A cluster control method for extending the average life of wind farm pitch bearings

Abstract

This invention discloses a cluster control method for extending the average lifespan of wind farm pitch bearings. The method reads wind farm dispatch instructions in real time to determine whether power restrictions are in effect. Based on different scenarios, the corresponding pitch bearing lifespan extension control strategy is automatically activated. This method consistently adheres to the principle of satisfying wind farm dispatch instructions without sacrificing overall farm power generation. It uses a fixed-pitch angle control algorithm in real time to change the minimum pitch angle of the wind turbine, thereby limiting the maximum pitch angle of the wind turbine and avoiding frequent pitch changes caused by wind speed fluctuations. By reducing the wind turbine's pitch stroke, wear on the pitch bearing is reduced, extending its service life.

Description

Technical Field

[0001] The present invention relates to the technical field of wind power generation measurement and control, and in particular to a cluster control method for extending the average life of variable pitch bearings in a wind farm. Background Art

[0002] Wind energy is an inexhaustible resource with low costs. Wind power generation is currently the most technologically mature, scalable, and promising renewable energy source. Countries around the world, especially developed countries, attach great importance to the development of wind power, viewing it as a crucial measure to adjust energy structures, protect the environment, rationally utilize resources, and achieve sustainable development. Currently, China is following the lead of developed countries and vigorously promoting wind power generation.

[0003] With favorable national policies, wind power generation in China is booming. The number of installed units and installed capacity are increasing annually. Pitch bearings are primarily used to adjust the orientation and pitch angle of wind turbine blades, ensuring they face the wind vertically and maintaining stable output power within a safe and efficient range. During wind turbine operation, the aerodynamic forces acting on the blades vary. Wind speed varies not only in magnitude but also in direction, generating gyroscopic torques. This inertial load acts on the blades, creating a variable inertial load. This load, in turn, acts as a source of vibration. This vibration causes the blades to vibrate. The blades are bolted to the inner race of the pitch bearing. Blade vibration can loosen the bolts, resulting in uneven force distribution on the blades. This can cause the inner and outer races of the bearing to misalign under load, changing the contact angle. This can lead to uneven loading on the rolling elements or generate edge forces between the rolling elements and the raceway, potentially causing the rolling elements to seize. Pitch bearings are mounted on the support base to ensure sufficient bearing rigidity and uniform stress distribution. Improper slewing bearing installation can lead to operational deformation during bearing operation. Wind turbine blades adjust their pitch to maximize wind energy and ensure power generation. However, due to constant and frequent operation, the lifespan of pitch bearings decreases over time. However, each turbine is independent, and the operating stroke data for pitch bearings is inconsistent. By implementing different strategies based on the specific conditions of the wind farm and wind turbines, we can reduce the total operating stroke of pitch bearings across the entire wind farm, thereby increasing the average lifespan of pitch bearings across the wind farm. Summary of the Invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present invention provides a cluster control method for extending the average service life of wind farm pitch bearings, which solves the problem of low average service life of wind farm pitch bearings.

[0005] In order to achieve the above-mentioned invention object, the technical solution adopted by the present invention is as follows: a cluster control method for extending the average life of wind farm pitch bearings, reading wind farm dispatch instructions in real time to determine whether power is limited;

[0006] In the case of unlimited power supply at a wind farm: select all fully-powered units in the farm and call the fixed-pitch control algorithm in real time based on the power and wind speed changes of the fully-powered units to reduce the number of pitch change operations for all units in the farm, thereby reducing the pitch change stroke of all units in the farm and extending the life of all pitch bearings in the farm;

[0007] In the case of power rationing at a wind farm: the pitch stroke status of all wind turbines in the farm is counted in real time, and they are sorted from large to small to determine whether all wind turbines are in a stable and full-power state. The life extension unit screening algorithm is called in real time according to the unit status, and the units entering the life extension control are selected. The fixed pitch angle control strategy is started for the units entering the life extension. During the fixed pitch angle control process, it is determined in real time whether the actual power generation of the current wind farm has the risk of over-generation. If over-generation occurs, the units entering the life extension control need to be reselected to ensure that the actual power generation of the current wind farm meets the dispatch instructions.

[0008] Furthermore: In the case of no power restriction at a wind farm, if the wind farm has not received a power restriction order, all fully-generated units in the farm will be screened out. The fully-generated units will change their minimum blade angles, and the power and wind speed will be judged according to a certain time period, and a closed-loop life extension control strategy will be implemented through the changes in power and wind speed.

[0009] Furthermore: the closed-loop life extension control strategy is specifically as follows: for the fully-generated unit, a minimum blade angle is set, the wind turbine is retracted to the minimum blade angle, and it is judged in real time whether the power of the current fully-generated unit has decreased. If the power has decreased and is less than the full power, it is further judged whether the wind speed has decreased synchronously. If the wind speed has not changed, the blade angle reduction algorithm is called, and the wind turbine is re-adjusted to meet the power generation power. If the wind speed has decreased synchronously, the minimum blade angle setting is released and the life extension control is exited. If the power has not changed, the current blade angle for a certain period of time is synchronously collected, and according to the blade angle situation, it is judged whether to call the blade angle increase algorithm.

[0010] Furthermore: the pitch angle reduction algorithm is specifically as follows: the fixed pitch angle with the smallest power loss under the current wind speed is predicted through the wind speed pitch angle prediction model, the wind speed pitch angle prediction model establishes a regression model through historical data, and uses the least squares method to determine the optimal regression coefficient, and the optimal pitch angle under the current wind speed is predicted by the optimal regression coefficient.

[0011] Furthermore: when the power does not change, the full power state is maintained, and the fan blade angle within a period of time is synchronously collected to form a blade angle sequence. By judging whether the blade angle in the blade angle sequence is greater than the set minimum blade angle, if it is greater, the blade angle raising algorithm is called to change the minimum blade angle and readjust the blade angle of the wind turbine group. If it is less than, the current minimum blade angle is maintained.

[0012] Furthermore: the pitch angle raising algorithm is specifically as follows: the angle increase is limited to the main control output pitch angle exceeding the current angle, and a vibration zero-state response simulation method is adopted. It is assumed that exceeding the pitch angle brings an impact to the wind turbine, and the impact residual capacity after a period of time is collected as the basis for judging the pitch angle increase.

[0013] Furthermore: in the case of power rationing at a wind farm, the pitch change strokes of all wind turbines in the wind farm are sorted from large to small, and different status marks are given to the wind turbines according to whether they are in a stable full-power state. The life extension wind turbine screening algorithm is called according to the status mark and the wind farm power rationing data. The specific wind turbines entering the life extension control are optimized through the algorithm, and the fixed pitch angle control strategy is started for the life extension wind turbines to prevent frequent pitch changes of the units and reduce the formation of pitch changes of the units. In this fixed pitch angle control process, the current power generation of the entire field is monitored in real time to see if there is a risk of over-generation, and the entire control process is closed according to the over-generation situation.

[0014] Furthermore: in a stable full power state, calculate the average wind speed over a period of time, which must be greater than or equal to the rated wind speed of the wind turbine, and calculate the average power over a period of time, which must be within a certain range above and below the rated power of the wind turbine.

[0015] Furthermore, the life extension wind turbine screening algorithm specifically performs power calculation for units in different states. When screening units for life extension, the total power generation of the entire site cannot exceed the dispatch power limit generation requirement. In order to minimize the operating stroke of the variable pitch bearings of the entire site, the life extension unit screening algorithm needs to select more wind turbines for life extension control, and the cumulative stroke of the wind turbines entering life extension control must be the largest.

[0016] Assume that the upper limit of the total power is TP, the wind turbines in the field are numbered 0, 1, 2, 3, 4, ..., n, where n is the number of wind turbines, the power of each unit is P1, P2, P3, ..., Pn, and the pitch stroke of each unit is S1, S2, S3, ..., Sn. Set up a matrix with a height of n and a width of j, where j + = 1 and j < = TP, and the following relationship exists:

[0017]

[0018] In the above formula, C[i][j] represents the maximum pitch stroke that can be obtained when the upper limit of the total power is j and the optional wind turbines are numbered from 0 to i, and Pi is the unit power when the wind turbine is numbered i.

[0019] Furthermore: the fixed pitch angle control strategy is specifically as follows: the wind turbines are screened out through the life extension unit screening algorithm, the current pitch angles of these units are read in real time, and the fixed pitch angle control is completed by adding a minimum pitch angle on the basis of the current pitch angle; at the same time, the closed loop calculates the current power generation of the entire field in real time to determine whether the power generation meets the scheduling requirements. If not, the life extension wind turbine screening algorithm is re-called and the screening is completed again until the power generation meets the scheduling requirements.

[0020] The beneficial effects of the present invention are as follows: the present invention integrates the power restrictions of wind farms and the power generation conditions of wind turbines, and automatically starts the corresponding variable pitch bearing life extension control strategy for different situations. The present invention always follows the principle of meeting the wind farm scheduling instructions and not losing the power generation of the entire field, and calls the fixed pitch angle control algorithm in real time to change the minimum pitch angle of the wind turbine, thereby limiting the maximum pitch angle of the wind turbine and avoiding frequent pitch changes caused by wind speed changes. By reducing the pitch stroke of the wind turbine, the wear of the variable pitch bearing is reduced and its service life is extended. It can effectively extend the service life of the variable pitch bearings of the entire wind farm, and can effectively reduce the operation and maintenance costs caused by damage to the variable pitch bearings, thereby reducing the cost per kilowatt-hour of the entire wind farm and maximizing the overall benefits of the wind farm. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 Flowchart of the present invention;

[0022] Figure 2 This is a control logic diagram for the present invention under the condition of unlimited power;

[0023] Figure 3 This is a control logic diagram under power-limiting conditions in the present invention. DETAILED DESCRIPTION

[0024] The specific embodiments of the present invention are described below to facilitate understanding of the present invention by those skilled in the art. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the appended claims, these changes are obvious, and all inventions and creations utilizing the concepts of the present invention are protected.

[0025] like Figure 1A cluster control method for extending the average life of wind farm pitch bearings is described. The method includes: integrating wind farm power curtailment and wind turbine generation conditions, and automatically initiating corresponding pitch bearing life extension control strategies based on different scenarios. This method consistently adheres to the principle of meeting wind farm scheduling instructions and ensuring no loss of overall power generation. It invokes a pitch angle control algorithm in real time to change the minimum pitch angle of the wind turbines, thereby limiting their maximum pitch angle and avoiding frequent pitch changes caused by wind speed fluctuations. By reducing the pitch stroke of the wind turbines, it reduces wear on the pitch bearings and extends their service life. First, the wind farm scheduling instructions are read in real time. If power curtailment is not in effect, fully-generated turbines are selected. Based on the power and wind speed fluctuations of these fully-generated turbines, the pitch angle control algorithm is invoked in real time to reduce the number of pitch changes across all turbines, thereby reducing the pitch stroke and extending the life of all pitch bearings. If power curtailment occurs, the pitch stroke of all wind turbines in the wind farm is counted in real time and sorted from highest to lowest. At the same time, it is necessary to determine whether all wind turbines are in a stable, fully powered state. Based on this state, the life extension unit screening algorithm is invoked in real time to select units for life extension control and activate the fixed pitch angle control strategy for these units. During the fixed pitch angle control process, it is determined in real time whether the actual wind farm power generation is at risk of over-generation. If over-generation is present, the units that enter life extension control are reselected to ensure that over-generation is avoided.

[0026] like Figure 2 As shown, if there is no power curtailment, all fully-fed turbines are further screened out. Based on the power and wind speed changes of these fully-fed turbines, the pitch angle control algorithm is invoked in real time to reduce the number of pitch changes required by all turbines, thereby shortening the pitch change stroke and extending the life of all pitch bearings. If the wind farm does not receive a curtailment order, all fully-fed turbines are screened out, while the remaining partially-fed turbines are excluded from the life extension control. The fully-fed turbines adjust their minimum pitch angles and, based on power and wind speed variations over a specified time period, implement a closed-loop life extension control strategy.

[0027] A fully-generated unit changes its minimum blade angle and determines the power and wind speed over a certain period of time, implementing a closed-loop life extension control strategy based on changes in power and wind speed. For a fully-generated unit, an appropriate minimum blade angle is initially set, and the wind turbine retracts its blades to the target blade angle. A real-time determination is made as to whether the current fully-generated unit power has decreased. If the power has decreased and is less than the full power, a further determination is made as to whether the wind speed has decreased synchronously. If the wind speed has not changed, a pitch angle reduction algorithm is invoked to re-adjust the wind turbine blades to meet the power generation capacity. If the wind speed has decreased synchronously, the minimum blade angle setting is released, and life extension control is exited. If the power has not changed, the current blade angle is synchronously collected over a certain period of time, and based on the blade angle, a determination is made as to whether the pitch angle increase algorithm should be invoked.

[0028] The pitch reduction algorithm uses wind speed and pitch angle prediction as a target for accurate positioning, avoiding sudden increases and decreases. The pitch reduction algorithm uses a wind speed and pitch angle prediction model to predict the pitch angle that minimizes power loss at the current wind speed. The key idea behind this model is to establish a regression model using historical data, determine the optimal regression coefficient using the least squares method, and then predict the optimal pitch angle for the current wind speed.

[0029] If the power remains unchanged, the current blade angle is collected over a specific period of time. Based on the blade angle, a decision is made as to whether to invoke the blade angle raising algorithm. If the power remains unchanged after the wind turbine blade angle is adjusted, full power generation is maintained. The blade angles of the wind turbines are collected over a suitable period of time to form a blade angle sequence. A determination is made as to whether a certain percentage of the blade angles in the sequence exceeds the set minimum blade angle. If so, the blade angle raising algorithm is invoked to change the minimum blade angle and readjust. If not, the current minimum blade angle is maintained.

[0030] The pitch angle increase algorithm increases the pitch angle until the master control output pitch angle exceeds the current angle. The first exceeding angle has less impact on the target, while the later exceeding angle has more impact on the target. The algorithm uses the vibration zero-state response simulation method, assuming that the exceeding pitch angle brings an impact to the system. The residual energy of the system impact after a suitable time is collected as a judgment method for increasing the pitch angle. The pitch angle increase algorithm needs to satisfy the following system equations:

[0031] System equations:

[0032] r(t)=e(t)*h(t)

[0033] like Figure 3 As shown, if power curtailment occurs, the pitch strokes of all wind turbines in the wind farm are counted in real time and sorted from largest to smallest. Simultaneously, it is necessary to determine whether all wind turbines are in a stable, full-power state. Based on this state, the life extension unit screening algorithm is invoked in real time to select turbines for life extension control and initiate the pitch angle control strategy for these turbines. The entire pitch angle algorithm invocation process requires real-time correlation with wind turbine power, wind speed, and other data. It also determines in real time whether all turbines in the farm are at risk of over-powering. If this risk is triggered, the closed-loop pitch angle algorithm and the life extension wind turbine screening algorithm are activated to reselect turbines and assign a new minimum pitch angle. The pitch strokes of all wind turbines in the wind farm are sorted from largest to smallest, and wind turbines are assigned different status flags based on whether they are in a stable, full-power state. Based on the status flags and wind farm power curtailment data, the life extension wind turbine screening algorithm is invoked to optimize the wind turbines that will enter life extension control. The pitch angle control strategy is then initiated for these wind turbines with extended life to prevent frequent pitch changes and reduce pitch stroke. During the entire control process, it is necessary to monitor in real time whether the current power generation of the entire field has the risk of over-generation, and close the entire control process based on the over-generation situation.

[0034] A stable full-power state is achieved by calculating the average wind speed over a suitable time period. This average wind speed must be greater than or equal to the rated wind speed of the wind turbine. Simultaneously, the average power is calculated over a suitable time period. The calculated average power must be within a certain range above and below the rated power, ensuring that the calculated average power remains close to the rated power.

[0035] The life extension unit screening algorithm performs power calculations for units in different states. During the screening process for units entering life extension, the total power generation of the entire site cannot exceed the dispatch power restriction requirements. In order to minimize the operating stroke of the variable pitch bearings in the entire site, the life extension unit screening algorithm needs to select as many wind turbines as possible to enter life extension control, and must meet the maximum cumulative stroke of wind turbines entering life extension control. The algorithm draws on the idea of ​​the 0-1 knapsack problem. Its main feature is that there is a container. The core problem to be solved is how to put things with different value combinations into this container to obtain the maximum value. Therefore, the problem is transformed into:

[0036] 1) The total power limit of the entire site is used as a limiting condition, and the total power of each unit does not exceed the total power limit;

[0037] 2) Each unit is either a life-extending fan or a non-life-extending fan;

[0038] 3) Use dynamic programming methods to select the unit combination with the highest pitch stroke possible within the full power limit.

[0039] Assume that the total power limit is TP, the wind turbines in the field are numbered 0, 1, 2, 3, 4, ..., n, the power of each unit is P0, P1, P2, P3, ..., Pn, and the pitch stroke of each unit is S0, S1, S2, S3, ..., Sn. Set up a matrix with a height of n and a width of j (j + = 1, j < = TP), where C[i][j] represents the maximum pitch stroke that can be obtained when the total power limit is j (j < = TP) and the optional wind turbines are numbered from 0 to i, so the following relationship is obtained:

[0040]

[0041] Formula 1 represents the current upper limit of the total power of the entire field, j, and the optional wind turbine numbers are 0 to i. If Pi is greater than j, then the wind turbines currently selected for life extension can only be obtained from 0 to i-1. If j is greater than Pi, then the current wind turbine group pitch stroke + the remaining available power pitch stroke C[i-1][jP i ]+ i , compare with the previous C[i-1][j] pitch stroke and take the larger value.

[0042] This system selects turbines for life extension control and activates the fixed pitch angle control strategy for these turbines. This system primarily uses the life extension turbine screening algorithm to read the current pitch angle of these turbines in real time. This system then adds an appropriate minimum pitch angle to the current pitch angle to achieve fixed pitch angle control. Simultaneously, a closed-loop system calculates the current total power generation in real time to determine whether it meets the dispatch requirements. If it does not, the life extension turbine screening algorithm is re-invoked and the screening is repeated until the dispatch requirements are met.

Claims

1. A cluster control method for extending the average life of wind farm pitch bearings, characterized in that: Read wind farm dispatch instructions in real time to determine whether power is being restricted; In the case of unlimited power supply at a wind farm: select all fully-powered units in the farm and call the fixed-pitch control algorithm in real time based on the power and wind speed changes of the fully-powered units to reduce the number of pitch change operations for all units in the farm, thereby reducing the pitch change stroke of all units in the farm and extending the life of all pitch bearings in the farm; In the event of power rationing at a wind farm: the pitch stroke status of all wind turbines in the farm is counted in real time, and sorted from largest to smallest to determine whether all wind turbines are in a stable and fully powered state. The life extension unit screening algorithm is called in real time based on the unit status, and the units entering life extension control are selected. The fixed pitch angle control strategy is activated for the units entering life extension. During the fixed pitch angle control process, it is determined in real time whether the actual power generation of the current wind farm has the risk of over-generation. If over-generation occurs, the units entering life extension control need to be reselected to ensure that the actual power generation of the current wind farm meets the dispatching instructions; Under the condition of unlimited power supply at a wind farm, if the wind farm has not received any power supply restriction order, all fully-generated units in the farm will be screened out. The fully-generated units will change their minimum pitch angles and judge the power and wind speed over a certain period of time, and implement a closed-loop life extension control strategy based on the changes in power and wind speed. The closed-loop life extension control strategy is specifically as follows: for a fully-generated unit, a minimum blade angle is set, the wind turbine is retracted to the minimum blade angle, and it is judged in real time whether the power of the current fully-generated unit has decreased. If the power has decreased and is less than the full power, it is further judged whether the wind speed has decreased synchronously. If the wind speed has not changed, the blade angle reduction algorithm is called, and the wind turbine is re-adjusted to meet the power generation capacity. If the wind speed has decreased synchronously, the minimum blade angle setting is released and the life extension control is exited. If the power has not changed, the current blade angle for a certain period of time is synchronously collected, and according to the blade angle situation, it is judged whether to call the blade angle increase algorithm.

2. The cluster control method for extending the average life of wind farm pitch bearings according to claim 1 is characterized in that: The pitch angle reduction algorithm is specifically as follows: a wind speed and pitch angle prediction model is used to predict the fixed pitch angle with the smallest power loss under the current wind speed. The wind speed and pitch angle prediction model establishes a regression model through historical data, uses the least squares method to determine the optimal regression coefficient, and predicts the optimal pitch angle under the current wind speed based on the optimal regression coefficient.

3. The cluster control method for extending the average life of wind farm pitch bearings according to claim 1, characterized in that: When the power does not change, the full power state is maintained, and the wind turbine blade angles within a period of time are synchronously collected to form a blade angle sequence. By judging whether the blade angle in the blade angle sequence is greater than the set minimum blade angle, if it is greater, the blade angle raising algorithm is called to change the minimum blade angle and readjust the blade angle of the wind turbine group. If it is less than, the current minimum blade angle is maintained.

4. The cluster control method for extending the average life of wind farm pitch bearings according to claim 1, characterized in that: The pitch angle raising algorithm is specifically as follows: the pitch angle is increased when the main control output pitch angle exceeds the current angle, a vibration zero-state response simulation method is used, it is assumed that exceeding the pitch angle brings an impact to the wind turbine, and the residual capacity of the impact after a period of time is collected as the basis for judging the pitch angle increase.

5. The cluster control method for extending the average life of wind farm pitch bearings according to claim 1, characterized in that: In the case of power rationing at a wind farm, the pitch strokes of all wind turbines in the wind farm are sorted from large to small, and different status marks are given to the wind turbines according to whether they are in a stable full-power state. The life extension wind turbine screening algorithm is called according to the status mark and the wind farm power rationing data. The specific wind turbines that enter the life extension control are optimized through this algorithm, and the fixed pitch angle control strategy is started for the life extension wind turbines to prevent the units from frequent pitch changes and reduce the pitch stroke of the units. In this fixed pitch angle control process, the current power generation of the entire field is monitored in real time to see if there is a risk of over-generation, and the entire control process is closed according to the over-generation situation.

6. The cluster control method for extending the average life of wind farm pitch bearings according to claim 5, characterized in that: Under a stable full-power state, calculate the average wind speed over a period of time. The average wind speed must be greater than or equal to the rated wind speed of the wind turbine. Also calculate the average power over a period of time. The average power must be within a certain range above and below the rated power of the wind turbine.

7. The cluster control method for extending the average life of wind farm pitch bearings according to claim 5, characterized in that: The life extension wind turbine screening algorithm specifically performs power calculations for units in different states. When screening units for life extension, the total power generation of the entire site cannot exceed the dispatch power limit generation requirements. To minimize the operating stroke of the variable pitch bearings of the entire site, the life extension unit screening algorithm needs to select more wind turbines for life extension control, and the cumulative stroke of the wind turbines entering life extension control must be the largest. Assume that the upper limit of the total power is The fan numbers of the whole site are , is the number of fans, and the power of each unit is , the pitch stroke of each unit is , set up a high , width is , , The matrix has the following relationship: In the above formula, Indicates that when the power limit of the entire field is , and the optional fan numbers are 0 to The maximum pitch stroke that can be obtained under the condition of The fan number is The unit power at that time.

8. The cluster control method for extending the average life of wind farm pitch bearings according to claim 5, characterized in that: The fixed pitch angle control strategy is specifically as follows: the wind turbines are screened out through the life extension unit screening algorithm, the current pitch angles of these units are read in real time, and the fixed pitch angle control is completed by adding a minimum pitch angle on the basis of the current pitch angle; at the same time, the current power generation of the entire field is calculated in a closed loop in real time to determine whether the power generation meets the scheduling requirements. If not, the life extension wind turbine screening algorithm is called again and the screening is completed again until the power generation meets the scheduling requirements.

Images