Adaptive geological disaster monitoring data smoothing filtering method
Through the smooth filtering method of adaptive geological disaster monitoring data, the fluctuation problem of monitoring data is solved through interval fitting and repeated analysis, and the accurate fit of monitoring data is achieved, providing more accurate data support for geological disaster warning.
Patent Information
- Application Number
- CN202411036923.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-31
- Publication Date
- 2024-11-22
AI Technical Summary
In the existing geological disaster monitoring and early warning technology, the time series curve of the monitoring data is easily affected by non-deterministic factors, resulting in large fluctuations and difficulty in displaying the development trend of the real monitoring data, affecting the accuracy of stability evaluation and instability warning analysis.
Adaptive geological disaster monitoring data smooth filtering method is used to fit linear or exponential functions through partition intervals, and interval division and repeated fit are performed according to the correlation coefficient between the fitting function and the original monitoring curve until all interval fitting functions meet the requirements. Remove the impact of interfering data and achieve accurate fit of monitoring data.
The monitoring curve fluctuations caused by error values are effectively avoided, and a smoother curve is obtained, which is in line with the development trend of monitoring data, laying the foundation for accurate warning of geological disasters.
Smart Images

Figure CN119010842A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of geological disaster monitoring and early warning, and in particular to an adaptive geological disaster monitoring data smoothing and filtering method. Background Art
[0002] At present, geological disaster monitoring and early warning work has generally used professional monitoring equipment to obtain a large amount of monitoring data. However, since geological disaster bodies are inevitably affected by uncertain factors such as rainfall, human engineering activities, and reservoir water level fluctuations, the time series monitoring curve may fluctuate greatly, making it difficult to show the true development trend of monitoring data, which in turn affects the accuracy of geological disaster stability evaluation and instability warning analysis results. Preprocessing operations such as smoothing and filtering of monitoring data are an important basis for professional monitoring and early warning of geological disasters. However, the commonly used methods such as three-point linear sliding smoothing can only slow down the fluctuations of the monitoring curve to a certain extent, and cannot completely remove the curve fluctuations caused by the error value. Therefore, how to effectively avoid the fluctuations of the monitoring curve caused by the error value, achieve accurate fitting of the development trend of monitoring data, and lay the foundation for accurate early warning of geological disasters is a problem that technicians in this field need to solve urgently. Summary of the invention
[0003] The purpose of the present invention is to provide an adaptive geological disaster monitoring data smoothing filtering method, which can remove the influence of interference data and obtain the real development trend of monitoring data to meet the needs of geological disaster monitoring and early warning analysis.
[0004] In order to solve the above technical problems, the technical solution adopted by the present invention is: an adaptive geological disaster monitoring data smoothing filtering method, comprising the following steps: S1. Obtain the monitoring data collected by the sensor and draw the original time series monitoring curve; S2, based on the adaptive smoothing filtering method, a linear function or an exponential function is used to fit the data between the partitions; S3. Determine whether the results of the fitting functions of each interval meet the requirements according to the correlation coefficient between the fitting functions of each interval and the original monitoring curve; S4, retaining the interval fitting function that meets the requirements, dividing the interval data that does not meet the requirements into two parts, and then performing data fitting respectively; S5. Determine whether the secondary partition interval fitting function meets the requirements according to the correlation coefficient between the secondary partition interval fitting function and the original monitoring curve; S6. Repeat the above steps of interval division, function fitting and correlation coefficient analysis until the fitting functions of all intervals meet the requirements.
[0005] Preferably, the method for dividing the intervals in step S2 is: ; Among them, M is the total amount of monitoring data, n is the number of divided intervals, m is the amount of data in the first n-1 intervals, and s is the amount of data in the nth interval.
[0006] Preferably, the selection basis of each interval linear function or exponential function in step S2 is:
[0007] in, Step length l The acceleration between y i is any monitoring value in each interval, y i-l Step length l If the previous monitoring value appears 5 times in a row within a certain interval , then the exponential function is used to fit the data in this interval; otherwise, the linear function is used to fit the data.
[0008] Preferably, in step S2, n The linear function fitting method for -1 interval is: ; No. n The linear function fitting method for the interval is: ; in, a 0 , a 1 All are linear functions The unknown constant, x i is any time value in the interval, y i is any monitoring value within the interval, j is any interval among the first n-1.
[0009] Preferably, in step S2, n The exponential function fitting method for -1 interval is: ; No. n The exponential function fitting method for the interval is: ; in, a 2 , a 3 Both are exponential functions The unknown constant, x mjFor the j The last time value of the interval, y mj For the j The last monitored value of the interval, x M is the final time value, y M is the last monitored value.
[0010] Preferably, the correlation coefficient R between the interval fitting function in steps S3, S5 and S6 and the original monitoring curve is ≥ 0.80, indicating that the interval fitting function result meets the requirements.
[0011] Preferably, the method of repeatedly dividing the intervals in step S6 is: The interval data that does not meet the requirements is divided into intervals for the zth time, which will form 2z-1 intervals. The data volume of the first 2z-1-1 intervals is: ; The amount of data in the 2z-1th interval is: ; Among them, the symbol represents the floor function, z is the number of times the interval is divided.
[0012] The present invention provides an adaptive smoothing filtering method for geological disaster monitoring data. Compared with other methods, the filtered curve of the present invention is smoother and more in line with the development trend of monitoring data. It can effectively avoid the ups and downs of the monitoring curve caused by the error value, realize the accurate fitting of the development trend of monitoring data, and lay the foundation for accurate early warning of geological disasters. BRIEF DESCRIPTION OF THE DRAWINGS
[0013] The present invention will be further described below in conjunction with the accompanying drawings and embodiments: Figure 1 is a flow chart of the method of the present invention; Figure 2 It is the original time series monitoring curve of the embodiment of the present invention; Figure 3 A monitoring curve fitted by the first partitioning interval of an embodiment of the present invention; Figure 4 A monitoring curve fitted by the second divided interval according to an embodiment of the present invention; Figure 5 This is a monitoring curve fitted by the third division interval in an embodiment of the present invention. DETAILED DESCRIPTION
[0014] like Figure 1As shown, a method for adaptive smoothing and filtering of geological disaster monitoring data comprises the following steps: S1. Obtain the monitoring data collected by the sensor and draw the original time series monitoring curve; S2, based on the adaptive smoothing filtering method, a linear function or an exponential function is used to fit the data between the partitions; S3. Determine whether the results of the fitting functions of each interval meet the requirements according to the correlation coefficient between the fitting functions of each interval and the original monitoring curve; S4, retaining the interval fitting function that meets the requirements, dividing the interval data that does not meet the requirements into two parts, and then performing data fitting respectively; S5. Determine whether the secondary partition interval fitting function meets the requirements according to the correlation coefficient between the secondary partition interval fitting function and the original monitoring curve; S6. Repeat the above steps of interval division, function fitting and correlation coefficient analysis until the fitting functions of all intervals meet the requirements.
[0015] Preferably, the method for dividing the intervals in step S2 is: ; Among them, M is the total amount of monitoring data, n is the number of divided intervals, m is the amount of data in the first n-1 intervals, and s is the amount of data in the nth interval.
[0016] Preferably, the selection basis of each interval linear function or exponential function in step S2 is:
[0017] in, Step length l The acceleration between y i is any monitoring value in each interval, y i-l Step length l If the previous monitoring value appears 5 times in a row within a certain interval , then the exponential function is used to fit the data in this interval; otherwise, the linear function is used to fit the data.
[0018] Preferably, in step S2, n The linear function fitting method for -1 interval is: ; No. n The linear function fitting method for the interval is: ; in, a 0 ,a 1 All are linear functions The unknown constant, x i is any time value in the interval, y i is any monitoring value within the interval, j is any interval among the first n-1.
[0019] Preferably, in step S2, n The exponential function fitting method for -1 interval is: ; No. n The exponential function fitting method for the interval is: ; in, a 2 , a 3 Both are exponential functions The unknown constant, x mj For the j The last time value of the interval, y mj For the j The last monitored value of the interval, x M is the final time value, y M is the last monitored value.
[0020] Preferably, the correlation coefficient R between the interval fitting function in steps S3, S5 and S6 and the original monitoring curve is ≥ 0.80, indicating that the interval fitting function result meets the requirements.
[0021] Preferably, the method of repeatedly dividing the intervals in step S6 is: The interval data that does not meet the requirements is divided into intervals for the zth time, which will form 2z-1 intervals. The data volume of the first 2z-1-1 intervals is: ; The amount of data in the 2z-1th interval is: ; Among them, the symbol represents the floor function, z The number of times the interval is divided Taking the GNSS monitoring data of a deformed body in the reservoir area of Baihetan Hydropower Station in the lower reaches of Jinsha River from July 20, 2021 to February 1, 2024 as an example, according to the attached Figure 1The process of an adaptive geological disaster monitoring data smoothing filtering method shown is as follows: (1) Draw the original time series monitoring curve of the deformed body from July 20, 2021 to February 1, 2024 as follows: Figure 2 As shown, the monitoring curve has large fluctuations up and down; (2) The total amount of GNSS monitoring data M of the deformed body from July 20, 2021 to February 1, 2024 is determined to be 19699. The number of intervals divided for the first time, n, is 20. The amount of data m of the first 19 intervals is 1000, and the amount of data s of the 20th interval is 699.
[0022] (3) The basis for selecting each interval linear function or exponential function in step S2 is: ; The step length l is 10, and the 19th interval and the 20th interval appear 5 times in a row. Therefore, the exponential function is used for data fitting in the 19th and 20th intervals, and the linear function is used for data fitting in the remaining intervals. The monitoring curve of the first interval fitting of the deformed body is shown in the attached figure. Figure 3 As shown; (4) The fitting function parameters of each interval divided for the first time and the correlation coefficient between the fitting function and the original monitoring curve are as follows:
[0023] (5) The correlation coefficient R between the fitting function and the original monitoring curve of the 2nd, 8th, 11th, 12th, 13th, 14th, 15th, 17th, 19th and 20th intervals is less than 0.80, which does not meet the requirements and needs to be divided into intervals for the second time. The data volume of the 2-1 interval, 2-2 interval, 8-1 interval, 8-2 interval, 11-1 interval, 11-2 interval, 12-1 interval, 12-2 interval, 13-1 interval, 13-2 interval, 14-1 interval, 14-2 interval, 15-1 interval, 15-2 interval, 17-1 interval, 17-2 interval, 19-1 interval and 19-2 interval is 500, the data volume of the 20-1 interval is 349, and the data volume of the 20-2 interval is 350; (6) The step length l is 10, depending on whether it occurs 5 times in a row According to the rule judgment, linear function is used for data fitting in intervals 2-1, 2-2, 8-1, 8-2, 11-1, 11-2, 12-1, 12-2, 13-1, 13-2, 14-1, 14-2, 15-1, 15-2, 17-1, 17-2 and 19-1. Exponential function is used for data fitting in intervals 19-2, 20-1 and 20-2. The fitting function results of the other intervals remain unchanged. The monitoring curve of the second interval fitting of the deformed body is shown in the attached figure. Figure 4 As shown; (7) The fitting function parameters of each interval of the second division and the correlation coefficient between the fitting function and the original monitoring curve are as follows:
[0024] (8) The correlation coefficient R between the fitting function of the 2-2 interval, 19-2 interval, 20-1 interval and 20-2 interval and the original monitoring curve is less than 0.80, which does not meet the requirements and needs to be divided for the third time. The data volume of the 2-2-1 interval, 2-2-2 interval, 19-2-1 interval and 19-2-2 interval is 250, the data volume of the 20-1-1 interval is 174, and the data volume of the 20-1-2 interval, 20-2-1 interval and 20-2-2 interval is 175; (9) The step length l is 10, depending on whether it occurs 5 times in a row According to the rule judgment, linear function is used for data fitting in the 2-2-1 interval and the 2-2-2 interval, exponential function is used for data fitting in the 19-2-1 interval, the 19-2-2 interval, the 20-1-1 interval, the 20-1-2 interval, the 20-2-1 interval and the 20-2-2 interval, and the fitting function results of the other intervals remain unchanged. The monitoring curve of the second interval fitting of the deformed body is shown in the attached figure. Figure 5 As shown; (10) The fitting function parameters of each interval of the third division and the correlation coefficient between the fitting function and the original monitoring curve are as follows:
[0025] (11) The correlation coefficient R between the fitting function and the original monitoring curve in all intervals is ≥ 0.80. The smoothing filtering of the GNSS monitoring data of the deformed body is completed and can be used for subsequent monitoring and early warning analysis.
[0026] The above embodiments are only preferred technical solutions of the present invention and should not be regarded as limiting the present invention. The protection scope of the present invention shall be the technical solutions recorded in the claims, including equivalent replacement solutions of the technical features in the technical solutions recorded in the claims. That is, equivalent replacement improvements within this scope are also within the protection scope of the present invention.
Claims
1. An adaptive geological disaster monitoring data smoothing filtering method, characterized in that: The following steps are involved: S1. Obtain the monitoring data collected by the sensor and draw the original time series monitoring curve; S2, based on the adaptive smoothing filtering method, a linear function or an exponential function is used to fit the data between the partitions; S3. Determine whether the results of the fitting functions of each interval meet the requirements according to the correlation coefficient between the fitting functions of each interval and the original monitoring curve; S4, retaining the interval fitting function that meets the requirements, dividing the interval data that does not meet the requirements into two parts, and then performing data fitting respectively; S5. Determine whether the secondary partition interval fitting function meets the requirements according to the correlation coefficient between the secondary partition interval fitting function and the original monitoring curve; S6. Repeat the above steps of interval division, function fitting and correlation coefficient analysis until the fitting functions of all intervals meet the requirements.
2. According to claim 1, a method for adaptive smoothing and filtering of geological disaster monitoring data is characterized in that: The method of dividing the intervals in step S2 is: ; Among them, M is the total amount of monitoring data, n is the number of divided intervals, m is the amount of data in the first n-1 intervals, and s is the amount of data in the nth interval.
3. According to claim 1, a method for adaptive smoothing and filtering of geological disaster monitoring data is characterized in that: The basis for selecting each interval linear function or exponential function in step S2 is: in, Step length l The acceleration between y i is any monitoring value in each interval, y i-l Step length l If the previous monitoring value appears 5 times in a certain interval , then the exponential function is used to fit the data in this interval; otherwise, the linear function is used to fit the data.
4. According to claim 1, the method for adaptive smoothing and filtering of geological disaster monitoring data is characterized in that: In step S2, n The linear function fitting method for -1 interval is: ; No. n The linear function fitting method for the interval is: ; in, a 0, a 1 are all linear functions The unknown constant, x i is any time value in the interval, y i is any monitoring value within the interval, j is any interval among the first n-1.
5. According to claim 1, the method for adaptive smoothing and filtering of geological disaster monitoring data is characterized in that: In step S2, n The exponential function fitting method for -1 interval is: ; No. n The exponential function fitting method for the interval is: ; in, a 2, a 3 are all exponential functions The unknown constant, x mj For the j The last time value of the interval, y mj For the j The last monitoring value of the interval, x M is the final time value, y M is the last monitored value.
6. The method of adaptive smoothing and filtering of geological disaster monitoring data according to claim 1, characterized in that: The correlation coefficient R between the interval fitting function in steps S3, S5 and S6 and the original monitoring curve is ≥ 0.80, indicating that the interval fitting function result meets the requirements.
7. The method for adaptive smoothing and filtering of geological disaster monitoring data according to claim 1, characterized in that: The method of continuously repeating the interval division in step S6 is: The interval data that does not meet the requirements is divided into intervals for the zth time, which will form 2z-1 intervals. The data volume of the first 2z-1-1 intervals is: ; The amount of data in the 2z-1th interval is: ; Among them, the symbol represents the floor function, z is the number of times the interval is divided.
