Active clearance control method and system for wind turbine generator

Abstract

The invention discloses an active clearance control method and system for a wind turbine generator, and relates to the field of clearance control, and the method comprises the steps: obtaining a sound source signal, wind speed data and rotating speed data of fan blades, and determining the position of a sound source based on the sound source signal; determining a clearance distance based on the sound source position; when the wind speed data and the rotating speed data of the fan blades are both larger than a set threshold value, and the clearance distance is smaller than or equal to a set early warning threshold value, control parameters are optimized through an improved tribe competition and member cooperation algorithm, and the optimized control parameters are obtained; and variable pitch operation control is conducted based on the optimized control parameters, so that active control over the clearance distance is achieved. According to the method, the problem of lack of a corresponding effective active safety control strategy can be solved, and meanwhile, the problems of high cost, low precision, poor environmental adaptability, non-real-time performance and the like of a current wind turbine generator clearance detection mode are solved.

Description

Technical Field

[0001] The present application relates to the field of clearance control, and in particular to a method and system for active clearance control of a wind turbine generator set. Background Art

[0002] As wind turbine capacity increases, especially in low-wind-speed regions, tall towers and long blades have become standard for low-wind-speed wind turbines (wind turbines, also known as wind turbines) to capture more wind energy. This has led to increased blade flexibility. This, in turn, increases the blade's deflection during operation, increasing the probability of collision with the tower.

[0003] When wind speeds are too high or operating power is too high, blades can strike the tower, potentially causing damage or even tower collapse. To prevent this, clearance distance detection is necessary. Currently, clearance distance detection is typically performed using video, Bluetooth, or millimeter-wave radar. However, video detection is subject to weather and time constraints and cannot function consistently. Millimeter-wave radar is expensive and has high detection costs. Bluetooth is affected by the complex electromagnetic environment of wind farms, making detection accuracy and safety unreliable.

[0004] In addition, existing clearance distance detection methods mostly focus on real-time monitoring and early warning of clearance distance, and lack relevant active safety control technology research and development. Although millimeter-wave radar can achieve good detection effects, its core function is still at the risk discovery stage, and has not yet been deeply coupled with the unit's main control system to achieve dynamic regulation. Summary of the Invention

[0005] The purpose of this application is to provide a wind turbine active clearance control method and system, which can overcome the lack of corresponding effective active safety control strategies while overcoming the problems of high cost, low precision, poor environmental adaptability and non-real-time performance of current wind turbine clearance detection methods.

[0006] To achieve the above objectives, this application provides the following solutions:

[0007] In a first aspect, the present application provides a method for active clearance control of a wind turbine generator set, comprising:

[0008] Acquire a sound source signal, wind speed data, and rotation speed data of a fan blade, and determine the sound source position based on the sound source signal; the sound source signal is a sound signal of a set point at the tip position of the fan blade;

[0009] determining a clearance distance based on the sound source location;

[0010] When the wind speed data and the rotation speed data of the wind turbine blades are both greater than a set threshold value, and the clearance distance is less than or equal to a set warning threshold value, the control parameters are optimized by using an improved tribe competition and member cooperation algorithm to obtain optimized control parameters;

[0011] Pitch operation control is performed based on the optimized control parameters to achieve active control of the clearance distance.

[0012] Optionally, an improved tribe competition and member cooperation algorithm is obtained by dynamically adjusting the inertia weight and cognitive factor of the tribe competition and member cooperation algorithm; the cognitive factor includes an individual cognitive factor and a tribe cognitive factor.

[0013] Optionally, the process of optimizing the control parameters by using the improved tribe competition and member cooperation algorithm includes:

[0014] Initializing the parameters of the improved tribe competition and member cooperation algorithm to obtain initialization parameters; the initialization parameters include: the number of tribes, the number of members of each tribe and the velocity vector of each member, the maximum number of iterations, loyalty, initial inertia weight and initial cognitive factor; using the control parameters to be optimized as the members;

[0015] Based on the initialization parameters and the position fitness value of each member, the velocity vector of each member in each tribe is continuously iterated and updated until the maximum number of iterations is reached, thereby obtaining the optimized control parameters.

[0016] Optionally, the dynamic adjustment process of the inertia weight includes:

[0017] The inertia weight in each iterative update is dynamically adjusted based on the maximum number of iterations, the initial inertia weight, and the current inertia weight.

[0018] Optionally, the dynamic adjustment process of the cognitive factor includes:

[0019] An exponential function is used to dynamically adjust the cognitive factor in each iterative update process based on the maximum number of iterations.

[0020] Optionally, the dynamically adjusted inertia weight is expressed as:

[0021]

[0022] Where ω(t) is the inertia weight after dynamic adjustment, ω max is the initial inertia weight, Max_iteration is the maximum number of iterations, l is the current number of iterations, ω min is the current inertia weight;

[0023] The dynamically adjusted cognitive factor is expressed as:

[0024]

[0025] Where c1 is the individual cognitive factor, c2 is the tribal cognitive factor, exp() is the exponential function, Max_iteration is the maximum number of iterations, and l is the current number of iterations.

[0026] In a second aspect, the present application provides a wind turbine active clearance control system, comprising:

[0027] A data acquisition unit, configured to collect sound source signals, wind speed data, and rotation speed data of fan blades; the sound source signal refers to the sound signal of a set point at the tip position of the fan blade;

[0028] an edge computing device, connected to the data acquisition unit, for determining a clearance distance based on the sound source location;

[0029] A host computer is connected to the edge computing device and is used to optimize the control parameters through an improved tribal competition and member cooperation algorithm when the wind speed data and the rotational speed data of the wind turbine blades are both greater than a set threshold and the clearance distance is less than or equal to a set warning threshold, to obtain optimized control parameters, and to perform pitch operation control based on the optimized control parameters to achieve active control of the clearance distance.

[0030] Optionally, the data acquisition unit includes:

[0031] A microphone array mechanism, connected to the edge computing device, for collecting sound source signals;

[0032] an anemometer, connected to the edge computing device, for collecting the wind speed data;

[0033] A tachometer is connected to the edge computing device and is used to collect speed data of the fan blades.

[0034] Optionally, the microphone array mechanism includes:

[0035] A microphone array includes a plurality of microphones arranged in a multi-dimensional three-dimensional cross structure;

[0036] A support structure is installed to drive the microphone array to slide.

[0037] Optionally, the mounting support structure includes: an electric pulley, a clamp, a slide rail, a displacement sensor, a camera and a processing unit;

[0038] The clamp is arranged around the tower of the wind turbine; the slide rail is horizontally fixed to the clamp by a plurality of support rods; a conductive slip ring is arranged inside the slide rail; the microphone array is mounted on an electric pulley through a bracket, and the electric pulley is used to drive the microphone array to slide on the conductive slip ring;

[0039] The displacement sensor is arranged on the electric pulley; the displacement sensor and the camera are both electrically connected to the processing unit; the processing unit is used to control the sliding of the electric pulley based on the image data captured by the camera and the position data detected by the displacement sensor to keep the microphone array consistent with the rotation plane of the wind turbine blades.

[0040] According to the specific embodiments provided in this application, this application has the following technical effects:

[0041] The present application provides a method and system for active clearance control of a wind turbine. By adopting an improved tribal competition and member cooperation algorithm to optimize the control parameters, and performing pitch operation control based on the optimized control parameters, it can overcome the problem of lack of corresponding effective active safety control strategy after the clearance distance reaches the warning threshold, and by performing pitch operation control, each wind turbine blade can be finely controlled to meet the clearance requirements.

[0042] In addition, in order to overcome the current problems of wind turbine clearance detection such as high cost, low precision, poor environmental adaptability and non-real-time performance, the present application designs a simple, efficient, safe, reliable and low-cost wind turbine blade clearance distance detection system including a data acquisition unit and an edge computing device. By using edge computing devices to execute the method provided above in this application in real time, dynamic detection of the clearance distance can be performed in a timely manner, and pitch operation control can be performed, which has higher robustness. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0044] Figure 1 A schematic flow chart of a method for active clearance control of a wind turbine generator system according to an embodiment of the present application;

[0045] Figure 2 A block diagram of active safety control of a wind turbine based on an improved tribe competition and member cooperation algorithm according to an embodiment of the present application;

[0046] Figure 3A schematic structural diagram of a wind turbine active clearance control system provided in one embodiment of the present application;

[0047] Figure 4 This is a diagram showing the overall installation position of a data acquisition unit provided in one embodiment of the present application;

[0048] Figure 5 A schematic diagram of the microphone array structure provided in one embodiment of the present application;

[0049] Figure 6 A schematic diagram of the configuration of the mounting support structure provided in one embodiment of the present application;

[0050] Figure 7 This is a schematic diagram of the overall flow of a wind turbine active clearance control method provided in one embodiment of the present application.

[0051] Reference numerals:

[0052] A-data acquisition unit, 1-camera, 2-microphone array, 3-tower, 4-conductive slip ring, 5-electric pulley, 6-slide rail. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0054] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0055] In an exemplary embodiment, the present application provides a method for active clearance control of a wind turbine generator system, such as Figure 1 As shown, the method includes:

[0056] Step 100: Acquire a sound source signal, wind speed data, and wind turbine blade rotation speed data, and determine the sound source location based on the sound source signal. The sound source signal refers to the sound signal of a set point at the tip of the wind turbine blade.

[0057] Step 101: Determine the clearance distance based on the sound source position.

[0058] Step 102: When the wind speed data and the rotation speed data of the wind turbine blades are both greater than the set thresholds, and the clearance distance is less than or equal to the set warning threshold, the control parameters are optimized by using the improved tribe competition and member cooperation algorithm to obtain the optimized control parameters.

[0059] Step 103: Perform pitch control based on the optimized control parameters to achieve active control of the clearance distance.

[0060] By implementing the above steps 100 to 103, the present application adopts an improved tribal competition and member cooperation algorithm to optimize the control parameters, which can overcome the problem of lack of corresponding effective active safety control strategy after the clearance distance reaches the warning threshold, and then use the optimized control parameters to finely control the pitch angle of each wind turbine blade to meet the clearance requirements.

[0061] In another exemplary embodiment of the present application, in order to achieve precise control of pitch operation, in this embodiment, taking PID control as an example, the implementation process of the active clearance control method of the wind turbine provided in the present application is described.

[0062] First, the transfer function P(s) between the pitch angle of the wind turbine blade and the clearance distance is established. The real-time pitch angle value obtained through the SCADA system is used as the independent variable, and the measured clearance distance is used as the dependent variable. Both are time series data. The cross-correlation spectrum φ between the pitch angle and the clearance distance is calculated. xy (ω) and the autocorrelation spectrum of the pitch angle φ xx (ω), the ratio of the two is the frequency response P(e) of the transfer function P(s) jω ),have:

[0063]

[0064] Secondly, the transfer function P(s) can be obtained by reasonably setting the number of zeros and poles and combining it with least squares fitting.

[0065] Furthermore, to simplify calculations, this application uses single-blade control instead of the more commonly used Coleman transform-based control. The advantages of single-blade control are simple implementation and decoupling of each blade. For example, for a three-blade wind turbine, cascading three identical single-input, single-output control systems can achieve the desired clearance control.

[0066] Based on the above description, during the implementation of step 102, if the predicted clearance distance L is greater than the upper limit of the set warning threshold (i.e., the clearance safety threshold), the pitch change operation is not performed and the wind turbine continues to operate. If the predicted clearance distance L is within the clearance safety threshold (i.e., less than or equal to the set warning threshold), the pitch change operation is required before the blades reach the tower. In this case, the optimized PID algorithm is used to perform the pitch change operation to increase the clearance distance between the blades and the tower, ensuring safe operation of the unit.

[0067] like Figure 2As shown in the figure, the PID control parameters given above are optimized by the improved tribe competition and member cooperation algorithm, including:

[0068] (1) The PID control parameters to be optimized are used as members to initialize the parameters of the improved tribe competition and member cooperation algorithm to obtain the initialization parameters. The initialization parameters include the number of tribes, the number of members in each tribe, the velocity vector V of each member, the maximum number of iterations, the loyalty (i.e., the random factor), the initial inertia weight, and the initial cognitive factor.

[0069] (2) Based on the initialization parameters and the position fitness value of each member, the velocity vector of each member in each tribe is continuously iterated and updated until the maximum number of iterations is reached, and the optimized control parameters are obtained. Among them, the velocity vector V of each member is expressed as:

[0070]

[0071] Where, denote the speed of the mth member of the nth tribe after t and t+1 iterations respectively, yes The inertia weight, is an individual experience term, It is an experience item within the tribe. They represent the optimal fitness position found by the member in this iteration and the entire cycle, respectively. c1 is the individual cognitive factor, representing the coefficient of following one's own experience and controlling the degree to which tribe members rely on their own experience. c2 is the tribal cognitive factor, representing the coefficient of following the instructions of the tribal chief and controlling the degree to which members trust the tribe's best experience. They represent an individual's loyalty to his or her own experience and the experience of the tribe, respectively.

[0072] The loyalty of an individual member can be described as:

[0073]

[0074] Where r t and r t-1 represents the loyalty of a single member after iterations t and t+1, where p is an integer index. is a set of integers.

[0075] Evaluate the fitness value of each member's current position and update each member's personal optimal position The optimal position of each tribe is The global optimal position of the entire population is X best .

[0076] The formula of the fitness function is as follows:

[0077]

[0078] Where, fitness(σ,t s ) is the fitness function, Represents the maximum deviation, adjustment time, time multiplied by absolute error integral, μ σ , μ ITAE Represent the weight coefficients of the three indicators respectively.

[0079] Since random conflicts may occur between tribes, the weaker tribes will flee while the stronger tribes will not be affected. Based on this, I = rand(n) represents a random conflict between n tribes. Under this conflict, a new speed will be refreshed, which is:

[0080]

[0081] Where, They represent the optimal fitness of the current tribe and the optimal fitness value found in a randomly selected competitor tribe, representing the fitness levels of different tribes. The competition factor is a key adjustment in the speed update process. It simulates behavior during competition and helps individuals adjust based on their competitors' optimal solutions. If the opponent's optimal solution is superior, the competition factor forces the individual to adjust their speed away from the opponent's optimal solution, manifesting as a "fleeing" behavior. c3 is the competition adjustment coefficient, which controls the impact of competition adjustment on speed updates. It is a random factor that simulates the chaotic retreat behavior of individuals in competition and increases the randomness and exploration ability of the algorithm.

[0082] Based on the above description, the member's speed update depends on the result of the conflict, by comparing the fitness and If the current tribe performs better, the existing optimization strategy is maintained and the local optimal solution is continued to be developed. If the competitor performs better, the optimal solution information of the competitor tribe is introduced and the speed direction is adjusted. After the maximum number of iterations is reached, the optimization ends and the optimal PID control parameter K is output. P ,K I ,K D Among them, the three PID control parameters K to be optimized P ,K I ,K D The value ranges can be (0.001, 1.0), (0.001, 0.3), and (0.001, 0.2) respectively.

[0083] In another exemplary embodiment of the present application, the improvement to the tribal competition and member cooperation algorithm is mainly reflected in the dynamic adjustment of the inertia weight and the dynamic adjustment of the individual cognitive factor and the tribal cognitive factor. That is, the present application obtains an improved tribal competition and member cooperation algorithm by dynamically adjusting the inertia weight and cognitive factors (individual cognitive factor and tribal cognitive factor) of the tribal competition and member cooperation algorithm. Based on this, combined with the above-mentioned formula for representing the velocity vector V of each member, we have:

[0084]

[0085] Where ω(t) is the inertia weight after dynamic adjustment, which is mainly The improved dynamic inertia factor of ω max is the initial inertia weight, which is usually a larger value (such as 0.9) to enhance the early global search capability. Max_iteration is the maximum number of iterations, l is the current number of iterations, ω min is the current inertia weight, which usually takes a smaller value (such as 0.4) to enhance the local development capability in the later stage.

[0086] By dynamically adjusting the inertia weight, in the early stages of the algorithm, a larger inertia weight can preserve the historical speed of tribe members, enhance exploration capabilities, and thus quickly cover the entire search space. In the later stages of the algorithm, a smaller inertia weight reduces the inertial impact of tribe members, allowing them to focus more on developing local areas and improve optimization accuracy. Therefore, the improved tribe competition and member cooperation algorithm set up in this application can adaptively adjust the search strategy, balance the needs of exploration and development at different stages, and avoid falling into local optimal solutions.

[0087] In the tribal competition and member cooperation algorithm, the individual cognitive factor and the tribal cognitive factor are key parameters that control the search of tribal members in different directions. The individual cognitive factor (c1) represents the strength of the tribe members' optimization based on their own experience, while the tribal cognitive factor (c2) represents the strength of the tribe members' compliance with the tribe's optimal instructions. The introduction of dynamic cognitive factors can adaptively adjust the weights of these two cognitive factors according to different iteration stages, thereby balancing global search and local development capabilities and improving algorithm performance. Based on this, the dynamically adjusted cognitive factor of this application is expressed as:

[0088]

[0089] Where c1 is the individual cognition factor, which encourages members to align with their historical best position. It is smaller in the early stages and larger in the later stages. c2 is the tribe cognition factor, which encourages members to align with the tribe's historical best position. It is larger in the early stages and smaller in the later stages. exp() is an exponential function. This nonlinear adjustment of the tribe cognition factor causes the improved tribe competition and member cooperation algorithm to focus on global search in the early stages, resulting in stronger group perception, while gradually strengthening local development in the later stages, resulting in a reliance on self-perception.

[0090] Based on the same inventive concept, embodiments of the present application also provide a wind turbine active clearance control system for implementing the aforementioned wind turbine active clearance control method. The solution provided by this system is similar to the solution described in the aforementioned method. Therefore, the specific limitations of one or more wind turbine active clearance control system embodiments provided below can be found in the aforementioned limitations of the wind turbine active clearance control method and will not be further elaborated here.

[0091] In an exemplary embodiment, in order to overcome the problems of high cost, low accuracy, poor environmental adaptability and non-real-time performance of current wind turbine clearance detection, the present application provides an active wind turbine clearance control system, such as Figure 3 and Figure 4 As shown, it includes: a data acquisition unit A, an edge computing device and a host computer. The edge computing device is connected to the data acquisition unit A. The host computer is connected to the edge computing device. Figure 3 The edge computing device and the host computer are not shown. Figure 4 As shown, the data acquisition unit A can be set outside the tower. The edge computing device can be set inside the tower, and the location of the host computer is determined according to actual needs.

[0092] The data acquisition unit A collects the sound source signal, wind speed data and the rotation speed data of the fan blade. The sound source signal refers to the sound signal of the set point at the tip position of the fan blade.

[0093] The edge computing device integrates data preprocessing and clearance distance algorithms to process the collected raw acoustic data in real time and calculate the precise clearance distance.

[0094] When the wind speed data and the speed data of the wind turbine blades are both greater than the set thresholds, and the clearance distance is less than or equal to the set warning threshold, the upper computer optimizes the control parameters through the improved tribe competition and member cooperation algorithm to obtain the optimized control parameters, which are then used to perform pitch operation control based on the optimized control parameters to achieve active control of the clearance distance.

[0095] In another exemplary embodiment of the present application, the data acquisition unit A is configured to include a structure including a microphone array and its corresponding mounting support (ie, a mounting support structure), an anemometer, and a tachometer.

[0096] For example, the microphone array may be composed of 5 microphones, each of which may be of YG-201 type. Figure 5 As shown, the structure of the microphone array can be a five-element three-dimensional cross structure, with two microphones placed on the y-axis with the origin as the midpoint and at equal intervals, two microphones placed on the z-axis with the origin as the midpoint and at equal intervals, and another microphone placed in the positive direction of the x-axis. i represents the i-th microphone, the arrangement spacing of each microphone is a, and the coordinates of each microphone are (a,0,0), (0,a,0), (0,-a,0), (0,0,a), (0,0,-a).

[0097] For example, in order to increase the service life of the microphone, a windscreen and a rainscreen may be installed on the microphone to attenuate wind noise and reduce the impact of meteorological conditions while minimizing the loss of the test sound signal.

[0098] In another exemplary embodiment of the present application, in order to achieve accurate measurement of the sound source signal on the blade, in this embodiment, the mounting support structure includes: an electric pulley 4, a clamp, a slide rail 6, a displacement sensor, a camera 1 and a processing unit.

[0099] like Figure 6 As shown, a clamp tightly surrounds the wind turbine tower 3. A slide rail 6 is horizontally secured to the clamp via multiple support rods. A conductive slip ring 4 is located within the slide rail 6. The microphone array 2 is mounted on a motorized trolley 5 via a bracket. The motorized trolley 5 drives the microphone array 2 to slide on the conductive slip ring 4.

[0100] For example, the mounting support structure is fixed on the tower at a set height (e.g., 5m) from the ground. The 5m height was obtained through actual engineering deployment and has a good audio data acquisition effect. Furthermore, since the clearance distance algorithm can determine the specific coordinate position of the sound source, even if the data acquisition unit is installed below the blade tip position, it will not have a significant impact on the final positioning accuracy. Based on this, the arrangement of this height position takes into account the feasibility in actual operation and the flexibility of equipment deployment.

[0101] The displacement sensor is mounted on the electric pulley. Both the displacement sensor and the camera are electrically connected to a processing unit. The processing unit controls the sliding of the electric pulley based on image data captured by the camera and position data detected by the displacement sensor to maintain alignment of the microphone array with the rotational plane of the wind turbine blades.

[0102] As an optional implementation, after a luminous reflective sheet is affixed to the bottom of the outer side of the cabin, a camera with a target tracking function (which can also be a pan-tilt camera) will lock onto this reflective mark in real time when the wind turbine yaws, and transmit the image data captured by the camera to the processing unit in the data acquisition unit. This processing unit is dedicated to performing target tracking and image analysis tasks, thereby avoiding increasing the burden on the edge computing equipment in the tower. The processing unit first extracts the pixel coordinates of the reflective sheet in the picture, and then calculates the deviation angle of the camera's shooting angle relative to the reflective sheet based on the offset between the coordinates of the mark point and the center of the image. The angle error signal is converted into a displacement instruction that drives the electric pulley to move along the circular track of the tower.

[0103] The electric pulley's controller incorporates a built-in PID algorithm, which uses the drive motor to continuously adjust the pulley's position to maintain its fixed position on the reflective marker. Simultaneously, the pulley's displacement sensor provides real-time position feedback, which is then combined with the wind turbine's yaw data for comprehensive analysis. If a deviation between the actual position and the expected position is detected, the electric pulley's controller immediately sends a correction command to the drive motor, adjusting the pulley's trajectory to ensure the microphone array remains aligned with the wind turbine blade's rotational plane.

[0104] Based on the above description, the conductive slip ring inside the slide rail solves the problem of line entanglement caused by rotation. The camera, microphone array, processing unit and electric pulley controller are all connected to the conductive slip ring, and the conductive slip ring is connected to the cable to supply power to the components in the data acquisition unit.

[0105] To further enhance overall safety and reliability, a low-impedance grounding grid will be constructed on the tower's data acquisition unit to ensure lightning is quickly directed underground, reducing the risk of lightning strikes. Furthermore, dedicated lightning rods or strips will be installed on the rails to effectively guide and disperse lightning, preventing accidents during thunderstorms.

[0106] As an optional implementation, the data acquisition unit can send the collected blade acoustic data (i.e., sound source signal) to the edge computing device at the bottom of the tower via optical fiber. During the preprocessing process, the edge computing device uses a filtering and noise reduction link to suppress the noise of the collected sound source signal, reduce background noise interference such as wind noise, and ensure the accuracy of the clearance distance measurement. The processed acoustic data is then run through the clearance distance algorithm to obtain the clearance distance. The details of the clearance distance algorithm are as follows:

[0107] Given that the main noise source of the wind turbine blade is concentrated at the blade tip, and in order to simplify the acoustic model for easier analysis, it is assumed that the noise contribution of the entire blade can be represented by a point sound source at the blade tip. Assume that the coordinates of the sound source T are (x, y, z), and the time when each microphone receives the sound source signal is t i, the sound source's sounding time is t0, and the parameters k and T are defined i ,have:

[0108]

[0109] Where c is the actual ambient sound speed, c0 is the default sound speed, k is the sound speed correction coefficient of the ratio of the actual sound speed to the default sound speed, T i It represents the time it takes for a sound wave to travel from the sound source to the i-th sensor receiving the signal, assuming the speed of sound is the default value. T0 represents the reference time (i.e., the starting time) when the sound source emits the sound, and T1 to T5 represent the specific reception time when the i-th microphone receives the sound source signal. The reason for introducing these six quantities is that this is an arrival time positioning model. It is necessary to mark both the emission time of the sound source (T0) and the reception time of each microphone (T1 to T5) in order to write five spherical equations to solve the sound source position. The spacing between each microphone is a. Substituting the microphone coordinate position into the basic geometric relationship model, we can obtain:

[0110]

[0111] In order to simplify the representation, we introduce τ i =T i -T0,τ i It represents the equivalent time delay of each sensor relative to the launch time. By solving the geometric relationship model, we can know that:

[0112]

[0113] Where D 32 It represents the square difference of the delay between microphone 3 and microphone 2, D 54 It represents the square difference of the delay between microphone No. 5 and microphone No. 4, S 123 It represents the weighted square sum of the delays of microphones 1, 2, and 3, S 23 It represents the sum of the squares of the delays of microphones 2 and 3. M, N, P, and Q are all intermediate parameters.

[0114] In actual application scenarios, the sound source is usually located outside the microphone array. By simplifying, the specific location of the sound source T can be obtained. After further conversion, the specific clearance distance can be obtained as follows:

[0115]

[0116] To ensure that the edge computing device has strong computing power, the processing unit can be configured as a high-performance multi-core processor, equipped with large-capacity memory and high-speed storage devices. The edge computing device is located in the tower, and the dry-type transformer can directly power it through the cable after voltage conversion. Based on this, Figure 7 As shown in the figure, when the wind speed is detected to be greater than 5m / s and the blade speed is greater than 5r / min, that is, the fan is in a high-speed state, the edge computing device will determine whether there is a tower sweep risk based on the real-time measurement results provided by the clearance distance algorithm, and detect whether the clearance distance reaches the set warning threshold.

[0117] When the system detects a tower sweep risk or the clearance distance is lower than the warning threshold, the edge computing device will immediately send a request to the host computer (which can also be a server) through the wireless communication module to start the active safety control unit, and at the same time transmit the warning information to the control cabinet at the bottom of the tower through optical fiber. The control cabinet is then connected to the control cabinet responsible for pitch operation in the cabin through optical fiber. Among them, the active safety control unit is an independent control module, and its execution priority is higher than the conventional pitch control. When the clearance distance is lower than the warning threshold, the cabin control cabinet will switch the power generation control loop used to ensure economy to the safety control loop. The cabin control cabinet will perform independent pitch operation according to the results of the control algorithm running on the host computer, providing immediate active safety protection for the wind turbine, thereby ensuring that the entire safety monitoring and early warning control process can have a high degree of real-time and reliability.

[0118] Based on the above description, in this application, the entire clearance detection data collection, clearance distance calculation and communication transmission process all have real-time design and safety guarantees, and can timely transmit clearance distance changes and warning information to the control system, meeting the real-time requirements of wind turbine safety monitoring.

[0119] In actual applications, based on the detection results of the edge computing device provided above, the wind turbine active clearance control method provided in this application can be used to implement active safety control of wind turbines. Based on this, the processing core of the system provided in this application is to use the real-time clearance distance data obtained by the edge computing device, transmit it to the active safety control unit via the host computer, and dynamically adjust the wind turbine pitch angle to ensure the safe operation of the unit.

[0120] As an optional implementation, the data acquisition unit can use dedicated cables to power various components. These cables are routed through ventilation holes in the tower grid plate and fed into dry-type transformers located inside the tower or on the ground. Voltage conversion devices can be added as needed to meet the power supply requirements of different components.

[0121] In summary, compared with the prior art, this application has the following advantages:

[0122] 1. Reduced cost and universal applicability: This application uses a microphone array, which can greatly reduce the cost compared to mainstream millimeter-wave radar and visual methods for clearance detection.

[0123] 2. High Accuracy and Reliability: By leveraging the propagation characteristics of sound waves, this application collects clearance data that is unaffected by external weather conditions, including clouds, fog, rain, and snow. The data source is highly accurate and reliable. Furthermore, the host computer incorporates multiple auxiliary signal processing methods to optimize and filter the collected data, ensuring data validity and reducing misjudgments.

[0124] 3. Intelligence and safety: This application introduces acoustic-based clearance detection technology and a supporting active safety control mechanism to improve the intelligence level of the fan and effectively ensure its safety.

[0125] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0126] This document uses specific examples to illustrate the principles and implementation methods of this application. The description of the above examples is only intended to help understand the method and core concept of this application. At the same time, for those skilled in the art, based on the concept of this application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A wind turbine active clearance control method, characterized in that: include: Acquire a sound source signal, wind speed data, and rotation speed data of the fan blades, and determine the sound source position based on the sound source signal; The sound source signal refers to the sound signal of a set point on the tip of the fan blade; determining a clearance distance based on the sound source location; When the wind speed data and the rotation speed data of the wind turbine blades are both greater than a set threshold value, and the clearance distance is less than or equal to a set warning threshold value, the control parameters are optimized by using an improved tribe competition and member cooperation algorithm to obtain optimized control parameters; Pitch operation control is performed based on the optimized control parameters to achieve active control of the clearance distance.

2. The active clearance control method for wind turbines according to claim 1, characterized in that: An improved tribal competition and member cooperation algorithm is obtained by dynamically adjusting the inertia weight and cognitive factors of the tribal competition and member cooperation algorithm; the cognitive factors include individual cognitive factors and tribal cognitive factors.

3. The wind turbine active clearance control method according to claim 2, characterized in that: The process of optimizing control parameters through the improved tribe competition and member cooperation algorithm includes: Initializing the parameters of the improved tribe competition and member cooperation algorithm to obtain initialization parameters; the initialization parameters include: the number of tribes, the number of members of each tribe and the velocity vector of each member, the maximum number of iterations, loyalty, initial inertia weight and initial cognitive factor; using the control parameters to be optimized as the members; Based on the initialization parameters and the position fitness value of each member, the velocity vector of each member in each tribe is continuously iterated and updated until the maximum number of iterations is reached, thereby obtaining the optimized control parameters.

4. The wind turbine active clearance control method according to claim 3, characterized in that: The dynamic adjustment process of the inertia weight includes: The inertia weight in each iterative update is dynamically adjusted based on the maximum number of iterations, the initial inertia weight, and the current inertia weight.

5. The wind turbine active clearance control method according to claim 3, characterized in that: The dynamic adjustment process of the cognitive factors includes: An exponential function is used to dynamically adjust the cognitive factor in each iterative update process based on the maximum number of iterations.

6. The wind turbine active clearance control method according to claim 4 or 5, characterized in that: The dynamically adjusted inertia weight is expressed as: Where ω(t) is the inertia weight after dynamic adjustment, ω max is the initial inertia weight, Max_iteration is the maximum number of iterations, l is the current number of iterations, ω min is the current inertia weight; The dynamically adjusted cognitive factor is expressed as: Where c1 is the individual cognitive factor, c2 is the tribal cognitive factor, exp() is the exponential function, Max_iteration is the maximum number of iterations, and l is the current number of iterations.

7. An active clearance control system for a wind turbine generator set, characterized in that: include: A data acquisition unit, configured to collect sound source signals, wind speed data, and rotation speed data of fan blades; the sound source signal refers to the sound signal of a set point at the tip position of the fan blade; an edge computing device, connected to the data acquisition unit, for determining a clearance distance based on the sound source location; A host computer is connected to the edge computing device and is used to optimize the control parameters through an improved tribal competition and member cooperation algorithm when the wind speed data and the rotational speed data of the wind turbine blades are both greater than a set threshold and the clearance distance is less than or equal to a set warning threshold, to obtain optimized control parameters, and to perform pitch operation control based on the optimized control parameters to achieve active control of the clearance distance.

8. The wind turbine active clearance control system according to claim 7, characterized in that: The data acquisition unit includes: A microphone array mechanism, connected to the edge computing device, for collecting sound source signals; an anemometer, connected to the edge computing device, for collecting the wind speed data; A tachometer is connected to the edge computing device and is used to collect speed data of the fan blades.

9. The wind turbine active clearance control system according to claim 8, characterized in that: The microphone array mechanism comprises: A microphone array includes a plurality of microphones arranged in a multi-dimensional three-dimensional cross structure; A support structure is installed to drive the microphone array to slide.

10. The wind turbine active clearance control system according to claim 9, characterized in that: The installation support structure includes: an electric pulley, a clamp, a slide rail, a displacement sensor, a camera and a processing unit; The clamp is arranged around the tower of the wind turbine; the slide rail is horizontally fixed to the clamp by a plurality of support rods; a conductive slip ring is arranged inside the slide rail; the microphone array is mounted on an electric pulley through a bracket, and the electric pulley is used to drive the microphone array to slide on the conductive slip ring; The displacement sensor is arranged on the electric pulley; the displacement sensor and the camera are both electrically connected to the processing unit; the processing unit is used to control the sliding of the electric pulley based on the image data captured by the camera and the position data detected by the displacement sensor to keep the microphone array consistent with the rotation plane of the wind turbine blades.

Images