Wind turbine cranking transmission direction optimization system based on big data analysis

Abstract

This invention relates to the field of wind power generation technology, specifically to a wind turbine turning gear transmission direction optimization system based on big data analysis, comprising: a state perception module and a fatigue matrix module, an optimization decision module, and a flexible execution module. This invention achieves a dimensional upgrade in turning gear transmission direction from macroscopic load compliance to microscopic tooth surface protection by constructing a tooth surface phase tracking matrix and coupling a multi-objective cost function optimization model. Based on the weighted optimization of the comprehensive static torque and the fatigue degree of specific gear phases calculated in real time, it can ensure smooth motor start-up and significantly reduce instantaneous mechanical stress during startup. Simultaneously, through adaptive balancing of wear amounts in different phases along the entire circumference of the gear, it effectively avoids excessive fatigue and pitting damage on local tooth surfaces. Combined with a reverse micro-motion preload control strategy, it eliminates the collision impact caused by tooth flank clearance from a physical mechanism perspective.

Description

Technical Field

[0001] This invention relates to the field of wind power generation technology, and more specifically to a wind turbine crankshaft drive direction optimization system based on big data analysis. Background Technology

[0002] During installation, commissioning, blade locking, or routine mechanical maintenance, wind turbine generators require a turning gear system to drive the transmission chain at low speeds. Existing turning gear control strategies typically employ fixed-direction drive or manual direction selection based on maintenance personnel experience, failing to effectively integrate wind turbine operating big data for real-time condition analysis.

[0003] First, when the wind turbine is shut down, the rotor is subjected to static aerodynamic loads caused by ambient wind conditions and gravitational eccentric loads caused by uneven blade mass distribution. The combined vector direction of these loads changes dynamically with the shutdown azimuth angle and real-time wind speed. If the turning gear motor drives in the opposite direction to the combined static load, the turning gear motor must overcome enormous reverse resistance at startup, which can easily trigger the inverter's overcurrent protection or subject the gear meshing surface to strong mechanical impact stress, leading to localized cracks or even tooth breakage.

[0004] Second, while the gear pairs in the main transmission system exhibit uniform wear characteristics under macroscopic operation, the existing control logic cannot detect and record the microscopic meshing phase between the pinion and the gear during low-speed, short-stroke quasi-static actions such as turning the gear. This leads to the possibility that, over long maintenance cycles, specific phases of the tooth surface on the circumference of the gear may be frequently used as starting loading points, resulting in asymmetric localized fatigue pitting or spalling, which in turn triggers early failure of the transmission chain.

[0005] Third, due to limitations in gear pair processing and assembly technology, there is inevitably a physical tooth flank clearance between the teeth. Traditional direct starting methods lack a flexible preload mechanism when crossing this clearance, which will generate a severe transient dynamic load impact at the moment of meshing. Summary of the Invention

[0006] To achieve the above objectives, the present invention provides the following technical solution: This invention provides a wind turbine turning gear transmission direction optimization system based on big data analysis, including: a state perception module, used to collect the comprehensive static torque parameters of the wind turbine generator set in the shutdown state in real time, and to analyze the real-time meshing phase angle between the current turning gear pinion and the main drive gear through a position acquisition device; The fatigue matrix module is communicatively connected to the state perception module and has a built-in tooth surface phase tracking matrix based on circumferential distribution. It is used to extract the local tooth surface fatigue parameters of the corresponding meshing position in the clockwise and counterclockwise transmission directions according to the real-time meshing phase angle. The optimization decision module is communicatively connected to the state perception module and the fatigue matrix module, and has a built-in multi-objective cost function model. It is used to input the comprehensive static torque parameter and the local tooth surface fatigue parameter into the multi-objective cost function model after normalization, comprehensively calculate the evaluation cost of the two candidate transmission directions, clockwise and counterclockwise, and lock the candidate transmission direction with the minimum evaluation cost as the optimal transmission direction. The flexible execution module is communicatively connected to the optimization decision module. It is used to receive the optimal transmission direction command and, before the formal start of the turning gear, output a preset micro-torque to the non-optimal transmission direction to actively disengage from the tooth surface and eliminate gear meshing clearance. Then, it outputs a smooth acceleration curve to drive the turning gear system according to the optimal transmission direction.

[0007] Furthermore, the comprehensive static torque parameter includes an aerodynamic torque component and a gravitational eccentricity torque component; The state perception module includes an anemometer and a rotor position sensor, used to collect real-time wind speed outside the nacelle, the relative angle between the nacelle and the wind direction, and the stopping azimuth angle of the wind turbine rotor, so as to calculate the aerodynamic torque component and the gravitational eccentric torque component. The position acquisition device includes a multi-turn absolute encoder installed on the input shaft of the turning motor or the main drive chain; The state perception module obtains the current rotor zero offset by reading the angle data of the multi-turn absolute encoder, and calculates the real-time meshing phase angle on the circumferential circumference of the current turning pinion and the main drive gear based on the fixed tooth ratio of the turning pinion and the main drive gear.

[0008] Furthermore, when acquiring the comprehensive static torque parameters, the state perception module uses a superimposed evaluation torque model; The superimposed evaluation torque model vector sums the aerodynamic torque component under the current shutdown azimuth angle and real-time wind speed with the gravitational eccentricity torque component caused by blade mass eccentricity, to obtain the direction and magnitude of the comprehensive static load currently acting on the transmission chain, which serves as the basic calculation input value for the optimization decision module.

[0009] Furthermore, the tooth surface phase tracking matrix in the fatigue matrix module divides the 360-degree circumference of the main drive gear into several discrete phase angle intervals. When extracting the local tooth surface fatigue parameters, the fatigue matrix module first maps the real-time meshing phase angle input by the state perception module to the corresponding phase angle interval, and then retrieves the historically stored clockwise cumulative wear coefficient and counterclockwise cumulative wear coefficient in parallel from the phase angle interval. The fatigue matrix module is also configured with an adaptive update sub-unit; After each wind turbine turning action is completed, the adaptive update subunit obtains the actual execution direction, actual running time, and load current or output torque of the turning motor for this turning action. It then generates the single fatigue increment for this operation through time integration calculation and adds the single fatigue increment to the historical cumulative wear coefficient of the corresponding phase angle interval to realize the dynamic iteration of the tooth surface phase tracking matrix.

[0010] Furthermore, when calculating the single fatigue increment, the adaptive update subunit also incorporates a weighted correction based on gearbox ambient temperature parameters or lubricating oil temperature parameters obtained from the SCADA system. When the ambient temperature parameter or lubricating oil temperature parameter is lower than the preset low temperature threshold, the calculated value of the single fatigue increment in the corresponding phase angle interval is increased by increasing the correction weight coefficient, so as to compensate for the additional mechanical wear caused by poor gear lubrication under low temperature conditions.

[0011] Furthermore, the multi-objective cost function model built into the optimization decision module includes an initiation resistance penalty term and a local fatigue penalty term; For any candidate transmission direction, the cost of the starting resistance penalty term is calculated based on the consistency of the vector direction of the candidate transmission direction with the comprehensive static torque parameter. When the vector directions are in the same direction, a smaller cost is output, and when the vector directions are in opposite directions, a larger cost is output. The cost of the local fatigue penalty term is positively correlated with the local tooth surface fatigue parameter corresponding to the candidate transmission direction. The optimization decision module performs a weighted summation of the starting resistance penalty term and the local fatigue penalty term to obtain the evaluation cost of the corresponding candidate transmission direction.

[0012] Furthermore, the optimization decision module is also equipped with a dynamic weight allocation unit, which is used to dynamically adjust the first weight coefficient of the starting resistance penalty term and the second weight coefficient of the local fatigue penalty term according to the external working conditions perceived in real time. When the real-time wind speed fluctuation rate is higher than the preset wind condition threshold, the dynamic weight allocation unit increases the first weight coefficient to prioritize adapting to aerodynamic load resistance; when the real-time wind speed fluctuation rate is lower than the preset wind condition threshold and the historical wear average value of the preset period is high, the second weight coefficient is increased to prioritize avoiding tooth surface wear accumulation.

[0013] Furthermore, the optimization decision-making module also includes an over-limit safety protection subunit; When comparing the evaluation cost values, if the evaluation cost values ​​of both the clockwise and counterclockwise candidate transmission directions exceed the preset safe start cost threshold, the over-limit safety protection subunit will intercept the issuance of the optimal transmission direction command and send an aerodynamic unloading request command to the upper-level pitch system. After the upper-level pitch system adjusts the blade pitch angle to reduce the comprehensive static torque parameter, the calculation of the multi-objective cost function model will be retried.

[0014] Furthermore, the flexible execution module includes a turning gear frequency converter and a turning gear motor; When the turning gear frequency converter drives the turning gear system in the optimal transmission direction, it adopts an S-type speed command control strategy. By smoothly adjusting the derivative of acceleration, it suppresses the torque pulsation of the turning gear motor during the start-up and acceleration phases, ensuring that the turning gear pinion and the main drive gear complete a flexible re-meshing without impact. When the flexible actuator outputs the preset micro-torque, it adopts a non-displacement torque preload control strategy. The turning gear inverter applies a limiting torque, which is a preset percentage of the rated torque of the turning gear motor, to the non-optimal transmission direction, and quickly blocks the output after maintaining a preset micro-motion duration; the magnitude of the limiting torque and the preset micro-motion duration are set such that the turning gear pinion is only driven to overcome static friction to produce a small elastic yield or gap separation between the tooth surfaces, without causing the main shaft of the wind turbine generator to deflect macroscopically.

[0015] Furthermore, the flexible execution module is also equipped with a start-up anti-jamming monitoring subunit; During the initial climbing phase when the turning motor accelerates in the optimal transmission direction, the start-up anti-jamming monitoring subunit monitors the output current or estimated torque of the turning frequency converter in real time. If the rate of change of the output current or estimated torque exceeds the preset hard contact sudden change threshold before the rated turning speed is reached, the flexible execution module immediately interrupts the current turning command and triggers the parking brake, while simultaneously sending an abnormal collision fault code back to the cloud.

[0016] Beneficial effects This invention upgrades the direction of the turning gear transmission from macroscopic load compliance to microscopic tooth surface protection by constructing a tooth surface phase tracking matrix and coupling a multi-objective cost function optimization model. Based on the weighted optimization of the comprehensive static torque and the fatigue degree of specific phases of the gears calculated in real time, this technical solution can ensure smooth motor start-up and significantly reduce the mechanical stress at the start-up moment. At the same time, it effectively avoids excessive fatigue and pitting damage on local tooth surfaces by adaptively balancing the wear amount of different phases of the gears in the whole circumference. Combined with the reverse micro-motion preload control strategy, it eliminates the collision impact caused by tooth flank clearance from the physical mechanism, significantly improving the mechanical service life and system reliability of the wind power transmission chain under off-grid operation and maintenance conditions. Attached Figure Description

[0017] Figure 1 This is a diagram illustrating the overall architecture and communication logic of the present invention. Figure 2 This is a flowchart of the directional optimization decision-making method based on multi-objective cost function of the present invention; Figure 3 This is a schematic diagram illustrating the discretization and closed-loop update principle of the tooth surface phase tracking matrix in this invention. Figure 4 This is a diagram showing the torque and speed dual-axis timing control curves of the turning gear control system of the present invention. Detailed Implementation

[0018] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0019] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but includes other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0020] The present invention will now be described in further detail with reference to the accompanying drawings: Example: like Figure 1-4 As shown, the wind turbine turning gear transmission direction optimization system based on big data analysis includes the following steps: The status perception module is used to collect the comprehensive static torque parameters of the wind turbine generator in the shutdown state in real time, and analyze the real-time meshing phase angle between the current turning pinion and the main drive gear through the position acquisition device; The fatigue matrix module and the communication connection status sensing module have a built-in tooth surface phase tracking matrix based on circumferential distribution. This matrix is ​​used to extract local tooth surface fatigue parameters in the clockwise and counterclockwise transmission directions of the corresponding meshing position based on the real-time meshing phase angle. The optimization decision module is connected to the state perception module and the fatigue matrix module respectively. It has a built-in multi-objective cost function model, which is used to input the comprehensive static torque parameter and the local tooth surface fatigue parameter into the multi-objective cost function model after normalization. It comprehensively calculates the evaluation cost of the two candidate transmission directions, clockwise and counterclockwise, and locks the candidate transmission direction with the minimum evaluation cost as the optimal transmission direction. The flexible execution module, which is connected to the optimization decision module, is used to receive the optimal transmission direction command and output a preset micro-torque to the non-optimal transmission direction before the turnaround is officially started to actively disengage from the tooth surface and eliminate gear meshing backlash. Then, it outputs a smooth acceleration curve to drive the turnaround system according to the optimal transmission direction.

[0021] Furthermore, the specific operation process of the state awareness module is as follows: The state awareness module, serving as the underlying data acquisition and physical quantity analysis hub of the system, is primarily deployed within the main control cabinet of the wind turbine nacelle. It establishes hardwired or industrial Ethernet communication with external meteorological sensors and mechanical sensors on the drivetrain. The core functions of the state awareness module are divided into two independent execution branches: a static torque analysis flow and a meshing phase angle mapping flow.

[0022] For the static torque analytical flow, the state-aware module aims to calculate the comprehensive static torque parameters acting on the drivetrain. When the wind turbine is shut down, the rotor is subjected to aerodynamic torque components caused by ambient wind conditions and gravitational eccentric torque components caused by uneven blade mass distribution. The state-aware module obtains real-time wind speed by reading an anemometer located on the top of the nacelle. The relative yaw error angle between the cabin and the wind direction Simultaneously, by reading the rotor position sensor installed on the main shaft or pitch system, the current stopping azimuth angle of the wind turbine rotor is obtained. and the real-time pitch angle of each blade. .

[0023] The state perception module is equipped with a microprocessor, which first solves for the aerodynamic torque component based on a preset aerodynamic load calculation model. The calculation formula is as follows:

[0024] In the formula, This represents the aerodynamic torque component. This represents the real-time air density measured by sensors of the external environment of the cabin. Indicates the area swept by the wind turbine. Indicates the radius of the wind turbine. This indicates the real-time wind speed. This is the torque coefficient matrix function under static conditions. This function is a known three-dimensional lookup table function based on the wind turbine blade airfoil data. The processor will provide the real-time stopping azimuth angle. Pitch angle and yaw error angle The static aerodynamic torque coefficient under the current state is obtained by three-dimensional linear interpolation when input into this function.

[0025] The microprocessor then calculates the gravitational eccentricity torque component caused by rotor mass eccentricity. Due to manufacturing tolerances of the three blades or uneven surface icing during shutdown, the wind turbine has an overall eccentric mass. The formula for calculating the gravitational eccentricity torque component is:

[0026] In the formula, This represents the gravitational eccentricity torque component. This refers to the rotor equivalent eccentric mass estimated through the fan's factory calibration or icing monitoring system. The acceleration due to gravity is constant. The radial eccentricity from the center of mass of the equivalent eccentric mass to the center of rotation of the spindle. The aforementioned measured current stopping azimuth angle, The initial phase zero-point offset angle of the equivalent eccentric mass center on the wind turbine plane.

[0027] In obtaining separately and Subsequently, the state-aware module uses a superimposed evaluation torque model to perform vector summation calculations. The microprocessor sets clockwise torque as positive and counterclockwise torque as negative, and then calculates the comprehensive static torque parameters. .Should The symbol represents the clockwise or counterclockwise rotation trend of the rotor under the influence of the external physical environment, and its absolute value represents the static load base value that the turning motor needs to overcome or adapt to when starting.

[0028] For the meshing phase angle mapping flow, the state sensing module aims to achieve microscopic tooth surface position positioning. The position acquisition device uses a multi-turn absolute encoder installed on the non-drive end of the turning gear motor input shaft. During the initial grid-connected commissioning of the wind turbine, the state sensing module sets a reference meshing point between the turning gear pinion and the main drive gear, at which point the output value of the multi-turn absolute encoder is recorded as the absolute zero position. The 360-degree physical space of the main drive gear's circumference is proportionally mapped into the controller's memory.

[0029] When the wind turbine stops and is ready to execute the turning gear command, the state sensing module reads the current angle raw feedback value from the multi-turn absolute encoder. The state perception module internally stores the fixed gear ratio of the turning gear transmission mechanism. The transmission ratio The ratio of the number of teeth on the main drive gear to the number of teeth on the turning gear is defined as the microprocessor's calculation of the real-time meshing phase angle using the following logical formula:

[0030]

[0031]

[0032] In the above mapping calculation model, This is the cumulative angular displacement of the rotary motor shaft relative to its absolute zero position; This is the effective angular displacement corresponding to the main drive gear after gear reduction mapping; The reference starting phase angle is set (usually set to ). ); This refers to the real-time meshing phase angle between the current turning pinion and the main drive gear on the circumferential circumference, calculated by mapping. This indicates that a modulo division operation is performed to ensure that the real-time engagement phase angle of the output is strictly distributed within... Within the closed interval. Through this transmission ratio dimensionality reduction mapping calculation, the state perception module accurately converts the high-speed multi-turn rotation data at the turning gear motor end into a unique, stationary physical contact phase coordinate at the large gear end, and then... With calculation The data is simultaneously sent to the optimization decision-making module as basic data input.

[0033] Furthermore, the specific operation procedure of the fatigue matrix module is as follows: The fatigue matrix module is deployed in the non-volatile storage unit of the engine room controller or in a cloud database that communicates with it. It is used to construct and maintain a digital fatigue characteristic distribution map that maps to the physical space of the main drive gear. The core of the fatigue matrix module is to transform the continuous mechanical wear process into a discrete data matrix and to achieve rolling updates of fatigue data through a closed-loop feedback mechanism.

[0034] The fatigue matrix module internally stores a tooth surface phase tracking matrix, which is logically defined as a... A two-dimensional data array. Among them, This represents the number of discretized phase angle intervals that proportionally divide the 360-degree circumferential physical space of the main drive gear. For example, in this embodiment, the phase angle step size is set to... ,but The value is 360, thus forming 360 discrete phase angle interval indices. The second dimension of the matrix represents the direction of transmission. ,in Corresponding to the tooth surface fatigue data in the clockwise direction, This corresponds to tooth surface fatigue data in the counter-clockwise direction. Each element in the matrix... It stores the dimensionless cumulative wear coefficient for the corresponding phase angle interval in the corresponding transmission direction.

[0035] Before performing the turning operation, the fatigue matrix module receives the real-time engagement phase angle input from the aforementioned state perception module. The fatigue matrix module determines the current phase angle interval index through a rounding function or lower bound mapping logic. Its mapping logic expression is:

[0036] Using this index, the fatigue matrix module extracts the clockwise local tooth surface fatigue parameters corresponding to the specific meshing position from the matrix in parallel. And counterclockwise local tooth surface fatigue parameters These two parameters, as raw quantified data, are transmitted in real time to the optimization decision module to participate in the subsequent cost function calculation.

[0037] The key feature of the fatigue matrix module lies in its built-in adaptive update sub-unit. After each turning motion, this sub-unit is responsible for calculating the single fatigue increment generated by that turning motion. To accurately simulate the microscopic damage to gears during power transmission, the adaptive update sub-unit employs a fatigue calculation model based on the load power consumption integral:

[0038] In the above fatigue calculation model, This represents the fatigue increment per instance. and These represent the start and end times of this turning operation, respectively. The preset material fatigue transformation constant is determined by the SN fatigue curve of the gear material; For the turning motor in The real-time output torque at any given moment is obtained by collecting the real-time current vector of the frequency converter and calculating it in conjunction with the torque coefficient. This represents the real-time angular velocity of the motor. Based on lubricating oil temperature Wear-weighted correction factor.

[0039] Wear-weighted correction factor It is a key parameter reflecting the physical properties of materials. Because the viscosity of lubricating oil increases as temperature decreases, it is difficult to form a continuous fluid lubricating oil film on the gear meshing surface at low temperatures, leading to increased boundary friction. The fatigue matrix module has a pre-stored temperature-wear correction coefficient curve. When... Below the preset low temperature threshold hour, The value of increases exponentially as the temperature decreases, thereby compensating for the additional mechanical spalling and pitting damage caused under low-temperature conditions.

[0040] When the fatigue increment is After the calculation is completed, the adaptive update sub-unit updates according to the actual execution direction of the turning gear. The phase tracking matrix of the tooth surface is updated by overlaying data across the phase angle interval it spans. The update formula is as follows:

[0041] Through the aforementioned iterative process, the fatigue matrix module can track the health status of every tiny physical part on the circumference of the large gear in real time. This storage and update mechanism based on the "position-direction-load-temperature" framework enables the system to identify specific tooth surfaces that have undergone excessive fatigue, thereby providing the optimization decision-making module with spatiotemporally relevant decision support data and avoiding the reloading of fatigued parts during gear start-up.

[0042] Furthermore, the specific operation process of the optimization decision-making module is as follows: The optimization decision-making module is deployed in the edge computing node of the main control console of the wind turbine generator's nacelle, and interacts with the state awareness module and fatigue matrix module via a high-speed internal bus. The core execution logic of the optimization decision-making module is to perform quantitative evaluation and optimization of two candidate transmission directions, clockwise and counterclockwise, based on a multi-objective cost function model.

[0043] Let the candidate transmission direction parameter be ,in It represents the clockwise direction of transmission. The direction of transmission is counterclockwise. The multi-objective cost function within the optimization decision module. Penalty term for starting resistance Local fatigue penalty item Composition. To eliminate the physical quantity of torque (unit: To resolve the dimensional conflict between the input parameters and fatigue physical quantities (dimensionless or length evolution units), the microprocessor first performs a normalization mapping calculation on the input parameters.

[0044] Penalty for starting resistance The optimization decision-making module extracts the comprehensive static torque parameters output by the state perception module. Clockwise torque is defined as positive, and counterclockwise torque as negative. The microprocessor calculates the vector consistency metric between the candidate transmission direction and the stationary torque. Its normalized mapping formula is:

[0045] In the formula, This is a sign function that outputs 1 when the variable is greater than zero and -1 when it is less than zero. Achieved candidate directions Mapped to physical vector symbols . This represents the limit torque constant that the inverter can overcome at its permissible output. When the candidate direction is in the same direction as the combined static torque vector (i.e., starting with the momentum), the resistance penalty term in that direction... It is mapped to a minimum value (close to 0); when the two are reversed (i.e., starting against the trend), the resistance penalty term increases accordingly. As the absolute value increases, it linearly amplifies, eventually approaching 1.

[0046] Targeting localized fatigue penalties The optimization decision module extracts the corresponding local tooth surface fatigue parameters output by the fatigue matrix module. The microprocessor employs an extremum normalization algorithm:

[0047] in, This is the preset critical value for tooth surface wear failure. This penalty term directly characterizes the current health and deterioration of the physical contact surface in this transmission direction.

[0048] After normalization, the optimization decision module performs a weighted summation operation:

[0049] Among them, the first weight coefficient With the second weighting coefficient It is not a fixed constant, but is adjusted in real time by the module's built-in dynamic weight allocation unit, and meets the constraints. The dynamic weight allocation unit operates within a set time window. Internally, data from external wind speed sensors is collected, and the variance of wind speed fluctuation rate deviation is calculated. .

[0050] when When the wind fluctuation exceeds the preset high-frequency fluctuation threshold, it indicates a severe external aerodynamic environment and an extremely high risk of sudden changes in wind load resistance. At this point, the dynamic weight allocation unit uses an exponential nonlinear function to increase... The value of (e.g., let) This causes the multi-objective cost function to be in a decision-making mode of "prioritizing compliance with aerodynamic loads" during optimization. Conversely, when When the wind frequency fluctuation threshold is less than the threshold and the system detects a high average historical wear value for the current gears, the dynamic weight allocation unit will... The value of is increased (e.g., let ) This puts the system in a decision-making mode that prioritizes avoiding fatigue accumulation.

[0051] Complete the clockwise cost With counterclockwise cost After the calculation, the microprocessor executes the comparison and determination instruction. Under normal operating conditions, the optimization decision module will output the minimum value. The corresponding candidate direction is used as the final optimal transmission direction command.

[0052] To ensure the absolute safety of the wind turbine drivetrain, the optimization decision module is also equipped with an over-limit safety protection subunit. Before the final direction command is issued, the over-limit safety protection subunit intercepts... and The value and the preset safe startup cost threshold Perform a logical comparison. If the system determines... and (That is, regardless of which direction the engine starts, it will face extremely high resistance overload or severely aggravate the wear of the critical tooth surface.) The over-limit safety protection subunit will directly trigger the hardware-level interrupt interception mechanism to prevent the optimal transmission direction command from being sent to the frequency converter.

[0053] Simultaneously, the over-limit safety protection subunit sends a pneumatic unloading request frame to the wind turbine's upper-level pitch control system via the nacelle's internal CAN bus. Upon receiving this request, the upper-level pitch control system drives the hydraulic or electric pitch mechanism to actuate, causing the pitch angle of the three blades to move in the feathering direction. Direction) deflection at a preset angle (e.g.) This actively reduces the blade's wind-catching area. After the pitch control operation is completed, the state awareness module recalculates the reduced overall static torque parameters. This triggers the multi-objective cost function calculation closed loop of the optimization decision module again, until the cost of at least one candidate transmission direction decreases to... Only then can the command issuance channel be unlocked. This cross-system closed-loop unloading mechanism ensures the self-protection capability of the turning gear system under extreme operating conditions.

[0054] Furthermore, the specific operation process of the flexible execution module is as follows: The flexible actuator module mainly consists of a dedicated variable frequency drive (VFD) for the turning gear, a servo drive controller, and a turning gear motor located in the engine compartment. The flexible actuator module communicates with the optimization decision module via a fieldbus (such as EtherCAT or PROFINET) to convert the optimal drive direction command at the software level into a low-level motor stator current vector control sequence, achieving compliant drive that completely eliminates the physical impact of gear meshing.

[0055] The control timing of the flexible execution module is strictly divided into a reverse pre-tightening disengagement stage and a forward flexible acceleration stage.

[0056] After receiving the optimal transmission direction command from the optimization decision module, the flexible execution module first enters the reverse pre-tightening disengagement stage. Because the wind turbine is shut down for a long time, the meshing surface of the main drive large gear and the turning gear has extremely high static contact stress under static load. Directly starting in the optimal transmission direction is very likely to cause torsional oscillation of the transmission chain due to sudden changes in static friction.

[0057] To this end, the servo drive controller first analyzes the non-optimal transmission direction, which is logically opposite to the optimal transmission direction, and then sends a micro-torque command to the dedicated frequency converter for turning gears. The frequency converter uses a field-oriented control (FOC) algorithm to inject transient torque polarization current (q-axis current) into the stator windings of the motor while keeping the rotor excitation current (d-axis current) constant, so as to output the preset micro-torque. The formula for calculating the fretting torque is:

[0058] in, This is the rated output torque of the turning gear motor; This is the fretting torque coefficient, and its value is strictly limited to a range of values ​​within 100°C. Between. The frequency converter maintains the micro-torque output for an extremely short time extreme value. (For example, 200 milliseconds), and then immediately reduce the output torque to zero.

[0059] The physical mechanism of this reverse micro-motion stage is as follows: by applying a transient torque that is much lower than the system's starting frictional resistance torque, the pinion of the turn gear will not undergo macroscopic continuous rotational displacement, but it is sufficient to overcome the local static biting force between the meshing tooth surfaces, forcing the originally tightly fitted stress contact surfaces to undergo microscopic slippage and actively disengage, thereby artificially creating a preset tooth flank clearance in the optimal transmission direction to be executed, and completely cutting off the rigid resistance transmission path at the moment of start-up.

[0060] After completing the reverse preload release phase, the flexible actuation module seamlessly switches to the adaptive flexible acceleration phase. The frequency converter begins to output a smooth acceleration curve in the optimal transmission direction. To ensure that the instantaneous impact kinetic energy of the turning pinion approaches zero when it re-engages with the main drive gear across the aforementioned tooth flank clearance, the servo drive controller incorporates a shock-free speed planning algorithm based on a fifth-order polynomial.

[0061] Programming function for the target angular velocity of the turning motor output by the frequency converter As shown below:

[0062] In the formula, This indicates the real-time angular velocity of the motor during acceleration. This represents the steady-state rated turning angular velocity required by the fan manufacturing process. The set total acceleration climb cycle time; For the real-time time variable when the self-acceleration phase begins ( ).

[0063] This fifth-order polynomial velocity programming function has extremely rigorous dynamic characteristics: the angular acceleration is obtained by taking its first derivative. The function, and the jerk (i.e., the rate of change of acceleration) function obtained by taking the second derivative, at the initial moment and end time The values ​​of all of them are strictly zero.

[0064] The engineering essence of this advanced S-shaped acceleration control logic is as follows: In the initial stage of startup, the motor's angular acceleration is slowly built up at an extremely low rate of change, causing the pinion gear to be in a flexible, free state of "extremely low speed - extremely low acceleration" when crossing the pre-manufactured tooth flank clearance; only after the tooth surfaces gently re-engage (clearing the clearance) does the acceleration gradually increase according to the curve characteristics, eventually smoothly transitioning to a constant value. This control process eliminates tooth surface collision and tooth knocking caused by acceleration step jumps at their source.

[0065] Once the turning gear system reaches the set target position or duration, the flexible execution module receives a stop command from the host computer. The frequency converter executes the corresponding deceleration and stop curve. After the motor comes to a complete stop, the flexible execution module sends a run-end feedback frame containing the actual execution direction, running torque integral, and running time to the fatigue matrix module via the control bus. This triggers the adaptive update subunit to perform closed-loop iterative calculations on the tooth surface phase tracking matrix, thereby completing the full lifecycle data closed loop of a single turning gear control cycle.

[0066] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A wind turbine turning gear transmission direction optimization system based on big data analysis, characterized in that, include: The status perception module is used to collect the comprehensive static torque parameters of the wind turbine generator in the shutdown state in real time, and analyze the real-time meshing phase angle between the current turning pinion and the main drive gear through the position acquisition device; The fatigue matrix module is communicatively connected to the state perception module and has a built-in tooth surface phase tracking matrix based on circumferential distribution. It is used to extract the local tooth surface fatigue parameters of the corresponding meshing position in the clockwise and counterclockwise transmission directions according to the real-time meshing phase angle. The optimization decision module is communicatively connected to the state perception module and the fatigue matrix module, and has a built-in multi-objective cost function model. It is used to input the comprehensive static torque parameter and the local tooth surface fatigue parameter into the multi-objective cost function model after normalization, comprehensively calculate the evaluation cost of the two candidate transmission directions, clockwise and counterclockwise, and lock the candidate transmission direction with the minimum evaluation cost as the optimal transmission direction. The flexible execution module is communicatively connected to the optimization decision module. It is used to receive the optimal transmission direction command and, before the formal start of the turning gear, output a preset micro-torque to the non-optimal transmission direction to actively disengage from the tooth surface and eliminate gear meshing clearance. Then, it outputs a smooth acceleration curve to drive the turning gear system according to the optimal transmission direction.

2. The wind turbine turning gear transmission direction optimization system based on big data analysis according to claim 1, characterized in that, The comprehensive static torque parameter includes aerodynamic torque components and gravitational eccentricity torque components; The state perception module includes an anemometer and a rotor position sensor, used to collect real-time wind speed outside the nacelle, the relative angle between the nacelle and the wind direction, and the stopping azimuth angle of the wind turbine rotor, so as to calculate the aerodynamic torque component and the gravitational eccentric torque component. The position acquisition device includes a multi-turn absolute encoder installed on the input shaft of the turning motor or the main drive chain; The state perception module obtains the current rotor zero offset by reading the angle data of the multi-turn absolute encoder, and calculates the real-time meshing phase angle on the circumferential circumference of the current turning pinion and the main drive gear based on the fixed tooth ratio of the turning pinion and the main drive gear.

3. The wind turbine turning gear transmission direction optimization system based on big data analysis according to claim 2, characterized in that, When acquiring the comprehensive static torque parameters, the state perception module uses a superimposed evaluation torque model. The superimposed evaluation torque model vector sums the aerodynamic torque component under the current shutdown azimuth angle and real-time wind speed with the gravitational eccentricity torque component caused by blade mass eccentricity, to obtain the direction and magnitude of the comprehensive static load currently acting on the transmission chain, which serves as the basic calculation input value for the optimization decision module.

4. The wind turbine turning gear transmission direction optimization system based on big data analysis according to claim 3, characterized in that, The tooth surface phase tracking matrix in the fatigue matrix module divides the 360-degree circumference of the main drive gear into several discrete phase angle intervals. When extracting the local tooth surface fatigue parameters, the fatigue matrix module first maps the real-time meshing phase angle input by the state perception module to the corresponding phase angle interval, and then retrieves the historically stored clockwise cumulative wear coefficient and counterclockwise cumulative wear coefficient in parallel from the phase angle interval. The fatigue matrix module is also configured with an adaptive update sub-unit; After each wind turbine turning action is completed, the adaptive update subunit obtains the actual execution direction, actual running time, and load current or output torque of the turning motor for this turning action. It then generates the single fatigue increment for this operation through time integration calculation and adds the single fatigue increment to the historical cumulative wear coefficient of the corresponding phase angle interval to realize the dynamic iteration of the tooth surface phase tracking matrix.

5. The wind turbine turning gear transmission direction optimization system based on big data analysis according to claim 4, characterized in that, When calculating the single fatigue increment, the adaptive update subunit also incorporates a weighted correction based on gearbox ambient temperature parameters or lubricating oil temperature parameters obtained from the SCADA system. When the ambient temperature parameter or lubricating oil temperature parameter is lower than the preset low temperature threshold, the calculated value of the single fatigue increment in the corresponding phase angle interval is increased by increasing the correction weight coefficient, so as to compensate for the additional mechanical wear caused by poor gear lubrication under low temperature conditions.

6. The wind turbine turning gear transmission direction optimization system based on big data analysis according to claim 5, characterized in that, The multi-objective cost function model built into the optimization decision module includes an initiation resistance penalty term and a local fatigue penalty term; For any candidate transmission direction, the cost of the starting resistance penalty term is calculated based on the consistency of the vector direction of the candidate transmission direction with the comprehensive static torque parameter. When the vector directions are in the same direction, a smaller cost is output, and when the vector directions are in opposite directions, a larger cost is output. The cost of the local fatigue penalty term is positively correlated with the local tooth surface fatigue parameter corresponding to the candidate transmission direction. The optimization decision module performs a weighted summation of the starting resistance penalty term and the local fatigue penalty term to obtain the evaluation cost of the corresponding candidate transmission direction.

7. The wind turbine turning gear transmission direction optimization system based on big data analysis according to claim 6, characterized in that, The optimization decision module is also equipped with a dynamic weight allocation unit, which is used to dynamically adjust the first weight coefficient of the starting resistance penalty term and the second weight coefficient of the local fatigue penalty term according to the external working conditions perceived in real time. When the real-time wind speed fluctuation rate is higher than the preset wind condition threshold, the dynamic weight allocation unit increases the first weight coefficient to prioritize adapting to aerodynamic load resistance. When the real-time wind speed fluctuation rate is lower than the preset wind condition threshold and the historical wear average value of the preset period is high, the second weighting coefficient is increased to preferentially avoid tooth surface wear accumulation.

8. The wind turbine turning gear transmission direction optimization system based on big data analysis according to claim 7, characterized in that, The optimization decision-making module also includes an over-limit safety protection subunit; When comparing the evaluation cost values, if the evaluation cost values ​​of both the clockwise and counterclockwise candidate transmission directions exceed the preset safe start cost threshold, the over-limit safety protection subunit will intercept the issuance of the optimal transmission direction command and send an aerodynamic unloading request command to the upper-level pitch system. After the upper-level pitch system adjusts the blade pitch angle to reduce the comprehensive static torque parameter, the calculation of the multi-objective cost function model will be retried.

9. The wind turbine turning gear transmission direction optimization system based on big data analysis according to claim 8, characterized in that, The flexible execution module includes a turning gear frequency converter and a turning gear motor; When the turning gear frequency converter drives the turning gear system in the optimal transmission direction, it adopts an S-type speed command control strategy. By smoothly adjusting the derivative of acceleration, it suppresses the torque pulsation of the turning gear motor during the start-up and acceleration phases, ensuring that the turning gear pinion and the main drive gear complete a flexible re-meshing without impact. When the flexible actuator outputs the preset micro-torque, it adopts a non-displacement torque preload control strategy. The turning gear inverter applies a limiting torque, which is a preset percentage of the rated torque of the turning gear motor, to the non-optimal transmission direction, and quickly blocks the output after maintaining a preset micro-motion duration; the magnitude of the limiting torque and the preset micro-motion duration are set such that the turning gear pinion is only driven to overcome static friction to produce a small elastic yield or gap separation between the tooth surfaces, without causing the main shaft of the wind turbine generator to deflect macroscopically.

10. The wind turbine turning gear transmission direction optimization system based on big data analysis according to claim 9, characterized in that, The flexible execution module is also equipped with a startup anti-jamming monitoring subunit; During the initial climbing phase when the turning motor accelerates in the optimal transmission direction, the start-up anti-jamming monitoring subunit monitors the output current or estimated torque of the turning frequency converter in real time. If the rate of change of the output current or estimated torque exceeds the preset hard contact sudden change threshold before the rated turning speed is reached, the flexible execution module immediately interrupts the current turning command and triggers the parking brake, while simultaneously sending an abnormal collision fault code back to the cloud.

Images