CN-116990857-B - Automatic pickup method for longitudinal wave arrival time of multi-channel seismic signals based on first arrival time window

Abstract

The invention relates to a multi-channel seismic signal longitudinal wave arrival time automatic pickup method based on a first arrival time window. The first arrival pickup of the earthquake longitudinal wave is a key technology for expanding the work in the fields of earthquake exploration, earthquake early warning and the like. In the face of the production requirement of larger and larger data acquisition volume, a plurality of seismic signals need to be processed in the actual detection process. The usual manual pick-up method is time consuming and not accurate enough. The STA/LTA method needs to input and adjust parameters of each channel respectively, and cannot realize full-automatic pickup of multiple first arrival times, so that the method for automatically picking up the longitudinal wave arrival times of the multiple seismic signals based on the first arrival time varying window is provided. And optimizing and selecting a first arrival time window according to the relation between the parameters and the waveforms. Compared with the existing natural seismic longitudinal wave pickup method, the method is not only beneficial to improving the positioning precision, but also high in algorithm instantaneity and accuracy, and can achieve simultaneous pickup of multiple longitudinal wave first arrivals under the condition of zero input parameters.

Inventors

• JIANG TAO
• LI SHUANG
• Pang Zhongqin
• WANG XIN
• WANG JINGYE
• CHAO YUNFENG

Assignees

• 吉林大学

Dates

Publication Date

20260508

Application Date

20230701

Claims (1)

1. 1. A multi-channel seismic signal longitudinal wave arrival time automatic pickup method based on a first arrival time window is characterized by comprising the following steps: a. Reading vertical component seismic signals X (t) received by a plurality of geophone arrays, wherein t is a sampling point sequence number, the total channel number is K, then X (t) is represented by a set as X (t) = { X 1 (t),x 2 (t),...,x K (t) }, wherein a column vector X k (t) represents a kth channel signal, k=1, 2,..K, t=1, 2,..N, N is the total number of sampling points of each channel of seismic signals, the sampling rate is Fs, and in addition, the number of sampling points of background noise acquired before the occurrence of an earthquake in X k (t) is denoted as b k ; b. Let k=1; c. Hilbert transform is performed on x k (t): wherein τ is an integral variable, and y (k)max and y (k)min are the maximum and minimum values of y k (t), respectively; d. Defining a peak detection threshold E. column vector C has a size of Kx1, and all element values of 0.05, defined as y k ″(t)＝y k ′(t)-C (5) If y k ' (1) <0, then y k ' (t) need not be processed, otherwise, the first value less than 0 in y k ' (t) is searched for, and the corresponding sample point number is m k ,m k e [1, N ] and F. binarizing y k '' (t) to define The positions where four and more than four '1's appear continuously in D k (t) are marked as the effective signal positions of x k (t), the starting and ending sampling point serial numbers s k and e k ,s k ,e k E (1, N) of the first effective signal are searched G. Searching the maximum value of y k '' (t) in the range of [ s k ,e k ], marking as P k , and corresponding to the sampling point number r k ; h. Defining a pickup range right boundary detection threshold Q k ＝0.5(P k +0.05)-0.05 (8) Searching a value which can meet all conditions of the formula (9) in y k '' (t), wherein the sampling point serial number corresponding to the value is u k ; i. Let Δn k ＝r k -u k ; j. Let z k =3, "·" denote the multiplication, calculate the number of samples v k in [ u k -z k ·Δn k ,u k -(z k -1)·Δn k -1], y k "(t) greater than 0.1Q k , If v k does not satisfy formula (10), z k ＝z k +1, recalculate v k and determine if formula (10) is satisfied until formula (10) is satisfied, if v k satisfies formula (10), executing step k; k. The first arrival time is within [ u k -z k ·Δn k ,u k ] defining a one-dimensional vector W k ＝[x k (u k -z k ×Δn k ),x k (u k -z k ×Δn k +1),...,x k (u k )] T (11) 1. Accurate picking using AIC function AIC k (h)＝h·lg(Var(W k [1,h]))+(f k -h-1)·lg(Var(W k [h+1,f k ])) (12) Wherein f k is the length of W k , f k ＝z k ·Δn k +1,h＝1,2,...,f k ,Var(W k [1, h ]) refers to the variance of the h sample point data between W k (1) and W k (h); m, searching a sampling point sequence number a k corresponding to the minimum value of the finite rational number in AIC k (h), and setting the sampling point sequence number of the first arrival of the longitudinal wave as (u k -z k ·Δn k +a k -1); n, definition of Here, time k is x k (t) longitudinal first arrival Time; And if K is less than K, making k=k+1, repeating the steps c-o, and if k=K, completing multi-channel seismic longitudinal wave pickup.

Description

Automatic pickup method for longitudinal wave arrival time of multi-channel seismic signals based on first arrival time window Technical Field The invention belongs to the field of seismic detection, and particularly relates to a multi-channel seismic signal longitudinal wave arrival time automatic pickup method based on a first arrival time window. Background In the face of the production requirement of larger and larger data acquisition quantity, a plurality of micro-seismic signals need to be processed in the actual detection process. For example, in natural earthquakes, large errors exist in calculating the arrival time difference between the tracks due to the property differences of the underground medium, in addition, the number of earthquake stations is very large and the distribution is uneven, the sampling rates among the stations are different, and the first arrival pickup is difficult. Manually picking up large amounts of seismic data would waste a lot of manpower. The ratio of long time window (STA/LTA), the envelope function method, the red pool information criterion (AIC) and the like are common longitudinal wave arrival time picking methods, and play an important role in practical application. The method can not only rapidly and effectively pick up the first arrival of the longitudinal wave, but also has small error and high accuracy compared with the manual picking result. However, aiming at the condition of first arrival pickup of multiple channels of seismic signals, the methods need to input and adjust parameters of each channel respectively, have low calculation efficiency and can not meet the actual requirements of pickup of the multiple channels of seismic signals. Therefore, in order to perform multi-channel first-arrival rapid pickup, each parameter needs to be dynamically changed according to the actual condition of the seismic waves, so that the multi-channel first-arrival pickup with automation, simplicity and high pickup precision is realized. CN103995290 proposes a high-precision automatic picking method for the first arrival of the microseismic P-wave seismic phase, which solves the envelope of the microseismic signal by using Hilbert transformation, sets the threshold of the envelope to perform rough picking of the first arrival of the longitudinal wave, selects a time window and precisely solves the first arrival time of the longitudinal wave by using AIC function. The method has the advantages that subjective factors in parameter selection have great influence on the final pick-up result, the selection of parameters is quantized as much as possible, meanwhile, the envelope of a signal is solved by utilizing Hilbert transformation to have an end effect, and the coarse pick-up result is seriously influenced on the problem of multi-channel automatic pick-up so as to influence the precision of first-arrival pick-up. In addition, the method can only effectively pick up one seismic signal by inputting one parameter, and can not pick up multiple longitudinal waves of the seismic signals at the same time. In summary, there is a need for a first arrival picking method that can overcome the influence of the end point effect and self-select parameters on the premise of ensuring the longitudinal wave picking accuracy. Therefore, the invention provides a multi-channel seismic signal longitudinal wave arrival time automatic pickup method based on a first arrival time window, which divides pickup into pre-pickup and accurate pickup, and pre-picking, selecting time windows with different sizes near the first arrival according to different channel signal characteristics, and accurately picking signals within the range of the time windows. In the multi-channel picking problem, the method adjusts the time window size along with the change of the characteristics of different channels of seismic signals, so that the picking accuracy and multi-channel simultaneous picking are considered, and the problem of low picking efficiency in the multi-channel microseismic longitudinal wave picking process in the prior art is solved. Disclosure of Invention The invention aims to overcome the defects of the prior art and provide a multi-channel seismic signal longitudinal wave arrival time automatic pickup method based on a first arrival time varying window. The method is characterized in that a single channel of seismic signals are subjected to Hilbert transformation to obtain envelope signals, a first peak point of the envelope signals is closest to a first arrival of a seismic, a peak detection threshold is utilized to find the first peak point, in the process, an end effect influences the selection of the peak point, therefore, a method for judging whether the end effect influences pickup and a method for smoothing the end effect are needed, the size of a pickup window is designed according to the characteristics of different channels of signals by utilizing the right boundary detection threshold of a pi