An electronic factory micro-vibration online monitoring and early warning system

Abstract

The present application is an electronic factory micro-vibration online monitoring and early warning system, which can monitor the clean room structure with micro-vibration requirements in real time and give early warning when the vibration exceeds the standard. It includes a signal acquisition module for collecting vibration signals, a monitoring host module, a cloud server module and a client module. The cloud server module processes and analyzes the collected vibration signals based on micro-vibration evaluation indexes, outputs 1 / 3 octave speed spectrum effective value graphics, 1 / 3 octave speed spectrum frequency division maximum time graphics and 1 / 3 octave speed spectrum peak frequency time graphics, and judges whether the working environment vibration exceeds the standard according to the set precision equipment micro-vibration threshold. If the on-site vibration monitoring value exceeds the set threshold, it will be pushed to the bound client module for early warning. It can monitor the clean room structure with micro-vibration requirements in real time and give early warning when the vibration exceeds the standard, and can be applied in industries with micro-vibration requirements such as semiconductors, panel displays and high-precision laboratories.

Description

TECHNICAL FIELD

[0001] The present application relates to an electronic factory micro-vibration online monitoring and early warning system, and belongs to the technical field of building micro-vibration monitoring. BACKGROUND

[0002] In a clean room such as a semiconductor or display panel, a large number of precision process equipment are generally provided, and these equipment have very high micro-vibration requirements. If the environmental vibration exceeds the micro-vibration limit of the equipment, the normal work of the equipment will be affected, and then the production yield of the equipment will be affected. Therefore, the micro-vibration level of high-tech industrial buildings must be kept within the limit at all times.

[0003] The conventional method of the prior art is to test the micro-vibration of the clean room before the equipment is moved in and during the trial operation stage, so as to confirm that the micro-vibration of the factory building meets the requirements. However, the fact that the micro-vibration meets the requirements during the trial operation stage does not mean that the micro-vibration will always meet the requirements during the formal production process, because the vibration environment inside and outside the factory building may change. The prior art generally ignores the changes in micro-vibration during the production of the factory building, and does not pay attention to whether the micro-vibration continues to meet the requirements.

[0004] There is an engineering structure health monitoring system in the prior art, which is used as a vibration monitoring system. The monitoring indexes thereof are structure deformation, stress, strain, deflection, crack, vibration displacement time domain peak value, vibration velocity time domain peak value, vibration velocity time domain peak value, etc. There is no data processing module specially for the anti-micro-vibration requirements of precision equipment. The anti-micro-vibration evaluation index of the clean factory building is usually the 1 / 3 octave velocity spectrum effective value. The existing technology system generally cannot calculate and analyze the 1 / 3 octave velocity spectrum effective value, and does not have the display function of the 1 / 3 octave velocity spectrum frequency maximum time waveform and the 1 / 3 octave velocity spectrum peak frequency time waveform calculation method. Therefore, it cannot be used for micro-vibration online monitoring. SUMMARY

[0005] The present application proposes an electronic factory micro-vibration online monitoring and early warning system, which aims to overcome the above-mentioned deficiencies in the prior art, ensure the normal work of precision equipment, improve the product pass rate, and monitor the vibration environment of precision equipment in real time, so as to meet the real-time monitoring requirements of the clean room micro-vibration.

[0006] The technical solution of the present application: an electronic factory micro-vibration online monitoring and early warning system for real-time monitoring of clean room structures with micro-vibration requirements and timely warning when vibration exceeds the standard. It includes a signal acquisition module for collecting vibration signals, a monitoring host module, a cloud server module, and a client module. The cloud server module processes and analyzes the collected vibration signals based on micro-vibration evaluation indicators, outputs 1 / 3 octave speed spectrum effective value graphics, 1 / 3 octave speed spectrum frequency division maximum time graphics, and 1 / 3 octave speed spectrum peak frequency time graphics, and determines whether the working environment vibration exceeds the standard according to the set precision equipment micro-vibration threshold. If the on-site vibration monitoring value exceeds the set threshold, it is pushed to the bound client module for early warning.

[0007] Preferably, the signal acquisition module includes a signal, an electrically connected vibration sensor, and a vibration signal acquisition instrument. The vibration sensor collects vibration signals and transmits them to the vibration signal acquisition instrument for preliminary storage of digital signals. The vibration signal acquisition instrument converts analog signals into digital signals, and the converted digital signals are transmitted to the monitoring host module. The vibration signal acquisition instrument has a trigger control mode.

[0008] The vibration sensor is a three-axis vibration sensor orthogonal three-direction acquisition system for collecting vibration signals in three directions.

[0009] Preferably, the filter uses a band-pass filter, and the micro-vibration evaluation vibration frequency band is 1~100Hz. According to the 1 / 3 octave algorithm, the lower limit frequency of the 1Hz frequency band is 1 / (2 1 / 6 )=0.89Hz; the upper limit frequency of the 100Hz frequency band is 100×(2 1 / 6 )=112Hz, and the upper and lower limit frequencies of the band-pass filter are set to 0.8Hz~120Hz by default.

[0010] The baseline correction uses a polynomial least squares method to eliminate the trend term.

[0011] The monitoring host module is also used for device power management, network outage emergency storage, power outage emergency storage, incoming call automatic start, edge computing, remote control protocol setting, and device parameter setting.

[0012] Preferably, the cloud server module is also used for signal data storage, device parameter remote control, micro-vibration threshold setting, signal time domain analysis, signal time frequency analysis, real-time comparison of 1 / 3 octave speed spectrum frequency division maximum value with the set micro-vibration threshold, and transmission to the client module for trigger early warning if there is an over-standard.

[0013] Preferably, the 1 / 3 octave speed spectrum analysis is 1 / 3 octave speed spectrum effective value calculation, including the following steps:

[0014] a. Cut the data in the analysis signal segment into several data blocks;

[0015] b. Window processing is performed on each data block, and the default window function is hanning window;

[0016] c. FFT is performed on each data block;

[0017] c. The effective values corresponding to the center frequencies of the 1 / 3 octave frequency spectrum of the data block are calculated in order from small to large;

[0018] e. The effective value calculation of the 1 / 3 octave frequency spectrum of all data blocks is completed;

[0019] f. The effective value of the 1 / 3 octave frequency spectrum of the entire analysis period is calculated.

[0020] Preferably, the step f is linear average, peak envelope or 1 standard deviation + mean calculation method,

[0021] Assuming that the speed effective values of the first data block at the center frequencies fc1~fcN are v1(fc1), v1(fc2), …, v1(fcN); the speed effective values of the second data block at the center frequencies fc1~fcN are v2(fc1), v2(fc2), …, v2(fcN); and the speed effective values of the Mth data block at the center frequencies fc1~fcN are vM(fc1), vM(fc2), …, vM(fcN);

[0022] When the linear average calculation method is used, the results at the center frequencies fc1~fcN of the entire signal analysis period are:

[0023] v(fc1) = mean(v1(fc1), v2(fc1), …, vM(fc1));

[0024] v(fc2) = mean(v1(fc2), v2(fc2), …, vM(fc2));

[0025] …

[0026] v(fcN) = mean(v1(fcN), v2(fcN), …, vM(fcN));

[0027] Wherein mean() is the mean function.

[0028] The peak envelope calculation mode is assumed that the speed effective value of the first data block at the center frequency fc1~fcN is v1(fc1), v1(fc2), …, v1(fcN); the speed effective value of the second data block at the center frequency fc1~fcN is v2(fc1), v2(fc2), …, v2(fcN); and the speed effective value of the Mth data block at the center frequency fc1~fcN is vM(fc1), vM(fc2), …, vM(fcN);

[0029] The results of the entire signal analysis period at the center frequency fc1~fcN are:

[0030] v(fc1) = max(v1(fc1), v2f(fc1), …, vM(fc1));

[0031] v(fc2) = max(v1(fc2), v2f(fc2), …, vM(fc2));

[0032] …

[0033] v(fcN) = max(v1(fcN), v2f(fcN), …, vM(fcN));

[0034] Wherein max() is a maximum value function.

[0035] The 1 standard deviation + mean calculation mode is assumed that the speed effective value of the first data block at the center frequency fc1~fcN is v1(fc1), v1(fc2), …, v1(fcN); the speed effective value of the second data block at the center frequency fc1~fcN is v2(fc1), v2(fc2), …, v2(fcN); and the speed effective value of the Mth data block at the center frequency fc1~fcN is vM(fc1), vM(fc2), …, vM(fcN);

[0036] The results of the entire signal analysis period at the center frequency fc1~fcN are:

[0037] v(fc1) = mean(v1(fc1), v2f(fc1), …, vM(fc1))+std(v1(fc1), v2f(fc1), …, vM(fc1));

[0038] v(fc2) = mean(v1(fc2), v2f(fc2), …, vM(fc2))+std(v1(fc2), v2f(fc2), …, vM(fc2));

[0039] …

[0040] v(fcN) = mean (v1(fcN), v2f(fcN),…, vM (fcN))+std(v1(fcN), v2f(fcN),…, vM (fcN));

[0041] Among them, mean() is the mean function; std() is the standard deviation function.

[0042] Preferably, the 1 / 3 octave velocity spectrum frequency division maximum value analysis and 1 / 3 octave velocity spectrum peak frequency analysis include the following steps:

[0043] a. Assume that the time interval between consecutive signal analysis segments in the time domain is Δt, and the signal analysis segment duration is tLength. Δt and tLength are set to be constants. Assume that the start time of the first signal analysis segment is t01 and the end time is t11; the start time of the second signal analysis segment is t02 and the end time is t12; and so on. The start time of the Nth signal analysis segment is t0N and the end time is t1N. The total signal segment duration tLength = t11 - t01 = t12 - t02 = t1N - t0N.

[0044] Let tLength>△t, then each signal segment must contain a portion of the newly acquired signal△t and a portion of the old signal that has been analyzed last time, that is, the overlapping signal, assuming tTemp, that is, tLength=△t+tTemp; then the condition tLength>△t means that the signal analysis can be fully covered during the entire acquisition time;

[0045] b. Calculate the 1 / 3 octave velocity spectrum of the vibration signal with a starting time of t01 and an end time of t11, and analyze the signal for a total duration of tLength. Then calculate the maximum frequency division value v(f)_t11 = max{v(fc1), v(fc1), …v(fcN)} and save the velocity spectrum peak frequency fc_t11. The microvibration evaluation value at time t11 is considered to be v(f)_t11, and the peak frequency is fc_t11.

[0046] c. Calculate the micro-vibration level at the next adjacent moment:

[0047] The start time of the next analysis signal segment is t02 = t01 + △t, and the end time is t12 = t11 + △t. The maximum value of the 1 / 3 octave velocity frequency division of the analysis signal frequency band is v(f)_t12 = max{v(fc1), v(fc2), … v(fcN)}, and the velocity spectrum peak frequency fc_t12 is saved. The micro-vibration evaluation value at time t12 is considered to be v(f)_t12, and the peak frequency is fc_t12.

[0048] d. Calculate the maximum value of the vibration velocity spectrum frequency division and the corresponding main frequency at each moment in turn, and save them in an array file RST with N rows and 3 columns. The first column of the RST file stores the time value, the second column stores the maximum value of the 1 / 3 octave velocity spectrum frequency division, and the third column stores the peak frequency of the 1 / 3 octave velocity spectrum, that is:

[0049] RST[1,:]=[t11,t12,t13,…,t1N];

[0050] RST[2,:]=[v(f)_t11, v(f)_t12,…, v(f)_t1N];

[0051] RST[3,:]=[fc_t11, fc_t12,…,fc_t1N];

[0052] e. Generate a 1 / 3 octave velocity spectrum RMS value graph, a 1 / 3 octave velocity spectrum RMS value frequency division maximum value time history waveform, and a 1 / 3 octave velocity spectrum RMS value frequency division maximum value corresponding to the center frequency time history waveform.

[0053] The advantages of the present invention are: reasonable design, real-time monitoring of clean room structures with micro-vibration requirements, timely warning when vibration exceeds the standard, and application in industries such as semiconductors, panel displays, and high-precision laboratories with micro-vibration prevention requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 This is a structural block diagram of the electronic factory micro-vibration online monitoring and early warning system of the present invention.

[0055] Figure 2 It is a discrete illustration of vibration signal segments. DETAILED DESCRIPTION

[0056] The present invention will be further described in detail below with reference to examples and specific implementation methods. Example

[0057] like Figure 1 As shown, an online monitoring and early warning system for micro-vibration in an electronic factory building includes a signal acquisition module, a monitoring host module, a cloud server module and a client module.

[0058] (1) Signal acquisition module:

[0059] A three-axis acquisition system is composed of three single-axis vibration sensors in orthogonal arrangement to collect vibration signals in three directions. The vibration sensors transmit the collected vibration signals to a data collection instrument (vibration signal collection instrument) through data lines for preliminary storage of digital signals. The data collection instrument supplies power to the sensors and can convert analog signals to digital signals (signal D / A conversion). The converted digital signals are transmitted to a monitoring host module through data lines or wireless networks. The data collection instrument has a trigger control mode.

[0060] The micro-vibration evaluation frequency range is usually 1-100 Hz. According to the signal sampling theorem, the signal sampling frequency is at least 200 Hz. Considering the aliasing phenomenon of the filter, the sampling rate is set to 256 Hz by default.

[0061] (2) Monitoring host module:

[0062] The basic functions of the monitoring host module include digital signal preprocessing, device power management, network outage emergency storage, power outage emergency storage, incoming call automatic start, edge computing, remote control protocol setting, and device parameter setting.

[0063] Digital signal preprocessing refers to filtering and baseline correction of vibration digital signals.

[0064] Filtering: A band-pass filter is used. The micro-vibration evaluation vibration frequency band is 1-100 Hz. According to the 1 / 3 octave algorithm, the lower limit frequency of the 1 Hz frequency band at the center frequency is 1 / (2 1 / 6 )=0.89 Hz; the upper limit frequency of the 100 Hz frequency band at the center frequency is 100×(2 1 / 6 )=112 Hz. Therefore, the upper and lower limit frequencies of the band-pass filter are set to 0.8 Hz-120 Hz by default.

[0065] Baseline correction: The vibration signal data collected in vibration testing often deviates from the baseline. The entire process of deviation from the baseline over time is called the trend item of the signal. The trend item directly affects the correctness of the signal and should be removed. The polynomial least squares method is used to eliminate the trend item.

[0066] (3) Cloud server module:

[0067] The monitoring host module transmits the preprocessed digital signals to the cloud server module through optical fiber, 4G or 5G wireless signals. The functions of the cloud server module include signal data storage, remote control of device parameters, micro-vibration threshold setting, signal time domain analysis, signal time-frequency analysis, 1 / 3 octave speed spectrum analysis, 1 / 3 octave speed spectrum frequency division maximum value analysis, and 1 / 3 octave speed spectrum peak frequency analysis.

[0068] 1 / 3 octave velocity spectrum analysis, 1 / 3 octave velocity spectrum sub-frequency maximum analysis, 1 / 3 octave velocity spectrum peak frequency analysis algorithm and implementation are as follows:

[0069] 1) 1 / 3 octave velocity spectrum effective value calculation, including the following steps:

[0070] a. The data in the analysis signal segment is divided into several data blocks;

[0071] b. Each data block is processed by windowing, and the default window function in this embodiment is hanning window;

[0072] c. Perform FFT on each data block;

[0073] c. Calculate the effective value corresponding to each center frequency of the 1 / 3 octave velocity spectrum of the data block from small to large;

[0074] e. Complete the 1 / 3 octave velocity spectrum effective value calculation of all data blocks;

[0075] f. Calculate the 1 / 3 octave velocity spectrum effective value of the entire analysis period.

[0076] In this embodiment, there are three ways to calculate the entire signal analysis period result from the data block result: linear average, peak envelope, and 1 standard deviation + mean.

[0077] Assume that the velocity effective values of the first data block at center frequencies fc1~fcN are v1(fc1), v1(fc2), …, v1(fcN); the velocity effective values of the second data block at center frequencies fc1~fcN are v2(fc1), v2(fc2), …, v2(fcN); and the velocity effective values of the Mth data block at center frequencies fc1~fcN are vM(fc1), vM(fc2), …, vM(fcN).

[0078] When linear average is used, the results at center frequencies fc1~fcN for the entire signal analysis period are:

[0079] v(fc1) = mean(v1(fc1), v2f(fc1), …, vM(fc1));

[0080] v(fc2) = mean(v1(fc2), v2f(fc2), …, vM(fc2));

[0081] …

[0082] v(fcN) = mean(v1(fcN), v2f(fcN), …, vM(fcN)).

[0083] Using peak envelope, the results of the entire signal analysis period at the center frequency fc1~fcN:

[0084] v(fc1) = max(v1(fc1), v2f(fc1), …, vM (fc1));

[0085] v(fc2) = max(v1(fc2), v2f(fc2), …, vM (fc2));

[0086] …

[0087] v(fcN) = max(v1(fcN), v2f(fcN), …, vM (fcN)).

[0088] Using mean + 1 times standard deviation, the results of the entire signal analysis period at the center frequency fc1~fcN:

[0089] v(fc1) = mean(v1(fc1), v2f(fc1), …, vM (fc1))+std(v1(fc1), v2f(fc1), …, vM (fc1));

[0090] v(fc2) = mean(v1(fc2), v2f(fc2), …, vM (fc2))+std(v1(fc2), v2f(fc2), …, vM (fc2));

[0091] …

[0092] v(fcN) = mean(v1(fcN), v2f(fcN), …, vM (fcN))+std(v1(fcN), v2f(fcN), …, vM (fcN)).

[0093] Note: mean() is the mean function; max() is the maximum value function; std() is the standard deviation function.

[0094] 2) 1 / 3 octave velocity spectrum frequency division maximum value, 1 / 3 octave velocity spectrum peak frequency calculation and analysis:

[0095] a. Assume that the time interval between consecutive signal analysis segments in the time domain is Δt, and the signal analysis segment duration is tLength. Once Δt and tLength are set, they remain constant. Assume that the first signal analysis segment starts at t01 and ends at t11; the second signal analysis segment starts at t02 and ends at t12; and so on, the Nth signal analysis segment starts at t0N and ends at t1N. The total signal segment duration, tLength = t11 - t01 = t12 - t02 = t1N - t0N.

[0096] If tLength > △t, each signal segment must contain a portion of the newly acquired signal (△t) and a portion of the previously analyzed signal (overlapping signal, assumed to be tTemp, meaning tLength = △t + tTemp). Therefore, the condition tLength > △t means that full coverage analysis of the signal within the entire acquisition time is possible.

[0097] b. Figure 2 As shown, the calculation starts at t01 and ends at t11, with the total analyzed signal duration being tLength. The calculation process for the 1 / 3-octave velocity spectrum of the vibration signal is described in "1) Calculation of the Effective Value of the 1 / 3-octave Velocity Spectrum." The maximum frequency division value v(f)_t11 = max{v(fc1), v(fc1), …v(fcN)} is then calculated, and the velocity spectrum peak frequency fc_t11 is saved. The microvibration evaluation value at time t11 is then considered to be v(f)_t11, with a peak frequency of fc_t11.

[0098] c. Calculate the micro-vibration level at the next adjacent moment:

[0099] The next analysis signal segment starts at t02 = t01 + Δt and ends at t12 = t11 + Δt. The maximum value of the 1 / 3 octave velocity frequency division of the analysis signal frequency segment is v(f)_t12 = max{v(fc1), v(fc2), …v(fcN)}. The velocity spectrum peak frequency fc_t12 is also saved. The microvibration evaluation value at time t12 is considered to be v(f)_t12, and the peak frequency is fc_t12.

[0100] d. The maximum value of the vibration velocity spectrum frequency division and the corresponding main frequency at each moment can be calculated in turn and saved in an array file RST with N rows and 3 columns. The first column of the RST file saves the moment value, the second column saves the maximum value of the 1 / 3 octave velocity spectrum frequency division, and the third column saves the 1 / 3 octave velocity spectrum peak frequency, that is:

[0101] RST[1,:]=[t11,t12,t13,…,t1N];

[0102] RST[2,:]=[v(f)_t11, v(f)_t12,…, v(f)_t1N];

[0103] RST[3,:]=[fc_t11, fc_t12,…,fc_t1N];

[0104] e. Generating 1 / 3 octave velocity spectrum effective value graph, 1 / 3 octave velocity spectrum effective value frequency division maximum time waveform, 1 / 3 octave velocity spectrum effective value frequency division maximum corresponding center frequency time waveform.

[0105] The cloud server module will also compare the 1 / 3 octave velocity spectrum effective value frequency division maximum with the set micro-vibration threshold in real time, and if it is over the threshold, it will trigger an early warning.

[0106] (4) Client module:

[0107] The client module accesses the cloud server module through a website, and the calculation results of the cloud server are presented to the client module in the form of a large screen. The client module can display the following contents in real time: 1. Signal time waveform, signal time-frequency cloud chart, 1 / 3 octave velocity spectrum graph, 1 / 3 octave velocity spectrum frequency division maximum time waveform, 1 / 3 octave velocity spectrum peak frequency time waveform, and 1 / 3 octave velocity spectrum frequency division maximum daily statistical table.

[0108] When the cloud service module data analysis finds that the current vibration exceeds the micro-vibration threshold of the precision equipment, it will send the over-standard information to the mobile client through SMS or WeChat for early warning.

[0109] The above only describes the preferred embodiments of the present application, and it should be noted that for those skilled in the art, without departing from the inventive concept, several modifications and improvements can be made, which are all within the scope of protection of the present application.

Claims

1. An online monitoring and early warning system for micro-vibration in electronic workshops, characterized in that: It includes a signal acquisition module, a monitoring host module, a cloud server module and a client module which are connected in sequence. The signal acquisition module is used to collect vibration signals, convert them into vibration digital signals and transmit them to the monitoring host module. The monitoring host module is used to pre-process the vibration digital signal transmitted by the signal acquisition module, that is, to filter and correct the baseline, and transmit it to the cloud server module after pre-processing. The cloud server module is used to perform 1 / 3 octave velocity spectrum analysis, 1 / 3 octave velocity spectrum frequency division maximum value analysis, and 1 / 3 octave velocity spectrum peak frequency analysis on the pre-processed vibration digital signal transmitted from the monitoring host module, and compare the 1 / 3 octave velocity spectrum frequency division maximum value with the set micro-vibration threshold in real time. If any exceeds the standard, it is transmitted to the client module to trigger an early warning; The client module is used to access the cloud server module through the website and display the calculation results of the cloud server module. The displayed content includes: signal time-history waveform, signal time-frequency cloud map, 1 / 3 octave velocity spectrum graph, 1 / 3 octave velocity spectrum frequency division maximum value time-history waveform, 1 / 3 octave velocity spectrum peak frequency time-history waveform, and 1 / 3 octave velocity spectrum frequency division maximum value daily statistics table; The 1 / 3 octave velocity spectrum analysis is a 1 / 3 octave velocity spectrum effective value calculation, comprising the following steps: a. Divide the data in the analysis signal segment into several data blocks; b. Perform windowing on each data block, with the default window function being the Hanning window; c. Perform FFT on each data block; c. Calculate the effective value corresponding to each center frequency of the 1 / 3 octave velocity spectrum of the data block in order from small to large; e. Complete the calculation of the effective value of the 1 / 3 octave velocity spectrum of all data blocks; f. Calculate the effective value of the 1 / 3 octave velocity spectrum for the entire analysis period; The step f is a linear average, peak envelope or 1 times standard deviation + mean calculation method.

2. The electronic factory micro-vibration online monitoring and early warning system according to claim 1, characterized in that: The signal acquisition module includes a signal, an electrically connected vibration sensor and a vibration signal collector. The vibration sensor transmits the collected vibration signal to the vibration signal collector for preliminary storage of the digital signal. The vibration signal collector converts the analog signal into a digital signal, and the converted digital signal is transmitted to the monitoring host module. The vibration signal collector has a trigger control mode; The vibration sensor is a three-direction acquisition system composed of three orthogonal single-axis vibration sensors, and is used for collecting vibration signals in three directions.

3. The electronic factory micro-vibration online monitoring and early warning system according to claim 1, characterized in that: The filtering uses a bandpass filter, the vibration frequency band of the micro-vibration evaluation is 1~100Hz, and according to the 1 / 3 octave algorithm, the calculation lower limit frequency of the center frequency 1Hz band is 1 / (2 1 / 6 )=0.89Hz; the upper limit frequency of the calculated frequency band with a center frequency of 100Hz is 100×(2 1 / 6 ) = 112Hz, the upper and lower limit frequencies of the bandpass filter are set to 0.8Hz~120Hz by default; The baseline correction uses a polynomial least squares method to eliminate trend terms; The monitoring host module is also used for device power management, network disconnection emergency storage, power outage emergency storage, automatic start-up during incoming calls, edge computing, remote control protocol settings, and device parameter settings.

4. The electronic factory micro-vibration online monitoring and early warning system according to claim 1, characterized in that: The cloud server module is also used for signal data storage, remote control of equipment parameters, micro-vibration threshold setting, signal time domain analysis, and signal time-frequency analysis.

5. The electronic factory micro-vibration online monitoring and early warning system according to claim 4, characterized in that: The step f is a linear average calculation method, comprising: Assume that the effective speed values ​​of the first data block at the center frequencies fc1~fcN are v1(fc1), v1(fc2), ..., v1(fcN); the effective speed values ​​of the second data block at the center frequencies fc1~fcN are v2(fc1), v2(fc2), ..., v2(fcN); the effective speed values ​​of the Mth data block at the center frequencies fc1~fcN are vM(fc1), vM(fc2), ..., vM(fcN; The results of the entire signal analysis period at center frequencies fc1~fcN are: v(fc1) =mean(v1(fc1), v2f(fc1),…, vM (fc1)); v(fc2) =mean(v1(fc2), v2f(fc2),…, vM (fc2)); … v(fcN) =mean(v1(fcN), v2f(fcN),…, vM (fcN)); Where mean() is the mean function.

6. The electronic factory micro-vibration online monitoring and early warning system according to claim 4, characterized in that: The step f is a peak envelope calculation method, comprising: Assume that the effective speed values ​​of the first data block at the center frequencies fc1~fcN are v1(fc1), v1(fc2), ..., v1(fcN); the effective speed values ​​of the second data block at the center frequencies fc1~fcN are v2(fc1), v2(fc2), ..., v2(fcN); the effective speed values ​​of the Mth data block at the center frequencies fc1~fcN are vM(fc1), vM(fc2), ..., vM(fcN; The results of the entire signal analysis period at center frequencies fc1~fcN are: v(fc1) =max(v1(fc1), v2f(fc1),…, vM (fc1)); v(fc2) = max (v1(fc2), v2f(fc2),…, vM (fc2)); … v(fcN) = max (v1(fcN), v2f(fcN),…, vM (fcN)); Where max() is the maximum value function.

7. The electronic factory micro-vibration online monitoring and early warning system according to claim 4, characterized in that: The step f is a 1 times standard deviation + mean calculation method, including: Assume that the effective speed values ​​of the first data block at the center frequencies fc1~fcN are v1(fc1), v1(fc2), ..., v1(fcN); the effective speed values ​​of the second data block at the center frequencies fc1~fcN are v2(fc1), v2(fc2), ..., v2(fcN); the effective speed values ​​of the Mth data block at the center frequencies fc1~fcN are vM(fc1), vM(fc2), ..., vM(fcN; The results of the entire signal analysis period at center frequencies fc1~fcN are: v(fc1) = mean (v1(fc1), v2f(fc1),…, vM (fc1))+std(v1(fc1), v2f(fc1),…, vM(fc1)); v(fc2) = mean (v1(fc2), v2f(fc2),…, vM (fc2))+std(v1(fc2), v2f(fc2),…, vM(fc2)); … v(fcN) = mean (v1(fcN), v2f(fcN),…, vM (fcN))+std(v1(fcN), v2f(fcN),…, vM(fcN)); Among them, mean() is the mean function; std() is the standard deviation function.

8. An electronic factory micro-vibration online monitoring and early warning system according to any one of claims 5 to 7, characterized in that: The 1 / 3 octave velocity spectrum frequency division maximum value analysis and 1 / 3 octave velocity spectrum peak frequency analysis include the following steps: a. Assume that the time interval between consecutive signal analysis segments in the time domain is Δt, and the signal analysis segment duration is tLength. Δt and tLength are set to be constants. Assume that the start time of the first signal analysis segment is t01 and the end time is t11; the start time of the second signal analysis segment is t02 and the end time is t12; and so on. The start time of the Nth signal analysis segment is t0N and the end time is t1N. The total signal segment duration tLength = t11 - t01 = t12 - t02 = t1N - t0N. Let tLength>△t, then each signal segment must contain a portion of the newly acquired signal△t and a portion of the old signal that has been analyzed last time, that is, the overlapping signal, assuming tTemp, that is, tLength=△t+tTemp; then the condition tLength>△t means that the signal analysis can be fully covered during the entire acquisition time; b. Calculate the 1 / 3 octave velocity spectrum of the vibration signal with a starting time of t01 and an end time of t11, and analyze the signal for a total duration of tLength. Then calculate the maximum frequency division value v(f)_t11 = max{v(fc1), v(fc1), …v(fcN)} and save the velocity spectrum peak frequency fc_t11. The microvibration evaluation value at time t11 is considered to be v(f)_t11, and the peak frequency is fc_t11. c. Calculate the micro-vibration level at the next adjacent moment: The start time of the next analysis signal segment is t02 = t01 + △t, and the end time is t12 = t11 + △t. The maximum value of the 1 / 3 octave velocity frequency division of the analysis signal frequency band is v(f)_t12 = max{v(fc1), v(fc2), … v(fcN)}, and the velocity spectrum peak frequency fc_t12 is saved. The micro-vibration evaluation value at time t12 is considered to be v(f)_t12, and the peak frequency is fc_t12. d. Calculate the maximum value of the vibration velocity spectrum frequency division and the corresponding main frequency at each moment in turn, and save them in an array file RST with N rows and 3 columns. The first column of the RST file stores the time value, the second column stores the maximum value of the 1 / 3 octave velocity spectrum frequency division, and the third column stores the peak frequency of the 1 / 3 octave velocity spectrum, that is: RST[1,:]=[t11,t12,t13,…,t1N]; RST[2,:]=[v(f)_t11, v(f)_t12,…, v(f)_t1N]; RST[3,:]=[fc_t11, fc_t12,…,fc_t1N]; e. Generate a 1 / 3 octave velocity spectrum RMS value graph, a 1 / 3 octave velocity spectrum RMS value frequency division maximum value time history waveform, and a 1 / 3 octave velocity spectrum RMS value frequency division maximum value corresponding to the center frequency time history waveform.

Images