A passive wireless EMU brake disc bolt preload real-time monitoring component, device and system

Abstract

This invention discloses a passive wireless EMU brake disc bolt preload real-time monitoring component, device, and system, belonging to the field of intelligent operation and maintenance safety monitoring technology for rail transit. The brake disc bolt preload real-time monitoring component includes a brake disc bolt body. A mandrel, arranged along the extension direction of the mandrel and used to sense changes in the preload of the brake disc bolt body, is encapsulated within the bolt body's screw. A sensor connected to the mandrel and a signal processor connected to the sensor are encapsulated within the bolt head of the bolt body. The passive wireless EMU brake disc bolt preload real-time monitoring component, device, and system provided by this invention can monitor the changes in brake disc bolt preload during operation in real time, offering high monitoring accuracy, high timeliness, low maintenance costs, and ease of operation and application.

Description

Technical Field

[0001] This invention belongs to the field of intelligent operation and maintenance safety detection technology for rail transit, specifically relating to a passive wireless EMU brake disc bolt preload real-time monitoring component, device and system. Background Technology

[0002] As a key component of the EMU braking system, the brake disc mainly consists of a hub, friction disc, and pressure ring, all fastened together by multiple circumferentially distributed bolts. During braking, the brake disc is subjected to the thermomechanical coupling of brake pad pressure, friction, and heat generated by friction. This causes drastic changes in the preload of the connecting bolts, resulting in varying degrees of loosening and fatigue damage, reducing the reliability of the brake disc and posing a serious threat to the operational safety of the EMU. In extreme cases, bolt breakage can lead to catastrophic accidents caused by the brake disc detaching. Therefore, in the rail transit sector, real-time and effective monitoring of the brake disc bolt preload is crucial for improving the efficiency of daily maintenance of the EMU braking system and ensuring operational safety.

[0003] Currently, common monitoring methods for bolt preload testing include image recognition, symbol marking, and vibration detection. However, these methods are difficult to use effectively for detecting the preload of brake disc bolts in high-speed trains. The main reasons are as follows:

[0004] 1) The service environment of the brake disc bolts of the EMU is harsh. During long-term service, the symbols marked by image recognition method, symbol marking method, etc. will be covered by dirt or fall off, making it difficult to use effectively.

[0005] 2) The brake disc of the EMU has a complex structure and strong vibration and noise. The vibration response signals generated by the brake disc bolts in the loose and tight states are very similar, making it difficult for vibration detection methods to distinguish the loose and tight states of the brake disc bolts.

[0006] 3) Under the influence of complex external environmental forces and heat, the variation law of bolt load is extremely complex. Currently, there is a lack of real-time online monitoring methods for the changes in brake disc bolt preload during operation. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of the prior art and provide a passive wireless EMU brake disc bolt preload real-time monitoring component, device and system, which has high monitoring accuracy, high real-time performance and low maintenance cost.

[0008] To achieve the above objectives, the present invention is implemented using the following technical solution:

[0009] In a first aspect, the present invention provides a passive wireless EMU brake disc bolt preload real-time monitoring component, including a brake disc bolt body, wherein a mandrel is encapsulated within the bolt body's screw, which is arranged along the extension direction of the screw and used to sense changes in the preload of the brake disc bolt body, and a sensor connected to the mandrel and a signal processor connected to the sensor are encapsulated within the bolt head of the bolt body.

[0010] Furthermore, from the screw head to the screw rod, there are interconnected encapsulation grooves and core grooves that are located on the same axis as the bolt body. The diameter of the encapsulation groove is larger than the diameter of the core groove, and a pre-reserved gap is left between the bottom of the core groove and the top of the screw rod.

[0011] The sensor and the signal processor are installed sequentially from the inside out in the encapsulation groove. The spindle is fitted into the core groove, which is coated with grease around its perimeter, and is connected to the sensor in the encapsulation groove. A protective sleeve that wraps around the sensor and the signal processor in a ring shape and a protective cover that matches the size of the groove opening and is flush with it are installed in the encapsulation groove.

[0012] In a first aspect, the present invention provides a passive wireless EMU brake disc bolt preload real-time monitoring device, comprising:

[0013] Display terminal,

[0014] Multiple repeaters, each interconnected with the others;

[0015] And multiple passive wireless EMU brake disc bolt preload real-time monitoring components as described in the first aspect, and the repeaters and the monitoring components are connected one-to-one. Each repeater receives the bolt preload status signal sent from the corresponding monitoring component and then converges it to the display terminal.

[0016] Furthermore, the repeater includes a thermoelectric power generation module, an energy storage module, a control module, and a remote transmission module;

[0017] The thermoelectric power generation module is connected to the energy storage module via a conversion circuit and to the control module via a trigger circuit; the control module is connected to the energy storage module, the monitoring component, and the remote transmission module respectively; the energy storage module is connected to the monitoring component and the remote transmission module respectively.

[0018] Furthermore, the trigger circuit determines whether the disc brake is in working condition based on whether the thermoelectric generator module generates electricity. If it is in working condition, the thermoelectric generator module generates electrical energy, and the trigger circuit sends a trigger signal to the control module to make the control module respond and enter the timed wake-up mode.

[0019] Furthermore, when the disc brake device is not in operation, the thermoelectric power generation module does not generate electricity, and the control module and the remote transmission module are in sleep mode.

[0020] When the disc brake device is in operation, the control module is in a timed wake-up mode, with T c For time-interval wake-up, transmit acquisition control commands to the sensor, and simultaneously use T s The wake-up control command and brake disc bolt preload data to be sent are transmitted to the remote sending module at time intervals, where T s =nT c n is a positive integer; so that the remote transmission module responds after receiving the wake-up control command and enters the transmission mode from the sleep mode, and after receiving the brake disc bolt preload data, sends out the brake disc bolt preload data, and enters the sleep mode again to wait for the next wake-up control command.

[0021] Furthermore, the monitoring component is configured with a signal receiving unit and a signal processing unit;

[0022] The signal receiving unit is used to receive the bolt preload state signal collected after the brake disc bolt undergoes axial deformation;

[0023] The signal processing unit performs pre-training, abnormal data filtering, sequence correlation and decomposition processing, wavelet packet decomposition, and frequency band energy filtering on the bolt preload state signal to obtain historical feature vectors; the historical feature vectors are input into the support vector machine classification model for training to obtain the bolt state detection model and obtain the brake disc bolt preload monitoring signal.

[0024] Furthermore, the process of selecting historical feature vectors and then inputting these historical feature vectors into a support vector machine classification model for training to obtain a bolt state detection model includes the following steps:

[0025] The sensor collects the resistance change signal caused by the axial deformation of the bolt in real time and converts it into a voltage signal of 0-5V. The signal is amplified by 100 times using an instrumentation amplifier. A low-pass filter is used to set the cutoff frequency to 2kHz to remove high-frequency noise. A 16-bit ADC module is used to convert the analog signal into a digital signal with a resolution of 0.076mV / LSB.

[0026] The mean μ and standard deviation δ of 10,000 data points within a window of 1 second are calculated. If a data point exceeds the range of μ ± 3δ, it is marked as an outlier and the outlier is replaced by linear interpolation.

[0027] The linear trend term was fitted using the least squares method and subtracted from the original signal; the subtracted digital signal was decomposed into 6 intrinsic mode functions (IMFs), and the low-frequency components IMF1-IMF3 related to the preload were extracted.

[0028] The preprocessed digital signal is decomposed into 16 sub-bands by 4-level wavelet packet decomposition. The energy value of each sub-band signal is calculated, and the top 3 frequency bands with the highest energy ratio are selected to form the feature vector.

[0029] Using the mean, variance, and peak factor of time-domain features, and selecting energy values ​​from three frequency bands based on frequency-domain features, a 6-dimensional historical feature vector F is constructed:

[0030]

[0031] In the formula, μ is the mean; δ is the standard deviation; and the peak factor is the peak value. Energy value x i The sub-band signal energy is represented by i, the sub-band number is represented by i, K takes the value [1, N], and N is the number of data points of each sub-band signal.

[0032] The training and test sets were divided in a 7:3 ratio. Five-fold cross-validation was used to optimize the hyperparameters of the support vector machine (SVM) model. The model was trained using the training set and evaluated using the test set: accuracy ≥ 98.5% and the diagonal proportion of the confusion matrix > 97%.

[0033] The real-time feature vector is input into the trained support vector machine (SVM) model, and combined with the preset alarm threshold, the bolt preload state is judged and output.

[0034] When the preload decreases by more than 15% of the rated value, the bolt preload is considered loose.

[0035] When the preload increases by more than 10% of the rated value, the bolt preload is considered overtight.

[0036] All other bolt preload conditions are normal.

[0037] When ≥500 new data sets are added within a period of time, the model is fine-tuned and the support vectors are updated; if the model accuracy is <95% for 3 consecutive tests, full retraining is initiated.

[0038] Thirdly, the present invention provides a brake disc system, comprising:

[0039] Hub,

[0040] Friction disc,

[0041] Pressure ring,

[0042] And the passive wireless EMU brake disc bolt preload real-time monitoring component as described in the first aspect, or the passive wireless EMU brake disc bolt preload real-time monitoring device as described in any one of the second aspects.

[0043] The hub, friction disc, and pressure ring are assembled together via the monitoring component, a gasket and nut that match the monitoring component, and are press-fitted onto the brake disc seat of the axle through an interference fit.

[0044] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:

[0045] 1) This invention uses direct measurement of bolt deformation to sense the preload signal of the brake disc bolts, avoiding transmission errors and environmental noise effects during indirect measurement.

[0046] 2) The control module in the brake disc bolt preload monitoring device enters a timed wake-up mode via a trigger signal from the trigger circuit. When the disc brake device stops working, the brake disc bolt preload monitoring device is in sleep mode, thus not consuming too much power when it is not necessary to collect and transmit brake disc bolt preload data. At the same time, as soon as the disc brake device enters the working state, the trigger circuit immediately sends a trigger signal to the control module, and then the brake disc bolt preload monitoring device transmits brake disc bolt preload data online in real time.

[0047] 3) The system power consumption of this invention in sleep mode is only a few microamps, which can effectively ensure that the power of the energy storage module of the brake disc bolt preload monitoring device is always kept within the voltage range that the system can work normally. In the timed wake-up working state, the electrical energy generated by the thermoelectric generator converting vibration heat energy can continuously charge the energy storage module, so that the brake disc bolt preload monitoring device can work for a long time.

[0048] 4) This type of brake disc bolt preload monitoring device uses wireless data transmission, is easy to install and disassemble, and has low maintenance costs. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of a relay network operation provided in an embodiment of the present invention.

[0050] Figure 2 This is a block diagram of a bolt preload real-time monitoring device provided in an embodiment of the present invention.

[0051] Figure 3 This is a flowchart of a brake disc bolt loosening status signal processing method provided in an embodiment of the present invention.

[0052] Figure 4 This is a structural diagram of a brake disc system provided in an embodiment of the present invention.

[0053] Figure 5 This is a structural diagram of a monitoring component provided in an embodiment of the present invention.

[0054] In the picture:

[0055] 1. Hub; 2. Friction disc; 3. Pressure ring; 4. Monitoring component; 5. Gasket; 6. Nut; 7. Protective cover; 8. Protective sleeve; 9. Signal processor; 10. Sensor; 11. Grease; 12. Mandrel; 13. Bolt body. Detailed Implementation

[0056] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0057] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0058] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0059] Example

[0060] To detect the preload of the brake disc bolts, the bolt elongation is monitored in real time. When the bolt elongation changes, the corresponding electrical signal is obtained after the bolt elongation is collected by the sensor. The collected electrical signal can characterize the bolt preload state.

[0061] The monitoring component provided in this embodiment of the invention directly monitors the preload state of the brake disc bolts by measuring their elongation. This effectively distinguishes the preload state of the brake disc bolts, and the detection accuracy is not affected by mechanical vibration loads or braking thermal loads, thus ensuring the reliability of brake disc bolt preload detection and consequently ensuring the safe and stable operation of the EMU.

[0062] like Figure 1 As shown, the real-time monitoring device includes multiple monitoring components, repeaters, and display terminals.

[0063] During brake disc bolt preload testing, a monitoring component is encapsulated within the bolts that securely connect to the brake disc. This component collects bolt preload data in real time and transmits it to the corresponding repeater. The repeater receives the bolt preload data from the monitoring component, transmits it via a remote transmission module, and aggregates it to the display terminal. The display terminal receives the bolt preload data transmitted from the repeater network and displays it on the screen.

[0064] In this embodiment, multiple monitoring components are provided to collect the bolt elongation in real time and send the brake disc bolt elongation data to the corresponding repeater.

[0065] The repeater includes a thermoelectric power generation module, an energy storage module, a control module, and a remote transmission module.

[0066] The repeater network is used for data transmission and consists of multiple repeaters. Each repeater receives bolt elongation data sent by the remote transmission module in its corresponding monitoring component and then aggregates it to the display terminal. Each repeater communicates only with its corresponding monitoring component, and the repeaters use an STM32F103C8T6 as the main controller.

[0067] In this embodiment, a response relationship is established between the repeater network and the display terminal. After the repeater N sends the brake disc bolt preload data to the display terminal, the display terminal will send a response command back to the repeater N to confirm that the brake disc bolt preload data has indeed been received and displayed. If the repeater N does not receive the response command, the repeater N will send the brake disc bolt preload data again.

[0068] The repeater network consists of multiple interconnected repeaters, and each repeater is interconnected with the others. When repeater N receives the brake disc bolt preload data sent by the brake disc bolt preload monitoring device N, repeater N sends the brake disc bolt preload data to repeater N1, and repeater N1 then sends the brake disc bolt preload data to repeater N2, until repeater 1 receives the brake disc bolt preload data.

[0069] like Figure 2As shown, the bolt preload monitoring device generates electricity from the thermoelectric generator module and stores it in the energy storage module. During system operation, it supplies power to the control module, monitoring components, and remote transmission module. The operating status of the monitoring components and the thermoelectric generator module is controlled according to the instructions of the control module.

[0070] In this embodiment, the thermoelectric power generation module is installed on the brake disc pressure ring, connected to the energy storage module through a conversion circuit, and connected to the control module through a trigger circuit.

[0071] The trigger circuit includes a rectifier bridge, a Zener diode, a front-end capacitor, and a switching circuit connected in sequence. The switching circuit is connected to the control module. The trigger circuit determines whether the disc brake is in operation based on whether the thermoelectric generator is generating electricity. If it is in operation, the thermoelectric generator produces electrical energy, and the trigger circuit sends a trigger signal to the control module.

[0072] The control module is connected to the energy storage module, monitoring components, and remote transmission module, respectively.

[0073] The control module includes two operating modes, specifically:

[0074] 1) Sleep mode

[0075] When the disc brake is not in operation, the control unit is in a low-power sleep mode.

[0076] 2) Timed wake-up mode

[0077] When the disc brake system is in operation, the control unit is in data acquisition mode, using T... c For time-interval wake-up, transmit acquisition control commands to the sensor, and simultaneously use T s The wake-up control command and brake disc bolt preload data to be sent are transmitted to the remote sending module at time intervals, where T s =nT c , where n is a positive integer.

[0078] The energy storage module is connected to both the monitoring component and the remote transmission module.

[0079] When the disc brake is not in operation, the thermoelectric generator does not generate electricity, and the control module and remote transmission module are in sleep mode. When the disc brake enters the working state, the internal control process of the brake disc bolt preload monitoring device includes the following steps:

[0080] The disc brake system begins operation, generating heat through friction between the brake discs, creating a temperature difference. The thermoelectric generator converts this temperature difference energy into electrical energy. Part of this converted electrical energy is stored in the energy storage module via a conversion circuit, powering the remote transmission module, monitoring components, and control module. The other part of the electrical energy is sent to the control module via a trigger circuit. Upon receiving the trigger signal, the control module enters a timed wake-up mode. In this timed wake-up mode, the control module remains in sleep mode, but... c The system wakes up at time intervals and transmits acquisition control commands to the sensors. After receiving the acquisition control commands, the sensors acquire the preload data of the brake disc bolts and return the bolt preload data to the control module after acquisition.

[0081] In the timed wake-up mode of the control module, simultaneously with T s The system transmits a wake-up control command and the brake disc bolt preload data to the remote transmission module at time intervals. After receiving the wake-up control command, the remote transmission module enters the transmission mode from the sleep mode. After receiving the brake disc bolt preload data, the remote transmission module sends out the brake disc bolt preload data and then enters the sleep mode again to wait for the next wake-up control command.

[0082] In this embodiment, the monitoring component is configured with a signal receiving unit and a signal processing unit.

[0083] The signal receiving unit is used to receive the bolt preload status signal collected after the brake disc bolts undergo axial deformation;

[0084] like Figure 3 As shown, during the monitoring of brake disc bolt preload, the signal processing unit performs pre-training, abnormal data filtering, sequence correlation and decomposition processing, wavelet packet decomposition, and frequency band energy filtering on the bolt preload signal to obtain historical feature vectors. The historical feature vectors are then input into the support vector machine classification model for training to obtain the brake disc bolt status detection results.

[0085] In this embodiment, the classification accuracy is improved by more than 20% by combining wavelet packet decomposition with SVM; the whole process time is less than 50ms, which meets the real-time monitoring requirements of EMU braking; the signal processor adopts dynamic voltage regulation (DVS) and the power consumption is less than 10mW.

[0086] In this embodiment, the brake disc bolt preload signal processing unit preprocesses the bolt preload signal and inputs the historical feature vector into the support vector machine classification model for training, including the following steps:

[0087] Step 1: Signal Acquisition and Preprocessing

[0088] Based on the vibration characteristics of the EMU brake disc, the sampling frequency was set to 10kHz, and a suitable sensor type was selected: MEMS piezoresistive sensor. The parameters were set as follows: sensitivity: 2mV / N, range: 0~20kN.

[0089] The sensor acquires the resistance change signal caused by the axial deformation of the bolt in real time and converts it into a voltage signal (0-5V). The signal is amplified using an instrumentation amplifier (gain: 100 times); at the same time, high-frequency noise is removed by a low-pass filter (cutoff frequency: 2kHz).

[0090] A 16-bit ADC module is used to convert analog signals into digital signals with a resolution of 0.076mV / LSB.

[0091] Step 2: Screening out abnormal data

[0092] A sliding window detection method is used, with a window length of 1 second (containing 10,000 data points). The mean (μ) and standard deviation (δ) of the data within the window are calculated. Anomaly detection is performed on the calculation results: if a data point exceeds the range of μ±3δ, it is marked as an anomaly. Linear interpolation is used to replace the outlier values ​​to ensure data continuity.

[0093] Step 3: Sequence Correlation and Decomposition Processing

[0094] The linear trend term was fitted using the least squares method and subtracted from the original signal. The signal was decomposed into six intrinsic mode functions (IMFs), and the low-frequency components (IMF1-IMF3) related to the preload were extracted.

[0095] Step 4: Wavelet packet decomposition and frequency band energy screening

[0096] Wavelet packet decomposition parameters were set as follows: the wavelet basis function was set to db4 (Daubechies 4th order); the decomposition layer was set to 4 layers, generating 16 frequency bands; the frequency bands were selected from 1.5 to 3 kHz, and experiments verified that the energy of this frequency band was strongly correlated with the change in bolt preload.

[0097] The preprocessed signal is decomposed into 4 levels to obtain 16 sub-bands. The energy value of the signal in each sub-band is calculated.

[0098]

[0099] Select the top 3 frequency bands by energy percentage (e.g., sub-bands 5, 7, and 9) to form a feature vector.

[0100] Step 5: Construction of Historical Feature Vectors

[0101] Select time-domain features: mean, variance, and peak factor (3 dimensions); select frequency-domain features: energy values ​​of 3 selected frequency bands (3 dimensions); construct a 6-dimensional historical feature vector F based on the time-domain and frequency-domain features:

[0102]

[0103] In the formula, μ is the mean; δ is the standard deviation; and the peak factor is the peak value. Energy value x i The sub-band signal energy is represented by i, the sub-band number is represented by i, K takes the value [1, N], and N is the number of data points of each sub-band signal, which takes the value 625.

[0104] Step 6: Support Vector Machine (SVM) Model Training

[0105] The parameters of the Support Vector Machine (SVM) model are set as the radial basis function (RBF, γ = 0.1) of the kernel function; the regularization parameter is C = 0.1; where the classification labels are 0 for normal pre-tightening, 1 for loosening, and 2 for over-tightening.

[0106] The training and test sets were divided in a 7:3 ratio. The training set comprised 70% of the data and contained 3,000 samples under different operating conditions; the test set comprised 30% of the data and contained 1,286 samples.

[0107] Five-fold cross-validation was used to optimize the model hyperparameters, and the model evaluation metrics were set as follows: accuracy ≥ 98.5% (test set) and the proportion of the diagonal line of the confusion matrix > 97%.

[0108] Step 7: Real-time monitoring and feedback

[0109] The real-time feature vector is input into the trained SVM model, which outputs the bolt preload status (normal / loose / overtight). The model is then judged according to a preset alarm threshold.

[0110] Loosening threshold: Preload reduction > 15% of rated value;

[0111] Overtightness threshold: Preload increases by more than 10% of the rated value.

[0112] After obtaining real-time monitoring results, the status data and alarm information are sent to the display terminal through the repeater network, with a refresh rate of 1Hz.

[0113] Step 8: Model Update and Maintenance

[0114] When new data (≥500 sets) is added each month, the model is fine-tuned and the support vectors are updated; if the model accuracy is <95% for 3 consecutive tests, full retraining is started.

[0115] like Figure 4 As shown, the brake disc system mainly includes the disc hub, friction disc, and pressure ring.

[0116] The hub, friction disc, and pressure ring are assembled, and then assembled into a brake disc system by means of a monitoring component, brake disc gasket, and brake disc nut. Finally, the system is press-fitted onto the brake disc mount on the axle with an interference fit.

[0117] like Figure 5 As shown in the embodiment of the present invention, the passive wireless EMU brake disc bolt preload real-time monitoring component includes a brake disc bolt body, a mandrel arranged along the extension direction of the mandrel and used to sense changes in the preload of the brake disc bolt body is encapsulated inside the bolt body, and a sensor connected to the mandrel and a signal processor connected to the sensor are encapsulated inside the bolt head.

[0118] From the screw head to the screw rod, there are interconnected encapsulation grooves and core grooves that are located on the same axis as the bolt body. The diameter of the encapsulation groove is larger than the diameter of the core groove, and there is a reserved gap between the bottom of the core groove and the top of the screw rod.

[0119] The sensor and signal processor are installed sequentially from the inside out in the encapsulation groove. The spindle is fitted into the core groove, which is coated with grease around its perimeter, and connects to the sensor in the encapsulation groove. A protective sleeve that wraps around the sensor and signal processor in a ring shape and a protective cover that matches the size of the groove opening and is flush with it are installed circumferentially in the encapsulation groove.

[0120] Specifically, the spindle undergoes tensile and compressive deformation along with the brake disc bolt body, senses the change in the preload of the brake disc bolt body, and transmits it to the sensor. The signal processor then processes it into a bolt preload status signal and transmits it to the repeater.

[0121] In conjunction with specific embodiments, the workflow of the passive wireless EMU brake disc bolt preload real-time monitoring device provided in this embodiment of the invention is as follows:

[0122] The brake disc bolts containing the monitoring components are installed in the brake disc. When the brake disc starts running, the brake disc bolts are subjected to external loads, and their preload changes. The monitoring components transmit the measured bolt preload status signal to the repeater. After the repeater's signal detection unit, signal receiving unit, and signal processing unit train the support vector machine classification model, the bolt preload status signal is obtained and displayed on the remote terminal.

[0123] The passive wireless brake disc bolt preload real-time monitoring component (including encapsulation groove, mandrel, sensor and signal processor) is installed in the EMU brake disc using a precision assembly process. The specific steps are as follows:

[0124] Step 1: Bolt sealing and installation

[0125] A core groove (2.5mm in diameter) is made along the axis inside the screw of the brake disc bolt body, and an encapsulation groove (5mm in diameter) extends from the screw head to the core groove, with the two coaxially connected; a MEMS piezoresistive sensor (range 0~20kN) and a signal processor (power consumption <10mW) are sequentially embedded into the encapsulation groove; the mandrel (material: 316L stainless steel) is coated with high-temperature resistant grease (model: XG-3000) and then inserted into the core groove for precise coupling with the sensor; a silicone protective sleeve (1.2mm thick) and a titanium alloy protective cover are installed inside the encapsulation groove to ensure that the component has a waterproof and dustproof rating of IP67.

[0126] The monitoring bolts are press-fitted to the disc hub, friction disc, and pressure ring using high-strength shims (8.8 grade) and anti-loosening nuts (preload torque 120 N·m), with a tolerance of H7 / g6.

[0127] Step 2: Dynamic monitoring and signal transmission of preload force

[0128] When the brake disc is running, the friction of the brake pads generates a thermo-mechanical coupling load (temperature range -40 to 300℃), which causes the bolts to deform axially (deformation amount 0.01 to 0.2 mm); the spindle is displaced with the deformation, triggering the sensor to output a resistance change signal (sensitivity 2 mV / N), which is amplified and filtered by the signal processor and converted into a digital signal (sampling rate 10 kHz, resolution 16 bit).

[0129] The signal is transmitted to the corresponding repeater via the LoRa wireless protocol (frequency band 433MHz, transmission distance ≤50m); the repeater's main control chip STM32F103C8T6 starts the thermoelectric power generation module (power generation efficiency 12%), the energy storage module (supercapacitor, capacity 10F) is powered, and the trigger control module enters the timed wake-up mode (Tc=1s, T...). s =5s).

[0130] Step 3. Signal Processing and State Classification

[0131] Outliers outside the μ±3δ range were removed using a sliding window method (window length 1s), and then repaired by interpolation.

[0132] Low-frequency components (0–500 Hz) of IMF1–IMF3 were extracted via EMD decomposition; the top three sub-bands (E5, E7, E9) with the highest energy proportions in the 1.5–3 kHz frequency band were selected through 4-level decomposition of db4 wavelet packets; and a 6-dimensional feature vector μ,δ was constructed. 2 C f E5, E7, E9 (C f (Peak factor).

[0133] Input 3,000 sets of training data (70% training set, 30% test set), using the RBF kernel function (γ = 0.1, C = 1.0);

[0134] Output classification labels: 0 (normal, preload fluctuation <5%), 1 (loose, preload decrease ≥15%), 2 (too tight, preload increase ≥10%);

[0135] The model's accuracy rate is ≥98.5%, and the response time is <50ms.

[0136] Step 4. Remote Terminal Display and Alarm

[0137] Input 3,000 sets of training data (70% training set, 30% test set), using the RBF kernel function (γ = 0.1, C = 1.0); the repeater network uploads the processed preload status data to the display terminal via the ZigBee protocol (transmission rate 250kbps); the terminal interface refreshes in real time (frequency 1Hz), displaying the preload change trend of each bolt as a three-dimensional curve, and marking it with color codes: 1) Green: normal state (class 0); 2) Yellow: loosening warning (class 1, triggering audible and visual alarm); 3) Red: overtightening fault (class 2, forced shutdown command).

[0138] Data is synchronously stored on a cloud server, supporting historical data retrieval and maintenance report generation.

[0139] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A passive wireless EMU brake disc bolt preload real-time monitoring device, characterized in that, include: Display terminal, Multiple repeaters, each interconnected with the others; And multiple passive wireless EMU brake disc bolt preload real-time monitoring components, and the repeaters and the monitoring components are connected one-to-one. Each repeater receives the bolt preload status signal sent from the corresponding monitoring component and then converges it to the display terminal. The monitoring component is equipped with a signal receiving unit and a signal processing unit; The signal receiving unit is used to receive the bolt preload state signal collected after the brake disc bolt undergoes axial deformation; The signal processing unit performs pre-training, abnormal data filtering, sequence correlation and decomposition processing, wavelet packet decomposition, and frequency band energy filtering on the bolt preload state signal to obtain historical feature vectors; and inputs the historical feature vectors into a support vector machine classification model for training to obtain a bolt state detection model, thereby obtaining the brake disc bolt preload monitoring signal. The process of obtaining historical feature vectors through screening and training a support vector machine classification model using these historical feature vectors to obtain a bolt state detection model includes the following steps: The sensor collects the resistance change signal caused by the axial deformation of the bolt in real time and converts it into a voltage signal of 0-5V. The signal is amplified by 100 times using an instrumentation amplifier. The cutoff frequency is set to 2 kHz through a low-pass filter to remove high-frequency noise. A 16-bit ADC module is used to convert the analog signal into a digital signal with a resolution of 0.076 mV / LSB. Calculate the mean of 10,000 data points within a window with a window length of 1 second. and standard deviation If a certain data point exceeds If the range is not specified, it is marked as an anomaly, and the anomaly value is replaced using linear interpolation. The linear trend term was fitted using the least squares method and subtracted from the original signal; the subtracted digital signal was decomposed into 6 intrinsic mode functions (IMFs), and the low-frequency components IMF1-IMF3 related to the preload were extracted. The preprocessed digital signal is decomposed into 16 sub-bands by 4-level wavelet packet decomposition. The energy value of each sub-band signal is calculated, and the top 3 frequency bands with the highest energy ratio are selected to form the feature vector. Using the mean, variance, and peak factor of time-domain features, and selecting energy values ​​from three frequency bands based on frequency-domain features, a 6-dimensional historical feature vector F is constructed: ； In the formula, The mean; Standard deviation; peak factor Energy value , x i Indicates the energy of the sub-band signal. i K represents the sub-band number, where K takes values ​​[1, N] and N is the number of data points for each sub-band signal. The training and test sets were divided in a 7:3 ratio. Five-fold cross-validation was used to optimize the hyperparameters of the support vector machine (SVM) model. The model was trained using the training set and evaluated using the test set: accuracy ≥ 98.5% and the diagonal percentage of the confusion matrix > 97%. The real-time feature vector is input into the trained support vector machine (SVM) model, and combined with the preset alarm threshold, the bolt preload state is judged and output. When the preload decreases by more than 15% of the rated value, the bolt preload is considered loose. When the preload increases by more than 10% of the rated value, the bolt preload is considered overtight. All other bolt preload conditions are normal. When ≥500 new data sets are added within a period of time, the model is fine-tuned and the support vectors are updated; if the model accuracy is <95% for 3 consecutive tests, full retraining is initiated.

2. The passive wireless EMU brake disc bolt preload real-time monitoring device according to claim 1, characterized in that, The repeater includes a thermoelectric power generation module, an energy storage module, a control module, and a remote transmission module; The thermoelectric power generation module is connected to the energy storage module via a conversion circuit and to the control module via a trigger circuit; the control module is connected to the energy storage module, the monitoring component, and the remote transmission module respectively; the energy storage module is connected to the monitoring component and the remote transmission module respectively.

3. The passive wireless EMU brake disc bolt preload real-time monitoring device according to claim 2, characterized in that, The trigger circuit determines whether the disc brake is in working condition based on whether the thermoelectric generator module generates electricity. If it is in working condition, the thermoelectric generator module generates electrical energy, and the trigger circuit sends a trigger signal to the control module to make the control module respond and enter the timed wake-up mode.

4. The passive wireless EMU brake disc bolt preload real-time monitoring device according to claim 3, characterized in that, When the disc brake device is not in operation, the thermoelectric power generation module does not generate electricity, and the control module and the remote transmission module are in sleep mode. When the disc brake device is in operation, the control module is in a timed wake-up mode, so that... T c To wake up at time intervals, transmit acquisition control commands to the sensors, and simultaneously... T s The system transmits wake-up control commands and brake disc bolt preload data to the remote transmission module at time intervals. T s = nT c , n The value is a positive integer; so that the remote transmission module responds to the wake-up control command and enters the transmission mode from the sleep mode, and after receiving the brake disc bolt preload data, it sends out the brake disc bolt preload data and enters the sleep mode again to wait for the next wake-up control command.

5. A brake disc system, characterized in that, include: Hub, Friction disc, Pressure ring, And a passive wireless EMU brake disc bolt preload real-time monitoring component, or a passive wireless EMU brake disc bolt preload real-time monitoring device as described in any one of claims 1 to 4. The hub, friction disc, and pressure ring are assembled together via the monitoring component, a gasket and nut that match the monitoring component, and are press-fitted onto the brake disc seat of the axle through an interference fit.

Images