CN-121656144-B - Harmful gas monitoring device and method for intelligent livestock farm

CN121656144BCN 121656144 BCN121656144 BCN 121656144BCN-121656144-B

Abstract

The application relates to the technical field of photoacoustic signal denoising, in particular to a harmful gas monitoring device and method for an intelligent livestock farm, wherein the method comprises the steps of collecting gas to be detected in the livestock farm and acquiring a photoacoustic signal of the gas; the characteristics of the photoacoustic signals affected by the noise overlapping signals are calculated to determine punishment factors when the photoacoustic signals are subjected to noise reduction processing, so that the photoacoustic signals are subjected to noise reduction, the noise-reduced photoacoustic signals are processed into photoacoustic signals, and the photoacoustic signals are used for detecting harmful gas concentration through a gas acoustic wave detection module in the harmful gas monitoring device. The application aims to improve the noise reduction effect on the photoacoustic signal, and finally can effectively improve the detection precision of the harmful gas monitoring device.

Inventors

YAN ZHILING
WANG YU
GUO XIAOLING
JI SHUHUI
LU YIJING
HUANG KAI
XU YUQING
ZHANG CHILEI
XIAO YANXIA
ZHANG JUN

Assignees

北京中科硕天科技有限公司

Dates

Publication Date: 20260508
Application Date: 20251210

Claims (10)

1. A method for monitoring harmful gases in an intelligent livestock farm, the method comprising the steps of: Collecting gas to be detected in a livestock farm and acquiring a photoacoustic signal of the gas; Calculating a first characteristic of the photoacoustic signal affected by the noise overlapping signal according to the energy distribution difference degree of each high energy region relative to all low energy regions in the photoacoustic signal and the energy distribution characteristics of the high energy regions; Calculating a second characteristic of the photoacoustic signal affected by the noise overlapping signal according to the bandwidth discreteness and the phase fluctuation discreteness of all frame signals in the photoacoustic signal; determining a punishment factor when the photoacoustic signal adopts a variable modal decomposition algorithm by combining the first characteristic and the second characteristic of the photoacoustic signal, which are influenced by the noise overlapping signal, and using the punishment factor for the photoacoustic signal to reduce noise; and processing the photoacoustic signal after noise reduction into a photoacoustic signal for detecting the concentration of harmful gas through a gas acoustic wave detection module in the harmful gas monitoring device.
2. A method for monitoring harmful gases in a smart livestock farm according to claim 1, wherein the method for determining the high energy region and the low energy region comprises: Converting the acquired photoacoustic signal into a mel-language spectrogram; calculating the local density of each coordinate point in each Mel spectrogram by using a density peak clustering algorithm, and marking the local density as local energy density; Dividing the Mel spectrogram into a plurality of regions by using a region growing segmentation algorithm, wherein the difference value of the local energy density between coordinate points is used as a measurement distance in the region growing process; And taking the maximum energy in each region as the input of the maximum inter-class variance, outputting an energy threshold, wherein the region larger than the threshold is a high-energy region, and the region larger than the threshold is a low-energy region.
3. A method of monitoring harmful gases in a smart livestock farm according to claim 1, wherein the first characteristic is calculated by summing the energy distribution differences of all high energy regions relative to all low energy regions, and multiplying the energy distribution characteristics of the high energy regions by the sum, as the first characteristic.
4. A method of monitoring harmful gases in a smart livestock farm as claimed in claim 3, wherein the degree of energy distribution difference is determined by the average of the total density energy between each high energy region and all low energy regions, wherein the total density energy is the sum of the energy of each coordinate point in the high/low energy region multiplied by the corresponding energy density.
5. A method of monitoring harmful gases in a smart livestock farm according to claim 3, wherein the energy distribution characteristics of the high energy region are calculated as: , representing the energy distribution characteristics of the high energy region, Representing the area of the high energy region in the mel-pattern, Representing the total area of the mel-language spectrogram, Representing the standard deviation of the distance between the high energy regions, Indicating the number of high energy regions.
6. A method of monitoring harmful gases in a smart livestock farm according to claim 1, wherein the second characteristic is calculated by: Calculating the bandwidth discreteness of all frame signals in the photoacoustic signals; calculating the average value of the phase fluctuation discreteness of all frame signals under the same frequency in the photoacoustic signals; calculating the average value of the phase fluctuation values of all frame signals under all frequency points in the photoacoustic signals; and taking the product of the bandwidth discreteness, the average value of the phase fluctuation discreteness and the average value of the phase fluctuation value as the second characteristic.
7. A method of monitoring harmful gases in a smart livestock farm according to claim 6, wherein the phase fluctuation analysis method is: acquiring phase values of all frame signals under the same frequency point p; According to the integral relation between frequency and phase, the ideal phase value of all frame signals under the same frequency point p can be calculated by utilizing the ideal frequency; and taking the absolute value of the difference value between the phase value of each frame signal at the same frequency point p and the ideal phase value as the phase fluctuation value of each frame signal at the frequency point p.
8. A method of monitoring harmful gases in a smart livestock farm as claimed in claim 6, wherein said discretization is determined by a discrete coefficient.
9. A method of monitoring harmful gases in a smart livestock farm as claimed in claim 1, wherein the penalty factor calculation formula is: wherein, the method comprises the steps of, A penalty factor is indicated and is indicated, Representing the penalty factor of the initial setting, 、 Representing the preset weight coefficient of the weight coefficient, 、 The first and second characteristics of the photoacoustic signal affected by the noise superimposed signal are represented, respectively.
10. Harmful gas monitoring devices of wisdom livestock-raising field, the device includes gas acquisition module, sound wave acquisition module, gaseous sound wave detection module and alarm module, its characterized in that: The gas collection module is used for collecting the gas to be detected in the livestock farm and filtering floating dust and water vapor in the gas through a filter in the module; The sound wave acquisition module is used for acquiring a photoacoustic signal of the gas to be detected through light absorption, thermal expansion and photoacoustic effect and denoising the photoacoustic signal, and the module realizes the harmful gas monitoring method of the intelligent livestock farm according to any one of claims 1-9; The gas acoustic wave detection module is used for analyzing the denoised photoacoustic signal and detecting the concentration of harmful gas; The alarm module is used for judging the monitoring result of the gas sound wave detection module and sending out alarm sound when the harmful gas in the gas to be detected exceeds the standard.

Description

Harmful gas monitoring device and method for intelligent livestock farm Technical Field The application relates to the technical field of photoacoustic signal denoising, in particular to a harmful gas monitoring device and method for an intelligent livestock farm. Background In recent years, with the development of livestock breeding to scale, intensification and intellectualization, a high-density centralized breeding mode in a house has become a common choice of farms, and too high concentration of harmful gas in the livestock house not only pollutes the air, but also can cause high incidence of livestock diseases and decline of production performance, seriously threatens the health condition of livestock, and healthy breeding is a precondition for sustainable development of livestock industry, so that the intelligent breeding air environment monitoring device is also gradually applied to production practice, gas data in the livestock house are collected in real time, and the concentration of the harmful gas is accurately monitored, thereby providing guarantee for the health of livestock. The existing common harmful gas detection method is a photoacoustic spectroscopy technology, sound waves generated after substances absorb light energy are utilized for analysis to realize harmful gas detection, but when the method is applied to livestock farms, the method is easily influenced by livestock and poultry sound, equipment operation and other sounds, so that photoacoustic signals acquired by the photoacoustic spectroscopy technology contain livestock and poultry noise, a photoacoustic signal overlapping phenomenon occurs, when noise-containing signals are decomposed by using a VMD (variable modal decomposition) algorithm, optimal punishment factors are difficult to determine due to complex and changeable environmental noise of the livestock houses, modal aliasing can be caused by smaller punishment factors, local information can be lost due to larger punishment factors, and the denoising effect of the photoacoustic signals directly influences the harmful gas monitoring effect. Disclosure of Invention In order to solve the technical problems, the application provides a harmful gas monitoring device and a harmful gas monitoring method for an intelligent livestock farm, and the adopted technical scheme is as follows: In a first aspect, an embodiment of the present application provides a method for monitoring harmful gases in a smart livestock farm, the method comprising the steps of: Collecting gas to be detected in a livestock farm and acquiring a photoacoustic signal of the gas; Calculating a first characteristic of the photoacoustic signal affected by the noise overlapping signal according to the energy distribution difference degree of each high energy region relative to all low energy regions in the photoacoustic signal and the energy distribution characteristics of the high energy regions; Calculating a second characteristic of the photoacoustic signal affected by the noise overlapping signal according to the bandwidth discreteness and the phase fluctuation discreteness of all frame signals in the photoacoustic signal; determining a punishment factor when the photoacoustic signal adopts a variable modal decomposition algorithm by combining the first characteristic and the second characteristic of the photoacoustic signal, which are influenced by the noise overlapping signal, and using the punishment factor for the photoacoustic signal to reduce noise; and processing the photoacoustic signal after noise reduction into a photoacoustic signal for detecting the concentration of harmful gas through a gas acoustic wave detection module in the harmful gas monitoring device. Preferably, the method for judging the high-energy region and the low-energy region comprises the following steps: Converting the acquired photoacoustic signal into a mel-language spectrogram; calculating the local density of each coordinate point in each Mel spectrogram by using a density peak clustering algorithm, and marking the local density as local energy density; Dividing the Mel spectrogram into a plurality of regions by using a region growing segmentation algorithm, wherein the difference value of the local energy density between coordinate points is used as a measurement distance in the region growing process; And taking the maximum energy in each region as the input of the maximum inter-class variance, outputting an energy threshold, wherein the region larger than the threshold is a high-energy region, and the region larger than the threshold is a low-energy region. Preferably, the first feature is calculated by adding the sum of the energy distribution differences of all the high energy regions relative to all the low energy regions and the product of the energy distribution features of the high energy regions. Preferably, the energy distribution difference degree is determined by an average value of total density energy between ea