A vibration detection unit, a monitoring system, a method for constructing a monitoring model, and a monitoring simulation system.

Abstract

This invention discloses a vibration monitoring simulation system, which includes a vibrator, a rotating worktable fixed on the vibrator, and multiple vibration detection units. Each vibration detection unit is disposed on the rotating worktable and includes: an acrylic base plate; a first copper thin film layer attached to the surface of the acrylic base plate; a PET film layer attached to the upper surface of the first copper thin film layer; a PTFE film layer attached to the upper surface of the PET film layer; a second copper thin film layer attached to the upper surface of the PTFE film layer; an acrylic top plate disposed on the upper surface of the second copper thin film layer; and at least three springs connecting the acrylic base plate and the acrylic top plate at equally spaced angles, with one end of each spring connected to the acrylic base plate and the other end connected to the acrylic top plate.

Description

Technical Field

[0001] This invention belongs to the field of mechanical testing technology, and specifically relates to a vibration detection unit, a monitoring system, a monitoring model construction method, and a monitoring simulation system. Background Technology

[0002] In the industrial and engineering sectors, many aging aerospace, civil, and mechanical systems remain in use. These systems may suffer from accumulated structural damage, such as fatigue, cracks, structural deformation, and loosening of connections due to vibration. This damage can lead to equipment failures, such as accidents caused by fatigue cracks in aircraft wings, bridge collapses due to wind-induced resonance, and structural cracking in ships. Therefore, accurate vibration monitoring systems are crucial. By collecting and analyzing mechanical condition information in a timely manner through monitoring systems, potential problems can be identified and appropriate maintenance measures can be taken.

[0003] Vibration monitoring typically relies on sensors to detect changes in vibration. Existing sensors are mostly based on piezoelectric, piezoresistive, or capacitive principles, but these methods often have drawbacks, such as requiring an external power source, high cost, and susceptibility to environmental conditions. Furthermore, existing vibration monitoring systems often perform poorly when capturing low-amplitude vibrations or a wide range of mechanical frequencies, failing to fully reflect the complexity of structural vibrations.

[0004] Public content

[0005] One embodiment of this disclosure discloses a vibration energy monitoring system based on the triboelectric effect. This system employs one or more vibration detection units as vibration sensors to acquire vibration signals from the monitored object. A vibration monitoring model is then used to monitor the vibration of the monitored object or structure. The multiple detection units can be arranged in a grid pattern and positioned at points of interest in the vibration monitoring of the monitored object.

[0006] The vibration detection unit, based on the triboelectric effect, is used for detecting vibration energy. This unit includes:

[0007] Acrylic base plate;

[0008] A first copper thin film layer is attached to the surface of the acrylic base plate;

[0009] A PET film layer is attached and bonded to the upper surface of the first copper film layer;

[0010] A PTFE film layer is attached to the upper surface of the PET film layer;

[0011] A second copper thin film layer is attached to the upper surface of the PTFE thin film layer;

[0012] An acrylic top panel is disposed on the upper surface of the second copper thin film layer;

[0013] At least three springs are connected to the acrylic base plate and the acrylic top plate in a mutually equal angle configuration. One end of each spring is connected to the acrylic base plate, and the other end is connected to the acrylic top plate.

[0014] This disclosure presents a vibration energy monitoring system based on the triboelectric effect, characterized by self-powered operation, strong adaptability, and efficient energy recovery. The system automatically generates electrical charges during structural vibration, making it suitable for remote and inaccessible environments, reducing maintenance costs and improving monitoring convenience. Compared to existing monitoring methods, the triboelectric effect-based system offers significant advantages in capturing multi-directional and multi-frequency vibration information, providing more accurate data support, and is particularly suitable for analyzing complex vibration fields and real-time monitoring of structural damage. Attached Figure Description

[0015] The above and other objects, features, and advantages of this disclosure will become readily apparent from the following detailed description of exemplary embodiments, taken in conjunction with the accompanying drawings. Several embodiments of this disclosure are illustrated in the drawings by way of example and not limitation, in which:

[0016] Figure 1 A flowchart of a vibration energy monitoring method based on triboelectric effect according to one embodiment of the present disclosure.

[0017] Figure 2 A schematic diagram comparing the energy distribution algorithm and simulation results according to one embodiment of this disclosure.

[0018] Figure 3 A three-dimensional structural schematic diagram based on the triboelectric effect according to one embodiment of the present disclosure.

[0019] Figure 4 A schematic diagram of sensor frequency identification according to one embodiment of the present disclosure.

[0020] Figure 5 A schematic diagram of a vibration sensor network according to one embodiment of the present disclosure.

[0021] 1—Acrylic base plate, 2—First copper film layer, 3—PET film layer, 4—PTFE film layer, 5—Second copper film layer, 6—Acrylic top plate, 7—Spring. Detailed Implementation

[0022] The purpose of this disclosure is to provide a comprehensive, multi-faceted vibration monitoring method and system based on the triboelectric effect, addressing the problems of incomplete vibration energy recovery and inaccurate monitoring of multidimensional vibrations in existing technologies. The principle of using the triboelectric effect for vibration energy monitoring is based on the coupling effect of triboelectric charging and electrostatic induction. When two materials with different triboelectric sequences come into contact and separate during vibration, charge transfer occurs, forming charges on the material surfaces. Under the influence of electrostatic induction, these charges create a potential difference between the electrodes of the materials, thereby generating a current. This vibration energy monitoring method based on the triboelectric effect can monitor the distribution and intensity of vibration energy by analyzing changes in current and potential difference. The performance of this method can also be optimized through simulation and experimental studies. For example, by adjusting parameters such as the triboelectric sequence of the materials, contact area, and vibration frequency, energy conversion efficiency and monitoring accuracy can be improved.

[0023] According to one or more embodiments, such as Figure 1 As shown, a vibration energy monitoring method based on the triboelectric effect is used for damage localization and structural reinforcement in industrial and engineering structures. The method includes the following steps: 101. First, a three-dimensional structure based on the triboelectric effect is constructed. This structure consists of upper and lower triboelectric layers; the upper layer is a PTFE and PET film, and the lower layer is covered with a copper film as a back electrode. An acrylic mass block is symmetrically suspended by three springs at a 120° angle to achieve the three-dimensional structure based on the triboelectric effect.

[0024] 102. Use a signal generator to generate signals of different frequencies to simulate the vibration environment at different frequencies. This step is to simulate the energy conversion process under actual vibration conditions.

[0025] 103. A rotary table is set up and fixed on the vibrator to collect vibration energy data from multiple angles, simulating vibration scenarios from different angles. The voltage signal output by the sensor is converted into frequency domain data to obtain the dominant frequency information. By comparing the input frequency and the dominant frequency information, accurate frequency identification is ensured. Here, the normal of the rotary table forms an angle with the vibration output direction of the vibrator; this angle refers to the angle between the two. This means that the vibration output direction of the vibrator does not need to be the same as the normal of the vibration monitoring unit.

[0026] 104. To analyze the influence of angle and frequency on energy output, the angle range for energy data acquisition was set to 0° to 90°, the frequency range to 2Hz to 11Hz, and the step size to 0.2Hz. Through three repeated experiments, a multidimensional array of 1500 data sets was generated.

[0027] 105. A CatBoost regression algorithm was used to construct a model relating angle, frequency, and energy recovery efficiency. Angle and frequency were used as independent variables, and energy recovery efficiency as the dependent variable for model training. Here, since the vibration input direction of the exciter (used to simulate actual single-point vibration input) does not need to be consistent with the normal direction of the detection unit, even if the power output of the exciter remains constant, the explicit power of the vibration detection unit will vary with angle differences. The ratio of this explicit power to the actual input power is the energy recovery efficiency. To ensure that the vibration energy detection accuracy of the vibration detection unit at a single point is not affected by directionality, its explicit power needs to be corrected.

[0028] 106. The model was validated using a test set, and its performance metrics were calculated using a 9-fold cross-validation method. This method enables accurate prediction of energy conversion efficiency under vibration scenarios with different frequencies and angles.

[0029] 107. The approach extends from single-point precision energy harvesting to grid-based multi-point energy harvesting. A 3x3 sensor network is deployed on a 1mm thick aluminum plate to acquire sensor data.

[0030] 108. Develop a multi-source hierarchical differential evolution algorithm to generate energy cloud maps. Based on the collected vibration energy data, local optimization is performed to estimate the energy distribution of a single vibration source. In the hierarchical multi-source optimization stage, the interaction between multiple vibration sources is considered, and the position and intensity of the vibration sources are adjusted. Finally, an energy cloud map is generated through global optimization.

[0031] The monitoring method disclosed herein is used to monitor and optimize the vibration energy conversion process based on the triboelectric effect. Through precise experimental design, data acquisition, model building, and optimization, the efficiency and accuracy of vibration energy conversion are improved.

[0032] According to one or more embodiments, a vibration energy monitoring method based on the triboelectric effect includes the following steps:

[0033] S1. A layer of polyethylene terephthalate (PET) film is attached to an acrylic base plate, and a copper film is then placed on top as the back electrode. During fabrication, high-precision bonding equipment is used to ensure that the copper film and PET film surfaces are flat and firmly attached. A layer of polytetrafluoroethylene (PTFE) film is adhered to the bottom of the top acrylic layer to form a friction pair with the copper electrode. Three springs are symmetrically suspended at a 120° angle from the top of the acrylic plate. During installation, it is ensured that the springs are perpendicularly connected to the acrylic mass block, allowing it to vibrate freely in both the vertical and planar directions, thus forming a three-dimensional structure based on the triboelectric effect, such as... Figure 3 As shown, the three-dimensional structure here is a three-dimensional triboelectric nanogenerator structure that can simultaneously respond to vibrations in the x, y, and z directions.

[0034] Figure 3 English terms in the text include:

[0035] PTFE—polytetrafluoroethylene, a plastic material with high chemical stability and a low coefficient of friction. PET—polyethylene terephthalate, a synthetic polymer commonly used in the manufacture of plastic bottles and other containers.

[0036] Copper is a metal with good electrical conductivity and is often used to make electrodes.

[0037] Acrylic – Acrylic is a transparent plastic material with good weather resistance and processing properties. Spring – Spring is a mechanical component that can store elastic potential energy and is used to provide vibration.

[0038] Figure 3 A three-dimensional structure based on the triboelectric effect is described for vibration energy monitoring. This structure comprises the following parts:

[0039] 1. Acrylic base plate – serving as the foundation of the entire structure, providing a sturdy platform to support other components.

[0040] 2. PET film – attached to an acrylic substrate, serving as a layer for triboelectric effect. PET film possesses excellent insulating properties and can form effective triboelectric pairing with copper film.

[0041] 3. Copper thin film – coated on a PET film, serving as the back electrode. The conductivity of the copper thin film allows it to effectively collect charges generated by the triboelectric effect.

[0042] 4. PTFE film – adhered to the bottom of the acrylic top layer, forming a triboelectric pair with the copper electrode. The low coefficient of friction and high chemical stability of PTFE film make it an ideal triboelectric material.

[0043] 5. Spring and acrylic mass – symmetrically suspended at a 120° angle from the top of the acrylic sheet. The spring provides the elastic force for vibration, while the acrylic mass contacts and separates from the PTFE film during vibration, generating triboelectric charge.

[0044] The three-dimensional structural unit based on the triboelectric effect includes: an acrylic base plate; a first copper thin film layer attached to the surface of the acrylic base plate; a PET film layer attached to the upper surface of the first copper thin film layer; a PTFE film layer attached to the upper surface of the PET film layer; a second copper thin film layer attached to the upper surface of the PTFE film layer; an acrylic top plate disposed on the upper surface of the second copper thin film layer; and at least three springs connecting the acrylic base plate and the acrylic top plate at equally spaced angles, with one end of each spring connected to the acrylic base plate and the other end connected to the acrylic top plate.

[0045] During vibration, triboelectric charges are generated due to the contact and separation between the PTFE film and the copper electrode. These charges can be collected by the copper electrode and converted into electrical signals, thereby enabling the monitoring of vibration energy. The vibration energy monitoring method based on the triboelectric effect can be used for applications such as structural health monitoring, vibration energy harvesting, and self-powered sensor systems.

[0046] S2 uses a signal generator to generate signals of different frequencies and transmits them to the exciter through a power amplifier to simulate the vibration environment at different frequencies.

[0047] S3 acquires the voltage signal output by the sensor, converts the time-domain data to frequency-domain data using Fourier transform, and obtains the dominant frequency information. By comparing the input frequency and the dominant frequency information, it ensures that the sensor can accurately identify the frequency (taking input frequencies of 5.4Hz and 9.3Hz as examples). Figure 4 As shown,

[0048] Figure 4 English terminology in Chinese:

[0049] Main frequency refers to the frequency component in a signal where energy is most concentrated.

[0050] Frequency (Hz) – Frequency (Hertz) represents the number of times a periodic event repeats per second.

[0051] Figure 4The diagram shows the frequency domain data obtained after Fourier transforming the voltage signal output by the sensor. The left figure shows the frequency domain data when the input frequency is 5.4 Hz. The most significant peak appears at 5.4 Hz, indicating that the sensor successfully identified the dominant input frequency. Other smaller peaks may represent noise or other frequency components in the signal. The right figure shows the frequency domain data when the input frequency is 9.3 Hz. Similarly, the most significant peak appears at 9.3 Hz, again proving that the sensor can accurately identify the dominant input frequency. Accurate frequency identification using sensors allows for real-time monitoring of the mechanical system's operating status and timely detection of abnormal vibrations. S4, a rotating worktable is designed and fixed to a vibrator to simulate vibration scenarios from different angles, ensuring that each vibration data acquisition covers multiple directions, forming comprehensive multi-angle vibration energy distribution data. The vibrator generates vibrations of different frequencies and amplitudes, thereby exciting the vibration response of the rotating worktable or other structures. In vibration energy monitoring research, the vibrator is used in conjunction with sensors and data acquisition systems to achieve accurate measurement and analysis of vibration energy.

[0052] S5, to analyze the influence of angle and frequency on energy output, the angle range for vibration energy data acquisition is set to an angle array of Angle = [0°, 10°, ..., 90°]. The frequency range is set to a frequency array of Frequency = [2, 2.2, ..., 11] with a step size of 0.2 Hz. In the experiment, three repeated measurements were performed to ensure the reliability and consistency of the data, ultimately resulting in a multidimensional dataset containing 1500 data points. Here, there are other ways to select angle and frequency within the target range, including uniformly selecting no fewer than 10 points each; therefore, the settings can be adjusted as needed.

[0053] S6. Using angle and frequency as independent variables and energy recovery efficiency as the dependent variable, an initial dataset is formed. The preprocessed dataset is then randomly divided into training and test sets in a 7:3 ratio. This ensures that the training set is used to build the model, and the test set is used to verify the model's generalization ability. The dataset division must be random and representative to avoid bias in the model during training.

[0054] S7. Before building the model, set the key parameters of the CatBoost regression algorithm (or other machine learning methods with regression capabilities). Specifically, this includes: setting the learning rate to control the model's convergence speed and accuracy; setting the maximum depth of the decision tree to control model complexity; defining the number of training iterations to ensure the model fully learns the data features; and setting regularization parameters to avoid overfitting. These parameters need to be fine-tuned experimentally to find the optimal combination to ensure the model's accuracy and generalization ability.

[0055] In step S8, after setting the parameters, the CatBoost regression model is trained using angles and frequencies from the training set as input features and energy recovery efficiency as the target variable. The model iterates to continuously optimize the decision tree, gradually improving its accuracy in predicting energy recovery efficiency. CatBoost employs a gradient boosting algorithm, updating the model weights in each iteration by minimizing the prediction error.

[0056] S9. Validate the trained model using the test set and calculate the model's performance metrics (such as R-squared). 2 The predictive ability of a model is evaluated using metrics such as MAE and RMSE. During the initial validation process, if the model performs poorly, parameters or feature encoding methods need to be readjusted. Cross-validation (e.g., 9x cross-validation) can be used to reduce the risk of overfitting. This method can accurately predict energy conversion efficiency under vibration scenarios of different frequencies and angles.

[0057] S10 expands from single-point precise energy harvesting to grid-based multi-point energy harvesting. This disclosure arranges a 3x3 sensor network on a 1mm thick aluminum plate (the number and placement of sensors can vary depending on the required harvesting accuracy and structural area) to acquire sensor data, such as... Figure 5 As shown, the sensor recorded the energy distribution data generated on the plate by different vibration sources;

[0058] S11, Based on the vibration field energy attenuation formula, set the position and intensity parameters of the vibration source, where the position of the vibration source is S=(X0,Y0,α), and the sensor is located at (X... i ,Y i The energy decay formula is as follows:

[0059]

[0060] Where α is the vibration source intensity and c is an adjustment factor, this formula is used to predict the energy attenuation at the sensor; S12, the vibration field region is divided into grid regions, assuming that there is only one vibration source in each grid region. A differential evolution algorithm is used for local optimization, and the optimization objective is the sum of squared errors between the actual measured value and the predicted value. The objective function of the local optimization is the actual sensor energy measurement value E. means Compared with the predicted value E pred Sum of squared errors:

[0061]

[0062] Where N represents the number of sensors. After local optimization, the optimal vibration source location and intensity within each grid are obtained;

[0063] S13, entering the hierarchical multi-source optimization stage, considering the interaction of multiple vibration sources. An interaction term is introduced:

[0064]

[0065] Where σ is the spatial scale of the interaction, (X i ,Y i ) and (X j ,Y j ( ) represents the coordinates of the two vibration sources. This interaction term is used to correct the energy coupling between the vibration sources and improve the overall fitting accuracy;

[0066] S14, based on the results of hierarchical optimization, enters the global optimization stage, inputs the initial parameters of all vibration sources into the differential evolution algorithm, and further minimizes the sum of squared energy prediction errors through global optimization:

[0067]

[0068] Where M represents the total number of vibration sources. The result of global optimization determines the final location and intensity of the vibration sources;

[0069] S15, Through the above steps, an energy cloud map of the vibration field is generated, showing the actual location and intensity distribution of the vibration source. The reliability can be verified by comparing the actual data with the simulated energy cloud map (taking vibration sources of (50.0, 160.0) and (165.0, 75.0) as examples). Figure 2 As shown. Figure 2 The English terms used in this context are as follows:

[0070] Energy refers to the energy distribution of a vibrational field within a specific region.

[0071] X (mm) — X coordinate (millimeters), representing the horizontal position on a two-dimensional plane.

[0072] Y(mm) — Y coordinate (millimeters), representing the position in the vertical direction on a two-dimensional plane.

[0073] Figure 2 The left figure shows the actual energy distribution on a two-dimensional plane. The shades of color represent energy levels, with darker colors indicating higher energy. The positions of two vibration sources are marked (40.7, 165.0) and (176.0, 69.3), obtained through actual measurements. The right figure shows the energy cloud map obtained through simulation. Similarly, the shades of color represent energy levels, and the positions of two vibration sources are marked (50.0, 160.0) and (165.0, 75.0), predicted by the simulation.

[0074] By comparing the energy distribution obtained from actual measurements (left figure) with the energy distribution predicted by simulation (right figure), the accuracy and reliability of the simulation model were verified.

[0075] According to the embodiments of this disclosure, the innovative energy recovery and vibration monitoring method based on the triboelectric effect proposed in this disclosure mainly solves the shortcomings of existing sensor technologies in terms of vibration monitoring accuracy and applicability. Existing technologies not only rely on external power sources, but also have limitations in energy recovery efficiency and vibration monitoring capabilities under complex vibration scenarios, making it difficult to comprehensively capture minute vibrations and multi-frequency vibration information, resulting in low monitoring accuracy and response speed. This method achieves precise vibration monitoring from multiple angles and frequencies through a three-dimensional structure and accurately locates energy concentration areas, giving the system higher sensitivity and adaptability. This disclosure significantly improves the accuracy of vibration energy recovery and vibration monitoring, and can be widely applied in aerospace, large equipment, bridge monitoring, and other fields, especially suitable for vibration monitoring and damage assessment of complex structures that are remote or difficult to access.

[0076] In summary, the vibration energy monitoring method based on triboelectric effect disclosed herein includes the following steps:

[0077] First, a three-dimensional structure based on a 120° symmetrical suspension spring is constructed. This structure achieves omnidirectional three-dimensional energy recovery by combining two operating modes: vertical contact separation and in-plane sliding. The simultaneous existence of these two operating modes satisfies the requirement for a single device to respond to vibrations in the x, y, and z directions.

[0078] To more accurately predict the vibration energy capture effect, this disclosure uses the CatBoost algorithm to construct a model relating angle, frequency and energy recovery efficiency.

[0079] Finally, this disclosure presents a multi-source hierarchical differential evolution algorithm for generating energy cloud maps and accurately locating energy concentration areas. This algorithm can effectively integrate energy information from multiple vibration sources to generate energy cloud maps reflecting the internal vibration energy distribution of the structure, thus providing reliable data support for vibration damage analysis of the structure.

[0080] Under this technical process, the present disclosure can realize comprehensive three-dimensional vibration energy recovery and accurately predict and locate vibration energy from multiple angles and frequencies, thereby improving the accuracy of structural damage analysis and the practicality of the energy recovery system, and has broad prospects for industrial and engineering applications.

[0081] Therefore, the core technical features of this disclosure include:

[0082] 1. Based on the triboelectric effect, a three-dimensional structure is used to combine vertical and in-plane sliding modes, effectively capturing vibration energy from multiple angles and frequencies, and achieving precise energy acquisition at a single point.

[0083] 2. Employing machine learning methods, a nonlinear relationship model between angle, frequency, and energy recovery efficiency is trained to accurately predict energy capture performance, making it suitable for complex industrial and engineering environments;

[0084] 3. By defining a vibration energy attenuation function and combining hierarchical local optimization and global correction strategies, an energy cloud map reflecting the energy distribution inside the structure is generated, accurately locating energy concentration areas and providing reliable structural damage data support.

[0085] 4. The execution process and implementation logic of the entire vibration monitoring and energy recovery method. Here, energy recovery means that when vibration occurs at the detection point, the three-dimensional triboelectric nanogenerator structure converts the vibration energy into electrical energy for output, thereby realizing the energy recovery process.

[0086] Accordingly, the beneficial effects of this disclosure include:

[0087] 1. A three-dimensional structure based on the triboelectric effect was realized, which effectively combined the capture of vertical and horizontal vibration energy, overcoming the limitations of traditional sensor technology in terms of directionality and significantly improving the comprehensiveness of energy recovery.

[0088] This is because, in the three-dimensional structure disclosed herein, the capacitance of the parallel-plate capacitor formed by PTFE and PET changes regardless of whether the vibration is vertical or horizontal, while the charge carried by these two friction materials remains unchanged due to friction. Therefore, the charge induced on the copper electrode will continuously change with its vertical and horizontal vibration, resulting in a current in the external circuit. This current output is detected and analyzed as a signal for vibration energy recovery.

[0089] 2. Energy recovery efficiency can be predicted using angle and frequency information, enabling high-precision prediction of vibration energy in complex environments, thus improving the intelligence and accuracy of vibration monitoring systems.

[0090] 3. By deploying sensors, a cloud map reflecting the energy distribution of the structural vibration field can be generated, providing reliable data support for the accurate location of structural vibration damage and enhancing the practical value of the system.

[0091] The above description is merely a specific embodiment of this disclosure, but the scope of protection of this disclosure is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this disclosure, and these modifications or substitutions should all be covered within the scope of protection of this disclosure. Therefore, the scope of protection of this disclosure should be determined by the scope of the claims.

Claims

1. A method for constructing a vibration monitoring model, characterized in that, The vibration monitoring model uses the CatBoost regression algorithm with angle and frequency as independent variables and energy recovery efficiency as the dependent variable to construct a nonlinear relationship model between angle, frequency, and energy recovery efficiency. The vibration monitoring model is constructed by collecting vibration data training and testing sample sets through a vibration monitoring simulation system. The vibration data includes energy recovery efficiency data at different angles and frequencies, where energy recovery efficiency is defined as the ratio of the explicit power of the vibration detection unit to the actual input power. The vibration monitoring simulation system includes a vibrator, a rotary table fixed on the vibrator, and one or more vibration detection units, which are disposed on the rotary table. The normal of the rotary table forms an angle with the vibration direction output by the exciter, with the angle range set from 0° to 90°. This is used to simulate vibration scenarios from different angles, collect vibration energy data from multiple angles, and form comprehensive multi-angle vibration energy distribution data. The vibration detection unit, based on the triboelectric effect, is used for detecting vibration energy and includes: Acrylic base plate; A first copper thin film layer is attached to the surface of the acrylic base plate; A PET film layer is attached and bonded to the upper surface of the first copper film layer; A PTFE film layer is attached to the upper surface of the PET film layer; A second copper thin film layer is attached to the upper surface of the PTFE thin film layer; An acrylic top panel is disposed on the upper surface of the second copper thin film layer; At least three springs are connected to the acrylic base plate and the acrylic top plate in an equally spaced angle. One end of each spring is connected to the acrylic base plate and the other end is connected to the acrylic top plate, so that the acrylic base plate and the acrylic top plate can vibrate freely in the vertical and planar directions, thereby forming a three-dimensional structure based on the triboelectric effect that simultaneously responds to vibrations in the x, y, and z directions.

2. The vibration monitoring model construction method according to claim 1, characterized in that, The vibration data obtained by the vibration detection unit is collected with an angle range of 0° to 90°, a frequency range of 2 Hz to 11 Hz, and a step size of 0.2 Hz. Three repeated measurements are performed to form a multidimensional dataset containing 1500 sets of data.

3. A vibration monitoring system, characterized in that, The vibration detection unit as described in claim 1 is used as a vibration sensor to acquire the vibration signal of the monitored object. The vibration monitoring model obtained by the vibration monitoring model construction method as described in claim 1 is used to perform vibration monitoring and energy recovery efficiency prediction on the monitored object.

4. The vibration monitoring system according to claim 3, characterized in that, The vibration monitoring system includes multiple vibration detection units as described in claim 1 arranged in a grid pattern. The vibration detection units are set at the points of interest for vibration monitoring of the monitored object. A 3x3 sensor network is arranged on a 1mm thick aluminum plate to acquire sensor data, expanding from single-point precise energy acquisition to grid-based multi-point energy acquisition.

5. The vibration monitoring system according to claim 4, characterized in that, Based on the collected grid-based multi-point vibration energy data, an energy cloud map is generated using a multi-source hierarchical differential evolution algorithm for structural vibration damage analysis. The algorithm includes the following steps: S11, Based on the vibration field energy attenuation formula, the position and intensity parameters of the vibration source are set, including the vibration source position and intensity parameters as follows: The sensor is located The energy decay formula is as follows: （1） in, For the source intensity, As an adjustment factor, this energy decay formula is used to predict the energy decay at the sensor. S12. Divide the vibration field region into grid regions, assuming that there is only one vibration source in each grid region. Use the differential evolution algorithm for local optimization. The optimization objective is the sum of squared errors between the actual measured value and the predicted value. The objective function of local optimization is the actual sensor energy measurement value. Compared with the predicted value Sum of squared errors: （2） in, Given the number of sensors, the optimal vibration source location and intensity within each grid are obtained after local optimization. S13, Entering the hierarchical multi-source optimization stage, considering the interaction of multiple vibration sources, and introducing an interaction term: （3） in, For the source intensity, For the spatial scale of interaction, and Let be the coordinates of the two vibration sources. This interaction term is used to correct the energy coupling between the vibration sources. S14, based on the results of hierarchical multi-source optimization, enters the global optimization stage, inputs the initial parameters of all vibration sources into the differential evolution algorithm, and further minimizes the sum of squared energy prediction errors through global optimization: （4） in, Given the total number of vibration sources, the result of global optimization determines the final location and intensity of the vibration sources; S15. Through the above steps, an energy cloud map of the vibration field is generated, showing the actual location and intensity distribution of the vibration source. The reliability is verified by comparing the actual data with the simulated energy cloud map.

6. The vibration monitoring system according to claim 5, characterized in that, The energy cloud map is used for accurate location of structural vibration damage. It compares and verifies the energy distribution obtained by actual measurement with the energy distribution predicted by simulation. The energy level is represented by the color intensity, with darker colors indicating higher energy, thereby identifying the damaged area inside the structure.

Images