A micro-vibration prediction system and method based on actual measurement
By deploying vibration acquisition points both inside and outside the building, calculating the vibration transfer function, and correcting the model in real time, the problem of inaccurate vibration prediction in existing technologies has been solved, achieving real-time and accurate vibration prediction and early warning.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHANGHAI ARCHITECTURAL DESIGN & RES INST
- Filing Date
- 2026-03-30
- Publication Date
- 2026-07-24
AI Technical Summary
Existing technologies struggle to establish accurate dynamic response models between vibration sources, paths, and receptors, making it impossible to predict the vibration state of internal sensitive points in real time and quantitatively, and they lack adaptive correction models.
By deploying vibration acquisition points both inside and outside the building, data is collected synchronously, the vibration transfer function is calculated, and the model is corrected using real-time monitoring data, thus enabling vibration prediction and early warning.
It achieves accurate vibration prediction based on measured data, reduces numerical simulation errors, transforms into a pre-prediction mode, and provides real-time early warning capabilities.
Smart Images
Figure CN122448348A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of environmental vibration detection technology, specifically a micro-vibration prediction system and method based on actual measurements. Background Technology
[0002] Currently, the control of environmental micro-vibrations mainly relies on pre-construction simulation prediction and post-construction passive monitoring. Traditional prediction methods are mostly based on empirical formulas or finite element numerical simulations. However, due to the complexity of geological conditions, the nonlinearity of soil-structure interaction, and the time-varying characteristics of vibration sources, simulation results often deviate significantly from actual field measurements. Existing technologies mainly suffer from the following shortcomings: 1. It is difficult to establish an accurate dynamic response model between the vibration source, path, and receiver; 2. It is impossible to use readily monitorable external environmental data to predict the vibration state of internal sensitive points in real time and quantitatively; 3. It lacks a mechanism that can adaptively correct the model according to environmental changes. Summary of the Invention
[0003] The purpose of this invention is to provide a micro-vibration prediction system and method based on actual measurements, so as to solve the problems mentioned in the background art.
[0004] To achieve the above objectives, the present invention provides the following technical solution:
[0005] A micro-vibration prediction system based on actual measurements, comprising:
[0006] The measuring point layout module is used to deploy external vibration acquisition points along the propagation path of vibration sources outside the building, as well as internal vibration acquisition points in sensitive areas inside the building where vibration control is required.
[0007] The data acquisition module is used to simultaneously acquire vibration response data from the external environment and vibration response data from internal sensitive points;
[0008] The transfer function calculation module is used to calculate the vibration transmission law from the external vibration source to the internal sensitive point through the geological layer and building structure based on the collected vibration response data, and generate the vibration transfer function.
[0009] The vibration prediction module is used to convert real-time monitored external environmental vibration response data into predicted vibration values for sensitive areas inside the building using the vibration transfer function.
[0010] The verification and correction module is used to compare the measured values of internal sensitive points with the predicted vibration values generated by the vibration prediction module, and to dynamically correct the vibration transfer function based on the comparison results.
[0011] As a further embodiment of the present invention: in the measuring point arrangement module,
[0012] External vibration acquisition points are located outside the building foundation, on the ground surface or in the soil layer between the main vibration source and the building, and are used to capture vibration source characteristics, the filtering effect of the site soil, and the input characteristics of road traffic vibration.
[0013] Internal vibration acquisition points are located on the mounting bases of sensitive equipment or on the structural bearing points inside the building, and are used to collect real vibration response data of sensitive areas inside the building.
[0014] As a further aspect of the present invention: in the transfer function calculation module,
[0015] The laws governing vibration transmission include: the coefficient of variation of vibration amplitude, the attenuation characteristics of frequency components, and phase changes;
[0016] The vibration transfer function is either the frequency response function established in the frequency domain through transmissibility analysis, or the impulse response function established in the time domain through a system identification algorithm.
[0017] As a further embodiment of the present invention: the verification and correction module is also used for:
[0018] Periodically or when the external environment changes, temporary measurements are taken at the internal vibration acquisition points. The measured data are compared with the predicted data of the same period for error analysis, and the vibration transfer function is iteratively updated using Kalman filtering or least squares method.
[0019] A method for predicting micro-vibrations based on actual measurements includes the following steps:
[0020] S1. Vibration control index determination and vibration source identification: Based on the building's process requirements, determine the vibration control standards for sensitive areas inside the building, and identify the main vibration source types and locations in the surrounding area;
[0021] S2. Optimization of measuring point layout: Based on the results identified in step S1, external vibration acquisition points are set up along the propagation path between the external vibration source and the building, and internal vibration acquisition points are set up in the sensitive areas inside the building.
[0022] S3. On-site measurement and synchronous data acquisition: Simultaneously use vibration response data from external and internal vibration acquisition points over a continuous time period;
[0023] S4. Signal Processing and Transmission Law Analysis: Process the collected vibration response data, calculate the vibration transfer function from the external measuring point to the internal measuring point, and construct a vibration prediction model;
[0024] S5. Vibration prediction model verification: Using the vibration transfer function obtained in step S4, external measured data is used as input to predict the internal response, and the prediction results are compared with the measured results of internal measuring points to verify the accuracy of the vibration prediction model.
[0025] S6. Deployment of Real-time Prediction System: After verification, the vibration transfer function will be embedded in the vibration prediction module. By monitoring the vibration of external measuring points in real time, the real-time predicted vibration value of the internal sensitive area will be automatically output.
[0026] S7. Dynamic Model Correction: Periodically or when the environment changes, the vibration transfer function is compared, calibrated, and dynamically updated using measured data from temporarily arranged internal measuring points.
[0027] As a further aspect of the present invention: in step S4, the specific steps of signal processing and transmission law analysis are as follows:
[0028] S41. Preprocess the synchronously acquired external environmental vibration response data and internal sensitive point vibration response data;
[0029] S42. Perform Fast Fourier Transform on the external environmental vibration response data and the internal sensitive point vibration response data to obtain their respective spectra, and then use... The method estimates the individual power spectra of each component, as well as the cross-power spectra between them;
[0030] S43. Calculate the vibration transfer function based on the self-power spectrum and the cross-power spectrum.
[0031] As a further aspect of the present invention: in step S5, the specific steps for verifying the vibration prediction model are as follows:
[0032] S51. Among the collected external environmental vibration response data and internal sensitive point vibration response data, a set of synchronous measured data that did not participate in the processing of step S4 is reserved as a verification set.
[0033] S52. Using the external environmental vibration response data in the validation set as input, predict the corresponding internal response through the vibration transfer function;
[0034] S53. Compare the predicted internal response with the measured internal response, and calculate the error evaluation index;
[0035] S54. Compare the calculated error evaluation index with the preset verification threshold. If the error is within the threshold range, the prediction model is deemed to have passed the verification. Otherwise, return to step S2 or S4 to re-optimize the measurement point layout or vibration transfer function parameters.
[0036] Compared with the prior art, the beneficial effects of the present invention are as follows:
[0037] This invention establishes a vibration transfer function based on field measurement data, which truly reflects the actual attenuation law of vibration source propagating to sensitive points through geology and structure, avoiding errors caused by parameter uncertainty in pure numerical simulation; and changes the monitoring mode from post-event detection to pre-event prediction, using external real-time data to predict the internal vibration state, which can issue early warnings before vibration events reach sensitive equipment, buying time for production scheduling or active vibration isolation. Attached Figure Description
[0038] Figure 1 This is a schematic diagram of a micro-vibration prediction system based on actual measurements;
[0039] Figure 2 This is a flowchart illustrating a micro-vibration prediction method based on actual measurements. Detailed Implementation
[0040] Please see Figure 1 In this embodiment of the invention, a micro-vibration prediction system based on actual measurements includes:
[0041] The measuring point layout module is used to deploy external vibration acquisition points along the propagation path of the vibration source outside the building, and internal vibration acquisition points in sensitive areas inside the building where vibration control is required. External vibration acquisition points are located outside the building foundation, on the ground surface or in the soil layer between the main vibration source and the building, to capture vibration source characteristics, the filtering effect of the site soil, and the input characteristics of road traffic vibration. Internal vibration acquisition points are located on the mounting bases or structural bearing points of sensitive equipment inside the building, to collect real vibration response data of sensitive areas inside the building.
[0042] The data acquisition module is connected to external and internal vibration acquisition points to synchronously acquire external environmental vibration response data and internal sensitive point vibration response data.
[0043] The transfer function calculation module is used to calculate the vibration transmission law from the external vibration source through the geological layer and building structure to the internal sensitive point based on the collected vibration response data, and generate the vibration transfer function; the vibration transmission law includes: the variation coefficient of vibration amplitude, the attenuation characteristics of frequency components, and phase change;
[0044] The vibration transfer function is either the frequency response function established in the frequency domain through transmissibility analysis, or the impulse response function established in the time domain through a system identification algorithm.
[0045] The vibration prediction module is connected to external vibration acquisition points in real time. It is used to convert the real-time monitored external environmental vibration response data into predicted vibration values for sensitive areas inside the building using the vibration transfer function.
[0046] The verification and correction module is used to compare the measured values of internal sensitive points with the predicted vibration values generated by the vibration prediction module, and to dynamically correct the vibration transfer function based on the comparison results; it is also used to periodically or when the external environment changes to perform temporary measurements on internal vibration acquisition points, perform error analysis on the measured data and the predicted data of the same period, and use Kalman filtering or least squares method to iteratively update the vibration transfer function.
[0047] Please see Figure 2 In this embodiment of the invention, a micro-vibration prediction method based on actual measurements includes the following steps:
[0048] S1. Vibration control index determination and vibration source identification: Based on the building's process requirements, determine the vibration control standards for sensitive areas inside the building, and identify the main vibration source types and locations in the surrounding area;
[0049] Suppose an integrated circuit manufacturing company is building a new 12-inch wafer fab, with the core area being the lithography zone, where a 28nm process lithography machine is planned to be installed. The lithography machine is extremely sensitive to micro-vibration environments, and its permissible vibration standard is VC-E level (velocity standard), meaning that the effective value of the vibration velocity at all one-third octave band center frequencies (1–80Hz) must not exceed 3.2 μm / s (micrometers per second), and the requirements are even stricter for some sensitive frequency bands (such as 8–25Hz) (must be controlled below 1.5 μm / s).
[0050] Based on the on-site survey, the main sources of vibration around the factory building include:
[0051] Vibration source 1: The subway line 2 tunnel, located 50m from the outer wall of the factory, is buried at a depth of 18m. The high frequency of train operation makes it the main source of vibration. Frequency characteristics: main frequency 20-40Hz, secondary frequency 50-63Hz.
[0052] Source 2: The main urban road 20m outside the factory boundary on the north side of the factory building, where heavy trucks frequently pass; Frequency characteristics: Main frequency 8-20Hz, caused by vehicle suspension and tire excitation;
[0053] S2. Optimization of measuring point layout: Based on the results identified in step S1, external vibration acquisition points are set up along the propagation path between the external vibration source and the building, and internal vibration acquisition points are set up in the sensitive areas inside the building.
[0054] Assuming we continue with the example from step S1,
[0055] 1) Two external vibration acquisition points are set up as follows:
[0056] Measurement point A: Located on the surface of the soil between the subway tunnel and the factory building, at the midpoint of the line connecting the projection point directly above the subway tunnel and the outer wall of the factory building, with a burial depth of 0.5m, used to capture the characteristics of the subway vibration source and the filtering effect of the site soil.
[0057] Measurement point B: Located on the ground surface inside the factory boundary wall, 5m from the road shoulder, to capture the input characteristics of road traffic vibration;
[0058] 2) An internal vibration acquisition point is set up as follows:
[0059] Measurement point C: Located on the concrete floor base of the lithography machine's intended installation area. This is the direct mounting base for the lithography machine, used to collect real vibration response data of sensitive areas inside the building, serving as a reference output for calibrating the vibration transmission law.
[0060] S3. On-site measurement and synchronous data acquisition: Simultaneously use vibration response data from external and internal vibration acquisition points over a continuous time period;
[0061] Assuming, following the example in step S2, select a period of time (22:00-24:00) during which the subway is operating normally at night and road traffic flow is moderate to conduct continuous synchronous data collection for 2 hours;
[0062] 1) The parameters collected by the sensor are as follows:
[0063] Sampling frequency: (Meets the requirements for vibration analysis up to 100Hz);
[0064] Recording duration: 2 hours (7200 seconds);
[0065] Synchronization method: All sensors are synchronized via the same data acquisition unit to ensure time consistency;
[0066] 2) The collected data is as follows:
[0067] Vibration velocity time history at measuring point A Vibration velocity time history at measuring point B: Vibration velocity time history at measuring point C: ;
[0068] S4. Signal Processing and Transmission Law Analysis: The collected vibration response data is processed to calculate the vibration transfer function from external measuring points to internal measuring points, and a vibration prediction model is constructed. The specific implementation steps are as follows:
[0069] S41. Response data to synchronously acquired external environmental vibration data. Vibration response data of internal sensitive points Preprocessing includes: removing DC components (e.g., subtracting the mean from the data of each channel to eliminate sensor zero-point drift), bandpass filtering (e.g., performing zero-phase bandpass filtering on the data according to the frequency band of interest in building vibration control to retain the effective frequency band and filter out high and low frequency interference), and windowing and segmentation (e.g., dividing long-term series data into several time segments of equal length).
[0070] S42. External environmental vibration response data for each segment Vibration response data of internal sensitive points Perform a Fast Fourier Transform to obtain the spectrum. and and adopt Method for estimating the self-power spectrum of external environmental vibration response data Self-power spectrum of vibration response data of internal sensitive points and the cross-power spectrum between the two. ,in, for The complex conjugate;
[0071] S43. Calculate the vibration transfer function based on the self-power spectrum and cross-power spectrum. ,in, It is a complex frequency response function, which includes amplitude and phase angle; its amplitude This represents the amplitude transfer coefficient from external vibration to the internal sensitive point, and the phase angle. Indicates phase delay;
[0072] Assuming, following the example in step S3, the amplitude transfer coefficients from measuring point A to measuring point C, calculated using the vibration transfer function, are 0.35, 0.79, 0.12, and 0.95 at 10Hz, 25Hz, 50Hz, and 63Hz, respectively; that is:
[0073] The vibration at 10Hz is transmitted from measuring point A to measuring point C, with an amplitude attenuation of 35%.
[0074] The vibration attenuates to 79% at 25Hz, which is relatively small.
[0075] Vibration at 50Hz is significantly attenuated, indicating a noticeable filtering effect from the soil layer.
[0076] The attenuation is very small near 63Hz, close to a certain mode of the structure;
[0077] The coherence function coefficients are 0.91, 0.86, 0.94, and 0.79, respectively. It can be seen that the coherence function coefficient is <0.8 at 63Hz, which should be noted when using the transfer function at this point.
[0078] S5. Vibration Prediction Model Verification: Using the vibration transfer function obtained in step S4, external measured data is used as input to predict the internal response. The prediction results are then compared with the measured results at the internal measuring points to verify the accuracy of the vibration prediction model. The specific implementation steps are as follows:
[0079] S51. Among the collected external environmental vibration response data and internal sensitive point vibration response data, a set of synchronous measured data that did not participate in the processing of step S4 is reserved as a verification set.
[0080] S52. Using the external environmental vibration response data in the validation set as input, predict the corresponding internal response through the vibration transfer function;
[0081] S53. Compare the predicted internal response with the measured internal response, and calculate the error evaluation index; among which, the error evaluation index includes time domain indexes (such as root mean square error, peak relative error, waveform correlation coefficient) and frequency domain indexes (such as 1 / 3 octave spectrum relative error = (predicted 1 / 3 octave spectrum value - measured 1 / 3 octave spectrum value) / 1 / 3 octave center frequency);
[0082] S54. Compare the calculated error evaluation index with the preset verification threshold. If the error is within the threshold range, the prediction model is deemed to have passed the verification. Otherwise, return to step S2 or step S4 to re-optimize the measurement point layout or vibration transfer function parameters.
[0083] Assuming we continue with the example from step S4,
[0084] 1) The frequency domain performance comparison and error analysis are as follows:
[0085] At center frequencies of 8, 16, 25, 50 and 63 Hz, the relative errors of the 1 / 3 octave spectrum are -6.7%, -7.1%, -4.9%, +11.1% and -8.3%, respectively. The relative errors of the 1 / 3 octave spectrum at all frequencies are within ±15%.
[0086] 2) The comparison and error analysis of time-domain indicators are as follows:
[0087] The root mean square error is 0.023 μm / s (the background vibration level is approximately 0.2 μm / s).
[0088] Peak error: The measured peak value of the maximum vibration event was 0.82 μm / s, and the predicted peak value was 0.77 μm / s, with an error of 6.1%.
[0089] The waveform correlation coefficient is 0.91, indicating that the waveforms are highly similar;
[0090] 3) Verification conclusion:
[0091] If the prediction error is within the preset threshold (1 / 3 octave band error < 15%, correlation function coefficient > 0.8), the vibration prediction model is validated and can be deployed in a real-time prediction system.
[0092] S6. Deployment of Real-time Prediction System: After verification, the vibration transfer function will be embedded in the vibration prediction module. By monitoring the vibration of external measuring points in real time, the real-time predicted vibration value of the internal sensitive area will be automatically output.
[0093] Assuming, following the example in step S5, after successful verification, internal measurement point C is removed; the system enters real-time operation mode:
[0094] 1) Measuring points A and B are permanent monitoring points, with sensors buried and fixed. Data is transmitted to the central server in real time via a wired network. A vibration prediction module is deployed in the server to solidify the verified vibration transfer function.
[0095] 2) The server receives real-time data streams from measurement point A and measurement point B every second; every 20 seconds is a calculation cycle, and the spectrum of measurement point A and measurement point B within that period is calculated; the internal prediction value of each vibration source contribution is calculated separately, and the total prediction spectrum is obtained by superimposing them; and then converted into a 1 / 3 octave band spectrum.
[0096] 3) If the predicted value of any frequency band exceeds 80% of the limit, the system will issue a yellow warning; if it exceeds 100% of the limit, a red alarm will be issued, prompting the production department to suspend sensitive processes or check vibration isolation measures.
[0097] S7. Dynamic Model Correction: Periodically or when the environment changes, the vibration transfer function is compared, calibrated and dynamically updated using measured data from temporarily arranged internal measuring points.
[0098] Assuming we continue with the example from step S6,
[0099] 1) Periodically revise as follows:
[0100] Once per quarter, using the sensor interface reserved on the lithography machine base, temporary measurement point C is set up to conduct a 24-hour synchronous data acquisition; the latest data is compared with the predicted value of the same period to calculate the error; if the error exceeds the threshold (such as one-third octave band error > 15%), the vibration transfer function is recalculated using the new data to replace the original vibration prediction model;
[0101] 2) Environmental changes are corrected as follows:
[0102] When significant changes occur in the surrounding environment (such as the addition of a branch line or adjustment of the operating speed of a subway line; the commencement of large-scale construction near the factory area; structural renovation or addition of equipment inside the factory);
[0103] Within 7 days of the aforementioned event, a temporary verification test shall be initiated, and the vibration transfer function shall be updated if necessary.
[0104] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.
Claims
1. A micro-vibration prediction system based on actual measurements, characterized in that, include: The measuring point layout module is used to deploy external vibration acquisition points along the propagation path of vibration sources outside the building, as well as internal vibration acquisition points in sensitive areas inside the building where vibration control is required. The data acquisition module is used to simultaneously acquire vibration response data from the external environment and vibration response data from internal sensitive points; The transfer function calculation module is used to calculate the vibration transmission law from the external vibration source to the internal sensitive point through the geological layer and building structure based on the collected vibration response data, and generate the vibration transfer function. The vibration prediction module is used to convert real-time monitored external environmental vibration response data into predicted vibration values for sensitive areas inside the building using the vibration transfer function. The verification and correction module is used to compare the measured values of internal sensitive points with the predicted vibration values generated by the vibration prediction module, and to dynamically correct the vibration transfer function based on the comparison results.
2. The micro-vibration prediction system based on actual measurements according to claim 1, characterized in that, In the measuring point layout module, External vibration acquisition points are located outside the building foundation, on the ground surface or in the soil layer between the main vibration source and the building, and are used to capture vibration source characteristics, the filtering effect of the site soil, and the input characteristics of road traffic vibration. Internal vibration acquisition points are located on the mounting bases of sensitive equipment or on the structural bearing points inside the building, and are used to collect real vibration response data of sensitive areas inside the building.
3. The micro-vibration prediction system based on actual measurements according to claim 1, characterized in that, In the transfer function calculation module, The laws governing vibration transmission include: the coefficient of variation of vibration amplitude, the attenuation characteristics of frequency components, and phase changes; The vibration transfer function is either the frequency response function established in the frequency domain through transmissibility analysis, or the impulse response function established in the time domain through a system identification algorithm.
4. The micro-vibration prediction system based on actual measurements according to claim 1, characterized in that, The verification and correction module is also used for: Periodically or when the external environment changes, temporary measurements are taken at the internal vibration acquisition points. The measured data are compared with the predicted data of the same period for error analysis, and the vibration transfer function is iteratively updated using Kalman filtering or least squares method.
5. A micro-vibration prediction method based on actual measurements, applied to the system described in any one of claims 1 to 4, characterized in that, Includes the following steps: S1. Vibration control index determination and vibration source identification: Based on the building's process requirements, determine the vibration control standards for sensitive areas inside the building, and identify the main vibration source types and locations in the surrounding area; S2. Optimization of measuring point layout: Based on the results identified in step S1, external vibration acquisition points are set up along the propagation path between the external vibration source and the building, and internal vibration acquisition points are set up in the sensitive areas inside the building. S3. On-site measurement and synchronous data acquisition: Simultaneously use vibration response data from external and internal vibration acquisition points over a continuous time period; S4. Signal Processing and Transmission Law Analysis: Process the collected vibration response data, calculate the vibration transfer function from the external measuring point to the internal measuring point, and construct a vibration prediction model; S5. Vibration prediction model verification: Using the vibration transfer function obtained in step S4, external measured data is used as input to predict the internal response, and the prediction results are compared with the measured results of internal measuring points to verify the accuracy of the vibration prediction model. S6. Deployment of Real-time Prediction System: After verification, the vibration transfer function will be embedded in the vibration prediction module. By monitoring the vibration of external measuring points in real time, the real-time predicted vibration value of the internal sensitive area will be automatically output. S7. Dynamic Model Correction: Periodically or when the environment changes, the vibration transfer function is compared, calibrated, and dynamically updated using measured data from temporarily arranged internal measuring points.
6. The micro-vibration prediction method based on actual measurements according to claim 5, characterized in that, In step S4, the specific steps of signal processing and transmission law analysis are as follows: S41. Preprocess the synchronously acquired external environmental vibration response data and internal sensitive point vibration response data; S42. Perform Fast Fourier Transform on the external environmental vibration response data and the internal sensitive point vibration response data to obtain their respective spectra, and then use... The method estimates the individual power spectra of each component, as well as the cross-power spectra between them; S43. Calculate the vibration transfer function based on the self-power spectrum and the cross-power spectrum.
7. The micro-vibration prediction method based on actual measurements according to claim 5, characterized in that, In step S5, the specific steps for verifying the vibration prediction model are as follows: S51. Among the collected external environmental vibration response data and internal sensitive point vibration response data, a set of synchronous measured data that did not participate in the processing of step S4 is reserved as a verification set. S52. Using the external environmental vibration response data in the validation set as input, predict the corresponding internal response through the vibration transfer function; S53. Compare the predicted internal response with the measured internal response, and calculate the error evaluation index; S54. Compare the calculated error evaluation index with the preset verification threshold. If the error is within the threshold range, the prediction model is deemed to have passed the verification. Otherwise, return to step S2 or S4 to re-optimize the measurement point layout or vibration transfer function parameters.