A modulation and analysis method of GPS time stamp for microseismic data adaptive acquisition rate

Abstract

The application discloses a modulation and analysis method of GPS time stamp for microseismic data adaptive acquisition rate, and belongs to the field of microseismic data processing.The application combines the two problems of time correction and data separation, and modulates the GPS time correction data into the acquisition signal synchronously.Because the acquisition signal contains the time correction information, any segment of flow is valid data, and context-related information such as file start does not need to be considered, when the time of each segment of data is obtained, the pulse corresponding to the data is found, and the time stamp contained in the pulse is decoded.The application provides the modulation and analysis method of GPS time stamp for microseismic data adaptive acquisition rate, which is efficient, low in error, and makes the time synchronization mechanism of multiple stations more reasonable and effective.

Description

Technical Field

[0001] This invention belongs to the field of microseismic data processing, and in particular relates to a design that requires precise timing of data, specifically involving the modulation and parsing of GPS timestamps for microseismic data. Background Technology

[0002] Microseismic monitoring technology is the most timely and information-rich monitoring method during reservoir fracturing. Its real-time, dynamic, and continuous operation is crucial for feedback of mine operation information. In monitoring the propagation behavior of reservoir fracturing fractures, the source mechanism is studied and the location of the seismic source is determined, identifying the fracture formation process, fracture orientation, length, and other information. This has a significant impact on well optimization design, fracturing scheme improvement, and increased production efficiency. With the development of mine digitalization and informatization, microseismic monitoring technology is widely used in mines prone to rockbursts or rock bursts. While reasonable mine design is clearly a necessary measure to control rockbursts, conducting microseismic monitoring during mining operations is an even more important measure for the prediction and control of microseismic events and rockbursts. One of the most obvious characteristics of events in microseismic monitoring is their short duration, generally only a fraction of a second. Identifying a single microseismic event requires about 10 microseismic acquisition devices. Furthermore, the synchronization and alignment of time is of paramount importance in microseismic identification. Therefore, it is essential to reasonably control the time error of multiple devices to ensure the timeliness of microseismic signals.

[0003] Existing methods for calculating the time of microseismic data mostly involve continuous system time synchronization, storing the time in the file header, and then obtaining the time of the corresponding waveform data through offsetting. This method has high network requirements and is computationally intensive. Each time the time is acquired, a large amount of computation is required, and if the network is interrupted, the acquisition must be restarted. This method has many problems: First, the time of the first data entry stored in the file is often obtained using the system time, which is often affected by the network and prone to errors. Second, it is impossible to obtain the data time at any random position; the time must be accumulated from the beginning each time data is acquired. Finally, when the data volume is large, the resulting error will become increasingly larger. While there may be small, fixed errors due to hardware such as the acquisition card, these small errors can be amplified by accumulating a large amount of data, causing significant impact. Summary of the Invention

[0004] To address the problems of existing technologies, this invention proposes a method for modulating and resolving GPS timestamps for microseismic data, which is essential. This method theoretically achieves an accuracy error at the microsecond level. Furthermore, it employs a pulse mechanism to integrate GPS timestamps into the data stream, ensuring that each point and each data point has a corresponding time. This allows for the acquisition of the time of random data, and the calculation of the time for each data point does not require accumulation, resulting in high efficiency and low error. This makes the synchronization mechanism of multiple stations more reasonable and effective.

[0005] This invention is achieved through the following technical solution:

[0006] Step 1: Obtain the time of each pulse from the GPS module, convert the pulse time information into a decimal timestamp, and then convert the decimal timestamp into a 32-bit binary format.

[0007] Step 2: The 32-bit binary timestamp is compiled into a data stream and written into the acquisition card and analysis program along with the acquired signal stream;

[0008] Step 3: Explicitly represent the 32-bit binary data using wide and narrow voltage levels: Treat points with voltage values ​​around 20,000 as voltage level points, and the wide voltage level point is twice the narrow voltage level point.

[0009] Step 4: Write a pulse position at the beginning of each timestamp stream. The pulse position represents the precise position of each timestamp.

[0010] Step 5: Integrate the time-stream data into the acquired data stream: The acquired data stream is stored and sent to the computing end, where all data needs to be parsed out;

[0011] Step 5-1 declares two buffers A and B to store the level points of pulses and timestamps;

[0012] Step 5-2 stores the number of points with wide and narrow levels at the current sampling rate through the B buffer, and then divides by 32 to obtain the intermediate threshold between the wide and narrow levels;

[0013] Step 5-3 determines whether a level point has appeared in the pulse data channel. If it has, the level point is stored in buffer B; if it has not, the determination of whether a level point has appeared continues.

[0014] Step 5-4: When buffer B is not empty, determine whether a level point has appeared in the timestamp data channel. If a level point appears, store all the level points at this moment in buffer A. When this level point ends, count the number of points in buffer A and then determine whether it is a wide level or a narrow level.

[0015] Step 5-5 determines whether it is a wide level or a narrow level. If it is a wide level, it is stored in a container with 1 as the value; if it is determined to be a narrow level, it is stored in a container with 0 as the value.

[0016] Steps 5-6 use the above steps to repeatedly check the storage. If the container is 32-bit, it means that the 32-bit timestamp has been parsed. Otherwise, continue to execute step 5-5.

[0017] Steps 5-7 output the container's timestamp data according to the last-in-first-out principle. Once the output is complete, a 32-bit binary timestamp is obtained.

[0018] Steps 5-8: Once the timestamp parsing is complete, clear the containers for A and B and wait for the next timestamp parsing.

[0019] When parsing the pulse timestamp position using the method in step 5 above in step 6, if you want to find the time between two pulses, you only need to use the parsed current pulse time and the time T required for the N sampling points after superimposing the pulses, where T = 1 / sampling rate * N.

[0020] The beneficial effects of this invention are:

[0021] This invention creates a GPS timestamp parsing method with adaptive acquisition rate, which solves the problem of the high real-time requirements of the data required for microseismic positioning. Furthermore, this invention can adaptively adapt to different acquisition frequencies, meaning it can be "plug and play" when the acquisition rate needs to be changed. Attached Figure Description

[0022] Figure 1 Flowchart for parsing GPS-encoded timestamps;

[0023] Figure 2 A satellite point map of the area surrounding the mining area;

[0024] Figure 3 A diagram showing the timestamp and pulse waveform;

[0025] Figure 4 This is a waveform representation of the event;

[0026] Figure 5 This is a diagram illustrating the GPS timestamp. Detailed Implementation

[0027] Step 1: Obtain the time of each pulse from the GPS module, convert the pulse time information into a decimal timestamp, and then convert the decimal timestamp into a 32-bit binary form.

[0028] Step 2 involves compiling a 32-bit binary timestamp into a data stream and writing it along with the acquired signal stream into the acquisition card and analysis program.

[0029] Step 3: Explicitly represent the 32-bit binary data using wide and narrow voltage levels: Treat points with voltage values ​​around 20,000 as voltage level points, and consider wide voltage level points as twice the number of narrow voltage level points.

[0030] Step 4: Write a pulse position at the beginning of each timestamp stream. The pulse position represents the precise position of each timestamp.

[0031] Step 5: Integrate the time-stream data into the acquired data stream: The acquired data stream is stored and sent to the computing end, where all data needs to be parsed out.

[0032] Step 5-1 declares two buffers A and B to store the level points of pulses and timestamps;

[0033] Step 5-2 stores the number of points with wide and narrow levels at the current sampling rate through the B buffer, and then divides by 32 to obtain the intermediate threshold between the wide and narrow levels;

[0034] Step 5-3 determines whether a level point has appeared in the pulse data channel. If it has, the level point is stored in buffer B; if it has not, the determination of whether a level point has appeared continues.

[0035] Step 5-4: When buffer B is not empty, determine whether a level point has appeared in the timestamp data channel. If a level point appears, store all the level points at this moment in buffer A. When this level point ends, count the number of points in buffer A and then determine whether it is a wide level or a narrow level.

[0036] Step 5-5 determines whether it is a wide level or a narrow level. If it is a wide level, it is stored in a container with 1 as the value; if it is determined to be a narrow level, it is stored in a container with 0 as the value.

[0037] Steps 5-6 use the above steps to repeatedly check the storage. If the container is 32-bit, it means that the 32-bit timestamp has been parsed. Otherwise, continue to execute step 5-5.

[0038] Steps 5-7 output the container's timestamp data according to the last-in-first-out principle. Once the output is complete, a 32-bit binary timestamp is obtained.

[0039] Steps 5-8: Once the timestamp parsing is complete, clear the containers for A and B, and wait for the next timestamp parsing.

[0040] When parsing the pulse timestamp position using the method in step 5 above in step 6, if you want to find the time between two pulses, you only need to use the parsed current pulse time and the time T required for the N sampling points after superimposing the pulses, where T = 1 / sampling rate * N.

[0041] Example 1:

[0042] To test the processing effectiveness and performance of this invention, as an example, 10 microseismic acquisition systems were deployed in a coal mine in Ordos City, Inner Mongolia Autonomous Region. The station location map is shown below. Figure 2 As shown, the deployment principles are: first, it should include the underground working face; second, it should include the working area on the ground; and finally, the selected locations should have a sense of hierarchy and overlap.

[0043] In these stations, we introduced methods and modules for GPS timestamp modulation and parsing.

[0044] First, the satellite time is obtained through the GPS module. Then, the satellite time obtained by the GPS module is processed by the STM32F103 core board. This satellite time is first converted into a 32-bit binary timestamp, and then marked with a pulse to represent the time at a specific point in time. This pulse is then represented by 1s and 0s using wide and narrow voltage levels, as shown in the waveform diagram. Figure 3 As shown.

[0045] Then, the timing of its vibration events is synchronized. For example... Figure 4 As shown, points with higher amplitudes represent microseismic events. To accurately determine the time of a microseismic event, we find the most recent pulse corresponding to the event's occurrence time T. Then, we use analytical methods to calculate the pulse time M at that moment. If the event occurred before the microseismic event, we shift the calculated pulse time backward by a certain number of points N. Using the acquisition rate X, we obtain the average time for each acquisition point, and then sum these average times to calculate the current time, as shown in the formula.

[0046]

[0047] This method reduces errors compared to traditional head-based time synchronization: Firstly, head-based time synchronization relies heavily on network conditions, but station deployments are often located around mines where signal base stations are scarce and networks are unstable. Secondly, head-based time synchronization often requires larger offsets. When a micro-vibration occurs at the end of the file, head-based time synchronization often needs to continuously offset that moment's time to the end. Since the sensors in the acquisition system have fixed single-digit deviations, these small errors accumulate and generate significant errors. The time synchronization method presented in this paper uses the time acquired by GPS and sets a pulse every two seconds throughout the entire file. For judging micro-vibration events, it only needs to find the preceding and following pulses, eliminating the need for offset calculations from the head and reducing the overhead of significant offset operations. The timestamp parsing is as follows... Figure 5 As shown.

Claims

1. A method for modulating and resolving GPS timestamps based on adaptive acquisition rates for microseismic data, characterized in that, The steps are as follows: Step 1: Obtain the time of each pulse from the GPS module, convert the pulse time information into a decimal timestamp, and then convert the decimal timestamp into a 32-bit binary form. Step 2: The 32-bit binary timestamp is compiled into a data stream and written into the acquisition card and analysis program along with the acquired signal stream; Step 3: Explicitly represent the 32-bit binary data using wide and narrow voltage levels: Treat points with voltage values ​​around 20,000 as voltage level points, and the wide voltage level point is twice the narrow voltage level point. Step 4: Write a pulse position at the beginning of each timestamp stream. The pulse position represents the precise position of each timestamp. Step 5: Integrate the time-stream data into the acquired data stream: The acquired data stream is stored and sent to the computing end, where all data needs to be parsed out; When parsing the pulse timestamp position using the method in step 5 above in step 6, if you want to find the time between two pulses, you only need to use the parsed current pulse time and the time T required for the N sampling points after superimposing the pulse, where T = 1 / sampling rate * N.

2. The method for modulation and parsing GPS timestamps for adaptive acquisition rate of microseismic data according to claim 1, characterized in that, In step 5, the specific method is as follows: Step 5-1 declares two buffers A and B to store the level points of pulses and timestamps; Step 5-2 stores the number of points with wide and narrow levels at the current sampling rate through the B buffer, and then divides by 32 to obtain the intermediate threshold between the wide and narrow levels; Step 5-3 determines whether a level point has appeared in the pulse data channel. If it has, the level point is stored in buffer B. If it does not appear, continue to determine whether a level point appears; Step 5-4: When buffer B is not empty, determine whether a level point has appeared in the timestamp data channel. If a level point appears, store all the level points at this moment in buffer A. When this level point ends, count the number of points in buffer A and then determine whether it is a wide level or a narrow level. Step 5-5 determines whether it is a wide level or a narrow level. If it is a wide level, it is stored in a container with 1 as the value; if it is determined to be a narrow level, it is stored in a container with 0 as the value. Steps 5-6 use the above steps to repeatedly check the storage. If the container is 32-bit, it means that the 32-bit timestamp has been parsed. Otherwise, continue to execute step 5-5. Steps 5-7 output the container's timestamp data according to the last-in-first-out principle. Once the output is complete, a 32-bit binary timestamp is obtained. Steps 5-8: Once the timestamp parsing is complete, clear the containers for A and B, and wait for the next timestamp parsing.

Images