Device for predicting service life of connecting bolt

Abstract

The utility model discloses a life prediction device for a connecting bolt, which relates to the technical field of wind driven generators, and comprises a connecting bolt and a strain device arranged at the root of the connecting bolt, the strain device comprises four strain gauges, the four strain gauges are connected with an acquisition network card through output lines, and the acquisition network card is connected with the connecting bolt. The acquisition network card is wirelessly connected with the pyroelectric sensor and the upper computer respectively; the stress condition of the bolt is collected in real time through the strain device, the four strain gauges capture axial, bending and shearing strain at the same time, the damage degree of the connecting bolt can be accurately predicted, and the monitoring precision and reliability of the connecting bolt are remarkably improved; the working temperature and the vibration condition of the connecting bolt are obtained through the pyroelectric sensor, the prediction comprehensiveness is improved, then the accuracy of fatigue life prediction is improved, and the stability and the safety of operation of the wind turbine generator system are improved.

Description

Technical Field

[0001] This utility model relates to the field of wind turbine technology, and in particular to a life prediction device for connecting bolts. Background Technology

[0002] In the field of wind power generation, wind turbine generator sets, with their characteristics of long speed, high load and low speed, place high demands on the strength of mechanical structures. The operating status of the mechanical equipment under service conditions directly affects the safety of the equipment. As a key part connecting various parts, the fatigue life of the connecting bolts of the wind turbine generator set is directly related to the stability and reliability of the wind turbine under long-term, high-load operating conditions.

[0003] In the existing technology, the fatigue damage detection of bolts in wind farms is carried out by manual regular inspections every six months or once a year or by irregular top-level inspections. Due to the harsh operating environment of wind turbine units, the connecting bolts will bear the non-stable loads brought by the wind turbine blades. At the same time, due to their long operating time, the temperature of the connecting bolts will also change. The thermal stress caused by the temperature rise will also increase the probability of fatigue damage.

[0004] Non-stationary random impacts and thermal stresses ultimately lead to a certain degree of randomness in the fatigue life of wind turbine connecting bolts. When fatigue-damaged bolts break and fall off in the rotating hub, it can cause damage to limit sensors, electrical cabinets, cables, and pitch gears, seriously affecting the safety of wind turbine operation.

[0005] Therefore, a life prediction device for connecting bolts is provided to solve the above problems. Utility Model Content

[0006] The purpose of this invention is to provide a life prediction device for connecting bolts, which can effectively predict the fatigue life of wind turbine bolts and provide timely warnings.

[0007] To achieve the above objectives, this utility model provides a life prediction device for connecting bolts, including a connecting bolt and a strain device disposed at the root of the connecting bolt. The strain device includes four strain gauges, all of which are connected to a data acquisition network card via output lines. The data acquisition network card is wirelessly connected to a pyroelectric sensor and a host computer, respectively.

[0008] Preferably, the four strain gauges are respectively disposed around the root of the connecting bolt, and the strain gauges are bonded to the connecting bolt by transparent tape or glue.

[0009] Preferably, the strain gauge includes a base layer and a cover layer, and a resistance wire is disposed between the base layer and the cover layer, with both ends of the resistance wire connected to the output line.

[0010] Preferably, the pyroelectric sensor includes a base and a cap disposed on the base. A preamplifier circuit is disposed on the top of the base, and a pyroelectric detector element is disposed on the preamplifier circuit. A filter is disposed on the top inner side of the cap.

[0011] Preferably, the acquisition network card integrates an access control unit, a digital-to-analog converter, a transformer, and a transceiver unit.

[0012] Preferably, the host computer integrates a data processing unit, a deep learning unit, an optimization unit, and a threshold alarm unit.

[0013] Therefore, the life prediction device for connecting bolts using the above-described structure of this utility model has the following beneficial effects:

[0014] (1) This solution obtains the working temperature and vibration of the connecting bolts through pyroelectric sensors, thereby improving the comprehensiveness of the prediction and thus improving the accuracy of fatigue life prediction.

[0015] (2) This solution collects the stress on the bolts in real time through a strain gauge. Four strain gauges simultaneously capture axial, bending and shear strains, which can accurately predict the damage to the connecting bolts and significantly improve the accuracy and reliability of connecting bolt monitoring.

[0016] The technical solution of this utility model will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0017] Figure 1 This is a structural diagram of a life prediction device for connecting bolts according to the present invention;

[0018] Figure 2 This is a structural diagram of the strain gauge of this utility model;

[0019] Figure 3 This is a structural diagram of the pyroelectric sensor of this utility model;

[0020] Figure 4 This is a structural diagram of the data acquisition network card of this utility model;

[0021] Figure 5 This is a structural diagram of the host computer of this utility model;

[0022] Figure 6 This is a flowchart of the prediction process for the prediction device of this utility model.

[0023] The components include: 1. Connecting bolts; 2. Strain gauge; 3. Strain gauge; 301. Base layer; 302. Covering layer; 303. Resistance wire; 4. Output line; 5. Data acquisition network card; 501. Access control unit; 502. Digital-to-analog conversion unit; 503. Transformer; 504. Transceiver unit; 6. Pyroelectric sensor; 601. Base; 602. Sealing cap; 603. Preamplifier circuit; 604. Pyroelectric detector element; 605. Filter; 7. Host computer; 701. Data processing unit; 702. Deep learning unit; 703. Optimization unit; 704. Threshold alarm unit. Detailed Implementation

[0024] The technical solution of this utility model will be further described below with reference to the accompanying drawings and embodiments.

[0025] Unless otherwise defined, the technical or scientific terms used in this utility model shall have the ordinary meaning understood by one of ordinary skill in the art to which this utility model pertains. The terms "first," "second," and similar terms used in this utility model do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship will also change accordingly.

[0026] Example

[0027] like Figure 1 As shown, this utility model provides a life prediction device for connecting bolts, including a connecting bolt 1 and a strain device 2 disposed at the root of the connecting bolt 1. The strain device 2 includes four strain gauges 3, which are respectively disposed around the root of the connecting bolt 1. The strain gauges 3 are bonded to the connecting bolt 1 by transparent tape or glue.

[0028] All four strain gauges 3 are connected to the data acquisition network card 5 via output lines 4. The data acquisition network card 5 is wirelessly connected to the pyroelectric sensor 6 and the host computer 7, respectively.

[0029] like Figure 2 As shown, the strain gauge 3 includes a base layer 301 and a cover layer 302. A resistance wire 303 is disposed between the base layer 301 and the cover layer 302, and both ends of the resistance wire 303 are connected to the output line 4.

[0030] like Figure 3As shown, the pyroelectric sensor 6 includes a base 601 and a cap 602 disposed on the base 601. A preamplifier circuit 603 is disposed on the top of the base 601, and a pyroelectric detector element 604 is disposed on the preamplifier circuit 603. A filter 605 is disposed on the top inner side of the cap 602.

[0031] like Figure 4 As shown, the acquisition network card 5 integrates an access control unit 501, a digital-to-analog conversion unit 502, a transformer 503, and a transceiver unit 504.

[0032] like Figure 5 As shown, the host computer 7 integrates a data processing unit 701, a deep learning unit 702, an optimization unit 703, and a threshold alarm unit 704.

[0033] like Figure 6 As shown, the prediction method of the prediction device is specifically set as follows:

[0034] S1. Data Acquisition: Collect strain data of the wind turbine connecting bolts throughout their entire lifespan in the laboratory, convert the strain data into stress, normalize it, and clarify the mapping relationship between stress data and remaining lifespan.

[0035] S11. Five sets of tests were conducted on the connecting bolts of the wind turbine generator set. Axial periodic cyclic stress, radial periodic cyclic stress, axial non-periodic cyclic stress, radial non-periodic cyclic stress and random stress were applied to one end of the bolt. The applied stress was kept within the bolt load range. All five sets of tests were run until the bolt completely failed.

[0036] S12. The strain condition of the connecting bolts of the wind turbine is determined using resistance strain gauges. Based on the various parameters of the bolts, the stress and lifespan of the bolts are determined.

[0037] S13. Normalize the full lifespan of the wind turbine connecting bolts, set the lifespan label of the first set of stress data to 1, and the lifespan label of the last set of stress data to 0, and divide the data between 0 and 1 equally.

[0038] S14. Invalid stress data of wind turbine connecting bolts are discarded;

[0039] S15. Map the stress data of the wind turbine connecting bolts to the normalized full life duration to clarify the stress data at different remaining life time points.

[0040] S2. Model Training: A basic fatigue life model is constructed based on stress data using a liquid time constant neural network algorithm. The fatigue life prediction model is tested and optimized to determine a deep learning model for bolt fatigue life that meets the accuracy threshold requirements.

[0041] S21. Analyze the full-life stress data of the connecting bolts of the wind turbine, select seven characteristic indicators of the stress data: maximum value, mean value, peak-to-peak value, absolute mean value, root mean square value, root mean square amplitude, and margin index, and construct a statistical characteristic group of the full-life stress data.

[0042] S22. Analyze the statistical characteristics of the entire life cycle of the connecting bolts of the wind turbine, select the mutation points of seven indicators for analysis, sort the mutation points of the seven indicators, select the minimum value as the theoretical degradation point, and use the maximum value as the data boundary between the training set and the test set.

[0043] S23. Divide the full-life stress data of the connecting bolts of the wind turbine into two parts. Use the data before the boundary point of the dataset as the training set and the data after the boundary point as the test set.

[0044] S24. Store the divided training and test sets for later use;

[0045] S25. Input the five sets of test full-life stress data of the wind turbine connecting bolts into the liquid time constant neural network, and adjust the various hyperparameters of the network according to the actual situation.

[0046] S26. After training the neural network for the liquid time constant of the wind turbine connecting bolts, the network model is tested. The root mean square error is selected as the evaluation index, and the root mean square value of 0.001 is used as the accuracy threshold. The accuracy of the test result is less than 0.001 as the sign that the model training is complete.

[0047] S3. Optimization and Prediction: Measure the stress and temperature data of the connecting bolts of the wind turbine under service conditions, optimize the basic fatigue life model, set up a bolt remaining life warning, and establish a bolt fatigue life prediction and early warning system.

[0048] S31. The stress of the connecting bolts of the wind turbine under service conditions is measured by a pyroelectric sensor, and the temperature of the connecting bolts of the wind turbine is measured by a non-contact sensor.

[0049] S32. Convert real-time bolt strain data into stress and temperature data as supplementary test data and input them into the trained model for testing, while simultaneously performing supplementary training on the model.

[0050] S33. When the model predicts that the remaining life will reach the degradation point of the connecting bolts, a remaining life warning will be issued to notify maintenance personnel to conduct testing and repair.

[0051] Therefore, the present invention provides a bolt life prediction device with the above-mentioned structure. By using a strain gauge to collect the bolt stress in real time, and four strain gauges to simultaneously capture axial, bending and shear strain, it can accurately predict the degree of damage to the bolt, significantly improving the accuracy and reliability of bolt monitoring. By using a pyroelectric sensor to obtain the bolt's working temperature and vibration, it improves the comprehensiveness of the prediction, thereby improving the accuracy of fatigue life prediction.

[0052] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of this utility model and not to limit it. Although the utility model has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solution of this utility model, and these modifications or equivalent substitutions cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of this utility model.

Claims

1. A device for predicting the lifespan of connecting bolts, characterized in that, The device includes a connecting bolt and a strain gauge located at the root of the connecting bolt. The strain gauge includes four strain gauges, all of which are connected to a data acquisition network card via output lines. The data acquisition network card is wirelessly connected to a pyroelectric sensor and a host computer, respectively.

2. The life prediction device for connecting bolts according to claim 1, characterized in that, The four strain gauges are respectively disposed around the root of the connecting bolt, and the strain gauges are bonded to the connecting bolt by transparent tape or glue.

3. The life prediction device for connecting bolts according to claim 1, characterized in that, The strain gauge includes a base layer and a cover layer, with a resistance wire disposed between the base layer and the cover layer, and both ends of the resistance wire being connected to the output line.

4. The life prediction device for connecting bolts according to claim 1, characterized in that, The pyroelectric sensor includes a base and a cap disposed on the base. A preamplifier circuit is disposed on the top of the base, and a pyroelectric detector element is disposed on the preamplifier circuit. A filter is disposed on the top inner side of the cap.

5. The life prediction device for connecting bolts according to claim 1, characterized in that, The acquisition network card integrates an access control unit, a digital-to-analog converter, a transformer, and a transceiver unit.

6. The life prediction device for connecting bolts according to claim 1, characterized in that, The host computer integrates a data processing unit, a deep learning unit, an optimization unit, and a threshold alarm unit.

Images