An intelligent active explosion suppression monitoring method based on multi-sensor fusion

Abstract

The present application relates to the field of multi-sensor fusion, and particularly relates to an intelligent active explosion suppression monitoring method based on multi-sensor fusion. The content includes: obtaining original sensor signals and carrying out denoising processing, based on the denoised sensor signals, extracting time domain standardized features and spatial correlation features, and generating enhanced feature values of the sensor signals; based on the denoised sensor signals, calculating the standardized response intensity of the sensor unit, and obtaining a reference point; based on the enhanced feature values of the sensor signals and the reference point, generating a comprehensive hazard index, and judging the explosion precursor. The problems of poor quality of original sensor signals, difficulty in extracting explosion precursor features, deviation caused by directly fusing different types of sensor signals, affecting the monitoring effect of explosion precursor signals, neglecting the spatial correlation between sensors, difficulty in capturing the propagation law of signals, and not considering the distance relationship between sensors and explosion sources, leading to deviation in explosion precursor judgment, are solved.

Description

Technical Field

[0001] This invention relates to the field of multi-sensor fusion, and in particular to an intelligent active explosion suppression monitoring method based on multi-sensor fusion. Background Technology

[0002] With the continuous improvement of industrial automation and intelligence, safety protection issues in fields such as hazardous chemical storage and transportation, coal mining operations, and metallurgical production have received increasing attention. Among these, explosion accidents, due to their high-energy instantaneous release and enormous destructive power, often lead to serious casualties and property losses. Traditional explosion monitoring methods mostly rely on single-type sensors, such as temperature sensors, pressure sensors, flame sensors, or gas concentration detectors. While a single sensor may have a certain ability to detect pre-explosion signs under specific conditions, it also has many limitations, such as slow response speed, weak anti-interference ability, insufficient adaptability to environmental changes, low data redundancy, and high false alarm rate. In complex industrial environments, such as when dust concentration changes drastically and multiple interference sources are present, the accurate detection of pre-explosion signals is difficult, the fusion and analysis of multi-source heterogeneous data is insufficient, and the intelligent triggering mechanism of explosion suppression devices is inadequate.

[0003] In recent years, with the rapid development of sensor manufacturing technology and artificial intelligence algorithms, multi-sensor fusion technology has become an important means to improve safety monitoring. Through multi-sensor fusion, more accurate, comprehensive and reliable perception and early warning capabilities can be achieved. Especially in explosion monitoring scenarios, it can realize real-time, three-dimensional and dynamic capture of explosion precursors, providing a basis for decision-making for proactive explosion suppression. Summary of the Invention

[0004] This invention provides an intelligent active explosion suppression monitoring method based on multi-sensor fusion to address the following technical problems in scenarios such as chemical plant tank areas or coal mine roadways: sensor signals are often affected by noise interference such as mechanical vibration and background thermal radiation, resulting in poor original signal quality and difficulty in extracting pre-explosion features; differences in the range and units of signals from different types of sensors lead to deviations when directly compared or fused, affecting the monitoring effect of pre-explosion signals; pre-explosion signals have propagation characteristics in space, and traditional explosion suppression monitoring methods ignore the spatial correlation between sensors, making it difficult to capture the propagation law of the signal; and the distance relationship between the sensor and the explosion source is not considered, and the weight allocation lacks physical basis, leading to deviations in the judgment of pre-explosion signals.

[0005] The present invention provides an intelligent active explosion suppression monitoring method based on multi-sensor fusion, which specifically includes the following technical solutions: An intelligent active explosion suppression monitoring method based on multi-sensor fusion includes the following steps: S1. Acquire the original sensor signal and perform denoising processing to obtain the denoised sensor signal; based on the denoised sensor signal, introduce a multi-sensor array enhanced spatial correlation feature extraction algorithm to obtain time-domain normalized features and spatial correlation features, and generate enhanced feature values ​​of the sensor signal. S2. Based on the denoised sensor signal, calculate the standardized response intensity of the sensor unit and obtain the reference point; based on the enhanced feature value of the sensor signal and the reference point, generate a comprehensive hazard index and judge the pre-explosion signs.

[0006] Preferably, S1 specifically includes: Different types of sensor arrays are deployed, and each sensor unit in each type of sensor array has unique three-dimensional spatial coordinates.

[0007] Preferably, S1 specifically includes: In the implementation of the multi-sensor array enhanced spatial correlation feature extraction algorithm, the denoised sensor signal is subjected to time-domain normalization processing to obtain time-domain normalized features.

[0008] Preferably, S1 specifically includes: In the implementation of the multi-sensor array enhanced spatial correlation feature extraction algorithm, based on the denoised sensor signal, a signal propagation delay is introduced to calculate the cross-correlation coefficient between sensor units of the same type, and the cross-correlation coefficient between sensor units of the same type is used as the quantification of spatial correlation features; the signal propagation delay is calculated based on the spatial distance between sensor units and the signal propagation speed.

[0009] Preferably, S1 specifically includes: By introducing adjustment parameters, the time-domain normalized features and spatial correlation features are weighted and fused to obtain the enhanced feature values ​​of the sensor signal.

[0010] Preferably, S2 specifically includes: The normalized response intensity of the sensor unit is calculated based on the absolute value of the difference between the denoised sensor signal at the current moment and the denoised sensor signal at the previous moment. The normalized response intensity of all sensor units is compared, and the spatial position of the sensor unit with the largest normalized response intensity is selected as the reference point.

[0011] Preferably, S2 specifically includes: Based on the distance from the sensor unit to the reference point and combined with the exponential decay function, dynamic weights are calculated; based on the dynamic weights, the enhanced feature values ​​of all sensor signals are weighted and fused to generate a comprehensive hazard index.

[0012] Preferably, S2 specifically includes: The comprehensive hazard index is compared with the preset explosion threshold to determine the precursor to an explosion, and a binary signal is output to indicate whether the explosion suppression device is triggered.

[0013] The beneficial effects of the technical solution of the present invention are: 1. By deploying different types of sensor arrays in the target monitoring area, including acoustic sensors, infrared sensors, and pressure sensors, the comprehensiveness and reliability of monitoring are significantly improved.

[0014] 2. By using a multi-sensor array to enhance the spatial correlation feature extraction algorithm, combined with time-domain normalization processing and cross-correlation coefficient calculation between similar sensor units, the feature extraction of explosion precursor signals in both time and space was achieved. This ensures that abnormal features such as pressure changes, high temperatures, or sound wave diffusion can be captured, significantly improving the ability to extract key signals from complex environments and providing a reliable basis for accurate assessment of explosion risks.

[0015] 3. By calculating the standardized response intensity of each sensor unit and selecting the spatial location of the sensor unit with the largest standardized response intensity as the reference point, the potential explosion source can be quickly located, enhancing the spatial resolution capability of the monitoring system and providing important support for timely early warning of explosion risks.

[0016] 4. Based on the enhanced feature values ​​and reference points of sensor signals, a comprehensive hazard index is generated. Combined with the dynamic weight design of distance attenuation, the degree of explosion risk is comprehensively characterized, realizing the quantitative assessment of explosion precursors and improving the scientific nature and real-time performance of risk assessment. Attached Figure Description

[0017] Figure 1 This is a flowchart of an intelligent active explosion suppression monitoring method based on multi-sensor fusion as described in this invention. Detailed Implementation

[0018] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0020] The following description, in conjunction with the accompanying drawings, details a specific scheme for an intelligent active explosion suppression monitoring method based on multi-sensor fusion provided by the present invention.

[0021] See attached document Figure 1 The diagram illustrates a flowchart of an intelligent active explosion suppression monitoring method based on multi-sensor fusion, according to an embodiment of the present invention. The method includes the following steps: S1. Acquire the original sensor signal and perform denoising processing to obtain the denoised sensor signal; based on the denoised sensor signal, introduce a multi-sensor array enhanced spatial correlation feature extraction algorithm to obtain time-domain normalized features and spatial correlation features, and generate enhanced feature values ​​of the sensor signal. Within the target monitoring area (such as a chemical plant tank area, a coal mine roadway, or a natural gas pipeline station), different types of sensor arrays are deployed. These sensor types include acoustic sensors, infrared sensors, and pressure sensors. Each type of sensor array contains multiple sensor units, and each sensor unit has unique three-dimensional spatial coordinates. , Indicates the sensor type, Indicates the sensor unit index; Each sensor unit acquires the raw sensor signal in real time at a fixed sampling frequency. The selection of the sampling frequency must ensure that it can capture the rapid change characteristics of the pre-explosion signal, such as the high-frequency components of sound waves or pressure waves. The sampling frequency is set according to expert experience and can be set to 1000Hz, that is, the time window is set to 1 second, containing 1000 sampling points for subsequent statistical analysis. Since the original sensor signal may be affected by environmental noise, such as mechanical vibration and background thermal radiation, wavelet transform is used to denoise the original sensor signal in order to ensure the quality of the original sensor signal, resulting in a denoised sensor signal; the wavelet transform is a well-known technique in the art and will not be described in detail here. A multi-sensor array-enhanced spatial correlation feature extraction algorithm is used to perform time-domain normalization on the denoised sensor signals to eliminate differences in sensor units and ranges. Combined with the cross-correlation coefficients between similar sensor units, enhanced feature values ​​of the sensor signals are generated, aiming to extract features of pre-explosion signals from complex environments. The implementation process of the multi-sensor array-enhanced spatial correlation feature extraction algorithm is as follows: First, the denoised sensor signal is subjected to time-domain normalization. Specifically, for the denoised sensor signal of each sensor unit, the mean and standard deviation of the denoised sensor signal are calculated within a time window. The mean represents the average level of the denoised sensor signal, and the standard deviation reflects the fluctuation amplitude of the denoised sensor signal. The mean is subtracted from the denoised sensor signal and divided by the standard deviation to obtain the time-domain normalized feature. The time-domain normalized feature reflects the degree of deviation of the denoised sensor signal from the normal fluctuation level, capturing the instantaneous changes of the denoised sensor signal. Precursor signals of explosion (such as sudden pressure changes or high temperatures) will show significant deviations in feature values, thus effectively capturing anomalies. Furthermore, the cross-correlation coefficients between sensor units of the same type are calculated to quantify spatial correlation characteristics, reflect the propagation characteristics of pre-explosion signals in space, and capture the propagation law of denoised sensor signals, such as sound waves spreading outward from the explosion source. The calculation of the cross-correlation coefficient between the same type of sensor units takes into account the signal propagation delay, which is calculated based on wave propagation theory and combined with the spatial distance between sensor units and the signal propagation speed. Furthermore, the time-domain normalized features and spatial correlation features are weighted and fused to generate enhanced feature values ​​of the sensor signals, which are used to comprehensively characterize the spatiotemporal characteristics of the explosion precursor. The formula for calculating the enhanced characteristic value of the sensor signal is: ; in, Indicates time No. Class 1 Enhanced characteristic values ​​of individual sensor signals; It represents the time-domain normalization characteristics, used to fairly compare signal changes of different types of sensor units, highlighting the significance of pre-explosion signals; Indicates time No. Class 1 The denoised sensor signal; This represents the mean value of the denoised sensor signal within the time window, reflecting the baseline level of the denoised sensor signal. The method for calculating the mean value is a well-known technique in the art and will not be elaborated here. The standard deviation represents the denoised sensor signal within the time window. It is used to measure the fluctuation of the denoised sensor signal. The calculation method of the standard deviation is a well-known technique in the art and will not be elaborated here. This represents an adjustment parameter used to balance the weights of time-domain normalized features and spatial correlation features. Through experimental calibration, its value range is [value range missing]. ; Indicates the first class and the second The spatial correlation characteristics between individual sensor units and other sensor units of the same type are quantified by the cross-correlation coefficient of signals between sensor units of the same type, so as to reflect the propagation characteristics of pre-explosion signals. Indicates the first Class 1 The sensor and the first Class 1 The cross-correlation coefficient of two sensor signals is used to measure the similarity between them. The calculation method of the cross-correlation coefficient is a well-known technique in the art and will not be elaborated here. Indicates time No. Class 1 The denoised sensor signal; Indicates the first Number of sensor types; Indicates the first Class 1 The sensor and the first Class 1 The signal propagation delay between the sensors is used to correct the time difference in the propagation of the denoised sensor signals, ensuring the accuracy of the cross-correlation coefficient calculation. The calculation formula is as follows: , and They represent the first Class 1 The first sensor unit and the first Class 1 Spatial coordinates of each sensor unit Indicates the first Class 1 The first sensor unit and the first Class 1 The spatial distance between the individual sensor units reflects the spatial layout of the sensor array. This distance is calculated using Euclidean distance, a technique well-known to those skilled in the art, and will not be elaborated upon here. Indicates the first The propagation speed of signals from sensor-type sensors is used to reflect the characteristics of the signal propagation medium; Using only cross-correlation of sensor units of the same type ensures the physical rationality of cross-correlation coefficients, avoids noise interference from cross-type cross-correlation, improves feature extraction accuracy, enhances the ability to locate explosion sources and monitor risks, and combines time-domain standardized features and spatial correlation features to adapt to complex environments and ensure reliable operation in scenarios such as chemical plant tank areas and coal mine roadways.

[0022] S2. Based on the denoised sensor signal, calculate the standardized response intensity of the sensor unit and obtain the reference point; based on the enhanced eigenvalues ​​of the sensor signal and the reference point, generate a comprehensive hazard index and judge the pre-explosion signs. To identify potential explosion source locations, the standardized response strength of each sensor unit is calculated. This is achieved by calculating the absolute value of the difference between the denoised sensor signal at the current moment and the denoised sensor signal at the previous moment, reflecting the instantaneous change amplitude of the denoised sensor signal. The standardized response strength is obtained by dividing the absolute value of the difference by the standard deviation, as shown in the following formula: ; in, Indicates time No. Class 1 The standardized response intensity of each sensor unit is used to quantify the change amplitude of the denoised sensor signal and to locate the reference point. Indicates time No. Class 1 The denoised sensor signal; Indicates time and The absolute value of the difference between the denoised sensor signals is used to reflect the dynamic changes of the denoised sensor signals in order to capture the abrupt changes in the precursory features of an explosion. This indicates a time interval; a value of 0.1 is recommended. By comparing the normalized response intensities of all sensor units, the spatial location of the sensor unit with the largest normalized response intensity is selected as the reference point, as expressed by the following formula: ; ; in, Indicates time The estimated reference point location, used to reflect the location of potential explosion sources, is determined by selecting the spatial location of the sensor unit with the highest standardized response intensity. This indicates the spatial location of the sensor unit with the highest standardized response intensity; This indicates the type and index of the sensor unit with the highest normalized response strength; Based on the reference point, dynamic weights are calculated, and the enhanced feature values ​​of all sensor signals are weighted and fused to generate a comprehensive hazard index that reflects the degree of explosion risk. The dynamic weights are calculated based on the distance from the sensor unit to the reference point, and the exponential decay function reflects that the sensor unit closer to the reference point is assigned a higher weight, so as to reflect the physical law that the intensity of the explosion precursor signal decreases with distance. The formula for calculating the comprehensive risk index is: ; in, Indicates time The comprehensive hazard index, which integrates the enhanced characteristic values ​​of all sensor signals, is used to quantify the degree of explosion risk. This represents a double summation, which iterates through all sensor units and sums their contributions, thus combining the contributions of all sensor units. Indicates the first class and the second Each sensor unit in time Dynamic weights; This represents the sum of the exponential decay terms of all sensor units, and is used as a normalization factor to normalize the dynamic weights. This represents the exponential decay term, used to simulate the weakening of sensor signal strength with distance; This represents a parameter used to control the distance sensitivity of dynamic weights, determined experimentally, with a value range of [value missing]. ; Indicates the first Class 1 The distance from each sensor to a reference point is used to reflect the spatial relationship between the sensor and a potential explosion source; The explosion threshold is set based on expert experience. The comprehensive hazard index is compared with the explosion threshold to determine whether to trigger the explosion suppression device and output a binary signal (triggered or not triggered). Specifically, if the overall hazard index is greater than the explosion threshold, it indicates that there is a significant explosion risk in the current environment, and a trigger signal is output, i.e., the binary signal is 1, to activate the explosion suppression device, such as spraying extinguishing agent, releasing inert gas, or closing the isolation valve; if the overall hazard index is less than or equal to the explosion threshold, it indicates that the risk is insufficient to trigger the explosion suppression, and a non-trigger signal is output, i.e., the binary signal is 0. By comparing the comprehensive hazard index and the explosion threshold, it can determine in real time whether there are signs of an impending explosion (such as sound waves, heat radiation, or abnormal pressure), and quickly activate the explosion suppression device when signs of an impending explosion appear. This can effectively prevent an explosion from occurring and protect the safety of personnel and equipment. It is suitable for scenarios with high real-time requirements, such as chemical plants or coal mines.

[0023] In summary, an intelligent active explosion suppression monitoring method based on multi-sensor fusion has been developed.

[0024] The order of the embodiments is for illustrative purposes only and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0025] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0026] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. An intelligent active explosion suppression monitoring method based on multi-sensor fusion, characterized in that, Includes the following steps: S1. Acquire the original sensor signal and perform denoising processing to obtain the denoised sensor signal; based on the denoised sensor signal, introduce a multi-sensor array enhanced spatial correlation feature extraction algorithm to obtain time-domain normalized features and spatial correlation features, and generate enhanced feature values ​​of the sensor signal. S2. Based on the denoised sensor signal, calculate the standardized response intensity of the sensor unit and obtain the reference point; based on the enhanced feature value of the sensor signal and the reference point, generate a comprehensive hazard index and judge the pre-explosion signs.

2. The intelligent active explosion suppression monitoring method based on multi-sensor fusion according to claim 1, characterized in that, S1 specifically includes: Different types of sensor arrays are deployed, and each sensor unit in each type of sensor array has unique three-dimensional spatial coordinates.

3. The intelligent active explosion suppression monitoring method based on multi-sensor fusion according to claim 2, characterized in that, S1 specifically includes: In the implementation of the multi-sensor array enhanced spatial correlation feature extraction algorithm, the denoised sensor signal is subjected to time-domain normalization processing to obtain time-domain normalized features.

4. The intelligent active explosion suppression monitoring method based on multi-sensor fusion according to claim 3, characterized in that, S1 specifically includes: In the implementation of the multi-sensor array enhanced spatial correlation feature extraction algorithm, based on the denoised sensor signal, a signal propagation delay is introduced to calculate the cross-correlation coefficient between sensor units of the same type, and the cross-correlation coefficient between sensor units of the same type is used as the quantification of spatial correlation features; the signal propagation delay is calculated based on the spatial distance between sensor units and the signal propagation speed.

5. The intelligent active explosion suppression monitoring method based on multi-sensor fusion according to claim 4, characterized in that, S1 specifically includes: By introducing adjustment parameters, the time-domain normalized features and spatial correlation features are weighted and fused to obtain the enhanced feature values ​​of the sensor signal.

6. The intelligent active explosion suppression monitoring method based on multi-sensor fusion according to claim 2, characterized in that, S2 specifically includes: The normalized response intensity of the sensor unit is calculated based on the absolute value of the difference between the denoised sensor signal at the current moment and the denoised sensor signal at the previous moment. The normalized response intensity of all sensor units is compared, and the spatial position of the sensor unit with the largest normalized response intensity is selected as the reference point.

7. The intelligent active explosion suppression monitoring method based on multi-sensor fusion according to claim 6, characterized in that, S2 specifically includes: Based on the distance from the sensor unit to the reference point and combined with the exponential decay function, dynamic weights are calculated; based on the dynamic weights, the enhanced feature values ​​of all sensor signals are weighted and fused to generate a comprehensive hazard index.

8. The intelligent active explosion suppression monitoring method based on multi-sensor fusion according to claim 7, characterized in that, S2 specifically includes: The comprehensive hazard index is compared with the preset explosion threshold to determine the precursor to an explosion, and a binary signal is output to indicate whether the explosion suppression device is triggered.

Images