Early Warning and Life Assessment Method for Yaw Friction Plates of Wind Turbines Based on SCADA Data

Abstract

The wind turbine yaw friction plate early warning and life assessment method based on SCADA data includes the following steps: Step 1, setting standard data; Step 2, filtering yaw time data, filtering out the wind speed, cabin temperature, and cabin vibration within each yaw time period according to the timestamp, and performing the following calculations in sequence: average wind speed, RMS value, and peak-to-peak value of cabin vibration; Step 3, real-time alarm calculation; Step 4, trend alarm calculation; Step 5, cumulative alarm calculation. This method solves the current problem in the wind power industry where yaw friction plate contamination cannot be reported in advance; this algorithm does not require additional sensors and hardware, and can achieve early warning of yaw friction plate problems only through analysis of the unit's own operating data; it solves the current extensive operation and maintenance management model of yaw friction plates and achieves more accurate life statistics and estimation.

Description

Technical Field

[0001] The present invention relates to the field of wind turbine detection, and in particular to a wind turbine yaw friction plate early warning and life assessment method based on SCADA data. Background Art

[0002] The yaw system of a wind turbine (hereinafter referred to as a wind turbine) has the following functions: 1. It enables the turbine to track wind direction in real time according to the main control system's wind-following logic, thereby improving wind energy utilization; and 2. It provides reverse yaw unwinding when the cable is twisted, protecting the cable. A basic overview of the yaw system's operation: 1. When the turbine does not need to yaw, the yaw system maintains the nacelle's position. At this time, the yaw motor's electromagnetic brake engages, and the yaw brake is pressurized to the system's designed brake pressure. The yaw system resists the turbine's yaw torque through the friction torque provided by the yaw brake pads and the yaw drive motor, maintaining the nacelle's position. 2. When the main control system issues a yaw command, the yaw system first reduces the hydraulic station brake pressure. At this point, the yaw brake pressure does not drop to zero, but instead decreases to a set yaw brake pressure. The yaw motor then releases the electromagnetic brake and drives the yaw system to yaw. The friction torque provided by the yaw brake pads acts as a damping torque, maintaining a stable yaw of the yaw system.

[0003] Defects and shortcomings of existing technology:

[0004] (1) Currently, wind turbine yaw friction pads are equipped with wear alarms or wear indicators. When the friction pads are worn to a certain extent, the unit will automatically sound an alarm, or maintenance personnel will regularly check the wear indicator to understand the wear condition of the yaw friction pads. However, a common problem in the industry is that due to hydraulic pipeline oil leakage, wear carbon dust accumulation, sand, dust, and water vapor contamination, the yaw friction pads will form glaze on their working surface before they are worn to the set alarm thickness, causing the friction pads to operate abnormally. The long-term accumulation of this glaze layer will cause abnormal vibration and noise during the yaw process. Currently, the wind power industry does not have an alarm for yaw friction pad contamination. By the time large yaw vibrations and noise are discovered on site, the friction pads cannot be restored through cleaning. To better maintain the friction pads and extend their service life, wind farm operation and maintenance urgently need early warning of yaw contamination.

[0005] Currently, industry standards for yaw friction plate replacement are unscientific, often based on the turbine's operating life or when wear reaches a certain thickness. Due to the varying locations of individual turbines within a wind farm, the actual yaw time of wind turbines in different areas varies significantly. To optimize maintenance planning, more precise calculations of the yaw time for each turbine in each time period are required, and friction plate lifespan management based on this yaw time is then performed. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to provide a wind turbine yaw friction plate early warning and life assessment method based on SCADA data, so as to solve the current problem that the yaw friction plate is contaminated and an alarm cannot be given in time.

[0007] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0008] The wind turbine yaw friction plate early warning and life assessment method based on SCADA data includes the following steps:

[0009] Step 1: Set the calculation period of this algorithm to D days, the data call interval to S seconds, and the yaw brake pressure comparison standard data to P bar;

[0010] Step 2: Within the calculation period D days, filter out data with an interval of S seconds from the wind turbine's SCADA system based on the yaw speed variable: data with yaw speed > 0 and yaw brake pressure > P bar. Based on the timestamps of the data points, count the yaw moments into N consecutive time periods, each with a duration of t1, t2, ..., tn. At the same time, filter out the wind speed, cabin temperature, and cabin vibration within each yaw time period based on the timestamps, and perform the following calculations in sequence: average wind speed, RMS value, and peak-to-peak value of cabin vibration.

[0011] In the calculated period of D days, the total yaw time is T, T=t1+t2+…+tn;

[0012] Step 3: Compare the RMS value and vibration peak value of the cabin vibration in N consecutive time periods with the set standard value, determine whether to issue a real-time alarm, and output a corresponding alarm;

[0013] Step 4: Screen each wind turbine in each calculation cycle based on whether it is in the standard calculation wind speed range at the yaw moment. Record the RMS value, peak-to-peak value, and average value of the cabin vibration at the yaw moment in the data within the standard calculation wind speed range. The data is divided into bins according to temperature. The one-year data of each unit in the wind farm is divided into bins according to the above method. Perform statistical calculations on the data of each bin, calculate the average value of the RMS value of the cabin vibration in each bin, calculate the average value of the peak value of the cabin vibration, use this average value as the benchmark value of the normal friction plate, and compare the RMS value of the cabin vibration in each bin with the benchmark value of the corresponding multiple. According to the comparison results, determine whether the performance of the yaw friction plate has declined.

[0014] Step 5: Accumulate the yaw time of each wind turbine within each calculation cycle D days. When the accumulated yaw time exceeds the corresponding set value, check and maintain the friction plate or prepare spare parts for replacement during the low wind season.

[0015] Step 3: By filtering the data in the SCADA, the data during yaw is screened out by comparing with the set value threshold, and the total yaw time of each yaw period is calculated. The RMS value of the cabin vibration at each moment during the yaw time is used to detect whether there is any abnormality in the friction plate and output an alarm in real time;

[0016] Step 4: Classify the data according to the wind speed range within the yaw moment, detect the RMS value, peak-to-peak value of the cabin vibration, and the average cabin temperature within the standard calculated wind speed range, and divide the data into bins based on the temperature. Then, based on the bin data, determine the average value of the cabin vibration peak value of the normal friction plate and judge the performance trend of the friction plate.

[0017] Step 5: Determine the maintenance and replacement of the friction plate by accumulating time;

[0018] This method can be used to calculate the real-time status, trend and cumulative friction performance of the yaw friction plate, and judge the friction plate from multiple dimensions.

[0019] In the above step 1, the calculation period is D = 1 day, the data calling interval is S = 1 second, and the yaw brake pressure comparison standard data is P = 5 bar.

[0020] In step 3 above, if the RMS value of the cabin vibration at a certain moment in time is greater than 0.3m / s² and less than 0.5m / s² within N consecutive time periods, the system directly outputs a warning: the friction plate surface condition is abnormal; if the RMS value of the cabin vibration at a certain moment in time is greater than or equal to 0.5m / s², and the vibration peak value is greater than 1m / s², the system directly outputs an alarm: please replace the friction plate.

[0021] In step 4 above, the standard calculated wind speed range is 6-8 m / s. The data is divided into bins according to temperature as follows: the first bin: the temperature is less than or equal to -10°C; the second bin: the temperature is greater than -10°C and less than +10°C; the third bin: the temperature is ≥10°C.

[0022] The detailed method for comparing the cabin vibration RMS value of each compartment with the corresponding multiple of the reference value in the above step 4 is as follows: when the cabin vibration RMS value of each compartment is greater than 2.5 times the reference value or the cabin vibration peak value is greater than 2 times the reference value, it is determined that the yaw friction plate performance has deteriorated.

[0023] In the above step 5, when the yaw time reaches 2000 hours, an early warning is issued and the friction plate is inspected and maintained. When it reaches 4000 hours, a reminder is given to prepare spare parts and replace them during the low wind season.

[0024] The present invention provides a wind turbine yaw friction plate early warning and life assessment method based on SCADA data, which solves the current problem in the wind power industry of yaw friction plate contamination and the inability to provide early warning. The algorithm does not require additional sensors and hardware, and can achieve early warning of yaw friction plate problems only through analysis of the unit's own operating data. It solves the current extensive operation and maintenance management model of yaw friction plates and achieves more accurate life statistics and estimation. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The present invention will be further described below with reference to the accompanying drawings and examples:

[0026] Figure 1 Flowchart of the algorithm of the present invention. DETAILED DESCRIPTION

[0027] The following describes the technical solution of the present invention in detail with reference to the accompanying drawings and embodiments, taking the operating data of a wind farm as an example:

[0028] The algorithm uses 1 day as the calculation cycle and the data calling interval is in seconds. This algorithm uses 1-second interval data as an example to illustrate the algorithm.

[0029] First, the system uses the yaw speed variable to filter out 1-second data points at the yaw moment, such as data with yaw speed > 0 and yaw brake pressure > 5 bar. This data excludes the yaw release moment, when the yaw brake pressure is completely released to 0. Considering the residual pressure release, the filtering condition is set to 5 bar. The yaw moment is then counted into N consecutive time periods based on the timestamps of the data points. The yaw process is continuous and generally lasts for more than 30 seconds, with each time period lasting t1, t2, ..., tn. The algorithm also filters the wind speed, cabin temperature, and cabin vibration within each yaw time period based on the timestamps, and performs the following calculations: average wind speed, RMS value, and peak-to-peak value of cabin vibration.

[0030] Second, the algorithm calculates the total yaw time T of one day, T=t1+t2+…+tn;

[0031] 3. Real-time alarm algorithm: When the algorithm identifies that the RMS value of the cabin vibration during a certain yaw period is greater than 0.3m / s² and less than 0.5m / s², the system directly issues a warning: the friction plate surface condition is abnormal. When the algorithm identifies that the RMS value during a certain yaw period is greater than or equal to 0.5m / s² and the vibration peak is greater than 1m / s², the system directly issues an alarm: please replace the friction plate.

[0032] 4. Trend alarm: In order to conduct trend analysis, each wind turbine is classified according to the wind speed range at the yaw moment every day. Since yaw mainly occurs in the medium and low wind speed range, this algorithm selects 6-8m / s. When the average wind speed at each yaw moment is in the 6-8m / s wind speed range, the algorithm also records the RMS value, peak-to-peak value of the cabin vibration, and the average cabin temperature during this period. Due to changes in ambient temperature, the thermal expansion and contraction of various components of the unit also affect the vibration during the yaw process. To ensure consistency in trend analysis, the data is also divided into three bins according to temperature. This algorithm is divided into three bins: the first bin: temperature less than or equal to -10°C; the second bin: temperature greater than -10°C and less than +10°C; the third bin temperature ≥10°C. The algorithm divides the one-year data of each unit in the wind farm into bins according to the above algorithm, and performs statistical calculations on the data in each bin. The mean of the RMS value of the nacelle vibration in each bin is calculated, and the mean of the peak value of the nacelle vibration is calculated. This mean is used as the baseline value for normal friction plates, and the alarm line is set according to the following logic. When the RMS value of the nacelle vibration in each bin is greater than 2.5 times the baseline value or the peak value of the nacelle vibration is greater than 2 times the baseline value, the system outputs a degradation of yaw friction plate performance.

[0033] 5. Accumulate the yaw time every day. When the yaw time reaches 2000 hours, issue a warning and inspect and maintain the friction plate. When it reaches 4000 hours, prepare spare parts and replace them during the low wind season.

Claims

1. A wind turbine yaw friction plate early warning and life assessment method based on SCADA data, characterized in that: The following steps are involved: Step 1: Set the calculation period of this algorithm to D days, the data call interval to S seconds, and the yaw brake pressure comparison standard data to P bar; Step 2: Within the calculation period D days, filter out data with an interval of S seconds from the wind turbine's SCADA system based on the yaw speed variable: data with yaw speed > 0 and yaw brake pressure > P bar. Based on the timestamps of the data points, count the yaw moments into N consecutive time periods, each with a duration of t1, t2, ..., tn. At the same time, filter out the wind speed, cabin temperature, and cabin vibration within each yaw time period based on the timestamps, and perform the following calculations in sequence: average wind speed, RMS value, and peak-to-peak value of cabin vibration. In the calculated period of D days, the total yaw time is T, T=t1+t2+…+tn; Step 3: Compare the RMS value and vibration peak value of the cabin vibration in N consecutive time periods with the set standard value, determine whether to issue a real-time alarm, and output a corresponding alarm; Step 4: Screen each wind turbine in each calculation cycle based on whether the yaw moment is within the standard calculation wind speed range. Record the RMS value, peak-to-peak value, and average cabin temperature of the nacelle vibration at the yaw moment within the standard calculation wind speed range. Divide the data into bins according to temperature. Divide the one-year data of each unit in the wind farm into bins according to the above method. Perform statistical calculations on the data in each bin. Calculate the average RMS value of the nacelle vibration in each bin. Calculate the average value of the peak value of the nacelle vibration. Use this average value as the baseline value of the normal friction plate. Compare the RMS value of the nacelle vibration in each bin with the corresponding multiple of the baseline value. Determine whether the yaw friction plate performance has declined based on the comparison results. Step 5: Accumulate the yaw time of each wind turbine within each calculation cycle D days. When the accumulated yaw time exceeds the corresponding set value, check and maintain the friction plate or prepare spare parts for replacement during the low wind season.

2. The wind turbine yaw friction plate early warning and life assessment method based on SCADA data according to claim 1 is characterized in that: In the step 1, the calculation period D is 1 day, the data calling interval is S is 1 second, and the yaw brake pressure comparison standard data is P=5 bar.

3. The wind turbine yaw friction plate early warning and life assessment method based on SCADA data according to claim 1 is characterized in that: In step 3, if the RMS value of the cabin vibration at a certain moment in time is greater than 0.3m / s² and less than 0.5m / s² within N consecutive time periods, the system directly outputs a warning: the friction plate surface condition is abnormal; if the RMS value of the cabin vibration at a certain moment in time is greater than or equal to 0.5m / s² and the vibration peak is greater than 1m / s², the system directly outputs an alarm: please replace the friction plate.

4. The wind turbine yaw friction plate early warning and life assessment method based on SCADA data according to claim 1 is characterized in that: The standard calculated wind speed range in step 4 is 6-8m / s, and the data is divided into bins according to temperature as follows: first bin: temperature is less than or equal to -10°C; second bin: temperature is greater than -10°C and less than +10°C; third bin temperature is ≥10°C.

5. The wind turbine yaw friction plate early warning and life assessment method based on SCADA data according to claim 4 is characterized in that: The detailed method for comparing the cabin vibration RMS value of each compartment with the corresponding multiple of the reference value in step 4 is as follows: when the cabin vibration RMS value of each compartment is greater than 2.5 times the reference value or the cabin vibration peak value is greater than 2 times the reference value, it is determined that the yaw friction plate performance has deteriorated.

6. The wind turbine yaw friction plate early warning and life assessment method based on SCADA data according to claim 1, characterized in that: In the step 5, when the yaw time reaches 2000 hours, an early warning is issued and the friction plate is inspected and maintained. When it reaches 4000 hours, a reminder is given to prepare spare parts and replace them during the light wind season.

Images