Mouse earthquake early warning and accurate positioning system and device based on voice broadcast

Abstract

The invention discloses a mouse earthquake early warning and accurate positioning system and device based on voice broadcast, and belongs to the technical field of earthquake early warning, and the system comprises an earthquake monitoring module, a background noise filtering module, a vibration data analysis module, an earthquake prediction module, an earthquake accurate positioning module, a voice synthesis module and a communication module. The data acquired by the earthquake monitoring module is preprocessed through the background noise filtering module, non-earthquake-related background noise generated in daily life is recognized and filtered out, the signal-to-noise ratio of earthquake signals is remarkably increased, the false alarm rate is reduced, and therefore the sensitivity and accuracy of the system are improved, and the reliability of the system is improved. A dynamic threshold setting method enables the system to adapt to different environmental conditions, a vibration data analysis module analyzes filtered vibration data in real time, potential seismic signals can be rapidly identified, multiple key features are extracted, a comprehensive index is defined to evaluate whether the current signal is an earthquake precursor or not, and the method has the advantages of being simple in structure and convenient to operate. And the accuracy of earthquake detection is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of earthquake early warning, and specifically refers to a mouse earthquake early warning and precise positioning system and device based on voice broadcast. Background Technique

[0002] With the development of technology, people's demand for earthquake early warning is increasing day by day. Traditional earthquake early warning systems usually rely on specialized earthquake detection equipment and send alerts to the public through means such as the Internet or radio. However, the popularization rate of such systems is limited by factors such as cost and installation complexity. At the same time, for individual users, there is a lack of direct and intuitive early warning methods. Some existing earthquake positioning technologies have limitations in accuracy, especially in quickly and accurately positioning at the initial stage of an earthquake;

[0003] However, the existing earthquake early warning and precise positioning systems still have certain defects. The existing earthquake early warning and precise positioning systems lack an effective background noise filtering mechanism, resulting in non-earthquake-related background noise interfering with the detection of earthquake signals. They lack real-time analysis of vibration data, resulting in a delay in the recognition of potential earthquake signals and an increase in the false alarm rate. They rely on historical earthquake data for prediction while ignoring the impact of real-time vibration data, resulting in low prediction accuracy. They only issue alerts in the form of text or images, lacking an intuitive and easy-to-understand alert method. Therefore, a mouse earthquake early warning and precise positioning system and device based on voice broadcast are proposed. Summary of the Invention

[0004] The purpose of the present invention is to provide a mouse earthquake early warning and precise positioning system and device based on voice broadcast to solve the problems raised in the above background technique.

[0005] To achieve the above purpose, the present invention provides the following technical solution: A mouse earthquake early warning and precise positioning system based on voice broadcast, including an earthquake monitoring module, a background noise filtering module, a vibration data analysis module, an earthquake prediction module, an earthquake precise positioning module, a voice synthesis module, and a communication module;

[0006] The earthquake monitoring module obtains earthquake data information in real time through the internal sensor device of the mouse and the earthquake monitoring network;

[0007] The background noise filtering module is used to preprocess the earthquake data obtained by the earthquake monitoring module and identify and filter non-earthquake-related background noise generated in daily life;

[0008] The vibration data analysis module is used to perform real-time analysis on the filtered vibration data to identify potential earthquake signals;

[0009] The earthquake prediction module is used to make predictions based on historical earthquake data and the analysis results of the vibration data analysis module;

[0010] The earthquake precise positioning module is used for precise positioning through positioning technology;

[0011] The speech synthesis module is used for generating an alarm message text according to the results of the earthquake prediction module and the earthquake precise positioning module, and converting the alarm message text into a voice signal;

[0012] The communication module is used for data transmission between each module and the mouse interaction module.

[0013] Among them, the earthquake monitoring module obtains earthquake data information in real time through the internal sensor device of the mouse and the earthquake monitoring network; continuously monitors ground vibrations through the triaxial accelerometer sensor inside the mouse, records the obtained ground vibration information, obtains the latest earthquake activity data through the earthquake monitoring network, and fuses the vibration data collected by the mouse with the data of the earthquake monitoring network. The fused earthquake data information is stored in the cloud.

[0014] Among them, the background noise filtering module preprocesses the data fused by the earthquake monitoring module, identifies and filters the non-earthquake-related background noise generated in daily life, smooths and normalizes the fused vibration data. Let the vibration data be x, and feature extraction is performed through multi-scale wavelet transform. By changing the scale parameter a, features are extracted in different frequency ranges. The implementation formula is:

[0015]

[0016] In the formula, W(a, b) represents the wavelet transform coefficient, a represents the scale parameter, b represents the displacement parameter, x(t) represents the time series signal of the input vibration data, ψ represents the wavelet function, and ψ * represents the complex conjugate of the wavelet function.

[0017] Among them, after extracting the vibration features according to the multi-scale wavelet transform coefficients, background noise model training is performed. The background noise model is defined as:

[0018]

[0019] In the formula, L represents the total loss, N represents the number of samples, y i represents the true label of the i-th sample, p i represents the probability that the i-th sample of the background noise model belongs to the background noise, λ represents the regularization coefficient, and R represents the regularization term; the extracted vibration features are input into the trained background noise model to obtain the probability p of each data point, and the threshold T = μ pd + kσ pd, if p > T, the data point is background noise; otherwise, it is a valid signal.

[0020] When the data point is background noise, noise filtering is performed. Let the original data be x[n], and the filtered data be x′[n]. The implementation formula is:

[0021]

[0022] In the formula, p[n] represents the probability that the nth data point of the background noise model is background noise, T represents the dynamically set threshold, W represents the window size, and the local average value is calculated. represents the sum of the data points within the window.

[0023] Among them, the vibration data analysis module receives the vibration data after background noise filtering and performs analysis, including energy, main frequency, signal mutation, and cumulative absolute difference.

[0024] Energy: The energy of the vibration signal reflects the vibration intensity. The implementation formula is:

[0025]

[0026] In the formula, N represents the calculation window size, n represents the current time, x′[n] represents the value of the vibration data after noise filtering at time point n, and the formula represents the sum of the squares of all x′[n] from n - N + 1 to n, obtaining the signal energy within a time period.

[0027] Main frequency: Estimate the main frequency through short-time Fourier transform, and select the frequency corresponding to the maximum amplitude in the amplitude spectrum as the main frequency. Let the window length be M, the overlapping part be L, and the signal of the mth window be x′ m [k] = x′[k + mL], calculate the Fourier transform of each window to obtain the frequency spectrum X m (f), and the implementation formula is:

[0028]

[0029] In the formula, x′ m [k] represents the signal of the mth window, w[k] represents the window function, f represents the frequency, M represents the window length, j represents the imaginary unit, and find the frequency with the largest amplitude according to the frequency spectrum. The implementation formula is:

[0030] f dom [n] = arg max f |X m (f)|,

[0031] In the formula, |X m (f)| represents the amplitude of the frequency spectrum, arg max fDenote finding the frequency f that maximizes |X m (f)|;

[0032] Signal mutation rate: The speed of signal change, and the implementation formula is:

[0033]

[0034] In the formula, x′[n] represents the vibration data at the current time point n, x′[n - 1] represents the vibration data at the previous time point n - 1, and Δt represents the sampling interval;

[0035] Cumulative absolute difference: Measuring the degree of signal fluctuation, and the implementation formula is:

[0036]

[0037] In the formula, the formula represents the sum of the absolute differences between all adjacent time points x′[n] and x′[n - 1] from n - N + 1 to n, measuring the degree of signal fluctuation within a period of time;

[0038] According to the extracted features, define a comprehensive index to evaluate whether the current signal is a seismic precursor. The comprehensive index formula is:

[0039] I(n) = w1·E[n] + w2·f dom [n] + w3·R[n] + w4·CD[n];

[0040] Set a threshold T for judgment. If I(n) > T, it indicates that a potential seismic precursor is detected, otherwise not.

[0041] Among them, the seismic prediction module makes a prediction based on historical earthquake data and the analysis results of the vibration data analysis module. Assume that the historical earthquake data includes the occurrence time and location, and analyze the historical earthquake data. The implementation formula is:

[0042]

[0043] In the formula, P hist (t, loc) represents the historical earthquake probability at the given time t and location loc, T represents the time window length, N represents the number of earthquake events in the past period of time, t i represents the time of the i-th earthquake occurrence, loc i represents the location of the i-th earthquake, α represents the attenuation factor, reducing the influence of earlier earthquakes on the current probability, and δ represents determining whether an earthquake occurs at the specified location;

[0044] Conduct real-time vibration data analysis according to the characteristics of real-time vibration data. The implementation formula is:

[0045]

[0046] In the formula, E[n] represents the energy of the vibration signal at time n, x′[n] represents the vibration data of the current time point n after filtering, and L represents the time window length for calculating the energy;

[0047] Combined prediction is carried out based on historical earthquake data analysis and real-time vibration data analysis, and the implementation formula is:

[0048] P prd (t, loc) = f(P hist (t, loc), E[n]),

[0049] In the formula, P prd represents the predicted earthquake probability, and f represents the input of P hist (t, loc) and E[n].

[0050] Among them, the earthquake precise positioning module is used to perform precise positioning through positioning technology. According to the prediction result of the earthquake prediction module, the geographical location of the user is obtained through the GPS module inside the mouse, the arrival times of the P-wave and S-wave are identified and recorded, the time difference between the arrival of the P-wave and S-wave is calculated, the epicenter distance is estimated by combining the velocity difference, and the epicenter position is more accurately determined through the relative time and position relationship between multiple devices. The obtained epicenter position is matched with the GIS data to determine the specific geographical location of the epicenter.

[0051] Among them, the voice synthesis module is used to generate an alarm message text according to the results of the earthquake prediction module and the earthquake precise positioning module, and convert the alarm message text into a voice signal; receive the prediction information from the earthquake prediction module and the epicenter position of the earthquake precise positioning module, fill the received earthquake prediction and positioning data into the alarm template to generate a specific alarm message text, preprocess the generated alarm message text, convert the alarm message text into a voice signal through voice synthesis technology, and send the converted voice signal to the mouse through the communication module.

[0052] Among them, the mouse earthquake early warning and precise positioning device based on voice broadcast includes: a three-axis accelerometer sensor, an earthquake sensing module, a central processing unit, a low-power Bluetooth module, a wireless communication module, a GPS module, an LED indicator, an operational amplifier module, a speaker, and a power supply module;

[0053] The triaxial accelerometer sensor is communicatively connected to the seismic induction module, the seismic induction module is communicatively connected to the central processor, the central processor is communicatively connected to the low-power Bluetooth module, the central processor is communicatively connected to the wireless communication module, the central processor is communicatively connected to the LED indicator, the central processor is communicatively connected to the power module, the central processor is communicatively connected to the GPS module, the central processor is communicatively connected to the operational amplifier module, and the operational amplifier module is communicatively connected to the speaker.

[0054] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0055] 1. The present invention preprocesses the data obtained by the seismic monitoring module through the background noise filtering module, identifies and filters out the background noise unrelated to earthquakes generated in daily life, significantly improves the signal-to-noise ratio of seismic signals, reduces the false alarm rate, and can more accurately extract the characteristics of seismic signals through signal processing technologies such as multi-scale wavelet transform, thereby improving the sensitivity and accuracy of the system. The method of dynamically setting the threshold enables the system to adapt to different environmental conditions, further enhancing the practicality.

[0056] 2. The present invention performs real-time analysis on the filtered vibration data through the vibration data analysis module, can quickly identify potential seismic signals, extract a variety of key features, and defines a comprehensive index to evaluate whether the current signal is a seismic precursor, improving the accuracy of earthquake detection.

[0057] 3. The present invention makes predictions through the earthquake prediction module by combining historical earthquake data and the results of the vibration data analysis module, can provide more accurate and reliable earthquake warnings. By analyzing the occurrence time and location of historical earthquakes and combining real-time vibration data, it can evaluate the probability of an earthquake occurring, not only considering historical earthquake patterns but also using current vibration characteristics, thereby improving the accuracy of prediction. The introduction of the attenuation factor reduces the influence of earlier earthquakes on the current prediction, enabling the system to better respond to new seismic activities.

[0058] 4. The present invention accurately locates the earthquake epicenter through the earthquake precise positioning module by obtaining the geographical location of the user through the GPS module and combining the arrival time difference between the P wave and the S wave, can achieve precise positioning of the earthquake epicenter. This method not only improves the accuracy of positioning but also can provide specific geographical location information for users. The voice synthesis module generates an alarm message text according to the results of the earthquake prediction module and the earthquake precise positioning module and converts it into a voice signal, providing an intuitive and easy-to-understand earthquake warning method. Through voice broadcast, even if the user is not in front of the computer or unable to view the screen, they can receive the warning information in a timely manner, improving the response speed of the warning. Description of the Drawings

[0059] Figure 1 This is a schematic structural diagram of the mouse earthquake early warning and precise positioning system based on voice broadcast of the present invention;

[0060] Figure 2 This is a flowchart of the operation of the background noise filtering module of the mouse earthquake early warning and precise positioning system based on voice broadcast of the present invention;

[0061] Figure 3 This is a flowchart of the operation of the vibration data analysis module of the mouse earthquake early warning and precise positioning system based on voice broadcast of the present invention;

[0062] Figure 4 This is a flowchart of the operation of the earthquake precise positioning module of the mouse earthquake early warning and precise positioning system based on voice broadcast of the present invention;

[0063] Figure 5 This is a schematic structural diagram of the mouse earthquake early warning and precise positioning device based on voice broadcast of the present invention. Detailed implementation manners

[0064] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0065] Embodiment

[0066] Please refer to Figures 1-5 As shown, the present invention provides a technical solution: including an earthquake monitoring module, a background noise filtering module, a vibration data analysis module, an earthquake prediction module, an earthquake precise positioning module, a voice synthesis module, and a communication module;

[0067] The earthquake monitoring module obtains earthquake data information in real time through the internal sensor device of the mouse and the earthquake monitoring network;

[0068] The background noise filtering module is used to preprocess the earthquake data obtained by the earthquake monitoring module, and identify and filter the background noise unrelated to earthquakes generated in daily life;

[0069] The vibration data analysis module is used to perform real-time analysis on the filtered vibration data to identify potential earthquake signals;

[0070] The earthquake prediction module is used to make predictions based on historical earthquake data and the analysis results of the vibration data analysis module;

[0071] The earthquake precise positioning module is used to perform precise positioning through positioning technology;

[0072] The voice synthesis module is used to generate an alarm message text according to the results of the earthquake prediction module and the earthquake precise positioning module, and convert the alarm message text into a voice signal;

[0073] The communication module is used for data transmission between each module and the mouse interaction module.

[0074] Among them, the earthquake monitoring module obtains earthquake data information in real time through the internal sensor device of the mouse and the earthquake monitoring network; continuously monitors the ground vibration through the triaxial accelerometer sensor inside the mouse, records the obtained ground vibration information, obtains the latest earthquake activity data through the earthquake monitoring network, and fuses the vibration data collected by the mouse with the data of the earthquake monitoring network. The fused earthquake data information is stored in the cloud.

[0075] Among them, the background noise filtering module preprocesses the data fused by the earthquake monitoring module, identifies and filters the background noise unrelated to earthquakes generated in daily life, smooths and standardizes the fused vibration data. Let the vibration data be x, and perform feature extraction through multi-scale wavelet transform. By changing the scale parameter a, features are extracted in different frequency ranges. The implementation formula is:

[0076]

[0077] In the formula, W(a, b) represents the wavelet transform coefficient, a represents the scale parameter, b represents the displacement parameter, x(t) represents the time series signal of the input vibration data, ψ represents the wavelet function, and ψ * represents the complex conjugate of the wavelet function.

[0078] Among them, after extracting the vibration features according to the multi-scale wavelet transform coefficients, background noise model training is performed. The background noise model is defined as:

[0079]

[0080] In the formula, L represents the total loss, N represents the number of samples, y i represents the true label of the i-th sample, p i represents the probability that the i-th sample of the background noise model belongs to the background noise, λ represents the regularization coefficient, and R represents the regularization term; the extracted vibration features are input into the trained background noise model to obtain the probability p of each data point. According to the probability distribution output by the background noise model, the threshold T = μ pd + kσ pd is dynamically set. If p > T, the data point is background noise, otherwise it is a valid signal;

[0081] When the data point is background noise, noise filtering is performed. Let the original data be x[n] and the filtered data be x′[n]. The implementation formula is as follows:

[0082]

[0083] In the formula, p[n] represents the probability that the nth data point of the background noise model is background noise, T represents the dynamically set threshold, W represents the window size, and the local average value is calculated. represents the sum of the data points within the window.

[0084] Among them, the vibration data analysis module receives the vibration data after background noise filtering and performs analysis, including energy, main frequency, signal mutation, and cumulative absolute difference.

[0085] Energy: The energy of the vibration signal reflects the vibration intensity. The implementation formula is as follows:

[0086]

[0087] In the formula, N represents the calculation window size, n represents the current time, x′[n] represents the value of the vibration data after noise filtering at time point n, and the formula represents the sum of the squares of all x′[n] from n - N + 1 to n, obtaining the signal energy within a time period.

[0088] Main frequency: Estimate the main frequency through short-time Fourier transform, and select the frequency corresponding to the maximum amplitude in the amplitude spectrum as the main frequency. Let the window length be M, the overlapping part be L, and the signal of the mth window be x′ m [k] = x′[k + mL], calculate the Fourier transform of each window to obtain the frequency spectrum X m (f). The implementation formula is as follows:

[0089]

[0090] In the formula, x′ m [k] represents the signal of the mth window, w[k] represents the window function, f represents the frequency, M represents the window length, j represents the imaginary unit, and find the frequency with the largest amplitude according to the frequency spectrum. The implementation formula is as follows:

[0091] f dom [n] = arg max f |X m (f)|,

[0092] In the formula, |X m (f)| represents the amplitude of the frequency spectrum, and arg max f represents finding the frequency f that makes |X m (f)| the largest.

[0093] Signal mutation rate: The speed of signal change, and the implementation formula is:

[0094]

[0095] In the formula, x′[n] represents the vibration data at the current time point n, x′[n - 1] represents the vibration data at the previous time point n - 1, and Δt represents the sampling interval;

[0096] Cumulative absolute difference: Measuring the degree of signal fluctuation, and the implementation formula is:

[0097]

[0098] In the formula, the formula represents the sum of the absolute differences between all adjacent time points x′[n] and x′[n - 1] from n - N + 1 to n, measuring the degree of signal fluctuation within a period of time;

[0099] According to the extracted features, define a comprehensive index to evaluate whether the current signal is a seismic precursor. The comprehensive index formula is:

[0100] I(n) = w1·E[n] + w2·f dom [n] + w3·R[n] + w4·CD[n];

[0101] Set a threshold T for judgment. If I(n) > T, it indicates that potential seismic precursors are detected, otherwise not.

[0102] Among them, the earthquake prediction module makes predictions based on historical earthquake data and the analysis results of the vibration data analysis module. Assume that the historical earthquake data includes the occurrence time and location. Analyze the historical earthquake data, and the implementation formula is:

[0103]

[0104] In the formula, P hist (t, loc) represents the historical earthquake probability at the given time t and location loc, T represents the time window length, N represents the number of earthquake events in the past period of time, t i represents the time of the i-th earthquake occurrence, loc i represents the location of the i-th earthquake, α represents the attenuation factor, reducing the impact of earlier earthquakes on the current probability, and δ represents determining whether an earthquake occurs at the specified location;

[0105] Conduct real-time vibration data analysis based on the characteristics of real-time vibration data, and the implementation formula is:

[0106]

[0107] In the formula, E[n] represents the energy of the vibration signal at time n, x′[n] represents the filtered vibration data at the current time point n, and L represents the length of the time window for calculating the energy;

[0108] Combined prediction is performed based on historical seismic data analysis and real-time vibration data analysis, and the implementation formula is:

[0109] P prd (t, loc) = f(P hist (t, loc), E[n]),

[0110] In the formula, P prd represents the predicted earthquake probability, and f represents the input of P hist (t, loc) and E[n].

[0111] Among them, the earthquake precise positioning module is used to perform precise positioning through positioning technology. According to the prediction result of the earthquake prediction module, the geographical location of the user is obtained through the GPS module inside the mouse, the arrival times of P-waves and S-waves are identified and recorded, the time difference between the arrivals of P-waves and S-waves is calculated, the epicenter distance is estimated by combining the velocity difference, the epicenter position is more precisely determined through the relative time and position relationships between multiple devices, and the obtained epicenter position is matched with the GIS data to determine the specific geographical location of the epicenter.

[0112] Among them, the voice synthesis module is used to generate an alarm message text according to the results of the earthquake prediction module and the earthquake precise positioning module, and convert the alarm message text into a voice signal; receive the prediction information from the earthquake prediction module and the epicenter position of the earthquake precise positioning module, fill the received earthquake prediction and positioning data into the alarm template to generate a specific alarm message text, preprocess the generated alarm message text, convert the alarm message text into a voice signal through voice synthesis technology, and send the converted voice signal to the mouse through the communication module.

[0113] Among them, the mouse earthquake early warning and precise positioning device based on voice broadcast includes: a three-axis accelerometer sensor, a seismic induction module, a central processing unit, a low-power Bluetooth module, a wireless communication module, a GPS module, an LED indicator, an operational amplifier module, a speaker, and a power module;

[0114] The three-axis accelerometer sensor is communicatively connected to the seismic induction module, the seismic induction module is communicatively connected to the central processing unit, the central processing unit is communicatively connected to the low-power Bluetooth module, the central processing unit is communicatively connected to the wireless communication module, the central processing unit is communicatively connected to the LED indicator, the central processing unit is communicatively connected to the power module, the central processing unit is communicatively connected to the GPS module, the central processing unit is communicatively connected to the operational amplifier module, and the operational amplifier module is communicatively connected to the speaker.

[0115] In this example, specifically: the earthquake monitoring module obtains earthquake data information in real time through the internal sensor device of the mouse and the earthquake monitoring network; continuously monitors ground vibrations through the triaxial accelerometer sensor inside the mouse, records the obtained ground vibration information, obtains the latest earthquake activity data through the earthquake monitoring network, and fuses the vibration data collected by the mouse with the data of the earthquake monitoring network. The fused earthquake data information is stored in the cloud.

[0116] The background noise filtering module preprocesses the data fused by the earthquake monitoring module, identifies and filters the non-earthquake-related background noise generated in daily life, smooths and standardizes the fused vibration data. Let the vibration data be x, and feature extraction is performed through multi-scale wavelet transform. By changing the scale parameter a, features are extracted in different frequency ranges. The implementation formula is:

[0117]

[0118] In the formula, W(a, b) represents the wavelet transform coefficient, a represents the scale parameter, b represents the displacement parameter, x(t) represents the time series signal of the input vibration data, ψ represents the wavelet function, and ψ * represents the complex conjugate of the wavelet function.

[0119] After extracting the vibration features based on the multi-scale wavelet transform coefficients, background noise model training is performed. The background noise model is defined as:

[0120]

[0121] In the formula, L represents the total loss, N represents the number of samples, y i represents the true label of the i-th sample, p i represents the probability that the i-th sample of the background noise model belongs to the background noise, λ represents the regularization coefficient, and R represents the regularization term; the extracted vibration features are input into the trained background noise model to obtain the probability p of each data point. According to the probability distribution output by the background noise model, the threshold T = μ pd + kσ pd is dynamically set. If p > T, the data point is background noise, otherwise it is a valid signal;

[0122] When the data point is background noise, noise filtering is performed. Let the original data be x[n], and the filtered data be x′[n]. The implementation formula is:

[0123]

[0124] In the formula, p[n] represents the probability that the nth data point of the background noise model is background noise, T represents a dynamically set threshold, W represents the window size, and the local average value is calculated. represents the sum of the data points within the window.

[0125] In this example, specifically: The vibration data analysis module receives the vibration data filtered by background noise for analysis, including energy, main frequency, signal mutation, and cumulative absolute difference.

[0126] Energy: The energy of the vibration signal reflects the vibration intensity, and the implementation formula is:

[0127]

[0128] In the formula, N represents the calculation window size, n represents the current time, x′[n] represents the value of the vibration data filtered by noise at time point n, and the formula represents the sum of the squares of all x′[n] from n - N + 1 to n, obtaining the signal energy within a time period.

[0129] Main frequency: Estimate the main frequency through short-time Fourier transform, and select the frequency corresponding to the maximum in the amplitude spectrum as the main frequency. Let the window length be M, the overlapping part be L, and the signal of the mth window be x′ m [k] = x′[k + mL], calculate the Fourier transform of each window to obtain the frequency spectrum X m (f), and the implementation formula is:

[0130]

[0131] In the formula, x′ m [k] represents the signal of the mth window, w[k] represents the window function, f represents the frequency, M represents the window length, j represents the imaginary unit, and find the frequency with the largest amplitude according to the frequency spectrum. The implementation formula is:

[0132] f dom [n] = arg max f |X m (f)|,

[0133] In the formula, |X m (f)| represents the amplitude of the frequency spectrum, and arg max f represents finding the frequency f that makes |X m (f)| the largest;

[0134] Signal mutation rate: The speed of signal change, and the implementation formula is:

[0135]

[0136] In the formula, x′[n] represents the vibration data at the current time point n, x′[n - 1] represents the vibration data at the previous time point n - 1, and Δt represents the sampling interval;

[0137] Cumulative absolute difference: Measures the degree of signal fluctuation. The implementation formula is:

[0138]

[0139] In the formula, the formula represents the sum of the absolute differences between all adjacent time points x′[n] and x′[n - 1] from n - N + 1 to n, measuring the degree of signal fluctuation over a period of time;

[0140] Based on the extracted features, a comprehensive index is defined to evaluate whether the current signal is a seismic precursor. The comprehensive index formula is:

[0141] I(n) = w1·E[n] + w2·f dom [n] + w3·R[n] + w4·CD[n];

[0142] Set a threshold T for judgment. If I(n) > T, it indicates that a potential seismic precursor is detected; otherwise, there is no seismic precursor.

[0143] Working principle: The seismic monitoring module continuously monitors the ground vibration information through the triaxial accelerometer sensor inside the mouse, obtains the latest seismic activity data through the seismic monitoring network, fuses the vibration data collected by the mouse with the data of the seismic monitoring network to form more comprehensive seismic data information, and stores it in the cloud. The background noise filtering module preprocesses the fused seismic data, identifies and filters out the non-seismic related background noise generated in daily life, extracts features through multi-scale wavelet transform, and conducts background noise model training on the extracted vibration features. Input the extracted vibration features into the trained background noise model, dynamically set the threshold according to the probability distribution output by the model, filter out the background noise, and retain the effective signal;

[0144] The vibration data analysis module receives the vibration data after background noise filtering and performs real-time analysis. The analysis content includes features such as energy, main frequency, signal mutation rate, and cumulative absolute difference. A comprehensive index is defined based on these features to evaluate whether the current signal is a seismic precursor. If the comprehensive index exceeds the set threshold, it indicates that a potential seismic precursor is detected. The earthquake prediction module makes a prediction based on historical earthquake data and the analysis results of the vibration data analysis module. It analyzes the historical earthquake data to obtain the historical earthquake probability at a given time and location, performs real-time vibration data analysis based on the features of the real-time vibration data, and combines the historical earthquake data analysis and the real-time vibration data analysis to obtain the predicted earthquake probability. The earthquake precise positioning module, according to the prediction result of the earthquake prediction module, obtains the user's geographical location through the GPS module inside the mouse, identifies and records the arrival times of the P-wave and S-wave, calculates the time difference between the arrivals of the P-wave and S-wave, estimates the epicenter distance in combination with the velocity difference, and more precisely determines the epicenter location through the relative time and position relationships between multiple devices. The obtained epicenter location is matched with the GIS data to determine the specific geographical location of the epicenter. The voice synthesis module receives the prediction information from the earthquake prediction module and the epicenter location from the earthquake precise positioning module, fills the received earthquake prediction and positioning data into the alert template to generate the specific alert message text, preprocesses the generated alert message text, converts the alert message text into a voice signal through voice synthesis technology, and sends the converted voice signal to the mouse through the communication module for the broadcast of the alert information through the voice broadcast function of the mouse.

[0145] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

[0146] The above describes the present invention and its implementation manners. This description is not restrictive, and what is shown in the drawings is only one of the implementation manners of the present invention. The actual structure is not limited thereto. Generally speaking, if those of ordinary skill in the art are inspired by it and, without departing from the purpose of the present invention, creatively design a structural manner and an embodiment similar to this technical solution, they shall fall within the protection scope of the present invention.

Claims

1. The mouse earthquake early warning and precise positioning system based on voice broadcast is characterized by: It includes earthquake monitoring module, background noise filtering module, vibration data analysis module, earthquake prediction module, earthquake precise positioning module, speech synthesis module and communication module; The earthquake monitoring module acquires earthquake data information in real time through the internal sensor equipment of the mouse and the earthquake monitoring network; The background noise filtering module is used to pre-process the seismic data acquired by the seismic monitoring module, and identify and filter the non-seismic related background noise generated in daily life; The vibration data analysis module is used to perform real-time analysis on the filtered vibration data to identify potential earthquake signals; The earthquake prediction module is used to make predictions based on historical earthquake data and the analysis results of the vibration data analysis module; The earthquake precise positioning module is used for precise positioning through positioning technology; The speech synthesis module is used to generate an alarm message text according to the results of the earthquake prediction module and the earthquake precise positioning module, and convert the alarm message text into a speech signal; The communication module is used for data transmission between each module and the mouse interaction module.

2. The mouse earthquake early warning and precise positioning system based on voice broadcast according to claim 1 is characterized in that: The earthquake monitoring module obtains earthquake data information in real time through the mouse's internal sensor equipment and the earthquake monitoring network; continuously monitors ground vibration through the three-axis accelerometer sensor inside the mouse, records the ground vibration information, obtains the latest earthquake activity data through the earthquake monitoring network, and fuses the vibration data collected by the mouse with the data of the earthquake monitoring network, and stores the fused earthquake data information in the cloud.

3. The mouse earthquake early warning and precise positioning system based on voice broadcast according to claim 1 is characterized in that: The background noise filtering module performs preprocessing based on the fused data of the earthquake monitoring module, identifies and filters the non-earthquake related background noise generated in daily life, and smoothes and standardizes the fused vibration data. Let the vibration data be x, and perform feature extraction through multi-scale wavelet transformation. By changing the scale parameter a, features are extracted within different frequency ranges. The implementation formula is: In the formula, W(a, b) represents the wavelet transform coefficient, a represents the scale parameter, b represents the displacement parameter, x(t) represents the time series signal of the input vibration data, ψ represents the wavelet function, and ψ * represents the complex conjugate of the wavelet function.

4. The mouse earthquake early warning and precise positioning system based on voice broadcast according to claim 3 is characterized in that: After extracting the vibration features based on the multi-scale wavelet conversion coefficients, the background noise model is trained and defined as: In the formula, L represents the total loss, N represents the number of samples, and y i represents the true label of the i-th sample, p i represents the probability that the i-th sample of the background noise model belongs to the background noise, λ represents the regularization coefficient, and R represents the regularization term; the extracted vibration features are input into the trained background noise model to obtain the probability p of each data point, and the threshold T = μ is dynamically set according to the probability distribution output by the background noise model pd +kσ pd , if p>T, the data point is background noise, otherwise it is a valid signal; When the data point is background noise, noise filtering is performed. Let the original data be x[n] and the filtered data be x′[n]. The implementation formula is: In the formula, p[n] represents the probability that the nth data point of the background noise model is background noise, T represents the dynamically set threshold, W represents the window size, and the local average is calculated. Represents the sum of the data points within the window.

5. The mouse earthquake early warning and precise positioning system based on voice broadcast according to claim 1 is characterized in that: The vibration data analysis module receives the vibration data after background noise filtering and performs analysis, including energy, main frequency, signal mutation and cumulative absolute difference; Energy: The energy of the vibration signal reflects the vibration intensity, and the realization formula is: In the formula, N represents the calculation window size, n represents the current time, x′[n] represents the value of the noise-filtered vibration data at time point n, and the formula represents the sum of the squares of all x′[n] from n-N+1 to n, which gives the signal energy within a time period; Main frequency: Estimate the main frequency through short-time Fourier transform, select the maximum corresponding frequency in the amplitude spectrum as the main frequency, set the window length to M, the overlapping part to L, and the signal of the mth window to x′ m [k] = x′[k+mL], calculate the Fourier transform of each window to get the frequency spectrum X m (f), the implementation formula is: In the formula, x′ m [k] represents the signal of the mth window, w[k] represents the window function, f represents the frequency, M represents the window length, j represents the imaginary unit, and the frequency with the largest amplitude is found according to the frequency spectrum. The implementation formula is: f dom [n]=argmax f |X m (f)|, In the formula, |X m (f)| represents the amplitude of the frequency spectrum, argmax f Indicates that |X is found m (f)|maximum frequency f; Signal mutation rate: the speed at which the signal changes. The implementation formula is: In the formula, x′[n] represents the vibration data at the current time point n, x′[n-1] represents the vibration data at the time point n-1, and Δt represents the sampling interval; Cumulative absolute difference: measures the degree of signal fluctuation, and the implementation formula is: In the formula, the formula represents the sum of the absolute differences between all adjacent time points x′[n] and x′[n-1] from n-N+1 to n, which measures the degree of fluctuation of the signal over a period of time; Based on the extracted features, a comprehensive index is defined to evaluate whether the current signal is an earthquake precursor. The comprehensive index formula is: I(n)=w1·E[n]+w2·f dom [n]+w3·R[n]+w4·CD[n]; A threshold T is set for judgment. If I(n)>T, it means that a potential earthquake precursor is detected, otherwise there is no earthquake precursor.

6. The mouse earthquake early warning and precise positioning system based on voice broadcast according to claim 1 is characterized in that: The earthquake prediction module predicts based on the historical earthquake data and the analysis results of the vibration data analysis module. Assuming that the historical earthquake data includes the occurrence time and location, the historical earthquake data is analyzed and the realization formula is: In the formula, P hist (t,loc) represents the historical earthquake probability at a given time t and location loc, T represents the length of the time window, N represents the number of earthquake events in the past period of time, t i represents the time when the i-th earthquake occurred, loc i represents the location of the ith earthquake, α represents the attenuation factor, which reduces the impact of earlier earthquakes on the current probability, and δ represents the factor used to determine whether an earthquake occurs at a specified location; According to the characteristics of real-time vibration data, real-time vibration data analysis is performed, and the implementation formula is: In the formula, E[n] represents the energy of the vibration signal at time n, x′[n] represents the filtered vibration data at the current time point n, and L represents the time window length for calculating the energy; Based on the combined prediction of historical earthquake data analysis and real-time vibration data analysis, the formula is: P prd (t,loc)=f(P hist (t,loc),E[n]), In the formula, P prd represents the predicted earthquake probability, and f represents P hist The input of (t,loc) and E[n].

7. The mouse earthquake early warning and precise positioning system based on voice broadcast according to claim 1 is characterized by: The earthquake precise positioning module uses positioning technology to perform precise positioning. According to the prediction results of the earthquake prediction module, the user's geographical location is obtained through the internal GPS module of the mouse, the arrival time of the P wave and the S wave is identified and recorded, the time difference between the arrival of the P wave and the S wave is calculated, and the distance to the epicenter is estimated based on the speed difference. The epicenter position is determined more accurately through the relative time and position relationship between multiple devices, and the obtained epicenter position is matched with the GIS data to determine the specific geographical location of the epicenter.

8. The mouse earthquake early warning and precise positioning system based on voice broadcast according to claim 1 is characterized by: The speech synthesis module is used to generate an alarm message text according to the results of the earthquake prediction module and the earthquake precise positioning module, and convert the alarm message text into a speech signal; receive the prediction information of the earthquake prediction module and the epicenter position of the earthquake precise positioning module, fill the alarm template with the received earthquake prediction and positioning data, generate a specific alarm message text, pre-process the generated alarm message text, convert the alarm message text into a speech signal through speech synthesis technology, and send the converted speech signal to the mouse through the communication module.

9. Mouse earthquake early warning and precise positioning device based on voice broadcast, including: Three-axis accelerometer sensor, earthquake sensing module, central processing unit, low-power Bluetooth module, wireless communication module, GPS module, LED indicator, op amp module, speaker and power module; The three-axis accelerometer sensor is communicatively connected to the seismic sensing module, the seismic sensing module is communicatively connected to the central processing unit, the central processing unit is communicatively connected to the low-power Bluetooth module, the central processing unit is communicatively connected to the wireless communication module, the central processing unit is communicatively connected to the LED indicator light, the central processing unit is communicatively connected to the power module, the central processing unit is communicatively connected to the GPS module, the central processing unit is communicatively connected to the operational amplifier module, and the operational amplifier module is communicatively connected to the speaker.

Images