500 extracts the chordal aerodynamic shear stress and spanwise rotational centrifugal force distributed along the orthogonal physical directions, performs vector superposition according to the parallelogram rule of fluid dynamics space, and synthesizes a two-dimensional migration direction vector characterizing the macroscopic flow trend of residual liquid water on the blade surface. The corresponding two-dimensional vector synthesis formula is as follows:

[0090]

[0091] In the formula, For spanwise rotational centrifugal force, the two-dimensional migration direction vector characterizes the superposition effect of airflow shearing and rotational centrifugal force; ρ w d is the density constant of liquid water. film Ω represents the water film thickness; r represents the rotor speed; and r represents the spanwise rotation radius of the discrete blade element node. It is the physical unit vector pointing towards the blade tip along the span.

[0092] S630, the vector synthesis module 500 extracts the chordal aerodynamic shear stress and spanwise rotational centrifugal force distributed along orthogonal physical directions, performs vector superposition operations according to the parallelogram rule of fluid dynamics space, and synthesizes a two-dimensional migration direction vector characterizing the macroscopic flow trend of residual liquid water on the blade surface. The corresponding two-dimensional vector synthesis formula is as follows:

[0093]

[0094] In the formula, The synthesized two-dimensional migration direction vector represents the dual physical effects of airflow shear drag and rotational centrifugal pull, and exhibits dynamic spatial deflection at different time steps due to fluctuations in ambient wind speed and rotational speed. This represents the chordal aerodynamic shear stress vector. This is the spanwise rotational centrifugal force vector.

[0095] In specific implementations, step S700 provided by the present invention may include the following steps:

[0096] S710, the instruction generation module 600 constructs a dynamic power allocation matrix corresponding to the arrangement of the electric heating grid array in the physical hardware, extracts the physical impact boundary interval of the supercooled water droplets defined by the boundary mapping module 300, marks all the electric heating grid arrays covered by the physical impact boundary interval as the main evaporation area in the matrix, and uniformly assigns the set basic evaporation target temperature to the main evaporation area. The basic evaporation target temperature is set to be 10 to 15°C higher than the critical temperature for water freezing.

[0097] S720, the instruction generation module 600 uses the spatial coordinates of the main evaporation zone as the origin reference, and performs spatial addressing operation along the direction of the two-dimensional migration direction vector output by the vector synthesis module 500. It locates the physical electric heating grid array downstream of the main evaporation zone through which the unevaporated liquid water film flows along the two-dimensional migration direction vector, and marks all the physical grids that are addressed as dynamic flow guiding antifreeze layers in the dynamic power allocation matrix.

[0098] S730, the instruction generation module 600 generates asymmetric temperature values ​​for the electrically heated grid array marked as a dynamic flow-guiding antifreeze layer. It extracts the physical geometric center coordinates of the grid as the grid centroid and constructs a spatial exponential decay function based on the physical spatial distance between the grid centroid and the leading edge stagnation point. This causes the target temperature of the dynamic flow-guiding antifreeze layer to decrease exponentially along the fluid migration direction, approaching the freezing point critical value. This forms a spatial thermal gradient distribution that blocks secondary icing of the water film during flow. The corresponding target temperature decay formula is as follows:

[0099]

[0100] In the formula, For the dynamic flow-guiding antifreeze layer, the distance is xd The grid at that location is assigned the target temperature; T ice ΔT is the absolute temperature constant for water freezing, taken as 273.15 Kelvin; ΔT is the initial margin for the antifreeze temperature difference, taken as the difference between the temperature of the main evaporation zone and the critical freezing point; k s The preset spatial attenuation coefficient is set to a value range of 0.1 to 0.5; x d The vector space scalar distance between the centroid of the current addressing grid and the stagnation point at the leading edge is set to 0. For the remaining redundant heating grids in the dynamic power allocation matrix that are neither marked as the main evaporation zone nor as the dynamic flow-guiding antifreeze layer, the target temperature is set to 0, maintaining a dormant state with no heat output.

[0101] See attached document Figure 6 In specific implementations, step S800 provided by the present invention may include the following steps:

[0102] S810 refers to the relative wind speed amplitude extracted by the parameter calculation module 500 of the vector synthesis module to calculate the chordal aerodynamic shear stress acting on the surface of the liquid water film. A fluid boundary layer tribology model is constructed by combining the external air density and the surface friction coefficient to calculate the aerodynamic force components along the physical chord of the airfoil and pointing towards the trailing edge of the blade. The aerodynamic shear stress is represented by the physical force driving the water film on the blade surface to migrate towards the trailing edge. The corresponding formula for calculating the chordal aerodynamic shear stress is as follows:

[0103]

[0104] In the formula, h c N is the convective heat transfer coefficient of this grid. u Let k be a Nusselt number. a The thermal conductivity of air is obtained by referring to the standard atmospheric parameter table based on real-time ambient temperature, L. C is the physical characteristic chord length of the current discrete leaf element node.

[0105] The S820 command generation module 600 calculates the basic sensible heat flux density required to counteract the convective heat dissipation of external cold air based on the target temperature of the activated grid in the dynamic power allocation matrix and the current ambient temperature, combined with the convective heat transfer coefficient of the corresponding grid. It then superimposes the latent heat flux density of the corresponding grid nodes and considers the physical coverage area and heat transfer efficiency of the electric heating grid to calculate the target electric power command required for each independent electric heating grid array within the current control cycle. This calculation process transforms aerodynamic and thermodynamic parameters into quantifiable control parameters for the electrical system hardware. The corresponding electric power command calculation formula is as follows:

[0106]

[0107] In the formula, P cmd For target electrical power command; Azone h represents the physical area of ​​a single electrically heated grid array. C T represents the convective heat transfer coefficient of the corresponding grid. tgt The target temperature of the grid assigned to the dynamic power allocation matrix; T a The external ambient temperature; q lat η is the latent heat flux density of phase change; th The heat transfer efficiency constant from the blade shell to the surface is calibrated by the thermal impedance test of the blade composite material and is taken as 0.8 to 0.95. In order to avoid calculating negative values ​​that violate the actual heat dissipation, a lower limit cutoff mechanism is set in the formula. Once the calculated heat flux density is negative, the system directly treats it as 0.

[0108] S830, the heating execution module 700 receives the target electrical power command from the command generation module and converts it into a pulse width modulation signal recognizable by the underlying hardware. Based on the rated maximum output power of the heating unit, it calculates the physical duty cycle of the high-level state within a time period. The generated pulse width modulation signal is transmitted via the electrical control bus to the solid-state relay group built into the blade root, driving the solid-state relay group to perform high-frequency closing and opening actions within a set time step. This adjusts the effective value of the transient current input to the multi-channel independent electric heating grid array on the blade surface. The corresponding duty cycle calculation formula is as follows:

[0109]

[0110] In the formula, D pwm P represents the duty cycle of the pulse width modulation signal; cmd For target electrical power command; P max To correspond to the inherent maximum physical output power constant of the electric heating grid array under rated voltage, the formula introduces a minimum value function to constrain the physical upper limit of the duty cycle to 100%; min means taking the minimum value among all terms in parentheses.

[0111] Application Examples:

[0112] To aid in understanding the technical solution of this invention, the following provides an application example and test example of an intelligent anti-icing system for wind turbine blades.

[0113] The data acquisition module 100 obtains the current external ambient temperature as -5℃ and relative humidity as 92%. When the input meteorological data reaches the physical boundary conditions of the icing temperature threshold being below 0℃ and the icing humidity threshold being above 85%, the system kernel generates a hardware interrupt signal to activate the anti-icing control calculation process. The parameter calculation module 200 extracts the spanwise rotation radius of a certain discrete blade element node as 30 meters. Combined with the hub wind speed of 8 meters per second and the wind turbine rotation speed of 15 revolutions per minute, the relative wind speed amplitude is calculated to be 38.5 meters per second based on spatial vector synthesis. The synchronous calculation outputs an airflow angle of attack of 6.2 degrees. The boundary mapping module 300 inputs the airflow angle of attack of 6.2 degrees into the airfoil aerodynamic database to perform interpolation mapping, delineating the impact boundary interval of supercooled water droplets, covering a narrow area from 0.06 meters on the upper surface of the blade leading edge to 0.03 meters on the lower surface.

[0114] The mass-thermal observation module 400 retrieves environmental temperature and humidity parameters to calculate the liquid water content of the incoming air as 0.45 g / m³. It introduces the water droplet collection efficiency variable to calculate the liquid water mass flux per unit surface area and calculates the latent heat flux density of phase change required to completely evaporate the captured supercooled water droplets as 3500 W / m². The vector synthesis module 500 extracts the water film thickness value of 0.15 mm at the current time step and solves for the two-dimensional migration direction vector synthesized by the chordal aerodynamic shear stress and the spanwise rotational centrifugal force, which points towards the tip of the trailing edge of the blade. The instruction generation module 600 marks the three electrically heated grid arrays covered by the leading edge impact boundary as the main evaporation zone and assigns a base evaporation target temperature of 10°C. It locates the four adjacent downstream grids along the two-dimensional migration direction vector and marks them as dynamic flow guiding antifreeze layers. The corresponding target temperatures are asymmetrically assigned to 6°C, 3°C, and 1°C according to the spatial exponential decay function.

[0115] The heating execution module 700 extracts the target temperature and the convective heat transfer coefficient of the corresponding grid, calculates the pulse width modulation signal of the main evaporation zone grid with a duty cycle of 85%, and the duty cycles of the downstream dynamic flow guiding antifreeze layer grid decrease sequentially to 42%, 21%, and 8%. The duty cycle of the unmarked redundant area is set to 0%. The experiment enters the icing wind tunnel simulation test verification stage. A scaled-down blade prototype with two-dimensional extrusion airfoil characteristics is extracted as the experimental object. A 30-channel independently controllable electric heating grid array is arranged on the surface of the prototype and placed in the core test section of the icing wind tunnel. The experimental environment is set with a wind speed of 15 m / s and a temperature of -8℃. The external liquid water content input parameter is set to 0.6 g / m³.

[0116] Test example:

[0117] The experimental subjects were divided into a control group implementing a full-area constant power heating strategy and an experimental group connected to an intelligent anti-icing system. The wind tunnel generator was started and maintained in a continuous icing spray test state for 60 minutes. The multi-channel power analyzer read the transient electrical power of the two systems in real time and performed time integration to output the cumulative energy consumption data. The three-dimensional laser scanner performed surface topology scanning on the redundant area without heating coverage at the rear edge of the prototype every 10 minutes, and output the physical measurement value of the maximum ice ridge thickness formed by the secondary freezing of overflow water. When the test data set was output, the edge sampling points corresponding to the spatial coordinates of the wind tunnel wall were removed. The real-time operation result data of the two test prototypes were extracted. The test results are shown in the table below.

[0118] Comparison of energy consumption and secondary icing thickness between traditional uniform flow heating and dynamic anti-icing systems.

[0119]

[0120] From the table above and appendix Figure 7 and attached Figure 8 It can be seen that the cumulative energy consumption of the experimental group during the 60-minute continuous test cycle was 5.02 kWh, while the cumulative energy consumption of the control group during the same cycle was 11.08 kWh. The maximum ice ridge thickness formed in the redundant area at the trailing edge of the prototype after 60 minutes of operation in the experimental group was 1.15 mm, while the maximum ice ridge thickness formed at the same location in the control group was 21.03 mm.

[0121] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A smart anti-icing system for wind turbine blades, characterized in that, include: The data acquisition module obtains external environmental meteorological data and mechanical status data of the unit operation; The parameter calculation module is communicatively connected to the data acquisition module and calculates the relative wind speed amplitude and airflow angle of attack of the discrete leaf element nodes of the blade based on the mechanical state data. The boundary mapping module is communicatively connected to the parameter calculation module, and maps the airflow angle of attack to the physical chord coordinates of the stagnation point on the leading edge surface, thus defining the physical impact boundary range of the liquid water droplets. The mass-thermal observation module is communicatively connected to the data acquisition module and the parameter calculation module, respectively. It establishes a mass conservation equation in combination with the environmental meteorological data, calculates the current liquid water mass flux, and iteratively updates the water film thickness value on the blade surface. The vector synthesis module is communicatively connected to the parameter calculation module and the mass-thermal observation module, respectively. Based on the relative wind speed amplitude and the water film thickness, it calculates the chordal aerodynamic shear stress and the spanwise rotational centrifugal force to synthesize a two-dimensional migration direction vector. The instruction generation module generates a dynamic power allocation matrix based on the physical impact boundary interval and the two-dimensional migration direction vector, and outputs an electric power instruction in combination with the convective heat transfer coefficient. The heating execution module receives the electrical power command and drives the electric heating grid array set on the blade surface to generate heat.

2. The intelligent anti-icing system for wind turbine blades according to claim 1, characterized in that: When the parameter calculation module calculates the relative wind speed amplitude and the airflow angle of attack, it extracts the axial induction factor and circumferential induction factor when the air flows through the plane of the rotating blade, and combines the hub wind speed and the wind turbine rotation speed to solve the axial relative velocity and tangential relative velocity of the discrete blade element node. The inflow angle and relative wind speed amplitude of the blade element node are obtained by spatial vector synthesis based on the axial relative velocity and the tangential relative velocity. The airflow angle of attack is then derived and calculated by combining the cross-sectional twist angle of the discrete blade element node and the pitch angle of the pitch system.

3. The intelligent anti-icing system for wind turbine blades according to claim 1, characterized in that: When defining the physical impact boundary interval, the boundary mapping module uses the airflow angle of attack as an addressing index to input the physical chord coordinates of the leading edge surface stagnation point into the airfoil aerodynamic database; Using the stagnation point on the leading edge surface as a spatial reference, a supercooled water droplet impact trajectory calculation model is introduced to obtain the spatial motion trajectory of the water droplet in the flow field. The physical coordinate limits of the farthest upper surface and the farthest lower surface of the liquid water droplet attached to the leading edge of the blade are established. The chord length interval covered by the physical coordinate limits of the farthest upper surface and the farthest lower surface is defined as the physical impact boundary interval.

4. The intelligent anti-icing system for wind turbine blades according to claim 1, characterized in that: The mass-heat observation module calculates the liquid water content of the incoming airflow, and calculates the current liquid water mass flux by combining the relative wind speed amplitude with the water droplet collection efficiency of the corresponding leaf element node; by superimposing the latent heat released by water freezing and the latent heat absorbed by liquid water vaporization, the phase change latent heat flux density required to completely evaporate the captured supercooled water droplets is calculated.

5. The intelligent anti-icing system for wind turbine blades according to claim 4, characterized in that: When iteratively updating the water film thickness value, the mass-thermal observation module extracts the recorded value of the electric power command sent to the electric heating grid array in the previous time step, retrieves the convective heat transfer coefficient and ambient temperature of the corresponding grid in the previous time step, calculates the excess heat flux density remaining after offsetting surface convective heat dissipation, and converts the excess heat flux density into the water mass flux consumed by evaporation in the previous time step. By combining the current liquid water mass flux with the water mass flux consumed by evaporation in the previous time step, the water film thickness value is iteratively updated in the discrete time domain based on the mass conservation equation.

6. The intelligent anti-icing system for wind turbine blades according to claim 1, characterized in that: The vector synthesis module combines the external air density and surface friction coefficient to calculate the chordal aerodynamic shear stress driving the water film migration based on the relative wind speed amplitude; it introduces the liquid water density constant and the wind turbine speed to calculate the spanwise rotational centrifugal force based on the water film thickness value. By using the parallelogram rule to perform vector superposition operation on the chordal aerodynamic shear stress and the spanwise rotational centrifugal force, the two-dimensional migration direction vector characterizing the macroscopic flow trend of residual liquid water is synthesized.

7. The intelligent anti-icing system for wind turbine blades according to claim 1, characterized in that: When generating the dynamic power allocation matrix, the instruction generation module marks all the electric heating grid arrays covered by the physical impact boundary interval as the main evaporation zone in the matrix and assigns them a basic evaporation target temperature. Taking the spatial coordinates of the main evaporation zone as the origin reference, the module locates the downstream adjacent electric heating grid arrays along the direction of the two-dimensional migration direction vector and marks them as dynamic flow guiding antifreeze layers in the matrix. For the electric heating grid arrays marked as dynamic flow guiding antifreeze layers, the module makes their target temperature decrease exponentially along the direction of the two-dimensional migration direction vector.

8. The intelligent anti-icing system for wind turbine blades according to claim 7, characterized in that: The instruction generation module extracts the physical geometric center coordinates of the grid in the dynamic flow-guiding antifreeze layer as the grid centroid, and constructs a spatial exponential decay function based on the physical spatial distance between the grid centroid and the stationary point on the leading edge surface to perform asymmetric temperature assignment in order to form a spatial thermal gradient distribution.

9. The intelligent anti-icing system for wind turbine blades according to claim 7, characterized in that: The instruction generation module calculates the Reynolds number based on the relative wind speed amplitude and characteristic chord length, and introduces the Prandtl number to derive the Nusselt number to solve the convective heat transfer coefficient of the corresponding grid. The basic sensible heat flux density is calculated based on the target temperature, external ambient temperature, and convective heat transfer coefficient assigned in the dynamic power allocation matrix. The latent heat flux density of phase change calculated by the mass-thermal observation module is superimposed, and the heat transfer efficiency is integrated to calculate the electric power command required to be output by each electric heating grid array.

10. The intelligent anti-icing system for wind turbine blades according to claim 1, characterized in that: The heating execution module includes a solid-state relay group connected via an electrical control bus. The heating execution module converts the electrical power command into a pulse width modulation signal, calculates the duty cycle of the pulse width modulation signal based on the inherent maximum physical output power constant of the electric heating grid array under rated voltage, drives the solid-state relay group to perform high-frequency closing and opening actions, and adjusts the effective value of the transient current input to multiple channels of the electric heating grid array.

Images