Embedded wireless passive bolt tightness monitoring method based on surface acoustic wave technology

Abstract

The present application relates to bolt state monitoring technical field, disclose a kind of embedded wireless passive bolt tightness state monitoring method based on surface acoustic wave technology, specifically includes the following steps: based on the real-time acquisition frequency drift amount of surface acoustic wave resonator, and the looseness risk of bolt is judged in real time, generates looseness warning signal and safety signal;Based on warning signal, the frequency drift amount before generating warning signal is analyzed at change rate, and the change rate degree is judged;Based on the change rate degree judging result, the time point when frequency drift amount reaches frequency drift amount risk value is predicted.The present application can judge bolt looseness risk in real time, generates warning signal in time when bolt appears looseness sign, reminds relevant personnel to pay attention to bolt state, avoid bolt looseness further development and not be perceived, to reduce the possibility of equipment failure, accident.

Description

Technical Field

[0001] This invention relates to the field of bolt condition monitoring technology, specifically to an embedded wireless passive bolt tightness monitoring method based on surface acoustic wave technology. Background Technology

[0002] The surface acoustic wave resonator is prefabricated in the bolt body. When the stress state of the bolt changes, it will cause the oscillation loop parameters of the surface acoustic wave resonator to change, which will cause its center frequency to drift.

[0003] When the system is working, the integrated terminal device sends an excitation signal to excite the surface acoustic wave resonator to vibrate. When the excitation signal stops, the reader switches to receiving mode and receives the free oscillation signal with continuously decaying amplitude emitted by the sensor. By measuring the frequency of this signal with a frequency meter, the frequency drift can be obtained, which can be used to measure the change in the stress state of the bolt and thus determine the tightness of the bolt.

[0004] Using wireless radio frequency excitation and echo signal frequency detection, it eliminates the need for wiring and external power supply, solving the problems of difficult wiring and power supply required by traditional monitoring methods. It can realize wireless passive monitoring of bolt tightness in harsh environments such as high and low temperatures and dust, as well as in scenarios where wiring and power supply are difficult to achieve, such as large machinery.

[0005] However, during long-term observation and research on the practical application scenarios of bolt condition monitoring using surface acoustic wave technology, it was found that while the frequency drift could be used to determine whether a bolt had entered a loosening warning state, the rate of loosening could not be accurately judged for bolts that had already issued warning signals. In the analysis of multiple equipment failure cases, it was found that some bolts became severely loose shortly after issuing warning signals, leading to equipment failure. Maintenance personnel, lacking a basis for judging the rate of loosening, failed to take timely measures. Summary of the Invention

[0006] The purpose of this invention is to provide an embedded wireless passive bolt tightness monitoring method based on surface acoustic wave technology, so as to solve at least one of the above-mentioned problems in the prior art.

[0007] This invention provides an embedded wireless passive bolt tightness monitoring method based on surface acoustic wave technology, which specifically includes the following steps: real-time acquisition of frequency drift based on surface acoustic wave resonator, real-time judgment of bolt loosening risk, and generation of loosening warning signal and safety signal;

[0008] Based on the warning signal, the rate of change of the frequency drift before the warning signal is generated is analyzed, and the degree of change rate is judged.

[0009] Based on the judgment result of the change rate degree, the time point when the frequency drift reaches the frequency drift risk value is predicted.

[0010] The change rate degree judgment change result includes generating a change rate fast signal and a change rate slow signal.

[0011] As a further scheme of the present application, the process of real-time judging the loosening risk of the bolt is:

[0012] The real-time frequency drift of the bolt is obtained, the frequency drift warning value and the frequency drift risk value are set, and the frequency drift warning value is less than the frequency drift risk value.

[0013] The real-time frequency drift is compared with the frequency drift warning value, and if the real-time frequency drift is greater than or equal to the frequency drift warning value, a warning signal is generated.

[0014] As a further scheme of the present application, the process of judging the change rate degree is:

[0015] The frequency drift is analyzed, and the change rate average ratio, the extreme value change ratio and the change rate number ratio are calculated.

[0016] The change rate average ratio, the extreme value change ratio and the change rate number ratio are weighted and summed to obtain the rate characteristic value.

[0017] The rate characteristic threshold is set, wherein the rate characteristic threshold is set by the person skilled in the art based on historical bolt loosening change data.

[0018] If the rate characteristic value is greater than the rate judgment characteristic threshold, it indicates that the rate change degree is fast, and a change rate fast signal is generated.

[0019] If the rate characteristic value is less than or equal to the rate judgment characteristic threshold, it indicates that the rate change degree is slow, and a change rate slow signal is generated.

[0020] As a further scheme of the present application, the process of obtaining the change rate average ratio and the extreme value change ratio is:

[0021] The time point of generating the warning signal is obtained, which is marked as the warning time point.

[0022] The frequency drift sequence before the warning time point is obtained: , wherein, represents the frequency drift corresponding to the i-th sampling time, i=1, 2, …, n; and the corresponding sampling time sequence: ;

[0023] The change rate of the adjacent two frequency drifts is calculated , the calculation formula is: , the change rate sequence is calculated: ;

[0024] The values in the change rate sequence are summed and averaged to obtain the change rate average, and the change rate average is calculated by ratio with the change rate limit value to obtain the change rate average ratio.

[0025] The change rate maximum value and the change rate minimum value are extracted, and the difference is calculated to obtain the change rate extreme value, and the change rate extreme value is calculated by ratio with the change rate minimum value to obtain the extreme value change ratio.

[0026] As a further scheme of the application: the change rate quantity ratio acquisition process is:

[0027] The change rate is compared with the change rate threshold value, the change rate greater than the change rate threshold value is extracted, and the number thereof is counted, and then the ratio calculation is performed with the total number of rate change values to obtain the high change rate quantity ratio.

[0028] As a further scheme of the application: the process of predicting the time point when the frequency drift reaches the frequency drift risk value comprises:

[0029] Based on the change rate fast signal, the change rate maximum value is obtained, and the change rate maximum value is used as the predicted change rate;

[0030] The frequency drift corresponding to the generated warning signal is obtained, and is marked as the warning frequency drift;

[0031] The frequency drift risk value is calculated by difference with the warning frequency drift, and the predicted change rate is calculated by ratio to obtain the predicted time value.

[0032] As a further scheme of the application: the process of predicting the time point when the frequency drift reaches the frequency drift risk value further comprises:

[0033] Based on the change rate slow signal, the warning frequency drift is obtained, and the mean value of the frequency drift risk value and the warning frequency drift is taken as the analysis frequency drift;

[0034] After the warning time point, the frequency drift is continuously acquired in real time to obtain the frequency drift between the warning frequency drift and the analysis frequency drift, which is marked as the to-be-analyzed sequence;

[0035] It is judged whether the to-be-analyzed sequence is linearly changed;

[0036] Based on the linear change, a fitting model is used The time point of the frequency drift amount reaching the frequency drift amount risk value is predicted, that is, the frequency drift amount risk value is substituted into the fitting model, and the output x value is the predicted time value;

[0037] Based on the nonlinear change, a radial basis function neural network (RBFNN) model is used to construct a prediction model;

[0038] The frequency drift amount risk value is input into the model, and the output result of the model is the predicted time value.

[0039] As a further scheme of the application, the process of determining whether the sequence to be analyzed is linearly changed is:

[0040] The values in the sequence to be analyzed are fitted by the least square method to obtain a fitting model ;

[0041] Based on the fitting model, the goodness of fit is calculated , and the calculation formula is: Wherein, m is the number of data points, is the jth actual value, represents the jth predicted value, represents the average value of the actual value;

[0042] A goodness of fit threshold is set, if the goodness of fit is greater than the goodness of fit threshold, the values in the sequence to be analyzed are linearly changed, otherwise, they are nonlinearly changed.

[0043] As a further scheme of the application, it also includes: based on the predicted time value, adjusting the monitoring frequency of the frequency drift amount;

[0044] The predicted time value corresponding to the generation of the fast change rate signal and the predicted time value corresponding to the generation of the slow change rate signal are obtained, and a first time threshold and a second time threshold are set;

[0045] The current monitoring frequency of the frequency drift amount is obtained;

[0046] The predicted time value corresponding to the generation of the fast change rate signal is compared with the first time threshold;

[0047] If the predicted time value corresponding to the generation of the fast change rate signal is less than the first time threshold, the predicted time value corresponding to the generation of the fast change rate signal is adjusted;

[0048] The predicted time value corresponding to the generation of the slow change rate signal is compared with the second time threshold;

[0049] If the predicted time value corresponding to the generation of the slow change rate signal is less than the second time threshold, the predicted time value corresponding to the generation of the fast change rate signal is adjusted.

[0050] As a further scheme of the present application: the process of adjusting the monitoring frequency of the frequency drift amount is:

[0051] The corresponding prediction time value when the fast change rate signal is generated is subtracted from the first time threshold, then the first time threshold is multiplied by the current monitoring frequency to obtain a frequency adjustment amount, and the current monitoring frequency is subtracted from the frequency adjustment amount to obtain a monitoring frequency adjustment value.

[0052] The corresponding prediction time value when the slow change rate signal is generated is subtracted from the second time threshold, then the second time threshold is multiplied by the current monitoring frequency to obtain a frequency adjustment amount, and the current monitoring frequency is subtracted from the frequency adjustment amount to obtain a monitoring frequency adjustment value.

[0053] The beneficial effects of the present application are:

[0054] The present application can judge the bolt loosening risk in real time, generate an alarm signal in time when the bolt shows signs of loosening, remind the relevant personnel to pay attention to the bolt state, avoid the further development of the bolt loosening without being detected, and thus reduce the possibility of equipment failure and accidents;

[0055] The rate representation value obtained by the present application through multi-dimensional analysis of the frequency drift amount change rate can more accurately represent the dynamic process of bolt loosening, improve the accuracy and reliability of the judgment of the bolt loosening change rate, enable the maintenance personnel to give early warning before the bolt loosening develops to a serious degree, facilitate early planning of maintenance work, reduce equipment failure, accidents and personnel casualties caused by bolt loosening, and also enable personnel to be allocated in advance according to the early warning;

[0056] The present application can predict the time when the frequency drift amount reaches the risk value, the maintenance personnel can plan the maintenance work in advance, reasonably arrange the time for replacing or tightening the bolt, avoid equipment failure, accidents and personnel casualties caused by bolt loosening, also can allocate personnel in advance according to the prediction time, and can adjust the monitoring frequency according to the prediction time value, maintain the existing frequency when the prediction time value is large, balance the monitoring cost and the demand for equipment operation state monitoring, increase the monitoring frequency in time when the prediction time value is small, guarantee the safe operation of the equipment, and further improve the reliability and safety of the equipment operation. BRIEF DESCRIPTION OF DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0058] Figure 1 is a flow chart of the embedded wireless passive bolt tightness monitoring method based on the surface acoustic wave technology of the present application;

[0059] Figure 2 is an architecture diagram of the embedded wireless passive bolt tightness monitoring system based on the surface acoustic wave technology of the present application. DETAILED DESCRIPTION

[0060] In order for those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should be within the scope of protection of the present application.

[0061] Embodiment one

[0062] Figure 1 The flow chart of the embedded wireless passive bolt tightness monitoring method based on the surface acoustic wave technology provided for the embodiment one of the present application, the present application embodiment can be applicable to the bolt state monitoring case, the embedded wireless passive bolt tightness monitoring method based on the surface acoustic wave technology can be executed by the embedded wireless passive bolt tightness monitoring system based on the surface acoustic wave technology, the embedded wireless passive bolt tightness monitoring system based on the surface acoustic wave technology can be realized by software and / or hardware, the embedded wireless passive bolt tightness monitoring system based on the surface acoustic wave technology can be configured in the embedded wireless passive bolt tightness monitoring device based on the surface acoustic wave technology. Optionally, the embedded wireless passive bolt tightness monitoring device based on the surface acoustic wave technology can be an electronic device, which can be a notebook, a desktop computer and a smart tablet, etc., and the present application embodiment does not limit this.

[0063] The embedded wireless passive bolt tightness monitoring method based on the surface acoustic wave technology provided by the present application embodiment specifically includes the following steps:

[0064] Step one: based on the real-time acquisition of the frequency drift of the surface acoustic wave resonator, the loosening risk of the bolt is judged in real time, and the loosening warning signal and the safety signal are generated;

[0065] In some embodiments, when the force state of the bolt changes, the oscillation loop parameter of the surface acoustic wave resonator changes, and the center frequency of the surface acoustic wave resonator drifts. When the system is working, the integrated terminal device sends an excitation signal to excite the surface acoustic wave resonator to vibrate. When the excitation signal stops, the reader enters a receiving state to receive the free oscillation signal with a continuously decaying amplitude from the sensor. The frequency of the signal is measured by a frequency meter to obtain the frequency drift amount, which is used to measure the change of the force state of the bolt to determine the tightness of the bolt.

[0066] The real-time frequency drift amount of the bolt is obtained, and a frequency drift amount warning value and a frequency drift amount risk value are set, and the frequency drift amount warning value is less than the frequency drift amount risk value.

[0067] The frequency drift amount warning value and the frequency drift amount risk value are set by technicians in the field according to test data in the system debugging stage. In the system debugging stage, standard bolt samples with different tightness are used to obtain the frequency values of the surface acoustic wave resonator corresponding to each standard state. By comparing the reference frequency of the resonator when the bolt is in a tight standard state, the frequency drift amount under different tightness is calculated, and an initial corresponding relationship database between the tightness of the bolt and the frequency drift amount is established.

[0068] For example, 10 standard bolts with different pre-tightening forces (representing different tightness) are tested, and the corresponding frequencies are recorded. The frequency drift amount data is obtained by comparing with the tightness reference frequency.

[0069] The real-time frequency drift amount is compared with the frequency drift amount warning value. If the real-time frequency drift amount is less than the frequency drift amount warning value, a safety signal is generated. If the real-time frequency drift amount is greater than or equal to the frequency drift amount warning value, a warning signal is generated.

[0070] Step 2: Based on the warning signal, the frequency drift amount before the warning signal is generated is analyzed for change rate, and the change rate is judged.

[0071] In some embodiments, the time point when the warning signal is generated is obtained, which is marked as a warning time point.

[0072] The frequency drift amount sequence before the warning time point is obtained: , wherein, represents the frequency drift amount corresponding to the i-th sampling time, i = 1, 2, …, n; and the corresponding sampling time sequence is: ;

[0073] The change rate of the adjacent two frequency drift amounts is calculated , and the calculation formula is: , and the change rate sequence is calculated as: ;

[0074] Summing up the values in the change rate sequence to get the average change rate, and calculating the ratio of the average change rate to the change rate limit value to get the average change rate ratio;

[0075] Extracting the maximum and minimum change rates, calculating the difference to get the change rate extreme value, and calculating the ratio of the change rate extreme value to the minimum change rate to get the extreme change ratio;

[0076] Comparing the change rate with the change rate threshold value, extracting the change rate greater than the change rate threshold value, and calculating the ratio of the number of change rates greater than the change rate threshold value to the total number of change rate values to get the high change rate number ratio;

[0077] Weighted sum calculation of the average change rate ratio, extreme change ratio and change rate number ratio to get the rate representation value;

[0078] The rate representation value can more accurately represent the dynamic process of bolt loosening through multi-dimensional analysis, improving the accuracy and reliability of the judgment;

[0079] Because the rate of bolt loosening change can be more accurately judged, the fast change rate signal can be sent in time for the fast change rate, which makes it possible to give early warning before the bolt loosening develops to a serious degree, so as to gain more time for maintenance measures;

[0080] The rate representation value is used to judge the change rate of frequency drift amount, which provides support for predicting the time when the frequency drift amount reaches the risk value of frequency drift amount after the frequency drift amount reaches the warning value of frequency drift amount;

[0081] Set the rate representation threshold value, wherein the rate representation threshold value is set by the skilled person in the art based on historical bolt loosening change data;

[0082] If the rate representation value is greater than the rate judgment representation threshold value, it means that the rate of change is fast, and a fast change rate signal is generated;

[0083] If the rate representation value is less than or equal to the rate judgment representation threshold value, it means that the rate of change is slow, and a slow change rate signal is generated;

[0084] The technical scheme of the embodiment is: based on the surface acoustic wave resonator, the frequency drift amount of the bolt is acquired in real time to determine the loosening risk of the bolt, and the specific operation is to set a frequency drift amount warning value and a risk value, compare the real-time frequency drift amount with the warning value, if the real-time frequency drift amount is less than the warning value, a safety signal is generated, and if the real-time frequency drift amount is greater than or equal to the warning value, a warning signal is generated; after the warning signal is generated, the frequency drift amount sequence and the sampling time sequence before the warning time point are acquired, the change rate of adjacent frequency drift amounts is calculated to obtain a change rate sequence, and then a change rate average ratio, an extreme value change ratio and a high change rate quantity ratio are calculated, the three ratios are weighted and summed to obtain a rate characteristic value, and the change rate degree of the frequency drift amount is determined according to the comparison result of the rate characteristic value and a rate characteristic threshold value, and a change rate fast signal or a change rate slow signal is generated;

[0085] Therefore, the loosening risk of the bolt can be determined in real time, the warning signal is generated in time when the bolt shows signs of loosening, relevant personnel are reminded to pay attention to the bolt state, the bolt loosening is prevented from further developing without being detected, the possibility of equipment failure and accidents is reduced, the rate characteristic value obtained by analyzing the change rate of the frequency drift amount in multiple dimensions can more accurately represent the dynamic process of the bolt loosening, the accuracy and reliability of the determination of the change rate of the bolt loosening are improved, the maintenance personnel can be warned in advance before the bolt loosening develops to a serious degree, more time is obtained for taking maintenance measures, maintenance work is planned in advance, equipment failure, accidents and personnel casualties caused by the bolt loosening are reduced, and personnel can be allocated in advance according to the warning.

[0086] Embodiment two

[0087] Based on the above embodiment, as shown in Figure 1 The bolt loosening state monitoring method based on the surface acoustic wave technology embedded wireless passive bolt provided by the embodiment of the application specifically includes the following steps:

[0088] Step three: based on the determination result of the change rate degree, the time point when the frequency drift amount reaches the frequency drift amount risk value is predicted;

[0089] In some embodiments, based on the change rate fast signal, the maximum value of the change rate is acquired, and the maximum value of the change rate is used as the predicted change rate;

[0090] The frequency drift amount corresponding to the generated warning signal is acquired and marked as a warning frequency drift amount;

[0091] The frequency drift amount risk value and the warning frequency drift amount are calculated by difference, and the predicted change rate is calculated by ratio, to obtain a predicted time value;

[0092] The calculated predicted time value is the time for the frequency drift amount to reach the risk value of the frequency drift amount, so that the maintenance personnel can plan the maintenance work in advance, and avoid equipment failure, accidents and even personnel casualties caused by bolt loosening, and personnel can be allocated in advance according to the predicted time value;

[0093] The reason for taking the maximum change rate as the predicted change rate is that, in the case of a fast bolt loosening rate, the maximum change rate is used as the predicted change rate to provide the most secure prediction, thereby minimizing the risk caused by bolt loosening. During the bolt loosening process, if the frequency drift amount change rate is fast, it means that the bolt state deteriorates rapidly. By using the maximum change rate for prediction, the time for the frequency drift amount to reach the risk value can be estimated as quickly as possible. This prediction method enables the maintenance personnel to carry out maintenance work, replace or tighten the bolt, and ensure the safe and stable operation of the equipment before the bolt loosening problem develops to a dangerous level, thereby avoiding equipment failure, accidents and even personnel casualties;

[0094] Based on the slow change rate signal, the warning frequency drift amount is obtained, and the average of the risk value of the frequency drift amount and the warning frequency drift amount is taken as the analysis frequency drift amount;

[0095] After the warning time point, the frequency drift amount is continuously obtained in real time to obtain the frequency drift amount between the warning frequency drift amount and the analysis frequency drift amount, which is marked as a to-be-analyzed sequence: , wherein, is the warning frequency drift amount, is the analysis frequency drift amount;

[0096] The values in the to-be-analyzed sequence are fitted by the least square method to obtain a fitting model ;

[0097] Based on the fitting model, the goodness of fit is calculated , and the calculation formula is: , wherein m is the number of data points, is the jth actual value, represents the jth predicted value, represents the average value of the actual value;

[0098] The goodness of fit threshold is set. If the goodness of fit is greater than the goodness of fit threshold, it means that the fitting degree of the data and the fitting straight line is higher, and the values in the to-be-analyzed sequence change linearly, otherwise, they change nonlinearly;

[0099] Based on the linear change, the fitting model is used to predict the time point for the frequency drift amount to reach the risk value of the frequency drift amount, that is, the risk value of the frequency drift amount is substituted into the fitting model, and the output x value is the predicted time value;

[0100] Based on the nonlinear change, a prediction model is constructed using other models, and the specific process is as follows:

[0101] In one possible embodiment, a radial basis function neural network (RBFNN) model is used to construct the prediction model;

[0102] The values in the sequence to be analyzed are normalized, for example, the minimum-maximum normalization method is used to map the data to the [0, 1] interval;

[0103] The normalized values are divided into a training set and a validation set according to a ratio of 3:7;

[0104] The input layer, hidden layer and output layer are set;

[0105] The number of input layer neurons depends on the number of input data features. Since the main focus is on predicting the time to risk value based on the frequency drift amount over time, the number of input layer neurons is 1, i.e. the input is the current frequency drift amount, and the number of output layer neurons is also 1, which outputs the predicted time value;

[0106] Different numbers of hidden layer neurons are tested through experiments to observe the performance of the model on the validation set, including but not limited to mean square error and mean absolute error;

[0107] The number of hidden layer neurons that optimizes the performance of the model is selected. Generally, a smaller number can be tried first, such as 5, and then gradually increased until the model performance no longer improves or overfitting occurs;

[0108] The radial basis function of the initialized hidden layer neurons needs to determine the center vector. N different data points (N is the number of hidden layer neurons) are randomly selected from the training set as the initial center vector, so that the center vector has a certain representativeness in the data space;

[0109] The width parameter is initialized, which determines the width of the radial basis function and affects the local approximation ability of the model. A fixed value can be used for initialization, for example, an initial value D is set first, and then the width parameter of all hidden layer neurons is set to D. The width parameter can be adjusted through optimization algorithms during the model training process. An empirical method to determine D is to calculate the average distance d between all data points in the training set, and then take where N is the number of hidden layer neurons;

[0110] Output layer weight initialization: the output layer weight (p represents the hidden layer neuron, and l represents the output layer neuron) is usually randomly initialized, and the initial weight value can be randomly generated within a certain range (such as [-1, 1]);

[0111] The model is trained using the training set data, and the goal of the training is to adjust the model parameters so that the error between the predicted output of the model and the actual time value in the training set is minimized;

[0112] The commonly used training algorithm is gradient descent and its variants, such as stochastic gradient descent (SGD) and mini-batch gradient descent (Mini-Batch SGD);

[0113] The trained RBFNN model is evaluated using the test set data, and the performance indicators of the model on the test set are calculated, such as mean square error (MSE) and mean absolute error;

[0114] The prediction model with satisfactory performance is obtained, and the frequency drift risk value (normalized value) is input into the model, and the output result of the model is the predicted time value;

[0115] The output is the normalized time value, which needs to be de-normalized to the actual time scale, and finally the predicted time value is obtained;

[0116] For example, if the minimum-maximum normalization method is used;

[0117] Step four: based on the predicted time value, adjust the monitoring frequency of the frequency drift;

[0118] In some embodiments, the predicted time value corresponding to the generation of the fast change rate signal and the predicted time value corresponding to the generation of the slow change rate signal are obtained, and a first time threshold and a second time threshold are set;

[0119] The current monitoring frequency of the frequency drift is obtained;

[0120] The predicted time value corresponding to the generation of the fast change rate signal is compared with the first time threshold;

[0121] If the predicted time value corresponding to the generation of the fast change rate signal is greater than or equal to the first time threshold, the current monitoring frequency is maintained;

[0122] If the predicted time value corresponding to the generation of the fast change rate signal is less than the first time threshold, the predicted time value corresponding to the generation of the fast change rate signal is adjusted;

[0123] Specifically, the predicted time value corresponding to the generation of the fast change rate signal is calculated by difference with the first time threshold, and then calculated by ratio with the first time threshold to obtain an adjustment coefficient, the adjustment coefficient is multiplied by the current monitoring frequency to obtain a frequency adjustment amount, and the current monitoring frequency is calculated by difference with the frequency adjustment amount to obtain a monitoring frequency adjustment value;

[0124] The corresponding prediction time value when the change rate slow signal is generated is compared with the second time threshold value;

[0125] If the corresponding prediction time value when the change rate slow signal is greater than or equal to the second time threshold value, the current monitoring frequency is maintained;

[0126] If the corresponding prediction time value when the change rate slow signal is less than the second time threshold value, the corresponding prediction time value when the change rate fast signal is adjusted;

[0127] Specifically, the corresponding prediction time value when the change rate slow signal is generated is calculated by difference and ratio with the second time threshold value to obtain an adjustment coefficient, the adjustment coefficient is multiplied by the current monitoring frequency to obtain a frequency adjustment amount, and the current monitoring frequency is calculated by difference with the frequency adjustment amount to obtain a monitoring frequency adjustment value;

[0128] It should be noted that when the change rate slow signal is generated, the adjustment of the monitoring frequency is based on the adjustment of the frequency drift value after the analysis of the frequency drift amount, and therefore the second time threshold value is different from the first time threshold value;

[0129] Thus, when the prediction time value is large, the existing frequency can be maintained to balance the monitoring cost and the equipment operation state monitoring demand, and when the prediction time value is small, adjustment is performed, which conforms to the principle of timely strengthening monitoring and increasing the monitoring frequency to ensure safe operation of the equipment;

[0130] The technical scheme of the embodiment is as follows: based on the change rate fast and slow signals, the time point when the frequency drift amount reaches the risk value is predicted, when the change rate fast signal is received, the maximum change rate is taken as the predicted change rate, the prediction time value is calculated by the difference between the frequency drift amount risk value and the warning frequency drift amount and the predicted change rate, when the change rate slow signal is received, the mean value of the frequency drift amount risk value and the warning frequency drift amount is taken as the analysis frequency drift amount, the change characteristics are judged according to the fitting goodness of the frequency drift amount sequence obtained after the warning time point, the linear change is predicted by the fitting model, and the nonlinear change is predicted by the radial basis function neural network (RBFNN) model;

[0131] According to the obtained prediction time value, the frequency drift amount monitoring frequency is adjusted, the prediction time value is compared with the corresponding time threshold value, when the prediction time value is greater than or equal to the threshold value, the current monitoring frequency is maintained, and when the prediction time value is less than the threshold value, the frequency adjustment amount is obtained by calculating the adjustment coefficient, and then the monitoring frequency adjustment value is obtained;

[0132] Therefore, the maintenance personnel can plan the maintenance work in advance, reasonably arrange the time for replacing or tightening the bolt, avoid equipment failure, accidents and personnel casualties caused by bolt loosening, and can also allocate personnel in advance according to the predicted time, and can adjust the monitoring frequency according to the predicted time value, maintain the existing frequency when the predicted time value is large, balance the monitoring cost and the equipment operation state monitoring demand, increase the monitoring frequency when the predicted time value is small, ensure the safe operation of the equipment, and further improve the reliability and safety of the equipment operation.

[0133] Embodiment three

[0134] Based on the above embodiments, as shown in Figure 2 The embedded wireless passive bolt loosening state monitoring system based on the surface acoustic wave technology provided by the embodiments of the present application specifically comprises:

[0135] The real-time judgment module: based on the surface acoustic wave resonator, the frequency drift amount is obtained in real time, the loosening risk of the bolt is judged in real time, and the loosening warning signal and the safety signal are generated;

[0136] The change rate judgment module: based on the warning signal, the change rate of the frequency drift amount before the warning signal is generated is analyzed, and the change rate degree is judged;

[0137] The risk prediction module: based on the change rate degree judgment result, the time point when the frequency drift amount reaches the frequency drift amount risk value is predicted;

[0138] The frequency adjustment module: based on the predicted time value, the monitoring frequency of the frequency drift amount is adjusted.

[0139] The above formulas are all dimensionless numerical calculations, and the formulas are obtained by software simulation of a large amount of data to obtain a formula of the nearest real situation, and the preset parameters in the formula are set by the person skilled in the art according to the actual situation.

[0140] The above describes one embodiment of the present application in detail, but the content described is only the preferred embodiment of the present application, and cannot be considered as limiting the scope of the present application. Any equivalent changes and improvements made according to the scope of the present application should still belong to the patent coverage range of the present application.

Claims

1. A method for monitoring the tightness of embedded wireless bolts based on surface acoustic wave technology, characterized in that, Includes the following steps: The frequency drift is acquired in real time, and the risk of bolt loosening is assessed in real time to generate a warning signal. Based on the warning signal, the rate of change of the frequency drift before the warning signal is generated is analyzed and judged to generate a fast-rate signal and a slow-rate signal. Based on the assessment of the rate of change, the time point at which the frequency drift reaches the risk value is predicted. Based on the slow rate of change signal, the average value of the frequency drift risk value and the warning frequency drift value is marked as the analysis frequency drift value. The frequency drift value between the warning frequency drift value and the analysis frequency drift value is obtained and marked as the sequence to be analyzed. It is then determined whether the sequence to be analyzed exhibits a linear change. If the change is linear, the frequency drift risk value is substituted into the linear fitting model, and the output value is the predicted time value. If the change is nonlinear, a radial basis function neural network model is used to construct the prediction model and output the predicted time value. If the prediction time value corresponding to the slow rate of change signal is less than the second time threshold, the prediction time value corresponding to the slow rate of change signal is adjusted. The difference between the predicted time value corresponding to the generation of a slow-changing signal and the second time threshold is calculated, and the ratio between the predicted time value and the second time threshold is calculated. Then, the product of the ratio with the current monitoring frequency is calculated to obtain the frequency adjustment amount. The difference between the current monitoring frequency and the frequency adjustment amount is calculated to obtain the monitoring frequency adjustment value.

2. The embedded wireless passive bolt tightness monitoring method based on surface acoustic wave technology according to claim 1, characterized in that, The process of real-time assessment of the risk of bolt loosening is as follows: The real-time frequency drift of the bolt is obtained. If the real-time frequency drift is greater than or equal to the frequency drift warning value, a warning signal is generated.

3. The embedded wireless passive bolt tightness monitoring method based on surface acoustic wave technology according to claim 1, characterized in that, The process of judging the degree of change rate is as follows: The frequency drift is analyzed and calculated to obtain the mean ratio, extreme value ratio, and number ratio of the rate of change. The mean ratio, extreme value ratio, and number ratio of the rate of change are then weighted to obtain the rate characterization value. If the rate of change is greater than the threshold for rate judgment, a fast rate of change signal is generated; otherwise, a slow rate of change signal is generated.

4. The embedded wireless passive bolt tightness monitoring method based on surface acoustic wave technology according to claim 3, characterized in that, The process for obtaining the ratio of the mean rate of change and the ratio of the extreme rate of change is as follows: Obtain the frequency drift sequence before the warning signal is generated, calculate the rate of change of two adjacent frequency drifts, and obtain the rate of change sequence. The average value of the values ​​in the rate of change sequence is processed, and then the ratio of the average rate of change is calculated with the rate of change limit to obtain the rate of change average ratio. Calculate the deviation between the maximum and minimum rates of change, and then calculate the ratio between the maximum and minimum rates of change to obtain the extreme value change ratio.

5. The embedded wireless passive bolt tightness monitoring method based on surface acoustic wave technology according to claim 4, characterized in that, The process for obtaining the ratio of the rate of change is as follows: Extract the rate of change greater than the rate of change threshold, count the number of such rates, and then calculate the ratio of the rate of change to the total number of rate changes to obtain the rate of change ratio.

6. The embedded wireless passive bolt tightness monitoring method based on surface acoustic wave technology according to claim 1, characterized in that, The process of predicting the time point at which the frequency drift reaches the frequency drift risk value includes: Based on the fast-changing signal, the maximum value of the changing rate is obtained, and the maximum value of the changing rate is used as the predicted changing rate. The frequency drift corresponding to the generated warning signal is marked as the warning frequency drift. The difference between the frequency drift risk value and the warning frequency drift value is calculated, and then the ratio of this difference to the predicted rate of change is calculated to obtain the predicted time value.

7. The embedded wireless passive bolt tightness monitoring method based on surface acoustic wave technology according to claim 1, characterized in that, The process of determining whether the sequence to be analyzed exhibits a linear change is as follows: The numerical values ​​in the sequence to be analyzed are fitted using the least squares method to obtain the fitted model; Based on the fitting model, its goodness of fit is calculated. If the goodness of fit is greater than the goodness of fit threshold, the values ​​in the sequence to be analyzed show a linear change; otherwise, they show a non-linear change.

8. The embedded wireless passive bolt tightness monitoring method based on surface acoustic wave technology according to claim 1, characterized in that, Also includes: Based on the predicted time value, adjust the monitoring frequency of the frequency drift. Obtain the prediction time value corresponding to the generation of a signal with a fast rate of change and the prediction time value corresponding to the generation of a signal with a slow rate of change, and set a first time threshold and a second time threshold. If the prediction time value corresponding to a signal with a fast rate of change is less than the first time threshold, the prediction time value corresponding to the signal with a fast rate of change is adjusted.

9. The embedded wireless passive bolt tightness monitoring method based on surface acoustic wave technology according to claim 8, characterized in that, The process of adjusting the monitoring frequency for frequency drift is as follows: The difference between the predicted time value corresponding to the generation of the fast-changing signal and the first time threshold is calculated, and the ratio between the first time threshold and the first time threshold is calculated. Then, the product of the ratio with the current monitoring frequency is calculated to obtain the frequency adjustment amount. The difference between the current monitoring frequency and the frequency adjustment amount is calculated to obtain the monitoring frequency adjustment value.

Images