Wind turbine blade based on streaming control, wind turbine and control method of wind turbine

Abstract

The invention discloses a wind turbine blade based on streaming control, the outer end of the wind turbine blade is a blade tip, the inner end of the wind turbine blade is a blade root fixed on a blade base, a streaming control unit is arranged on the trailing edge of the wind turbine blade in the spanwise direction, and the streaming control unit is provided with a streaming control unit base embedded in the trailing edge. An insertion hole is formed in the outer side of the streaming control unit base, a piezoelectric material is embedded in the insertion hole, and a vibration sheet is arranged at the outer end of the piezoelectric material. According to the invention, a device which vibrates by a piezoelectric material to drive air to vibrate so as to achieve the purpose of streaming control is designed and arranged on the tail edge of the wind turbine blade along the spanwise direction; according to the device, a vortex structure for generating noise radiation is destroyed at a higher frequency; the energy block radiated by eddy current noise is reduced, and the eddy current noise is quickly dissipated in the radiation process; and thus, the sound feeling of the receiver is reduced.

Description

Technical Field

[0001] This invention relates to the field of wind power generation technology. Background Technology

[0002] Wind energy, as a renewable and clean energy source, has been widely used in recent decades and is gradually becoming a mainstream alternative energy source. The horizontal axis three-blade wind turbine (hereinafter referred to as wind turbine) is the main equipment solution for utilizing wind energy. With the technological evolution towards larger scale, its power generation cost has become lower than that of coal-fired power generation, making it a highly competitive power source.

[0003] Noise is a significant factor affecting the installation and application of wind turbines, and the aerodynamic noise of their blades is the primary noise source. Blade aerodynamic noise is mainly generated by eddies in the trailing edge region of the blades. This type of noise exhibits broadband characteristics in the frequency domain, and is characterized by strong penetration, slow attenuation, and a large radiation range.

[0004] Large, commercially available wind turbine blades typically feature a serrated trailing edge on their main output section. Related research, testing, and practical applications have confirmed that serrated trailing edges can reduce broadband trailing edge noise without affecting the airfoil's aerodynamic performance. The basic principle is to break down the larger vortex structure at the trailing edge into smaller vortex structures, utilizing the dissipation of these vortices; essentially, it breaks down large noise radiation energy clusters into smaller structures, causing them to attenuate rapidly and thus reducing the intensity and distance of noise radiation.

[0005] Patent CN201620945792.4 discloses a gourd-shaped flat noise reduction component, which is similar to a serrated shape fixed to the trailing edge of a blade to achieve the purpose of reducing noise;

[0006] Patent CN202280033950.7 discloses a design for a cavity formed by the opening of the blade to reduce aerodynamic noise, thereby enabling wind turbines to have a wider range of applications and higher power generation.

[0007] Patent CN201410321398.9 discloses a noise reduction eddy current generator. By specially designing the shape and installation method of the eddy current generator, the momentum of the relative motion of the free flow is introduced into the boundary layer around the flow to achieve the purpose of noise reduction.

[0008] In summary, all of these designs are passive, fixed noise reduction structures. The problem with this design is that once installed and fixed, the noise reduction structure's impact on the flow pattern is fixed, meaning it is only effective against specific eddies. As a result, its noise reduction effect is significant at specific wind speeds, but poor at other wind speeds. Summary of the Invention

[0009] The technical problem to be solved by the present invention is to realize a flexible, simple, and actively controlled wind turbine noise control device.

[0010] To achieve the above objectives, the technical solution adopted by the present invention is as follows: a wind turbine blade based on flow control, wherein the outer end of the wind turbine blade is the blade tip, the inner end is the blade root fixed on the blade base, a flow control unit is arranged in the spanwise direction of the trailing edge of the wind turbine blade, the flow control unit is provided with a flow control unit base embedded in the trailing edge, the outer side of the flow control unit base is provided with a socket, a piezoelectric material is embedded in the socket, and a vibrating sheet is provided at the outer end of the piezoelectric material.

[0011] The flow control unit is continuously fixed to the trailing edge, forming a linear structure along the spanwise direction of the trailing edge. Each independent flow control unit is inserted into the groove of the trailing edge of the blade. Each flow control unit has the same width. The tangential direction of the suction surface of the blade element airfoil trailing edge at the center of the spanwise dimension of the wind turbine blade is the extension direction of the flow control unit. Since the airfoil profile is different at different positions of the wind turbine blade, the extension direction of the flow control unit is adjusted according to the position of the wind turbine blade.

[0012] The trailing edge of the wind turbine blade is blunt, and the length of the flow control unit extending beyond the trailing edge of the wind turbine blade is [missing information]. The length of the vibrating sheet extending beyond the piezoelectric material is The thickness of the trailing edge is The thickness of the trailing edge of the wind turbine blade gradually decreases from the blade root to the blade tip;

[0013] in: , .

[0014] The vibrating sheet is a reinforced acetate sheet with a width of l and a height of [missing information]. The vibration frequency control range of the vibrating sheet covers The gaps between the flow control units are all the same, g, where .

[0015] Each flow control unit base is connected to the wind turbine blade via wires. The flow control unit base is electrically connected to the piezoelectric material via contact electrodes. The blade portion extending from the blade root to the blade tip is sequentially divided into three base frequency blocks: base frequency block 1, base frequency block 2, and base frequency block 3. Each base frequency block of the wind turbine blade has an independent inverter circuit breaker and a base frequency block signal processor. The flow control unit base of each base frequency block is connected to the inverter circuit breaker and the base frequency block signal processor of its respective base frequency block via wires. Both the inverter circuit breaker and the base frequency block signal processor are fixed at the blade root position.

[0016] A wind turbine based on flow-around control includes a blade base, a drive shaft connected to the blade base, and a generator connected to the drive shaft. The wind turbine blades are fixed on the base, and an acoustic radiation sensor is provided inside the blade base.

[0017] A control method based on flow bypass control includes the following steps:

[0018] Step 1: Collect and process the wind turbine status information and record it as a status set. As inputs to the control system, the fan power and fan speed are collected. Information, and calculate the time step. ;

[0019] Step 2: Retrieve the dynamic state database and compare it with the input state set;

[0020] Step 3: If the state set successfully matches the dynamic database, retrieve the matching inverter control parameters from the GBF-CMAC function generator. ,exist During the specified time period, it serves as the control signal for the inverter to control the winding unit;

[0021] Step 4: If the state set does not successfully match the dynamic database, then... Divide into two time periods; in During the specified time period, the frequency converter is stopped; and signals from the overall machine's flow control unit are collected. , And enlarged. ;exist During the time period, the signal Input the frequency converter and execute the working action; at the same time, Input GBF-CMAC trainer;

[0022] Step 5: Using the publicly available Gaussian function, the cerebellar neural network is divided into two layers to construct the GBF-CMAC trainer;

[0023] Step 6: Using the GBF-CMAC trainer obtained in Step 5, with the training method described in Step 4... The value is used as the initial value; the control variable of the whole machine flow control unit is used as the initial value. As an optimization of the search space, the publicly available particle swarm optimization algorithm is used to optimize the optimal controllable parameters of the flow-around unit under the current overall machine state and environmental conditions: ,in

[0024] The objective is:

[0025] constraint: ,

[0026] Step 7: Store the optimized control parameters and corresponding environmental states in the memory for retrieval and use.

[0027] AC frequency within a single baseband block ,Voltage Current The driving frequency of the piezoelectric material vibration driven by the flow control unit extending from the blade root to the blade tip gradually increases, and the change of the driving frequency is achieved by changing the single-phase alternating frequency converter circuit.

[0028] The driving frequencies of adjacent flow control units differ by a certain amount. ,in The number of flow control units within a block; the three base frequency block control signals on a single wind turbine blade are:

[0029]

[0030] The control signals of the entire machine are expressed in matrix form as follows:

[0031]

[0032] in This is the control signal for the second blade. For the control signal of the third blade

[0033] , .

[0034] When the inverter is not working, the vibrating plate of the flow control unit is subjected to the trailing edge eddy current, which produces vibration characteristics consistent with the evolution of the trailing eddy structure. This drives the piezoelectric material to vibrate and generate alternating current, making the flow control unit a sensor for the evolution of the trailing edge flow eddy structure of the blade.

[0035] The fundamental frequency block receives the signal set of alternating current.

[0036]

[0037]

[0038]

[0039] in,

[0040] For blade 1 fundamental frequency block The set of control unit frequencies, in Hz;

[0041] For blade 1 fundamental frequency block The set of control unit voltage amplitudes, in V;

[0042] For blade 1 fundamental frequency block The set of control unit current amplitudes, in A;

[0043] This refers to the number of units controlled by this baseband block;

[0044] Computational Data Set , , The confidence intervals of frequency, voltage amplitude, and current amplitude are obtained, and their average values ​​are taken to obtain the statistical average of the output signal of the fundamental frequency block's flow control unit.

[0045]

[0046]

[0047]

[0048] The fundamental frequency block output signal of the first blade:

[0049] ,

[0050] The fundamental frequency block output signal of the second blade:

[0051] ,

[0052] The fundamental frequency block output signal of the third blade:

[0053] ,

[0054] The output signals of the wind turbine are:

[0055] ,

[0056] Output signal , , For a single flow control unit; , , It is a set of statistical means of fundamental frequency block signals;

[0057] Output signal , , As a discrete quantity, it serves as the input value for constructing the GBF-CMAC trainer. , , As a control signal for controlling the frequency converter.

[0058] Step 5 includes the following steps:

[0059] Step 5.1: The first layer is divided into three lobes with separate Gaussian functions and cerebellar neural networks for parallel processing, and the network structure is consistent.

[0060] First blade ,Will The variables are reorganized and written as follows: Input the number of dimensions The set of discrete quantities; output Dimensions Lenovo Space Layers State parameters corresponding to a single blade Number of blocks per layer in Lenovo Space Quantification level Construct the GBF-CMAC network structure for blade 1, where , Determined by the following formula;

[0061]

[0062]

[0063]

[0064] By analogy, the Gaussian functions of the second and third lobes are obtained into the cerebellar neural network.

[0065] Step 5.2 The second layer expresses the influence of the overall machine state and environmental state on the flow control unit;

[0066] Overall status Environmental conditions The second-layer GBF-CMAC Lenovo space is divided into 5 layers corresponding to the overall system status and environmental status; the quantification level is also taken as 1. The input is a three-bladed GBF-CMAC network. The output is... , , Two output power Unit [kW], sound pressure level [dBa];

[0067] Performance metrics:

[0068]

[0069]

[0070] in, , for The values ​​of the total power and noise radiation level of the inverter during actual operation over the specified time period; based on the above performance indicators... , The GBF-CMAC network training is completed by using gradient descent algorithm to correct the weight coefficients and train and adjust the weights of the GBF-CMAC trainer.

[0071] The dimensions of the flow control unit in this invention are designed and implemented based on the thickness of the airfoil trailing edge at different positions along the blade span. The flow control unit is arranged along the trailing edge of the blade span. The dimensions and arrangement angle of the flow control unit are determined according to the airfoil of the blade element section along the span, adapting to the three-dimensional flow physical characteristics of the wind turbine. According to the physical laws of air flow around the wind turbine, the flow control unit is arranged along the span and controlled in blocks. At the same time, the alternating current generated by the alternating force of piezoelectric materials is utilized to autonomously learn the flow control strategy according to the wind turbine installation environment and the wind turbine's operating state.

[0072] A device is designed and arranged along the spanwise direction at the trailing edge of a wind turbine blade, using piezoelectric material to vibrate and drive air vibration to achieve flow control. This device disrupts the eddy current structure that generates noise radiation at a higher frequency, reducing the energy of the eddy current noise radiation and dissipating it rapidly during the radiation process, thereby reducing the receiver's perception of the sound. Attached Figure Description

[0073] The following is a brief explanation of the content represented by each figure in this specification:

[0074] Figure 1 Arrangement of blades and flow control devices;

[0075] Figure 2(a) shows a three-dimensional magnification of part A;

[0076] Figure 2(b) is a magnified three-dimensional front view of part A in the y-direction;

[0077] Figure 3(a) shows the airfoil profile of section B;

[0078] Figure 3(b) shows a partial installation diagram of the flow control unit;

[0079] Figure 3(c) shows a partial installation breakdown of the flow control unit;

[0080] Figure 3(d) is a magnified view of part D along the z-axis.

[0081] Figure 4 Enlarged layout of the flow control unit spanwise;

[0082] Figure 5(a) shows the circuit connection of the active control working circuit of the fundamental frequency block flow control unit;

[0083] Figure 5(b) shows a schematic diagram of the active control circuit connection of a single blade flow control unit;

[0084] Figure 6 Layout intention of the frequency converter for the flow control unit around the wind turbine blades;

[0085] Figure 7(a) shows the connection circuit for generating alternating current in the baseband block control unit;

[0086] Figure 7(b) shows the non-active control circuit connection of a single blade flow control unit;

[0087] Figure 8(a) shows the adaptive noise control system for the fan;

[0088] Figure 8(b) shows the structure of the GBF-CMAC trainer;

[0089] Figure 8(c) shows the GBF-CMAC(B_1S) network structure. Detailed Implementation

[0090] The following description, with reference to the accompanying drawings, details the specific implementation of the present invention, including the shape and structure of each component, the relative positions and connections between the parts, the function and working principle of each part, the manufacturing process, and the operation and use methods, to help those skilled in the art to have a more complete, accurate, and in-depth understanding of the inventive concept and technical solution of the present invention.

[0091] The size of the flow control unit is designed and implemented based on the thickness of the airfoil trailing edge at different positions along the blade span; the flow control unit is arranged along the trailing edge of the blade span; the size and arrangement angle of the flow control unit are determined according to the airfoil of the blade spanwise element section, adapting to the three-dimensional flow physical characteristics of the wind turbine.

[0092] The present invention uses piezoelectric material to drive the normal vibration of a vibrating sheet at the trailing edge of a wind turbine blade, injecting energy into the flow structure around the blade; the principle is to utilize the reverse pressure characteristics of the piezoelectric material and use an AC point of a certain frequency to make the piezoelectric material vibrate, which drives the vibrating sheet fixed on it to vibrate normally, and to generate a disturbance to the vibrating sheet. The frequency and amplitude of the disturbance are controlled by the AC current received by the piezoelectric material, as shown in Figure 3(c).

[0093] Specifically, the flow control unit is composed of a vibrating sheet, piezoelectric material, and AC-AC frequency converter circuit carrier, as shown in Figure 3(b);

[0094] The flow control unit is fixed on the flow control unit base. The flow control unit is pluggable on the flow control unit base and is connected to the flow control unit base through electrical conductor terminals to form a current path. It should be noted that the pluggable design of the flow control unit facilitates operation and maintenance.

[0095] The flow control unit base is embedded in the trailing edge of the blade body, receiving and feeding back electrical signals, as shown in Figure 3(c). Due to manufacturing limitations, the trailing edge of the blade body is blunt; the direction from the blade root to the blade tip is defined as the spanwise direction, as shown in Figure 3(c). Figure 1 As shown in Figure 2(a) and (b), the control unit is arranged along the blade span. To ensure the effectiveness of the control unit in controlling the airflow around the blade, it needs to extend a certain length beyond the blade body. And the vibrating sheet extends to the specified size. As shown in Figure 3(d);

[0096] The position of the designed control unit in the blade spanwise direction The trailing edge thickness located at the spanwise center of the flow control unit (Unit: [mm]) is a dimensional reference length, as shown in Figure 3(c);

[0097] The recommended extension length of the flow control unit in this invention is:

[0098]

[0099] The recommended extension dimension of the vibrating sheet in this invention is:

[0100]

[0101] The control unit of the present invention extends to the recommended length dimension. and the extension size of the vibration plate of the control unit The method for determining the dimensions is one of the methods for dimension design;

[0102] The thickness of the trailing edge of a wind turbine gradually decreases from the blade root to the blade tip.

[0103] Depending on the location of the flow control unit along the blade span and the thickness of the blade trailing edge, the dimensions of the control unit will vary.

[0104] The flow control unit has the same width dimension in the blade spanwise direction, such as... Figure 4 As shown;

[0105] Width dimension l, height dimension Based on the material properties used in the control unit, the vibration frequency control range of the control unit is ensured to cover: The reinforced acetate sheet selected in this invention; wherein This is the main frequency range for blade trailing edge noise; the design clearance between the flow control units in the blade spanwise direction is:

[0106]

[0107] The objective of gap control between flow control units is to reduce the safety impact and aerodynamic interference between them; this recommended value meets the requirements of this patent; based on the position of the designed flow control unit in the blade spanwise direction. The tangential direction (direction angle alpha) of the blade element airfoil trailing edge suction surface at the center of the spanwise dimension of the flow control unit is used as the extension direction of the flow control unit extending out of the blade body, as shown in Figures 3(a) and (d).

[0108] Different airfoil profiles are typically used at different positions along the blade span, therefore the direction in which the flow control unit extends beyond the blade body is different when it is arranged. The airfoil profiles used at different positions along the blade span are not abrupt, but gradually change.

[0109] Based on the physical laws of airflow around a wind turbine, a flow control unit is arranged longitudinally and controlled in blocks; at the same time, the characteristics of piezoelectric materials generating alternating current under alternating force are utilized to act as a sensor for the evolution structure of the flow vortex at the trailing edge of the blade, serving as input control data for intelligent control.

[0110] Individual blade control units are arranged in several blocks along the spanwise direction for control, such as... Figure 1 As shown, this patent is divided into three control blocks, defined as baseband block 1, baseband block 2, and baseband block 3;

[0111] AC frequency within a single baseband block (Unit: Hz) Voltage (Unit: V) Current (Unit: A), controlled by an independent frequency converter; the frequency converter is placed in the cavity at the root of the blade; it is connected to the control unit of the base frequency block through a guide cable arranged in the cavity inside the blade body;

[0112] The control unit in the base frequency block is connected in parallel to the flow guide cable; the schematic connection circuit is shown in Figure 5(a); the width dimension of the flow control unit is a small amount relative to the spanwise length dimension of the blade, and a large number of them need to be arranged in the spanwise direction. The block division avoids the result of a huge number of wire harnesses for controlling the flow structure individually.

[0113] Within a single block, the driving frequencies of the piezoelectric materials driven by adjacent flow control units along the blade span differ by a certain value, and gradually increase along the span; within the fundamental frequency block 1 shown in Figure 5(a) of this patent, the driving frequencies of the flow control units differ by a certain value. (Unit: Hz), and the frequency value increases along the blade spanwise, among which This refers to the number of flow control units within the block; the driving frequency of the flow control units is changed through a commonly used single-phase AC-AC converter circuit.

[0114] The linear velocity of the blade rotating along the span increases gradually, and the eddy current detachment frequency that generates noise also gradually increases along the span. Therefore, the drive frequency variation of the flow control unit at different positions along the blade span in this patent is designed for this purpose. The number of frequency converters for a single blade corresponds to the number of base frequency blocks.

[0115] This patent divides a single blade into three base frequency blocks, which are controlled by three frequency converters. The control diagram is shown in Figure 5(b). The control signal for blade 1 is:

[0116]

[0117] It should be noted that different numbers of fundamental frequency blocks can be divided as needed, and controlled by a corresponding number of frequency converters; for the flow control around the wind turbine blades, a group of frequency converters is used to independently control the fundamental frequency block of the flow control unit for each blade, such as... Figure 6 As shown; the azimuth angle of the blades on the plane of rotation varies depending on the working state of the fan; for the sake of space and core points, the specific installation structure of the blade root cavity and the blade body cavity is not shown in the figure in this patent. Figure 6 The diagram illustrates the arrangement of the frequency converter in the flow control unit around the wind turbine blades; the control signals of the entire machine are expressed in matrix form as follows:

[0118]

[0119] in For the control signal of blade 2, Control signal for blade 3

[0120] ,

[0121] A base frequency block signal processing unit is designed at the inverter. When the inverter is not working, the connection circuit is shown in Figure 7(a). The base frequency block signal processing unit processes the alternating current signal. When the inverter is not working, the vibrating sheet of the flow control unit is subjected to the trailing edge eddy current, which generates vibration characteristics consistent with the evolution of the trailing eddy structure. This drives the piezoelectric material to vibrate and generate alternating current. In this case, the flow control unit acts as a sensor for the evolution structure of the trailing edge flow eddy current of the blade.

[0122] The fundamental frequency block signal processing unit uses a publicly available wavelet analysis algorithm to analyze and obtain the signal set of the alternating current.

[0123]

[0124]

[0125]

[0126] in,

[0127] For blade 1 fundamental frequency block The set of control unit frequencies, in Hz;

[0128] For blade 1 fundamental frequency block The set of control unit voltage amplitudes, in V;

[0129] For blade 1 fundamental frequency block The set of control unit current amplitudes, in A;

[0130] This refers to the number of units controlled by this baseband block;

[0131] Specifically, at a 95% confidence level ( Using a Gaussian distribution, statistical calculations are performed on the dataset. , , The confidence intervals of frequency, voltage amplitude, and current amplitude are obtained, and their average values ​​are taken to obtain the statistical average of the output signal of the fundamental frequency block's flow control unit.

[0132]

[0133]

[0134]

[0135] Specifically, the fundamental frequency block output signal of blade 1,

[0136] ,

[0137] The output signals of the wind turbine are:

[0138] ,

[0139] in, , This is the output signal for blade 2. , For the output signal of blade 3

[0140] ,

[0141] ,

[0142] It should be noted that the output signal For a single flow control unit; It is a set of statistical means of fundamental frequency block signals; As discrete quantities, they serve as input values ​​for building the GBF-CMAC trainer, thus avoiding the quantization mapping process; As a control signal for controlling the frequency converter.

[0143] Based on the main factors affecting wind turbine noise radiation, this invention uses a cerebellum neural network (Gaussian function) to design an adaptive noise control system for wind turbines that employs an autonomous learning flow-around control strategy to control noise radiation sources without altering the existing control logic of the wind turbine.

[0144] This invention involves installing an acoustic radiation sensor at the center of the front of the nacelle canopy; the sensor meets the requirements for acoustic measurement instruments in GB / 22516-2015; and its output measurement signal outputs a Class A equivalent sound pressure level signal according to the GB / 22516-2015 standard. The unit is [Hz].

[0145] Specifically, this invention designs a fan adaptive noise control system logic based on flow around control, as shown in Figure 8, and its specific operation is as follows:

[0146] Step 1: Collect and process the wind turbine status information and record it as a status set. As inputs to the control system, as shown in Table 1; fan power and fan speed are collected. (Unit: [rpm]) information, and calculate the time step. ;

[0147] Table 1

[0148]

[0149] Step 2: Retrieve the dynamic state database and compare it with the input state set;

[0150] Step 3: If the state set successfully matches the dynamic database, retrieve the matching inverter control parameters from the GBF-CMAC function generator. ,exist During the specified time period, it serves as the control signal for the inverter to control the winding unit;

[0151] Step 4: If the state set does not successfully match the dynamic database, then... Divide into two time periods; in During the specified time period, the frequency converter is stopped; and signals from the overall machine's flow control unit are collected. , And enlarged. ;exist During the time period, the signal Input the frequency converter and execute the working action; at the same time, Input GBF-CMAC trainer;

[0152] Step 5: GBF-CMAC trainer design, using the publicly available Gaussian basis function cerebellar neural network (GBF-CMAC, cerebellar model articulation controller with Gaussian basis function), which has a two-layer structure;

[0153] Step 5.1 The first layer is divided into three lobes with separate Gaussian functions and parallel processing by the cerebellar neural network; and the network structure is consistent.

[0154] Specifically, taking blade 1 as an example, ,Will The variables are reorganized and written as follows: Input the number of dimensions The set of discrete quantities; output Dimensions Lenovo Space Layers State parameters corresponding to a single blade Number of blocks per layer in Lenovo Space Quantification level As shown in Figure 8(c), the GBF-CMAC network structure of blade 1 is constructed, where , Determined by the following formula;

[0155]

[0156]

[0157]

[0158] Step 5.2 The second layer expresses the influence of the overall machine state and environmental state on the flow control unit; overall machine state Environmental conditions As shown in Table 2, the corresponding status descriptions are used; the second-layer GBF-CMAC Lenovo space is divided into 5 layers corresponding to the overall system status and environmental status; the quantization level is also taken as . The input is a three-bladed GBF-CMAC network. The output is... , , Two output power Unit [kW], sound pressure level [dBa];

[0159] Performance metrics:

[0160]

[0161]

[0162] in, , for The values ​​of the total power and noise radiation level of the inverter during actual operation over the specified time period; based on the above performance indicators... , The gradient descent algorithm is used to correct the weight coefficients, train and adjust the weights of the GBF-CMAC trainer, and complete the training of the GBF-CMAC network.

[0163] Step 6: The GBF-CMAC optimizer uses the GBF-CMAC network output in Step 5 as the target output system; based on the data from Step 4... The value is used as the initial value; the control variables of the whole machine flow control unit (3 blades * 3 fundamental frequency blocks * 3 current parameters, a total of 27 variables) are used in... As an optimization of the search space, the publicly available particle swarm optimization algorithm is used to optimize the optimal controllable parameters of the flow-around unit under the current overall machine state and environmental conditions: ,in

[0164] The objective is:

[0165] constraint: ,

[0166] Optimization logic: The control variables of the flow control unit change within a certain search range to find the control parameters with the minimum noise radiation, while ensuring that the output power of the whole machine does not decrease;

[0167] It should be noted that the particle swarm optimization algorithm is one of the optimization algorithms selected in this patent, and there are other better algorithms.

[0168] Step 7: Store the optimized control parameters and corresponding environmental states in the memory for retrieval and use;

[0169] This completes the self-learning process of noise control for the wind turbine bypass unit.

[0170] Table 2: State Variable Table of Fan Unit

[0171]

[0172] At the trailing edge of wind turbine blades, a device is designed and arranged along the spanwise direction using piezoelectric material to vibrate and drive air vibration, achieving flow control. This disrupts the vortex structure that generates noise radiation at a higher frequency, reducing the energy of the vortex noise radiation and causing it to dissipate rapidly during radiation; thus reducing the perceived sound for the receiver. (Aerodynamic noise, which cannot be eliminated, is addressed by actively dividing the energy of the aerodynamic noise source into smaller noise radiation particles, causing them to dissipate rapidly in the air.)

[0173] This invention presents an active noise reduction device based on flow-around control. The size of the flow-around control unit is designed and implemented according to the thickness of the airfoil trailing edge at different positions along the blade span. The flow-around control unit is arranged along the trailing edge of the blade span. The size and arrangement angle of the flow-around control unit are determined according to the airfoil of the blade's spanwise blade element section, adapting to the three-dimensional flow physics characteristics of the wind turbine and providing noise reduction across a wide range of applicable wind speeds. Furthermore, it innovatively employs an adaptive learning control strategy, autonomously learning noise reduction control for different wind turbine installation environments.

[0174] The present invention has been described above by way of example with reference to the accompanying drawings. Obviously, the specific implementation of the present invention is not limited to the above-described manner. Any non-substantial improvements made using the inventive concept and technical solution of the present invention, or the direct application of the inventive concept and technical solution of the present invention to other occasions without modification, are all within the protection scope of the present invention.

Claims

1. A wind turbine blade based on flow control, the wind turbine blade having an outer end being a tip and an inner end being a root fixed to a blade base, characterized in that: The trailing edge of the wind turbine blade is arranged with a flow control unit, the flow control unit is embedded in a flow control unit base embedded in the trailing edge, the outer side of the flow control unit base is provided with a socket, the socket is embedded with piezoelectric material, and the outer end of the piezoelectric material is provided with a vibrating sheet.

2. A wind turbine blade based on flow control according to claim 1, characterised in that: The flow control unit is continuously fixed on the trailing edge and forms a linear structure along the trailing edge, each independent flow control unit is inserted and matched with the groove of the trailing edge of the blade, the width of each flow control unit is the same, the tangential direction of the suction surface of the trailing edge of the blade element airfoil at the center position of the spanwise dimension of the wind turbine blade is the extension direction of the flow control unit, and the airfoil profile at different positions of the wind turbine blade is different, so the extension direction of the flow control unit is adjusted according to the position of the wind turbine blade.

3. A wind turbine blade based on flow control according to claim 1 or 2, characterised in that: The trailing edge of the wind turbine blade is a blunt body, the length of the flow control unit extending from the trailing edge of the wind turbine blade is L c2-1 , the length of the vibrating sheet extending from the piezoelectric material is L s2-1 , the thickness of the trailing edge is d 2-1 , the thickness of the trailing edge of the wind turbine blade gradually decreases from the root to the tip where: L c2-1 = 2.5 x d 2-1 , L s2-1 = 0.5 x L c2-1 .

4. A wind turbine blade based on flow control according to claim 3, characterised in that: The vibrating sheet is a reinforced acetate sheet with a width dimension of l and a height dimension of d f The vibrating sheet has a vibrating frequency control range of 750-2000 Hz, and the gap between the flow control units is g, where g = 1 x 4%.

5. A wind turbine blade based on flow control according to claim 1 or 4, characterised in that: The flow control unit base is connected by a wire in the wind turbine blade, the flow control unit base is electrically connected with the piezoelectric material through a contact electrode, the blade part extending from the blade root to the blade tip is sequentially a fundamental frequency block 1, a fundamental frequency block 2 and a fundamental frequency block 3, each fundamental frequency block of the wind turbine blade has an independent frequency converter action circuit breaker and a fundamental frequency block signal processor, the flow control unit base of each fundamental frequency block is connected with the frequency converter action circuit breaker and the fundamental frequency block signal processor of the fundamental frequency block through a wire, and the frequency converter action circuit breaker and the fundamental frequency block signal processor are fixed at the blade root position.

6. A flow control based windmill comprising a blade base, a transmission shaft connected to the blade base, and a generator connected to the transmission shaft, characterized in that: The base is fixed with the wind turbine blade as claimed in any one of claims 1-5, and the blade base is provided with an acoustic radiation sensor.

7. A control method based on flow control, characterized by, The method comprises the following steps: Step 1, collect and process the fan state information, record as state set Z: {E S , T S , B 1S , B 2S , B 3S}, as the control system input, collect the fan power collection fan speed Ω information, and calculate the time step Δt; Step 2, searching a state dynamic database and comparing with the input state set; Step 3, if the state set successfully matches the dynamic database, retrieve the matching inverter control parameters from the GBF-CMAC function generator as the control signal of the inverter control flow unit in the duration of Δt; Step 4, if the state set and the dynamic database are not successfully matched, Δt is divided into two time periods; in the 0-Δt time period, the frequency converter is stopped working; Collect the signal of the whole machine flow control unit And do amplification processing S'=1.1×S, S O′ =1.1×S O ; in the time period of Δt / 2~Δt, input the signal S O′ to the frequency converter, and execute the working action; at the same time, input S' to the GBF-CMAC trainer; Step 5, using a public Gaussian basis function cerebellar neural network divided into two layers to construct a GBF-CMAC trainer; Step 6, use the GBF-CMAC trainer derived from Step 5 to train the S O′ values as initial values; the control variable of the whole aircraft flow control unit is between 0.9*S O′ ~ 1.5*S O′ As an optimization search space, the published particle swarm algorithm is used to optimize the best flow control unit in the current whole aircraft state, environmental state, and the most control parameters: b c , where Objective: min(L Aeq (t)) Constraint: P ≥ P(t), f, U, I ∈ [0.9 * S O′ ~ 1.5 * S O′ ] Step 7, storing the optimized control parameters and the corresponding environmental state in a memory storage, and preparing for searching and using.

8. The control method based on flow control according to claim 7, characterized in that: The alternating current frequency f, voltage U and current I in a single fundamental frequency block, the driving frequency of the piezoelectric material vibration driven by the flow control unit extending from the blade root to the blade tip gradually increases, and the change of the driving frequency is realized by a single alternating frequency conversion circuit; The driving frequency of adjacent flow control units is different by Δf1, wherein n1 is the number of flow control units in the block, and the control signals of the three fundamental frequency blocks on a single wind turbine blade are: The control signals of the whole machine are expressed in the form of a matrix as follows: wherein is a control signal for the second blade, is a control signal for the third blade 9. The control method based on flow control according to claim 8, characterized in that: When the frequency converter does not work, the vibrating sheet of the flow control unit is affected by the trailing edge vortex, and generates vibration characteristics consistent with the evolution of the trailing edge vortex structure, which drives the piezoelectric material to vibrate to generate alternating current, so that the flow control unit acts as a sensor of the trailing edge flow vortex evolution structure of the blade; The signal set of the alternating current obtained by the fundamental frequency block is f 1-I : {f0, f1, f2,..., f n1} U 1-I : {U0, U1, U2,..., UN-1} n1} I 1-I : {I0, I1, I2,..., IN-1} n1} Wherein, f 1-I is the set of frequencies of the control unit of the fundamental block I of the blade 1, in Hz; U 1-I Vset is the set of control unit voltage amplitudes for the fundamental block I of the blade 1, in V. I 1-I is the set of current amplitudes of the control unit of the fundamental block I of the blade 1, in A; n1 is the number of units controlled by the fundamental frequency block; The calculation data set f 1-I , U 1-I , I 1-I Get the confidence interval of the frequency, voltage amplitude, current amplitude, and take the mean value to get the statistical mean value of the flow control unit output signal of the fundamental frequency block: The fundamental frequency block output signal of the first blade is: The fundamental frequency block output signal of the second blade is: The fundamental frequency block output signal of the third blade is: The output signal of the whole wind turbine is: The output signals S1, S2, S3 are a signal set of a single flow control unit; The mean value set is calculated for the base frequency block signals. The output signals S1, S2, S3 are discrete quantities, which are used as input values for constructing a GBF-CMAC trainer, As a control signal for controlling the frequency converter.

10. The control method based on flow control according to claim 7, 8 or 9, characterized in that: The step 5 comprises the following steps Step 5.1, the first layer is divided into three leaflets and is separately processed by a Gaussian basis function cerebellar neural network, and the network structure is consistent; First leaf The variables in S'1 are reorganized and written as: Input dimension number h i Discrete quantity set with = 9; output O 11 Dimension 1; number of layers h of associative space j = 4, corresponding to the state parameters of a single leaf Number of blocks h of each layer of associative space k = 3, quantization level Q1 = 7, construct the GBF-CMAC network structure of leaf 1, wherein g ijk , a jk Determined as follows; g ijk ∈(0,1] Similarly, the second leaflet and the third leaflet Gaussian basis function cerebellar neural network are obtained; Step 5.2, the second layer expresses the influence of the whole machine state and the environment state on the control of the flow control unit; Whole machine state T S , Environment state E S , The second layer GBF-CMAC association space is divided into 5 layers corresponding to the whole machine state and the environment state; The quantization level is also taken as Q2=7; The input is three leaf GBF-CMAC network output O 11 , O 12 , O 13 ; Two output quantities power P, unit [kw], sound pressure level L Aeq [dBa] Performance index: wherein, L Aeq (t), P(t) are the values of the actual working power and noise radiation level of the whole machine corresponding to the frequency converter in the time period of Δt / 2~Δt; through the above performance indicators E1∈[0, 10], E2∈[0, 0.5], using the correction of the weight coefficient, the gradient descent algorithm is used to train and adjust the GBF-CMAC trainer weight, and the GBF-CMAC network training is completed.

Images