A smart water affairs voice broadcasting method and system based on intelligent audio monitoring

Abstract

The present invention relates to the field of acoustic monitoring technology, specifically a method and system for intelligent water affairs voice broadcasting based on intelligent audio frequency monitoring, comprising the following steps: collecting sound pressure level data and sound wave frequency content through underwater acoustic sensors, identifying abnormal water flow sound wave characteristics, calculating the abnormal level of water affairs events, screening water flow sound signal categories, and dividing characteristic frequency band intervals to generate a water environment acoustic feature set. In the present invention, by extracting and analyzing sound pressure level data and sound wave frequency, it is possible to monitor and predict the abnormal level of water affairs events in real time and accurately, and to respond quickly to potential water affairs problems, effectively reducing risks to the environment and public safety, optimizing fundamental frequency, and improving the clarity and comprehensibility of voice broadcasting, especially in a multi-noise environment. By analyzing the background noise in detail, the sound signal is adjusted to adapt to the noise environment, effectively optimizing the sound propagation quality, and significantly reducing misunderstandings caused by noise interference.

Description

Technical Field

[0001] The present invention relates to the field of acoustic monitoring technology, and in particular to a smart water affairs voice broadcasting method and system based on intelligent audio monitoring. Background Art

[0002] The field of acoustic monitoring technology includes technical solutions for acquiring, analyzing and applying acoustic signals in the environment. Its core content involves extracting target information from sound wave signals and completing target event identification and response through signal analysis and processing methods. This technical field includes multiple links such as acoustic signal acquisition, spectrum analysis, feature extraction, pattern recognition, etc., and is widely used in scenarios such as environmental monitoring, safety protection, and equipment status diagnosis. In the field of acoustic monitoring technology, the signal acquisition quality is improved through microphone arrays, multi-channel signal processing, echo analysis, etc., and the acoustic event recognition capability is improved by combining time-frequency domain transformation, audio stream analysis, classifier modeling and other methods. Some application scenarios require the combination of signal enhancement, denoising, speech recognition and other technologies to meet the acoustic monitoring needs in complex environments.

[0003] Among them, the smart water voice broadcast method based on intelligent audio monitoring refers to the use of acoustic monitoring technology to optimize the voice broadcast system in water management. Technical matters include collecting and analyzing sound data of water flow and its related environment, judging the status and category of water events by identifying specific patterns in the sound data, and adjusting the content and form of voice broadcasts through characteristic analysis and frequency optimization of sound signals to ensure that voice broadcasts can clearly and effectively convey key water information.

[0004] Acoustic monitoring in existing technologies is often affected by the quality of acoustic signals in high-noise environments, resulting in impaired accuracy and response speed in sound event recognition. In particular, without the support of complex signal processing, it is impossible to accurately distinguish the sound source and the nature of the sound, which is especially difficult in underwater environment backgrounds. In addition, traditional methods fail to effectively utilize advanced analysis technologies such as time-frequency domain transformation in the process of sound feature extraction and pattern recognition, making the system poorly adaptable in a changing environment and having low recognition efficiency. This limitation leads to delays in the transmission of key information when responding to emergencies, missing the optimal response time, and increasing environmental and public safety risks. In terms of voice broadcasting, existing technologies fail to effectively deal with the interference of environmental noise on voice clarity, resulting in the broadcast content being difficult to clearly understand in a noisy environment, thereby affecting the timely communication and reception of emergency information. Summary of the Invention

[0005] In order to solve the problem that acoustic monitoring in the prior art is often affected by the quality of acoustic signals in high-noise environments, resulting in impaired accuracy and response speed of sound event recognition, especially in the absence of complex signal processing support, it is impossible to accurately distinguish the source of sound and the nature of sound, especially in the underwater environment background. In addition, the traditional method fails to effectively utilize advanced analysis technologies such as time-frequency domain transformation in the process of sound feature extraction and pattern recognition, which makes the system poorly adaptable in a changing environment and has low recognition efficiency. It is limited to delaying the transmission of key information when responding to emergencies, missing the optimal response time, and increasing environmental and public safety risks. In terms of voice broadcasting, the prior art fails to effectively deal with the interference of environmental noise on voice clarity, resulting in the broadcast content being difficult to be clearly understood in a noisy environment, thereby affecting the timely transmission and reception of emergency information. The embodiment of the present invention provides a smart water voice broadcasting method and system based on intelligent audio monitoring. The technical solution is as follows:

[0006] On the one hand, a smart water affairs voice broadcast method based on intelligent audio monitoring is provided, comprising the following steps:

[0007] S1: Underwater acoustic sensors are used to collect sound pressure level data and sound wave frequency content, identify abnormal water flow sound wave characteristics, calculate the abnormality level of water affairs events, filter the water flow sound signal categories, and divide the characteristic frequency band intervals to generate a water environment acoustic feature set;

[0008] S2: calling the water environment acoustic feature set, analyzing the fundamental frequency, formant and timbre characteristics, determining the main frequency range and formant distribution pattern, adjusting the formant position, and generating frequency optimization parameters;

[0009] S3: Based on the frequency optimization parameters, the peak value of the main frequency, energy distribution density and instantaneous change rate of the background noise are analyzed, the main frequency offset degree of the water flow sound signal and the ambient noise is calculated, and the masking ratio of the water flow sound signal is adjusted to form a noise adaptation spectrum;

[0010] S4: calling the noise adaptation spectrum, combining the water flow sound signal and turbulence characteristics, calculating the background noise coverage time interval, adjusting the pause interval of the water affairs voice broadcast according to the flow velocity fluctuation, and generating the broadcast rhythm control parameters;

[0011] S5: Call the broadcast rhythm control parameters, calculate the voice signal power, measure the energy density of the voice broadcast, compare the signal power and the noise energy, adjust the volume dynamic range, and generate a voice broadcast adjustment result.

[0012] On the other hand, the water environment acoustic feature set includes frequency characteristics, energy peaks, sound wave deviation pattern characteristics and sound signal classification indicators. The frequency optimization parameters include fundamental frequency correction value, resonance peak adjustment ratio and main frequency interval optimization range. The noise adaptation spectrum includes background noise main frequency peak offset parameter, signal masking ratio adjustment coefficient and spectrum energy compensation parameter. The broadcast rhythm control parameters include adjusted pause interval time, flow rate change sensitivity and rhythm synchronization index. The voice broadcast adjustment results include volume gain adjustment configuration, volume level configuration and environmental noise adjustment parameters.

[0013] On the other hand, the steps for obtaining the water environment acoustic feature set are specifically as follows:

[0014] S101: Acquire sound pressure level data and sound wave frequency content collected by underwater acoustic sensors, calculate the rate of change of sound wave amplitude, identify the periodic characteristics of sound pressure changes, extract the main frequency interval where energy is concentrated based on the sound wave frequency content, compare the pattern characteristics that deviate from normal water flow sound waves, and generate a sound pattern index;

[0015] S102: Analyzing the deviation of the sound pressure level data and the frequency content using the sound pattern index, determining the degree of abnormality of the acoustic pattern, determining the level of the water service event based on the abnormality intensity, and classifying the water flow acoustic signal to generate an abnormality level classification index;

[0016] S103: calling the abnormality level classification index, defining the key frequency range of each sound signal, calculating the concentration of energy distribution within the frequency range, adjusting the characteristic frequency band boundary according to the energy proportion, and obtaining the water environment acoustic feature set.

[0017] On the other hand, the step of obtaining the frequency optimization parameters is specifically as follows:

[0018] S201: extracting the fundamental frequency, formant, and timbre parameters of the water flow sound signal based on the water environment acoustic feature set, analyzing the frequency fluctuation characteristics of the water flow sound signal within the difference time window, calculating the distribution pattern of the formant in the frequency spectrum, and obtaining characteristic parameters of the water flow sound signal;

[0019] S202: calling the characteristic parameters of the water flow sound signal, analyzing the fundamental frequency variation trend of the water flow sound signal, screening the main frequency range of the water flow sound signal, adjusting the center value of the fundamental frequency according to the variation characteristics of the main frequency range, and generating a fundamental frequency adjustment value;

[0020] S203: calling the fundamental frequency adjustment value, analyzing the distribution characteristics of the resonance peaks of the water flow sound signal, calculating the adjustment ratio of the resonance peak frequency, adjusting the frequency positioning of the resonance peak, determining the optimized resonance peak range, and generating frequency optimization parameters.

[0021] On the other hand, the step of acquiring the noise adaptation spectrum is specifically as follows:

[0022] S301: Calling the frequency optimization parameters, obtaining power spectrum data of the water environment background noise, analyzing the noise main frequency peak, energy distribution density and instantaneous change rate, detecting the fluctuation characteristics of the noise energy in the difference time window, and identifying the frequency range where the energy fluctuation exceeds the standard, and establishing a noise characteristic baseline value;

[0023] S302: Calculate the degree of deviation between the main frequency of the water flow acoustic signal and the main frequency of the ambient noise based on the noise characteristic baseline value, analyze the energy distribution characteristics of the deviation interval, evaluate the impact of noise interference on the water flow acoustic signal, and measure the energy loss ratio of the main frequency signal, adjust the spectral structure of the water flow acoustic signal, and establish frequency band energy compensation parameters;

[0024] S303: Analyze the distribution of water flow sound signal energy in different frequency bands using the frequency band energy compensation parameters, determine the masking ratio of the water flow sound signal, calculate the noise interference intensity, judge the degree of interference of the ambient noise on the water flow sound signal, identify the frequency band range of the water flow sound signal affected by the noise, and establish a noise adaptation spectrum.

[0025] On the other hand, the noise interference intensity is calculated using the formula:

[0026] ;

[0027] Determine the degree of interference of environmental noise on water flow acoustic signals, identify the frequency band of water flow acoustic signals affected by noise, and establish a noise adaptation spectrum;

[0028] in, represents the noise interference intensity, Represents the water flow sound signal in the The power value of the frequency band, Represents the ambient noise in The power value of the frequency band, Represents the total number of frequency bands.

[0029] On the other hand, the steps for obtaining the broadcast rhythm control parameters are specifically as follows:

[0030] S401: Extracting the time domain variation characteristics of the water flow acoustic signal based on the noise adaptation spectrum, analyzing the energy distribution of the background noise in the differential time interval, calculating the power variation of the noise in the time window, identifying the continuous coverage time of the background noise, and generating a noise persistence analysis result;

[0031] S402: Based on the noise persistence analysis results, extract the turbulence characteristics of the water flow acoustic signal, calculate the instantaneous fluctuation range of the flow velocity, compare the matching degree between the noise coverage interval and the flow velocity fluctuation range, determine the interference degree of the flow velocity change on the background noise, and generate a flow velocity adjustment impact index;

[0032] S403: Call the flow rate adjustment impact index, compare the background noise coverage time with the pause interval of the water service voice broadcast, calculate the flow rate fluctuation adjustment coefficient, analyze the flow rate fluctuation trend, combine the fluctuation trend and the noise coverage time, dynamically correct the voice broadcast pause time, and establish the broadcast rhythm control parameters.

[0033] On the other hand, the flow rate fluctuation adjustment coefficient is calculated using the formula:

[0034] ;

[0035] Analyze the flow velocity fluctuation trend, combine the fluctuation trend and noise coverage time, dynamically correct the voice broadcast pause time, and establish the broadcast rhythm control parameters;

[0036] in, represents the flow rate fluctuation adjustment coefficient, Representative time period The instantaneous change of internal flow velocity, Representative The flow rate measurement value of a time period, Represents the average flow rate within the selected time range, Represents the total number of flow rate measurements within the selected time range, represents the background noise coverage time, A standard time interval representing the rhythm of a broadcast.

[0037] On the other hand, the steps for obtaining the voice broadcast adjustment result are specifically as follows:

[0038] S501: Based on the broadcast rhythm control parameters, volume change data of the broadcast content is obtained, distribution characteristics of the volume in the different speech segments are analyzed, the deviation range between the volume peak and the average value is measured, and the speech segments whose volume change amplitude exceeds the standard are identified to obtain volume distribution characteristic values;

[0039] S502: Calculate the voice signal power by calling the volume distribution characteristic value, measure the energy ratio of the signal power in the different frequency bands based on the energy density of the ambient noise, select the frequency band range where the signal-to-noise ratio is different from the standard, adjust the voice power according to the signal attenuation amplitude, and establish the signal-to-noise ratio optimization parameter;

[0040] S503: Calculate the volume gain variation range of the broadcast content using the signal-to-noise ratio optimization parameters, identify the volume adjustment area affected by ambient noise, and adjust the volume distribution ratio and dynamic range according to the volume variation pattern to obtain a voice broadcast adjustment result.

[0041] On the other hand, a smart water affairs voice broadcasting system based on intelligent audio monitoring is provided. The system is applied to a smart water affairs voice broadcasting method based on intelligent audio monitoring, including:

[0042] The acoustic data processing module obtains the sound pressure level data and sound wave frequency content collected by the underwater acoustic sensor, identifies the abnormal water flow sound wave characteristics, filters the water flow sound signal categories, divides the characteristic frequency band intervals, and generates the water environment acoustic feature set;

[0043] The frequency characteristic adjustment module analyzes the fundamental frequency, formant and timbre characteristics based on the water environment acoustic characteristic set, determines the main frequency range and formant distribution pattern, adjusts the formant position, and generates frequency optimization parameters;

[0044] The noise analysis and optimization module analyzes the main frequency peak, energy distribution density and instantaneous change rate of the background noise based on the frequency optimization parameters, calculates the main frequency offset degree of the water flow sound signal and the ambient noise, adjusts the masking ratio of the water flow sound signal, and generates a noise adaptation spectrum;

[0045] The broadcast rhythm configuration module calls the noise adaptation spectrum, calculates the background noise coverage time interval, adjusts the pause interval of the water affairs voice broadcast according to the flow velocity fluctuation, and generates the broadcast rhythm control parameters;

[0046] The volume adjustment module calls the broadcast rhythm control parameter, calculates the voice signal power, measures the energy density of the voice broadcast, adjusts the volume dynamic range, and generates a voice broadcast adjustment result.

[0047] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0048] By extracting and analyzing sound pressure level data and sound wave frequency, it is possible to monitor and predict the abnormal level of water events in real time and accurately, and to respond quickly to potential water problems, effectively reducing risks to the environment and public safety, optimizing the fundamental frequency, and improving the clarity and comprehensibility of voice broadcasts, especially in high-noise environments. Through detailed analysis of background noise, the sound signal is adjusted to adapt to the noise environment, effectively optimizing the sound propagation quality and significantly reducing misunderstandings caused by noise interference. This not only enhances the practicality of sound data, but also greatly improves the efficiency of information transmission by precisely controlling the broadcast rhythm and volume. This is particularly critical in emergency situations, such as floods or other water quality monitoring alarms, ensuring that key information can be conveyed to decision makers and the public in a timely and accurate manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0050] Figure 1 It is a flow chart of the main steps of the present invention;

[0051] Figure 2 is a flow chart of the steps of S1 of the present invention;

[0052] Figure 3 This is a flow chart of the steps of S2 of the present invention;

[0053] Figure 4 This is a flow chart of the steps of S3 of the present invention;

[0054] Figure 5 This is a flow chart of the steps of S4 of the present invention;

[0055] Figure 6 This is a flow chart of the steps of S5 of the present invention;

[0056] Figure 7 It is a system block diagram of the present invention. DETAILED DESCRIPTION

[0057] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0058] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as an "exemplary" in the present invention should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of the word "exemplary" is intended to present concepts in a concrete manner. Furthermore, in the embodiments of the present invention, "and / or" can mean both or either of the two.

[0059] In the embodiments of the present invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, when the distinction is not emphasized, the meanings they convey are the same. The terms "of," "corresponding," and "corresponding" may sometimes be used interchangeably. It should be noted that, when the distinction is not emphasized, the meanings they convey are the same.

[0060] In the embodiments of the present invention, sometimes a subscript such as W1 may be written as a non-subscript such as W1. When the difference is not emphasized, the meanings to be expressed are the same.

[0061] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0062] The embodiment of the present invention provides a smart water affairs voice broadcast method based on intelligent audio monitoring, such as Figure 1 As shown, the following steps are included:

[0063] S1: Underwater acoustic sensors are used to collect sound pressure level data and sound wave frequency content, identify pattern characteristics that are different from normal water flow sound waves, calculate the abnormality level index of water affairs events based on these characteristics, filter the categories of water flow sound signals, divide the distribution intervals of the characteristic frequency bands of water flow sound signals, and generate a water environment acoustic feature set;

[0064] S2: Call the water environment acoustic feature set to analyze the fundamental frequency, formant, and timbre characteristics of the water flow sound signal, determine the main frequency range and formant distribution pattern of the water flow sound signal, optimize the fundamental frequency based on the main frequency change trend, adjust the formant position, and generate frequency optimization parameters;

[0065] S3: Based on the frequency optimization parameters, the peak value of the main frequency, energy distribution density, and instantaneous change rate of the background noise in the water environment are analyzed. The degree of deviation between the main frequency of the water flow sound signal and the main frequency of the ambient noise is calculated. The masking ratio of the water flow sound signal is adjusted according to the signal energy distribution to optimize the propagation quality of the sound signal and generate a noise-adapted spectrum.

[0066] S4: Call the noise adaptation spectrum, combine the water flow sound signal and turbulence characteristics, calculate the background noise coverage time interval, adjust the pause interval of the water affairs voice broadcast according to the flow velocity fluctuation, and generate the broadcast rhythm control parameters;

[0067] S5: Call the broadcast rhythm control parameters, obtain the volume change data of the broadcast content, calculate the voice signal power, measure the energy density of the voice broadcast, compare the signal power and noise energy, adjust the volume dynamic range, optimize the impact of background noise during broadcasting, and generate the voice broadcast adjustment results.

[0068] The acoustic feature set of the water environment includes frequency characteristics, energy peaks, sound wave deviation pattern characteristics and sound signal classification indicators. The frequency optimization parameters include the fundamental frequency correction value, the resonance peak adjustment ratio and the main frequency interval optimization range. The noise adaptation spectrum includes the background noise main frequency peak offset parameter, the signal masking ratio adjustment coefficient and the spectrum energy compensation parameter. The broadcast rhythm control parameters include the adjusted pause interval time, flow rate change sensitivity and rhythm synchronization index. The voice broadcast adjustment results include the volume gain adjustment configuration, the volume level configuration and the environmental noise adjustment parameters.

[0069] like Figure 2 As shown in Figure 2, the steps for obtaining the water environment acoustic feature set are as follows:

[0070] S101: Acquire sound pressure level data and sound wave frequency content collected by underwater acoustic sensors, calculate the rate of change of sound wave amplitude, identify the periodic characteristics of sound pressure changes, extract the main frequency interval where energy is concentrated based on the sound wave frequency content, compare the pattern characteristics that deviate from normal water flow sound waves, and generate a sound pattern index;

[0071] Use underwater microphones or piezoelectric acoustic sensors to collect data, set the sampling frequency in the range of 44.1kHz to 96kHz to ensure that the main noise frequency range of the underwater environment is covered, and convert the collected sound pressure level data into the root mean square value of the sound pressure using the formula ,in, Represents the sound pressure level (unit: dB), is the root mean square value of sound pressure (unit: Pa), The sound pressure level is calculated for the reference sound pressure (1μPa) to obtain the underwater noise level at a specific moment, and the rate of change of the sound wave amplitude is calculated by analyzing the change in the sound pressure level in adjacent time windows. The formula is used. ,in, represents the rate of change of the sound wave amplitude, Represents the current time The sound pressure level at the moment, Represents the previous time The sound pressure level at the moment, Represents the time interval. In this example, if , , time interval seconds, then , analyze the periodic characteristics of sound pressure changes, convert the time domain signal into frequency domain data through short-time Fourier transform (STFT), obtain the change of spectrum over time, set the window length to 512 points, the step size to 128 points, calculate the spectrum of different time periods, use the autocorrelation function to calculate the periodicity of the signal, set the signal autocorrelation value greater than 0.8, define it as a significant periodic feature, if the calculated autocorrelation value of a certain period is 0.85, it is considered that there is a periodic change in this period, combine the sound wave frequency content to extract the main frequency interval where the energy is concentrated, and calculate the power spectral density (PSD). The formula is as follows: ,in, represents the power spectral density, Represents the signal duration, The Fourier transform result of the representative signal is used to calculate the power of each frequency component and filter the frequency range where the energy accounts for more than 70% of the total energy. If the main frequency energy range is 500Hz to 1.5kHz, the data in this frequency range is extracted and compared with the pattern characteristics of the deviation from the normal water flow sound wave. The degree of deviation is calculated using the Euclidean distance to calculate the degree of deviation between the current frequency distribution and the standard water flow sound wave frequency distribution. Assume that the main frequency distribution of the standard water flow sound wave is Hz, the main frequency distribution of the current detection data is Hz, the Euclidean distance is calculated as follows: , Generates sound pattern indicators.

[0072] S102: Using the sound pattern index, analyze the deviation degree of the sound pressure level data and frequency content, determine the abnormality degree of the acoustic pattern, determine the level of the water service event based on the abnormality intensity, and classify the water flow acoustic signal to generate an abnormality level classification index;

[0073] Set the sound pressure level reference range for a normal water flow environment. For example, when the flow rate is 1.5-3.0m / s, the corresponding sound pressure level is 65-85dB. Using this as the reference range, calculate the deviation of the sound pressure level of the current ambient noise. The calculation formula is as follows: ,in, Represents the deviation between the current environment and the reference sound pressure level. Represents the currently measured sound pressure level, Represents the reference sound pressure level of normal water flow state. If the current sound pressure level dB, reference sound pressure level dB, then , compare the changes in the sound pressure level offset value, combine the changes in the frequency content, judge the abnormality of the acoustic mode, calculate the spectrum offset, and use the offset amplitude of each frequency point to calculate ,in, Represents the frequency offset of the current detection signal, Represents the current main frequency, Represents the reference water flow sound frequency. If the current main frequency is 900Hz and the reference water flow sound frequency is 1200Hz, then , classify water affairs events according to abnormal intensity, set abnormal classification standards, if dB or Hz, it is classified as a high-level abnormality. dB or Hz, it is classified as a medium-level anomaly, otherwise it is classified as a low-level anomaly. In this case, dB, Hz, so it is determined to be a medium-level anomaly, and an anomaly level classification index is generated.

[0074] S103: Call the abnormal level classification index, define the key frequency range of each sound signal, calculate the concentration of energy distribution within the frequency range, adjust the characteristic frequency band boundary according to the energy proportion, and obtain the water environment acoustic feature set.

[0075] Analyze different types of water service abnormal signals, set classification standards, and divide them into low-level abnormalities (600-1200Hz), medium-level abnormalities (400-1400Hz), and high-level abnormalities (200-1600Hz). If the current classification index is determined to be medium-level abnormality, the corresponding main frequency range is 400-1400Hz. Calculate the concentration of energy distribution in this frequency range and use power spectrum density integral to calculate the energy proportion. ,in, Represents the energy proportion of the target frequency range, represents the power spectral density, Represents the main frequency range of the current signal, Represents the frequency range of the entire signal. It is calculated that the energy in the current frequency range accounts for 70%. The characteristic frequency band boundary is adjusted according to the energy proportion. If the energy proportion is lower than 60%, the frequency range is appropriately expanded. If it is higher than 80%, the frequency range is narrowed. In this example, 70% is within a reasonable range, and the boundary is not adjusted to obtain the water environment acoustic feature set.

[0076] like Figure 3 As shown, the steps for obtaining the frequency optimization parameters are as follows:

[0077] S201: extracting the fundamental frequency, formant, and timbre parameters of the water flow sound signal based on the water environment acoustic feature set, analyzing the frequency fluctuation characteristics of the water flow sound signal within the difference time window, calculating the distribution pattern of the formant in the frequency spectrum, and obtaining characteristic parameters of the water flow sound signal;

[0078] The time domain signal is converted into a frequency domain signal using the Fast Fourier Transform (FFT). The sampling rate is set to 44.1kHz and the number of FFT points is set to 2048. The power spectrum density is calculated, and the fundamental frequency is determined by the frequency point where the maximum correlation peak is located. A design example is given: if the main frequency component of the acquired signal is in the range of 520Hz to 1520Hz, and the autocorrelation function is calculated to obtain the maximum peak at 810Hz, then the fundamental frequency is determined to be 810Hz. The resonance peak is calculated and extracted using the Linear Predictive Coding (LPC) method. The prediction order is set to 12, and the main resonance peak positions are calculated to be 610Hz, 1230Hz, and 1810Hz. The timbre parameters are calculated using the Mel-Frequency Cepstral Coefficients (MFCCs). The number of filters is set to 26, and the first 13 MFCC coefficients are extracted as timbre parameters. The frequency fluctuation characteristics of the water flow sound signal within the difference time window are analyzed using the Short-Time Fourier Transform (STFT), with the window length set to 1024 points and the step size set to 256 points. The spectrum changes in different time windows are calculated, and the frequency standard deviation is calculated. ,in, represents the frequency standard deviation, Representative The center frequency of the time window, Represents the average frequency in the time window. If the frequency data for a certain period are 810Hz, 825Hz, 790Hz, 820Hz, and 800Hz respectively, then calculate the average frequency , calculate the standard deviation , calculate the distribution pattern of the resonance peak in the spectrum, and count the energy proportion of each resonance peak in the power spectrum. The calculation formula is as follows: ,in, Represents the energy ratio of the resonance peak, represents the current resonance peak power, Represents the overall spectrum power. Assuming that the current resonance peak powers of 610Hz, 1230Hz, and 1810Hz are 2.8mW, 3.2mW, and 1.7mW respectively, and the total power is 10.5mW, then , and obtain the characteristic parameters of the water flow sound signal.

[0079] S202: Calling characteristic parameters of the water flow sound signal, analyzing the fundamental frequency variation trend of the water flow sound signal, screening the main frequency range of the water flow sound signal, adjusting the center value of the fundamental frequency according to the variation characteristics of the main frequency range, and generating a fundamental frequency adjustment value;

[0080] Calculate the rate of change of the fundamental frequency in different time windows using the differential calculation method ,in, represents the fundamental frequency change rate, Represents the fundamental frequency of the current time window, represents the fundamental frequency of the previous time window, Represents the time interval. Suppose the fundamental frequency of a certain time window changes to 810Hz, 830Hz, and 790Hz. Calculate the rate of change: , filter the main frequency range of the water sound signal, count the distribution of the fundamental frequency in different time windows, calculate the mode, mean, and median of the fundamental frequency distribution, and assume that the fundamental frequencies in a certain period of time are 755Hz, 810Hz, 785Hz, 810Hz, and 830Hz respectively. Then the main frequency range is [755Hz, 830Hz]. According to the changing characteristics of the main frequency interval, calculate the adjustment value of the fundamental frequency and set the adjustment standard. If the fluctuation amplitude of the fundamental frequency is greater than 30Hz / s, adjust the center value and use the calculated mean as the adjusted center value. , generating a fundamental frequency adjustment value.

[0081] S203: calling the fundamental frequency adjustment value, analyzing the distribution characteristics of the resonance peaks of the water flow sound signal, calculating the adjustment ratio of the resonance peak frequency, adjusting the frequency positioning of the resonance peak, determining the optimized resonance peak range, and generating frequency optimization parameters.

[0082] The statistical calculation formula for the offset of the resonance peak frequency in different time windows is as follows: ,in, represents the shift of the resonance peak frequency, Represents the current formant frequency, Represents the base resonance frequency. If the current resonance frequency of 610Hz changes to 600Hz, then , calculate the adjustment ratio of the resonance peak frequency and set the adjustment coefficient is 0.75, and the adjustment formula is as follows: , set the current resonance peak offset Hz, then , adjust the frequency positioning of the resonance peak and set the adjustment threshold. If the adjusted resonance peak is more than 5% lower than the baseline value, it will be adjusted upward by 5%, otherwise it will remain unchanged. In this example, the adjusted value is 592.5Hz, which is 2.87% lower than 610Hz and does not reach the 5% adjustment threshold, so it remains unchanged. Determine the optimized resonance peak range and generate frequency optimization parameters.

[0083] like Figure 4 As shown in FIG, the steps for obtaining the noise adaptation spectrum are as follows:

[0084] S301: Calling frequency optimization parameters to obtain power spectrum data of the water environment background noise, analyzing the noise main frequency peak, energy distribution density and instantaneous change rate, detecting the fluctuation characteristics of noise energy in the difference time window, and identifying the frequency range where energy fluctuation exceeds the standard, and establishing the noise characteristic baseline value;

[0085] Use underwater microphones or hydrophones to collect noise signals in different time windows and perform fast Fourier transform (FFT). Set the sampling rate to 48kHz and the number of FFT points to 4096 to calculate the power spectral density (PSD). Use the peak detection method to identify the peak of the main frequency of the noise. Assuming that the main frequency components of the current noise signal are 460Hz, 910Hz, and 1360Hz, the corresponding powers are 2.3mW, 3.9mW, and 2.0mW, respectively, calculate the energy distribution density using the formula: ,in, represents the noise energy distribution density, Represents the power value of the current frequency, Represents the overall spectrum power. If the total power is 9.8mW, then , calculate the instantaneous rate of change, using the differential calculation method ,in, represents the instantaneous rate of change, Represents the power value of the current time window, Represents the power value of the previous time window, Represents the time interval. If the noise power changes in a certain period of time are 3.9mW, 4.3mW, and 3.6mW, then , detect the fluctuation characteristics of noise energy under the difference time window, and calculate the standard deviation ,in, represents the standard deviation of noise power, Representative The power value of a time window, Represents the average power in the time window. If the power data in a certain period are 3.7mW, 4.0mW, 3.6mW, 4.2mW, and 3.9mW respectively, then calculate the average power , calculate the standard deviation;

[0086] ;

[0087] ;

[0088] To identify the frequency range where energy fluctuations exceed the standard, the threshold is set to twice the standard deviation, i.e. , detect that the noise power variation amplitude in a certain period is 0.55mW, which is greater than the threshold value of 0.428mW, then the frequency range is identified as an abnormal fluctuation range, and the noise characteristic baseline value is established.

[0089] S302: Calculate the degree of deviation between the main frequency of the water flow acoustic signal and the main frequency of the ambient noise based on the noise characteristic baseline value, analyze the energy distribution characteristics of the deviation interval, evaluate the impact of noise interference on the water flow acoustic signal, and measure the energy loss ratio of the main frequency signal. Adjust the spectral structure of the water flow acoustic signal and establish frequency band energy compensation parameters.

[0090] Based on the noise characteristic baseline value, the degree of deviation between the main frequency of the water flow sound signal and the main frequency of the ambient noise is calculated. The Euclidean distance is used to calculate the degree of deviation between the main frequency distribution of the current water flow sound signal and the frequency distribution of the ambient noise. Assuming that the main frequencies of the water flow sound are 620Hz, 1240Hz, and 1860Hz, and the main frequencies of the noise are 460Hz, 910Hz, and 1360Hz, the degree of deviation is calculated.

[0091] ;

[0092] ;

[0093] ;

[0094] Analyze the energy distribution characteristics of the offset interval and calculate the energy ratio of the overlap interval between the main frequency signal and the noise frequency using the formula: ,in, Represents the energy proportion of the overlapping interval, represents the frequency range of the overlapping interval, The frequency range representing the overall signal is obtained. The energy proportion of the overlapping interval is calculated to be 67%. The impact of noise interference on the water flow acoustic signal is evaluated and the interference level is set. If the energy proportion of the overlapping interval is greater than 70%, it is considered that the noise interference on the water flow signal is strong. If it is 50%≤≤≤70%, the interference is moderate. If it is <50%, the interference is weak. In this case, the proportion is 67%, so it is judged to be moderate interference. The energy loss ratio of the main frequency signal is calculated using the following formula: ,in, Represents the energy loss ratio of the main frequency signal, represents the power of the overlapping interval, Represents the total signal power. Assuming the overlapping interval power is 6.3mW and the total power is 9.2mW, then , adjust the spectrum structure of the water flow sound signal and set the compensation ratio. If the energy loss ratio is greater than 60%, 30% is compensated. If 40%≤loss≤60%, 15% is compensated. If the loss is less than 40%, no compensation is made. In this example, the loss ratio is 68.5%, which meets the 30% compensation standard. Calculate the compensation parameters. , establish frequency band energy compensation parameters.

[0095] S303: Using the frequency band energy compensation parameters, analyze the distribution of the water flow sound signal energy in the different frequency bands, determine the masking ratio of the water flow sound signal, and calculate the noise interference intensity to determine the degree of interference of the environmental noise on the water flow sound signal. Identify the frequency band range of the water flow sound signal affected by the noise and establish a noise adaptation spectrum.

[0096] To calculate the noise interference intensity, use the formula:

[0097] ;

[0098] Determine the degree of interference of environmental noise on water flow acoustic signals, identify the frequency band of water flow acoustic signals affected by noise, and establish a noise adaptation spectrum;

[0099] in, represents the noise interference intensity, Represents the water flow sound signal in the The power value of the frequency band, Represents the ambient noise in The power value of the frequency band, represents the total number of frequency bands analyzed;

[0100] The sound pressure level data is calculated based on the sound pressure level data monitored by the underwater acoustic sensor. Detected by the microphone array and converted into sound power, the calculation method is as follows:

[0101] ;

[0102] in, Represents the sound power reference value, generally taken W, Representative The sound pressure level of the frequency band is measured in dB (decibel). In the actual monitoring data, the sound pressure level of the water flow sound signal is measured to be dB, dB, dB, substitute into the formula to calculate:

[0103] ;

[0104] ;

[0105] ;

[0106] parameter Calculated from the sound pressure level of background noise data, the calculation method is the same as Similarly, in the actual monitoring data, the sound pressure level of the ambient noise is measured to be dB, dB, dB, substitute into the formula to calculate:

[0107] ;

[0108] ;

[0109] ;

[0110] Calculate the total absolute deviation between the water flow sound signal and the ambient noise power:

[0111] ;

[0112] ;

[0113] ;

[0114] Calculate the sum of the squares of the power of each frequency band of the ambient noise:

[0115] ;

[0116] ;

[0117] ;

[0118] Calculating noise interference intensity :

[0119] ;

[0120] ;

[0121] ;

[0122] The results show that the interference degree between the water flow sound signal and the background noise is 0.88. The closer the value is to 1, the stronger the masking effect of noise on the water flow sound signal. Reducing noise interference improves the distinguishability of the water flow sound signal and establishes a noise adaptive spectrum.

[0123] like Figure 5 As shown, the steps for obtaining the broadcast rhythm control parameters are as follows:

[0124] S401: Extract the time domain variation characteristics of the water flow acoustic signal based on the noise adaptation spectrum, analyze the energy distribution of the background noise in the differential time interval, calculate the power variation of the noise in the time window, identify the continuous coverage time of the background noise, and generate the noise persistence analysis result;

[0125] The short-time Fourier transform (STFT) is used to analyze the time-frequency characteristics of the water flow acoustic signal. The window length is set to 1024 points and the step size is 256 points. The spectrum changes of the time domain signal in different time windows are calculated, and the fluctuation of the signal amplitude over time is extracted. The signal amplitudes in the current time window are set to 0.85Pa, 0.88Pa, 0.82Pa, 0.90Pa, and 0.86Pa respectively, and the amplitude mean is calculated.

[0126] ;

[0127] Calculate the standard deviation of the amplitude;

[0128] ;

[0129] ;

[0130] Analyze the energy distribution of background noise in the difference time interval, use short-time power spectrum density to calculate the change of background noise energy, calculate the power change of noise in the time window, and assume that the background noise power in the current time window is 2.5mW, 2.8mW, 2.4mW, 2.9mW, and 2.6mW, then calculate the mean noise power in the time window. , calculate the rate of change of noise power, ,in, represents the noise power change rate, represents the noise power of the current time window, represents the noise power of the previous time window, represents the time interval, and the noise power changes to 2.8mW, 2.5mW, and 2.6mW, then , identify the continuous coverage time of background noise, set the noise persistence threshold to 2.5mW, calculate the duration of the noise power greater than the threshold in the continuous time window, and assume that the number of time windows with background noise power above 2.6mW is 12, and the time window interval is 0.5 seconds, then , generate noise persistence analysis results.

[0131] S402: Based on the noise persistence analysis results, the turbulence characteristics of the water flow acoustic signal are extracted, the instantaneous fluctuation range of the flow velocity is calculated, the matching degree between the noise coverage interval and the flow velocity fluctuation range is compared, the interference degree of the flow velocity change on the background noise is determined, and the flow velocity adjustment impact index is generated;

[0132] Use a flow meter to obtain flow velocity data in different time windows. Assume that the flow velocity measurements in the current time window are 1.8m / s, 2.1m / s, 1.9m / s, 2.3m / s, and 2.0m / s, respectively. Calculate the mean flow velocity. , calculate the standard deviation of flow rate;

[0133] ;

[0134] ;

[0135] Compare the matching degree between the noise coverage range and the velocity fluctuation range, calculate the correlation between the two, and use the Pearson correlation coefficient ,in, represents the correlation coefficient between flow velocity fluctuation and noise power change, Representative The flow rate in a time window, Representative The noise power of a time window is calculated as follows: If the calculated correlation coefficient is 0.68, it means that the flow rate change has a greater impact on the noise. The degree of interference of the flow rate change on the background noise is judged and the interference level is set. If the correlation coefficient is greater than 0.7, the interference is strong. If 0.5≤correlation coefficient≤0.7, the interference is medium. If <0.5, the interference is weak. In this example, the correlation coefficient is 0.68, so it is judged as medium interference. The flow rate adjustment impact index is calculated. ,in, represents the flow rate adjustment impact index, represents the standard deviation of the noise power, let mW, then , generating the flow rate adjustment impact index.

[0136] S403: Call the flow rate adjustment impact index, compare the background noise coverage time with the pause interval of the water service voice broadcast, calculate the flow rate fluctuation adjustment coefficient, analyze the flow rate fluctuation trend, combine the fluctuation trend and noise coverage time, dynamically correct the voice broadcast pause time, and establish the broadcast rhythm control parameters.

[0137] Calculate the flow rate fluctuation adjustment coefficient using the formula:

[0138] ;

[0139] Analyze the flow velocity fluctuation trend, combine the fluctuation trend and noise coverage time, dynamically correct the voice broadcast pause time, and establish the broadcast rhythm control parameters;

[0140] in, represents the flow rate fluctuation adjustment coefficient, Representative time period The instantaneous change of internal flow velocity, Representative The flow rate measurement value of a time period, Represents the average flow rate within the selected time range, Represents the total number of flow rate measurements within the selected time range, represents the background noise coverage time, A standard time interval representing the rhythm of the broadcast;

[0141] Instantaneous change in flow rate The flow rate data of water flow in different time periods is monitored by flow meter or ultrasonic flow meter, and the absolute value of the velocity difference between adjacent time points is calculated. The sampling interval is 1 second. At time t, the flow rate is m / s, The flow rate at that moment is m / s, therefore: ;

[0142] The calculation of the standard deviation of flow rate is based on the flow rate data in the selected time window, with a sampling period of 5 seconds and a measurement per second. The sampling value is m / s, m / s, m / s, m / s, m / s, calculate the average flow velocity:

[0143] ;

[0144] Calculate the standard deviation of the flow rate:

[0145] ;

[0146] ;

[0147] ;

[0148] Background noise coverage time Measured by noise detection equipment, calculated based on the time when the ambient noise level exceeds 40dB. The background noise coverage time during this period is measured. Seconds. Standard time interval for broadcast rhythm Set by the announcement system to 2.5 seconds.

[0149] Set by the announcement system to 2.5 seconds.

[0150] Substitute the above values ​​into the formula to calculate :

[0151] ;

[0152] ;

[0153] The results show that the current flow rate fluctuation adjustment coefficient is 3.635. The higher the value, the more significant the impact of flow rate fluctuations on the voice broadcast rhythm. This coefficient is used to dynamically correct the voice broadcast pause time. Combined with the influence of background noise, the pause time is adjusted to optimize the broadcast rhythm.

[0154] like Figure 6 As shown, the steps for obtaining the voice broadcast adjustment result are as follows:

[0155] S501: Based on the broadcast rhythm control parameters, volume change data of the broadcast content is obtained, distribution characteristics of the volume in the different speech segments are analyzed, the deviation range of the volume peak and the average value is measured, and the speech segments with volume change amplitudes exceeding the standard are identified to obtain volume distribution characteristic values;

[0156] Use an audio signal processing tool to segment the broadcast audio, set the time window to 500ms, calculate the volume (sound pressure level) within each window, and use the formula to calculate the instantaneous volume: ,in, Represents the volume level (unit: dB), Represents the root mean square volume value in the current window (unit: Pa), Assume that the base sound pressure is 0.00002Pa, and the root mean square volume value in a certain time window is 0.1Pa, then , analyze the distribution characteristics of volume in different speech segments and calculate the mean volume of each speech segment: ,in, Represents the average volume. Representative The volume of each speech segment, Represents the number of speech segments. Suppose the volume measurement values ​​of a speech segment are 72dB, 75dB, 73dB, 78dB, and 76dB. , measure the deviation range of the volume peak and average value, and calculate the deviation of the volume peak and average value: ,in, Represents the maximum volume value. If the volume peak is 78dB, then , identify the speech segments whose volume variation exceeds the standard, set the threshold value to 3dB, if the deviation is greater than 3dB, it is determined that the volume variation of the segment exceeds the standard. , is greater than the threshold, so the paragraph is marked as abnormal and the volume distribution feature value is obtained.

[0157] S502: Calling the volume distribution characteristic value to calculate the voice signal power, measuring the energy ratio of the signal power in the different frequency bands based on the energy density of the ambient noise, screening the frequency bands where the signal-to-noise ratio is different from the standard, adjusting the voice power based on the signal attenuation amplitude, and establishing the signal-to-noise ratio optimization parameters;

[0158] Call the volume distribution characteristic value to calculate the voice signal power and use the energy calculation formula ,in, represents the voice signal power, Represents the signal duration, Represents the instantaneous volume. Assuming the signal duration is 2 seconds and the root mean square volume value is 0.1 Pa, then, According to the energy density of the ambient noise, the energy proportion of the signal power in the different frequency bands is measured, and the power distribution of different frequency bands is calculated using the power spectrum density calculation formula: ,in, Represents the energy proportion of a specific frequency band, Represents the target frequency band range, Represents the overall speech frequency range. Assuming the target frequency band energy is 0.003W and the total power is 0.005W, then , filter the frequency band range where the signal-to-noise ratio is different from the standard and calculate the signal-to-noise ratio: ,in, Represents the noise power. If the noise power is 0.001W, then , set the signal-to-noise ratio standard threshold to 10dB. If the signal-to-noise ratio is less than 10dB, the frequency band needs to be adjusted. In this example, SNR=6.99dB, which is lower than the standard, so it needs to be adjusted. Adjust the voice power according to the signal attenuation amplitude and calculate the compensation gain , , establish the signal-to-noise ratio optimization parameters.

[0159] S503: Calculate the volume gain variation range of the broadcast content using the signal-to-noise ratio optimization parameters, identify the volume adjustment area affected by ambient noise, and adjust the volume distribution ratio and dynamic range according to the volume variation pattern to obtain the voice broadcast adjustment result.

[0160] Adjust the voice signal gain and calculate the volume after gain. in, Represents the adjusted volume value, set the compensation gain , original volume Pa, then , identify the volume adjustment area affected by ambient noise, set the noise interference threshold, if the ambient noise exceeds 40dB, then the area is considered to require additional gain compensation. If the ambient noise in a certain area is 42dB, then the area needs gain adjustment, and calculate the final volume adjustment value: , adjust the volume distribution ratio according to the volume change mode, set the volume adjustment rules for different voice segments, if the signal-to-noise ratio is <10dB and the noise is >40dB, increase the gain by 3dB, otherwise keep the original volume. In this example, the conditions are met and the voice broadcast adjustment result is obtained.

[0161] like Figure 7 As shown, a smart water affairs voice broadcasting system based on intelligent audio monitoring includes:

[0162] The acoustic data processing module obtains the sound pressure level data and sound wave frequency content collected by the underwater acoustic sensor, identifies the abnormal water flow sound wave characteristics, filters the water flow sound signal categories, divides the characteristic frequency band intervals, and generates the water environment acoustic feature set;

[0163] The frequency feature adjustment module analyzes the fundamental frequency, formant, and timbre characteristics based on the acoustic feature set of the water environment, determines the main frequency range and formant distribution pattern, adjusts the formant position, and generates frequency optimization parameters;

[0164] The noise analysis and optimization module analyzes the main frequency peak, energy distribution density and instantaneous change rate of background noise based on frequency optimization parameters, calculates the main frequency offset between the water flow sound signal and the ambient noise, adjusts the masking ratio of the water flow sound signal, and generates a noise adaptation spectrum;

[0165] The broadcast rhythm configuration module calls the noise adaptation spectrum, calculates the background noise coverage time interval, adjusts the pause interval of the water affairs voice broadcast according to the flow rate fluctuation, and generates the broadcast rhythm control parameters;

[0166] The volume adjustment module calls the broadcast rhythm control parameters, calculates the voice signal power, measures the energy density of the voice broadcast, adjusts the volume dynamic range, and generates the voice broadcast adjustment result.

[0167] It should be understood that the term "and / or" as used herein simply describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A alone, A and B together, or B alone. A and B can be singular or plural. Furthermore, the character " / " as used herein generally indicates an "or" relationship between the associated objects, but it may also indicate an "and / or" relationship. For specific understanding, please refer to the context.

[0168] In this disclosure, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, "at least one of a, b, or c" can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or plural.

[0169] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0170] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0171] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0172] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interface, indirect coupling or communication connection of the device or unit, which can be electrical, mechanical or other forms.

[0173] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0174] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0175] If the functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or the portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage media include various media that can store program code, such as USB flash drives, mobile hard drives, read-only memories (ROM), random access memories (RAM), magnetic disks, or optical disks.

[0176] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A smart water affairs voice broadcasting method based on intelligent audio monitoring, characterized in that: The method comprises: S1: Underwater acoustic sensors are used to collect sound pressure level data and sound wave frequency content, identify abnormal water flow sound wave characteristics, calculate the abnormality level of water affairs events, filter the water flow sound signal categories, and divide the characteristic frequency band intervals to generate a water environment acoustic feature set; S2: calling the water environment acoustic feature set, analyzing the fundamental frequency, formant and timbre characteristics, determining the main frequency range and formant distribution pattern, adjusting the formant position, and generating frequency optimization parameters; S3: Based on the frequency optimization parameters, analyze the main frequency peak, energy distribution density and instantaneous change rate of the background noise, calculate the main frequency offset degree of the water flow sound signal and the ambient noise, adjust the masking ratio of the water flow sound signal, and generate a noise adaptation spectrum; S4: calling the noise adaptation spectrum, combining the water flow sound signal and turbulence characteristics, calculating the background noise coverage time interval, adjusting the pause interval of the water affairs voice broadcast according to the flow velocity fluctuation, and generating the broadcast rhythm control parameters; S5: Calling the broadcast rhythm control parameters, calculating the voice signal power, measuring the energy density of the voice broadcast, comparing the signal power with the noise energy, adjusting the volume dynamic range, and generating a voice broadcast adjustment result; The abnormal level of water service events is based on the sound pressure level deviation and frequency offset Divide, where or It is a high-level exception. and It is a medium-level abnormality, and the rest are low-level abnormalities.

2. The smart water affairs voice broadcasting method based on intelligent audio monitoring according to claim 1 is characterized in that: The water environment acoustic feature set includes frequency characteristics, energy peaks, sound wave deviation pattern characteristics and sound signal classification indicators; the frequency optimization parameters include fundamental frequency correction value, resonance peak adjustment ratio and main frequency interval optimization range; the noise adaptation spectrum includes background noise main frequency peak offset parameter, signal masking ratio adjustment coefficient and spectrum energy compensation parameter; the broadcast rhythm control parameters include adjusted pause interval time, flow rate change sensitivity and rhythm synchronization index; the voice broadcast adjustment result includes volume gain adjustment configuration, volume level configuration and environmental noise adjustment parameter.

3. The smart water affairs voice broadcasting method based on intelligent audio monitoring according to claim 1 is characterized in that: The steps for obtaining the water environment acoustic feature set are specifically as follows: S101: Acquire sound pressure level data and sound wave frequency content collected by underwater acoustic sensors, calculate the rate of change of sound wave amplitude, identify the periodic characteristics of sound pressure changes, extract the main frequency interval where energy is concentrated based on the sound wave frequency content, compare the pattern characteristics that deviate from normal water flow sound waves, and generate a sound pattern index; S102: Analyzing the deviation of the sound pressure level data and the frequency content using the sound pattern index, determining the degree of abnormality of the acoustic pattern, determining the level of the water service event based on the abnormality intensity, and classifying the water flow acoustic signal to generate an abnormality level classification index; S103: calling the abnormality level classification index, defining the key frequency range of each sound signal, calculating the concentration of energy distribution within the frequency range, adjusting the characteristic frequency band boundary according to the energy proportion, and obtaining the water environment acoustic feature set.

4. The smart water affairs voice broadcasting method based on intelligent audio monitoring according to claim 1 is characterized in that: The steps for obtaining the frequency optimization parameters are specifically as follows: S201: extracting the fundamental frequency, formant, and timbre parameters of the water flow sound signal based on the water environment acoustic feature set, analyzing the frequency fluctuation characteristics of the water flow sound signal within the difference time window, calculating the distribution pattern of the formant in the frequency spectrum, and obtaining characteristic parameters of the water flow sound signal; S202: calling the characteristic parameters of the water flow sound signal, analyzing the fundamental frequency variation trend of the water flow sound signal, screening the main frequency range of the water flow sound signal, adjusting the center value of the fundamental frequency according to the variation characteristics of the main frequency range, and generating a fundamental frequency adjustment value; S203: calling the fundamental frequency adjustment value, analyzing the distribution characteristics of the resonance peaks of the water flow sound signal, calculating the adjustment ratio of the resonance peak frequency, adjusting the frequency positioning of the resonance peak, determining the optimized resonance peak range, and generating frequency optimization parameters.

5. The smart water affairs voice broadcasting method based on intelligent audio monitoring according to claim 1 is characterized in that: The steps for obtaining the noise adaptation spectrum are specifically as follows: S301: Calling the frequency optimization parameters, obtaining power spectrum data of the water environment background noise, analyzing the noise main frequency peak, energy distribution density and instantaneous change rate, detecting the fluctuation characteristics of the noise energy in the difference time window, and identifying the frequency range where the energy fluctuation exceeds the standard, and establishing a noise characteristic baseline value; S302: Calculate the degree of deviation between the main frequency of the water flow acoustic signal and the main frequency of the ambient noise based on the noise characteristic baseline value, analyze the energy distribution characteristics of the deviation interval, evaluate the impact of noise interference on the water flow acoustic signal, and measure the energy loss ratio of the main frequency signal, adjust the spectral structure of the water flow acoustic signal, and establish frequency band energy compensation parameters; S303: Analyze the distribution of water flow sound signal energy in different frequency bands using the frequency band energy compensation parameters, determine the masking ratio of the water flow sound signal, calculate the noise interference intensity, judge the degree of interference of the ambient noise on the water flow sound signal, identify the frequency band range of the water flow sound signal affected by the noise, and establish a noise adaptation spectrum.

6. The smart water affairs voice broadcasting method based on intelligent audio monitoring according to claim 5 is characterized in that: The noise interference intensity is calculated using the formula: ； Determine the degree of interference of environmental noise on water flow acoustic signals, identify the frequency band of water flow acoustic signals affected by noise, and establish a noise adaptation spectrum; in, represents the noise interference intensity, Represents the water flow sound signal in the The power value of the frequency band, Represents the ambient noise in The power value of the frequency band, Represents the total number of frequency bands.

7. The smart water affairs voice broadcasting method based on intelligent audio monitoring according to claim 1 is characterized in that: The steps for obtaining the broadcast rhythm control parameters are specifically as follows: S401: Extracting the time domain variation characteristics of the water flow acoustic signal based on the noise adaptation spectrum, analyzing the energy distribution of the background noise in the differential time interval, calculating the power variation of the noise in the time window, identifying the continuous coverage time of the background noise, and generating a noise persistence analysis result; S402: Based on the noise persistence analysis results, extract the turbulence characteristics of the water flow acoustic signal, calculate the instantaneous fluctuation range of the flow velocity, compare the matching degree between the noise coverage interval and the flow velocity fluctuation range, determine the interference degree of the flow velocity change on the background noise, and generate a flow velocity adjustment impact index; S403: Call the flow rate adjustment impact index, compare the background noise coverage time with the pause interval of the water service voice broadcast, calculate the flow rate fluctuation adjustment coefficient, analyze the flow rate fluctuation trend, combine the fluctuation trend and the noise coverage time, dynamically correct the voice broadcast pause time, and establish the broadcast rhythm control parameters.

8. The method for intelligent water affairs voice broadcasting based on intelligent audio monitoring according to claim 7 is characterized in that: The flow rate fluctuation adjustment coefficient is calculated using the formula: ； Analyze the flow velocity fluctuation trend, combine the fluctuation trend and noise coverage time, dynamically correct the voice broadcast pause time, and establish the broadcast rhythm control parameters; in, represents the flow rate fluctuation adjustment coefficient, Representative time period The instantaneous change of internal flow velocity, Representative The flow rate measurement value of a time period, Represents the average flow rate within the selected time range, Represents the total number of flow rate measurements within the selected time range, represents the background noise coverage time, A standard time interval representing the rhythm of a broadcast.

9. The smart water affairs voice broadcasting method based on intelligent audio monitoring according to claim 1 is characterized in that: The steps for obtaining the voice broadcast adjustment result are specifically as follows: S501: Based on the broadcast rhythm control parameters, volume change data of the broadcast content is obtained, distribution characteristics of the volume in the different speech segments are analyzed, the deviation range between the volume peak and the average value is measured, and the speech segments whose volume change amplitude exceeds the standard are identified to obtain volume distribution characteristic values; S502: Calculate the voice signal power by calling the volume distribution characteristic value, measure the energy ratio of the signal power in the different frequency bands based on the energy density of the ambient noise, select the frequency band range where the signal-to-noise ratio is different from the standard, adjust the voice power according to the signal attenuation amplitude, and establish the signal-to-noise ratio optimization parameter; S503: Calculate the volume gain variation range of the broadcast content using the signal-to-noise ratio optimization parameters, identify the volume adjustment area affected by ambient noise, and adjust the volume distribution ratio and dynamic range according to the volume variation pattern to obtain a voice broadcast adjustment result.

10. A smart water affairs voice broadcast system based on intelligent audio monitoring, wherein the smart water affairs voice broadcast system based on intelligent audio monitoring is used to implement the smart water affairs voice broadcast method based on intelligent audio monitoring according to any one of claims 1 to 9, characterized in that: The system comprises: The acoustic data processing module obtains the sound pressure level data and sound wave frequency content collected by the underwater acoustic sensor, identifies the abnormal water flow sound wave characteristics, filters the water flow sound signal categories, divides the characteristic frequency band intervals, and generates the water environment acoustic feature set; The frequency characteristic adjustment module analyzes the fundamental frequency, formant and timbre characteristics based on the water environment acoustic characteristic set, determines the main frequency range and formant distribution pattern, adjusts the formant position, and generates frequency optimization parameters; The noise analysis and optimization module analyzes the main frequency peak, energy distribution density and instantaneous change rate of the background noise based on the frequency optimization parameters, calculates the main frequency offset degree of the water flow sound signal and the ambient noise, adjusts the masking ratio of the water flow sound signal, and generates a noise adaptation spectrum; The broadcast rhythm configuration module calls the noise adaptation spectrum, calculates the background noise coverage time interval, adjusts the pause interval of the water affairs voice broadcast according to the flow velocity fluctuation, and generates the broadcast rhythm control parameters; The volume adjustment module calls the broadcast rhythm control parameter, calculates the voice signal power, measures the energy density of the voice broadcast, adjusts the volume dynamic range, and generates a voice broadcast adjustment result.

Images