Earthquake signal-based debris flow early warning method

Abstract

The invention discloses a debris flow early warning method based on seismic signals. The debris flow early warning method comprises the following steps: S1, acquiring a seismic signal sequence of a target position; s2, calculating the ratio of the short-time mean square value to the long-time mean square value; s3, comparing the ratios obtained in the step S2, and when the ratios are greater than a first threshold value, extracting a signal in a set time period after the sampling point and dividing the signal according to a preset time length to obtain a plurality of continuous and equilong sub-segments; s4, calculating the mean square value of each sub-segment signal obtained in the S3, and performing normalization processing on the long-time mean square values at the sampling points to obtain a trend factor sequence; s5, counting the value of each trend factor in the trend factor sequence obtained in the S4 and the rising change frequency (namely Fj + 1gt; and when the value of each trend factor is greater than a first threshold value and the rising change frequency among the trend factors is greater than a second threshold value, giving a debris flow early warning alarm.

Description

Technical Field

[0001] This invention relates to the field of geological disaster monitoring technology, and in particular to a debris flow early warning method based on seismic signals. Background Technology

[0002] Debris flows, as a highly destructive and sudden geological disaster, seriously threaten the safety of transportation, buildings, and people in mountainous areas. Traditional debris flow monitoring methods mostly rely on rain gauges and video monitoring systems, primarily using meteorological data such as rainfall intensity and cumulative rainfall to determine the likelihood of debris flows, or using fixed camera equipment to observe surface debris flow conditions in real time. While these methods provide some early warning capabilities, they also have significant limitations. On the one hand, rainfall monitoring cannot directly reflect the state of surface movement and is highly susceptible to factors such as uneven regional rainfall or underground runoff, leading to a high false alarm rate or the risk of missed reports. On the other hand, video image monitoring depends on good lighting, field of view, and weather conditions, and is prone to monitoring blind spots at night, in foggy weather, or in complex terrain, severely restricting the continuity and reliability of the system.

[0003] In recent years, the rapid development of earthquake monitoring technology has provided new means for geological disaster early warning. Seismographs, as highly sensitive vibration sensors, possess advantages such as all-weather operation, long-distance range, and high penetration. They can capture weak vibration information from the surface or subsurface in real time, unaffected by external conditions such as lighting, visibility, or complex terrain. This makes them suitable for deployment in mountainous areas with limited visibility, and particularly applicable to the dynamic monitoring of sudden disasters such as debris flows. Therefore, seismographs have significant advantages in dynamic disaster monitoring in mountainous areas and other areas prone to debris flows. Studies have shown that debris flows generate significant seismic waves during their initiation and movement, and the signals often exhibit significant energy changes. Therefore, using seismic signals to monitor debris flow events has become possible. However, in practical applications, debris flow seismic signals are often submerged in complex background noise, easily confused with other disturbances (such as rainfall, human activity, and earthquakes), leading to frequent misjudgments and missed detections.

[0004] Therefore, there is an urgent need to develop a robust debris flow identification method based on seismic signals, which can not only accurately capture the vibration characteristics unique to debris flow events, but also effectively distinguish background disturbance signals, thereby improving the reliability and practicality of the early warning system.

[0005] Application content

[0006] The purpose of this invention is to overcome the problems of high false alarm rate, strong environmental dependence, and insufficient identification accuracy in existing debris flow early warning methods, and to propose a debris flow early warning method based on seismic signals. The specific technical solution is as follows:

[0007] A debris flow early warning method based on seismic signals includes: S1, acquiring a seismic signal sequence at the target location; S2, calculating the mean square value of the signal within a short-term window and a long-term window of a predetermined length ahead of each sampling point in the seismic signal acquired in S1, and calculating the ratio of the short-term mean square value to the long-term mean square value; S3, comparing the ratio obtained in S2, and when the ratio is greater than a first threshold, extracting the signal of a predetermined time period after the sampling point and dividing it according to a preset duration to obtain several continuous, equal-length sub-segments; S4, calculating the mean square value of the signal of each sub-segment obtained in S3, and normalizing it with the long-term mean square value at the sampling point to obtain a trend factor sequence; S5, statistically analyzing the values ​​of each trend factor in the trend factor sequence obtained in S4, as well as the number of increases between each trend factor, and issuing a debris flow early warning alarm when the values ​​of each trend factor are greater than a first threshold and the number of increases between each trend factor is greater than a second threshold.

[0008] The seismic signal sequence for the target location obtained in S1 includes: acquiring the original seismic signal obtained by the seismographs deployed at debris flow monitoring points in a certain area, with a sampling rate of 50-150Hz, and using this signal as input.

[0009] S2 Short-term window length T S Set to 20-40 seconds, long window length T L Set to 180-360 seconds.

[0010] The formula for calculating the ratio R of the short-time mean square value STA to the long-time mean square value LTA obtained in S2 is as follows:

[0011]

[0012] Where i represents the sampling point, N s N represents the number of samples within a short time window. L The sampling number is the number of samples within a long time window, and the sampling rate is f. s N s =T S ×f s N L =T L ×f s .

[0013] In S3, the first threshold γ1 is set to 2-3, and after extracting this sampling point, a time period T is set. A Set to 60-120 seconds, segment duration T D Set to 6-15 seconds, each segment contains N D =T D ×f s Number of samples.

[0014] Calculate the trend factor sequence F in S4j The formula for time is:

[0015]

[0016] Where j = 0, 1, 2, ..., M-1, M = T A / T D This represents the number of trend analysis segments.

[0017] In S5, the second threshold is set to be more than 30% of the number of segments.

[0018] The beneficial effects of this invention lie in its ability to rapidly identify debris flow vibration signals through a STA / LTA energy ratio triggering mechanism. Subsequent trend factor analysis leverages the continuous enhancement of debris flow signals over a period of time, thereby improving the continuity and stability of identification. Compared to traditional methods, this invention significantly reduces the false alarm rate and enhances the practicality and reliability of early warning systems in complex terrain and multi-interference environments. It also boasts significant advantages such as flexible deployment, high computational efficiency, and strong applicability.

[0019] Instruction manual illustrations

[0020] Figure 1 This is a schematic diagram of the process of the present invention. Specific Implementation

[0021] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments and the accompanying drawings. It should be understood that these descriptions are merely exemplary and not intended to limit the scope of the invention. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concept of the invention.

[0022] A debris flow early warning method based on seismic signals includes:

[0023] S1. Obtain the seismic signal sequence of the target location; specifically, collect the original seismic signal obtained by the seismographs deployed at debris flow monitoring points in a certain area, with a sampling rate of 50-150Hz, and use this signal as input.

[0024] S2. Calculate the mean square value of the signal within the short-time window and the long-time window of a predetermined length for each sampling point in the seismic signal acquired in S1, and calculate the ratio of the short-time mean square value to the long-time mean square value; the short-time window length T in S2. S Set to 20-40 seconds, long window length T L Set to 180-360 seconds. In practical applications, T can be selected. S For 30 seconds, T L It lasts for 240 seconds.

[0025] The formula for calculating the ratio R of the short-time mean square value STA to the long-time mean square value LTA obtained in S2 is as follows:

[0026]

[0027] Where i represents the sampling point, N s N represents the number of samples within a short time window. L The sampling number is the number of samples within a long time window, and the sampling rate is f. s N s =T S ×f s N L =T L ×f s .

[0028] S3. Compare the ratio obtained in S2. When the ratio is greater than the first threshold, extract the signal of the time period after the sampling point and divide it according to the preset duration to obtain several continuous, equal-length segments. In S3, the first threshold γ1 is set to 2-3, and in practical applications, 2.5 can be selected. The time period T after the sampling point is extracted is set. A Set to 60-120 seconds, but 90 seconds can be selected in practical applications. The segment duration T D Set to 6-15 seconds, but 10 seconds can be selected in practical applications. Each segment contains N. D =T D ×f s The number of samples is determined. When the ratio is less than the first threshold, it indicates that no debris flow has occurred at that sampling point, and the calculation proceeds to the next sampling point.

[0029] S4. Calculate the mean square value of each segment signal obtained in S3, and normalize the long-time mean square value at the sampling point to obtain the trend factor sequence; S4 calculates the trend factor sequence F. j The formula for time is:

[0030]

[0031] Where j = 0, 1, 2, ..., M-1, M = T A / T D This represents the number of trend analysis segments.

[0032] S5. Calculate the value of each trend factor in the trend factor sequence obtained in S4, and the number of upward changes among each trend factor (i.e., F). j+1 >F jThe system determines the number of times a trend factor changes. A debris flow warning is issued when the values ​​of all trend factors are greater than the first threshold and the number of increases between trend factors is greater than the second threshold. In S5, the second threshold is set to be more than 30% of the number of segments. In practical applications, when the number of segments is 9, the second threshold is set to 4.

[0033] By employing the STA / LTA energy ratio triggering mechanism, rapid initial judgment of debris flow vibration signals is achieved. Subsequent trend factor analysis leverages the continuous enhancement characteristics of debris flow signals over a period of time, enhancing the continuity and stability of identification. Compared to traditional methods, this invention significantly reduces the false alarm rate and improves the practicality and reliability of the early warning system in complex terrain and multi-interference environments. It also boasts significant advantages such as flexible deployment, high computational efficiency, and strong applicability.

Claims

1. A debris flow early warning method based on seismic signals, characterized in that, include: S1. Obtain the seismic signal sequence at the target location; S2. Calculate the mean square value of the signal within the short-time window and the long-time window of a set length in front of each sampling point in the seismic signal obtained in S1, and calculate the ratio of the obtained short-time mean square value to the long-time mean square value. S3. Compare the ratio obtained in S2. When the ratio is greater than the first threshold, extract the signal of the time period after the sampling point and divide it according to the preset duration to obtain several continuous and equal-length sub-segments. S4. Calculate the mean square value of each sub-segment signal obtained in S3, and normalize it with the long-time mean square value at the sampling point to obtain the trend factor sequence. S5. Calculate the value of each trend factor in the trend factor sequence obtained in S4, and the number of times each trend factor increases. When the value of each trend factor is greater than the first threshold and the number of times each trend factor increases is greater than the second threshold, issue a debris flow early warning alarm.

2. The debris flow early warning method based on seismic signals as described in claim 1, characterized in that, The process of obtaining the seismic signal sequence at the target location in S1 includes: acquiring the original seismic signal obtained by the seismograph deployed at a debris flow monitoring point in a certain area, with a sampling rate of 50-150Hz, and using this signal as input.

3. The debris flow early warning method based on seismic signals as described in claim 2, characterized in that, The short time window length T in S2 S Set to 20-40 seconds, long window length T L Set to 180-360 seconds.

4. The debris flow early warning method based on seismic signals as described in claim 3, characterized in that, The formula for calculating the ratio R of the short-time mean square value STA to the long-time mean square value LTA obtained in S2 is as follows: Where i represents the sampling point, N s N represents the number of samples within a short time window. L The sampling number is the number of samples within a long time window, and the sampling rate is f. s N s =T S ×f s N L =T L ×f s .

5. The debris flow early warning method based on seismic signals as described in claim 4, characterized in that, In S3, the first threshold γ1 is set to 2-3, and a time period T is set after extracting the sampling point. A Set to 60-120 seconds, segment duration T D Set to 6-15 seconds, each segment contains N D =T D ×f s Number of samples.

6. The debris flow early warning method based on seismic signals as described in claim 5, characterized in that, In S4, the trend factor sequence F is calculated. j The formula for time is: Where j = 0, 1, 2, ..., M-1, M = T A / T D This represents the number of trend analysis segments.

7. The debris flow early warning method based on seismic signals as described in claim 6, characterized in that, In S5, the second threshold is set to be more than 30% of the number of segments.

Images