Full-coverage slope safety monitoring and early warning method with active early warning function

Abstract

The invention belongs to the technical field of slope monitoring, and provides an active early-warning full-coverage slope safety monitoring and early-warning method, which comprises the following steps: according to the induction times and induction time of each induction factor in a plurality of historical periods, carrying out induction prediction analysis on abnormal alarms in a unit area in the current period, according to the prediction analysis result and in combination with the induction performance of the induction factors in a plurality of historical periods, whether sensors corresponding to the induction factors are arranged in the unit areas is judged, setting of sensor groups in the unit areas is completed, and by analyzing the prediction time point of slope abnormity alarm in the unit areas, the slope abnormity alarm is given out. According to the method, the occurrence types of slope abnormal alarms in the unit areas are identified and analyzed, uniform occurrence and non-uniform occurrence exist, then whether the trend of centralized occurrence exists or not is analyzed under the condition of non-uniform occurrence, and by predicting and analyzing the slope abnormal alarms in the unit areas, strategy optimization of slope abnormal monitoring and early warning is facilitated.

Description

Technical Field

[0001] The invention belongs to the technical field of slope monitoring, in particular to a full-coverage slope safety monitoring and early warning method with active early warning. Background Art

[0002] With the continuous advancement of technology, fiber optic sensor technology has been widely used in the field of slope safety monitoring. The integration of fiber optic sensing technology can further improve the safety of slope monitoring. However, in the existing technology, the application of fiber optic sensing technology in the field of slope safety monitoring is still insufficient. Therefore, it is of great significance to study an active warning full-coverage slope safety monitoring and early warning method.

[0003] In the prior art, when monitoring a fully covered slope, a variety of sensors are usually installed in the slope area to facilitate the collection of various slope data, such as soil temperature, moisture content, etc. However, in the prior art, there is a lack of targeted analysis of the slope area, that is, there is a lack of targeted analysis of various inducing factors when abnormal alarms occur in the slope area, so as to set up a reasonable sensor group by combining historical data and predictive analysis results, resulting in low monitoring efficiency. Secondly, after the sensor is set up, there is a lack of application of predictive analysis of abnormal alarms in the slope area to select the monitoring method.

[0004] To this end, the present invention provides a full-coverage slope safety monitoring and early warning method with active early warning. Summary of the invention

[0005] In order to make up for the deficiencies of the prior art, at least one technical problem raised in the background technology is solved.

[0006] The technical solution adopted by the present invention to solve the technical problem is as follows: In the first aspect, the present invention provides a full-coverage slope safety monitoring and early warning method with active early warning, comprising: Step 1: Divide the slope area into unit areas according to the monitoring range of the sensor, obtain the monitoring information of the unit areas in multiple historical periods through the slope monitoring report, process and analyze the monitoring information, and set an adaptive sensor group for the unit area according to the analysis results; Step 2: Process and analyze the predicted time points of the slope abnormality alarm in the unit area caused by various inducing factors corresponding to the sensor groups set in the unit area, and determine the predicted type of the slope abnormality alarm in the unit area according to the analysis results, wherein the prediction types include uniform occurrence type and non-uniform occurrence type, and the non-uniform occurrence type includes concentrated type and non-concentrated type; Step 3: Determine the monitoring method for the unit area according to the predicted type of slope abnormality alarm occurring in the unit area.

[0007] In a second aspect, the present invention provides a full coverage slope safety monitoring and early warning system with active early warning, comprising: Sensor configuration module: divide the slope area into unit areas according to the monitoring range of the sensor, obtain the monitoring information of the unit area in multiple historical periods through the slope monitoring report, process and analyze the monitoring information, and set the appropriate sensor group for the unit area according to the analysis results; Prediction and analysis module: Process and analyze the predicted time point of slope abnormality alarm in the unit area caused by various inducing factors corresponding to the sensor groups installed in the unit area, and determine the prediction type of slope abnormality alarm in the unit area according to the analysis results, where the prediction type includes uniform occurrence type and non-uniform occurrence type, and the non-uniform occurrence type includes concentrated type and non-concentrated type; Monitoring module: Determine the monitoring method of the unit area according to the predicted type of slope abnormality alarm occurring in the unit area.

[0008] The beneficial effects of the present invention are as follows: 1. The present invention processes and analyzes various inducing factors when abnormal alarm occurs in the unit area, and performs inducing prediction analysis on the abnormal alarm of the unit area in the current period according to the number of inducing factors and the inducing time in multiple historical periods. According to the prediction analysis results and in combination with the inducing performance of the inducing factors in multiple historical periods, it is determined whether to set sensors corresponding to the inducing factors in the unit area to complete the setting of the sensor group in the unit area.

[0009] 2. By analyzing the predicted time point of the slope abnormality alarm in the unit area, the occurrence type of the slope abnormality alarm in the unit area is identified and analyzed, and there are uniform occurrence and uneven occurrence. Secondly, under the condition of uneven occurrence, it is analyzed whether there is a concentrated occurrence trend. The present invention is beneficial to the optimization of the strategy of slope abnormality monitoring and early warning by predicting and analyzing the occurrence of slope abnormality alarm in the unit area. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The present invention will be further described below in conjunction with the accompanying drawings.

[0011] Figure 1 It is a flowchart of steps of a full-coverage slope safety monitoring and early warning method with active early warning according to an embodiment of the present invention; Figure 2 It is a flowchart of a full-coverage slope safety monitoring and early warning system with active early warning described in an embodiment of the present invention. DETAILED DESCRIPTION

[0012] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods. Example 1

[0013] like Figure 1 As shown, a full-coverage slope safety monitoring and early warning method with active early warning according to an embodiment of the present invention comprises the following steps: Step 1: Divide the slope area into unit areas according to the monitoring range of the sensor, obtain the monitoring information of the unit areas in multiple historical periods through the slope monitoring report, process and analyze the monitoring information, and set an adaptive sensor group for the unit area according to the analysis results; Among them, sensors include but are not limited to rain gauges, soil moisture sensors, etc.; Specifically, among all sensors, the monitoring ranges of each sensor are compared, the minimum monitoring range of the sensor is extracted, and the slope area is divided into several unit areas of equal area according to the minimum monitoring range of the sensor. The division restriction conditions are: the minimum monitoring range of the sensor is greater than or equal to the area range of the unit area and the monitoring range of the sensor is sufficient to cover the unit area; Obtain monitoring information of the unit area in multiple historical periods through the slope monitoring report, wherein the monitoring information includes the inducing factors when the slope abnormality alarm occurs in the unit area, wherein the inducing factors include but are not limited to factors such as rainfall, soil moisture and soil temperature; Based on any one of the predisposing factors in any unit area; The number of times the same inducing factor causes abnormal slope alarms in the unit area is counted, and the ratio is processed with the total number of abnormal slope alarms in the unit area to obtain the inducing number ratio, which is marked as YF; Through the slope monitoring report, the time point when the slope abnormality alarm occurs in the unit area due to the same inducing factor is obtained, and the predicted time point when the slope abnormality alarm occurs in the unit area due to the same inducing factor is determined by processing and analysis, specifically: The time point when the same inducing factor causes the abnormal slope alarm in the unit area is marked as the inducing time point, and the inducing time points are integrated to obtain the inducing time series; The induced time series is processed and analyzed using the rolling average method to determine whether the induced time series is stable, including: S1, acquisition of evoked time series; S2, set the rolling window, the rolling window is a historical period; S3, as the rolling window rolls, the mean of the data in the rolling window at each roll, i.e., the rolling average value, is calculated; S4, integrating all the rolling averages obtained to obtain a rolling average sequence; S5, extracting every two adjacent rolling averages in the rolling average sequence, integrating them into a rolling adjacent group, and calculating the change rate of the rolling adjacent group; For example, assume that the rolling average sequence is: (A1, A2, A3...A n ), where A n represents the rolling mean, n represents the number of rolling mean values, and each two adjacent rolling mean values ​​are extracted and integrated into a rolling adjacent group, i.e. (A1, A2), (A2, A3), (A3, A4)... (A n-1 ,A n ); Calculate the rate of change of the rolling adjacent group through the rate of change formula: , where BH is the rate of change; S6, comparing the obtained change rate with a preset change rate; If the change rate is greater than or equal to the preset change rate, the rolling adjacent group is marked as an abnormal rolling adjacent group; If the change rate is less than the preset change rate, the scrolling adjacent group is marked as a normal scrolling adjacent group; Count the number of normal scrolling adjacent groups and compare it with the preset number threshold; If the number of normal rolling adjacent groups is greater than or equal to the preset number threshold, it means that the induced time series is stable; If the number of normal rolling adjacent groups is less than the preset number threshold, it indicates that the induced time series is unstable; It should be noted that the preset number threshold is set by those skilled in the art. In judging the stability of a sequence, the stability of the sequence is usually judged based on the number of rolling adjacent groups whose change rate is lower than the preset change rate. If the induced time series is stable, the interval between every two adjacent induced time points in the induced time series is obtained, and the sum is taken to obtain the average to obtain the predicted interval time. The induced time series is expanded according to the predicted interval time to obtain the expanded induced time series. In the expanded induced time series, the expanded induced time point is the predicted time point at which the same induced factor causes the slope abnormality alarm in the unit area. For example, assume that the induced time series is (SJ1, SJ2, S J3 ......SJ i ), the expanded induced time series is (SJ1, SJ2, SJ3...SJ i , SJ i +c, SJ i +2c, SJ i +3c......), where i represents the number of historical cycles and c represents the length of the prediction interval; If the induced time series is unstable, the random forest method is used to expand the induced time series, specifically: A1, using feature extraction technology to extract common features of induced time series, including but not limited to lag features and moving averages; A2, smoothing the induced time series using a difference algorithm; A3, 70% of the induced time data in the processed induced time series are used as the training set, and 30% of the induced time data are used as the test set. The training set and the test set are used to train and test the random forest model to obtain the random forest model after training; A4, predict the induced time series through the trained random forest model to obtain the expanded induced time series; In the expanded induced time series, the expanded induced time point is the predicted time point when the same induced factor causes the slope abnormality alarm in the unit area; The predicted time point of slope abnormality alarm in the unit area caused by the same inducing factor is compared with the current cycle, and the number of predicted time points of slope abnormality alarm in the unit area caused by the same inducing factor in the current cycle is counted, and the ratio is processed with the highest number of slope abnormality alarms in the unit area in the historical cycle to obtain the alarm prediction number ratio, which is marked as CS; For example, assuming that the current cycle is January, the predicted time points at which the inducing factors in the current cycle cause the unit area to have an abnormal slope alarm are January 2, January 18, and January 28. The number of predicted time points at which the inducing factors in the current cycle cause the unit area to have an abnormal slope alarm is 3. Based on the historical cycle, the highest number of abnormal slope alarms in the unit area in the historical cycle is 5, and the alarm prediction number ratio is 3 / 5. According to the predicted time point of the slope abnormality alarm in the unit area caused by the same inducing factor and the mid-term time point of the current cycle, the predicted time deviation is calculated, and all the predicted time deviations are summed and averaged to obtain the mean of the predicted time deviation. The mean of the predicted time deviation is ratioed with the duration corresponding to the current cycle to obtain the alarm deviation value, which is marked as PL; The mid-term time point of the current cycle is the mid-term time point of the current cycle period. For example, the mid-term time point of the 1st to 30th is the 15th, and the mid-term time point of the 1st to 31st is 15.5; For example, assuming that the current cycle is January, the predicted time points for the occurrence of slope abnormality alarms in the unit area due to the same inducing factor in the current cycle are January 2, January 18, and January 28, and the mid-term time point is the 15th, then the predicted time deviations are 13, 3, and 13 respectively; The obtained induced frequency ratio YF, alarm prediction frequency ratio CS and alarm deviation value PL are processed by the formula: The alarm prediction value BJY is obtained, where s1, s2 and s3 are all preset proportional coefficients, s1 is 1.14, s2 is 1.26, and s3 is 1.47; It should be noted that the alarm prediction value BJY means: the alarm prediction value BJY is calculated by the induction frequency ratio YF, the alarm prediction frequency ratio CS and the alarm deviation value PL. The induction frequency ratio YF reflects the number of slope abnormality alarms in the unit area caused by the same inducing factor in the historical period. The alarm prediction frequency ratio CS reflects the time point of the slope abnormality alarm in the unit area caused by the same inducing factor in the historical period. The prediction analysis shows that in the current period, the same inducing factor will cause the predicted number of slope abnormality alarms in the unit area. The alarm deviation value PL reflects the position of the predicted time point of the slope abnormality alarm in the unit area caused by the same inducing factor in the time dimension of the current access period. For example, in the historical period, the same inducing factor causes a large number of slope abnormality alarms in the unit area, the same inducing factor causes a large number of predicted times of slope abnormality alarms in the unit area, and the predicted points of the slope abnormality alarms in the unit area caused by the same inducing factor are basically in the middle stage of the current period, which means that in the current period, the same inducing factor will cause a high risk of slope abnormality alarms in the unit area, and the corresponding sensors of the inducing factors are set in the unit area. comparing the alarm prediction value with the alarm prediction threshold; If the alarm prediction value is greater than or equal to the alarm prediction threshold, a sensor corresponding to the same inducing factor is set in the unit area. For example, if the same inducing factor is soil moisture, a soil moisture sensor is set in the unit area. If the alarm prediction value is less than the alarm prediction threshold, there is no need to set up sensors corresponding to the same inducing factor in the unit area; Repeat the above operation, after processing and analyzing various inducing factors in each unit area, integrate the sensors set in each unit area into a sensor group, and complete the adaptation setting of the sensor group in each unit area; The technical solution described in the embodiment of the present invention is: the present invention processes and analyzes the various inducing factors when the unit area abnormal alarm occurs, and according to the number of induction times and induction times of each inducing factor in multiple historical periods, performs inducing prediction analysis on the unit area abnormal alarm in the current period, and according to the prediction analysis results and in combination with the inducing performance of the inducing factors in multiple historical periods, determines whether to set sensors corresponding to the inducing factors in the unit area, and completes the setting of the sensor group in the unit area. Example 2

[0014] like Figure 1As shown, based on Example 1, a full-coverage slope safety monitoring and early warning method with active early warning described in an embodiment of the present invention includes the following steps: Step 2: Process and analyze the predicted time points of slope abnormality alarms in the unit area caused by various inducing factors corresponding to the sensor groups set in the unit area, and determine the predicted type of slope abnormality alarms in the unit area according to the analysis results, wherein the prediction types include uniform occurrence type and non-uniform occurrence type, wherein the non-uniform occurrence type covers concentrated type and non-concentrated type; Specifically, the predicted time point at which various inducing factors corresponding to the sensor groups set in the unit area cause the unit area to have a slope abnormality alarm is marked as an alarm prediction point; In the current cycle, the alarm prediction points are marked, the current cycle is divided into a number of continuous sub-unit cycles of equal length, the number of alarm prediction points in the current cycle is counted and the ratio is processed with the number of divided sub-unit cycles to obtain the average value of the alarm prediction points; Based on any subunit cycle; Count the number of alarm prediction points within the sub-unit cycle and compare and analyze them with the average value of alarm prediction points. Specifically: If the number of alarm prediction points in a sub-unit period is not equal to the mean value of alarm prediction points, the sub-unit period is marked as an unequal sub-period; If the number of alarm prediction points in a sub-unit period is equal to the mean value of alarm prediction points, the sub-unit period is marked as an equal sub-period; Count the number of equal sub-periods and compare them with the number of sub-unit periods to obtain the equal period ratio, which is marked as XD; Based on the unequal sub-periods, the number of alarm prediction points in each unequal sub-period is processed with the mean value of the alarm prediction points to obtain the number deviation of the alarm prediction points. The number deviations of all alarm prediction points are integrated into a number deviation group (PC1, PC2, PC3...PC j ), where PC j It represents the number deviation of alarm prediction points, and j represents the number of unequal sub-periods; The variance of the quantity deviation group is calculated using the formula: Get the variance value FC of the quantity deviation group, where PC 均 represents the mean of the quantitative deviation group; The equal period ratio XD and the variance value FC of the quantity deviation group are processed on Xining data through the formula: The predicted average value JY is obtained, where z1 and z2 are both preset proportional coefficients, z1 is 1.416, and z2 is 1.013; In some embodiments, the predicted average value JY is compared with a predicted average threshold value; If the predicted uniform value JY is greater than or equal to the predicted uniform threshold, it means that the prediction type of slope abnormality alarm in the unit area is uniform occurrence type; If the predicted uniform value JY is less than the predicted uniform threshold, it means that the predicted type of slope abnormality alarm in the unit area is non-uniform occurrence type; Based on the prediction type of slope abnormality alarm in the unit area being a non-uniform occurrence type, in the current cycle, the alarm prediction points in the current cycle are sorted in chronological order, and the first ranked alarm prediction point, which appears earliest, is subjected to difference processing with the start time point of the current cycle, and the absolute value of the difference is taken to obtain the start time deviation; the last ranked alarm prediction point, which appears last, is subjected to difference processing with the end time point of the current cycle, and the absolute value of the difference is taken to obtain the end time deviation, and the start time deviation and the end time deviation are summed to obtain the total time deviation, and the total time deviation is subjected to ratio processing with the time corresponding to the current cycle to obtain the time deviation ratio, which is marked as SP; Based on the alarm prediction point, the interval time between every two adjacent alarm prediction points is obtained, and the sum is taken to get the average of the interval time, and the average of the interval time is processed by ratio processing with the time corresponding to the current cycle to obtain the interval performance value, which is marked as JG; By formula: The concentrated performance value JZc is obtained, where a1 and a2 are both preset proportional coefficients, a1 is 1.314, and a2 is 1.286; It should be noted that the concentration performance value JZc is obtained through the time deviation ratio SP and the interval performance value JG. The concentration performance value JZc reflects whether the alarm prediction points are concentrated in terms of the occurrence time, while the interval performance value JG reflects the average interval time between adjacent alarm prediction points. If the alarm prediction points are concentrated in terms of the occurrence time and the average interval time between adjacent alarm prediction points is short, it means that the alarm prediction points are concentrated in the state of occurrence, which essentially reflects that in the current cycle, the occurrence of slope abnormal alarms in the unit area will appear in a concentrated manner, which is conducive to providing an optimization basis for the monitoring of slope abnormal alarms in subsequent unit areas based on the analysis of the alarm prediction points; In some embodiments, the concentration performance value is compared to a concentration performance threshold; If the concentration performance value is greater than or equal to the concentration performance threshold, it means that the prediction type of slope abnormality alarm in the unit area is concentrated; If the concentration performance value is less than the concentration performance threshold, it means that the prediction type of slope abnormality alarm in the unit area is non-concentrated; Step 3: Determine the monitoring method of the unit area according to the predicted type of slope abnormality alarm occurring in the unit area; Specifically, if the prediction type of the slope abnormality alarm occurring in the unit area is a uniform occurrence type, the unit area is monitored in real time during the current cycle; If the prediction type of slope abnormality alarm in the unit area is non-centralized, the unit area is monitored in real time during the current cycle; If the prediction type of slope abnormality alarm in the unit area is concentrated, then in the current cycle, the difference between the first ranked alarm prediction point, i.e. the earliest alarm prediction point, and the last ranked alarm prediction point, i.e. the last alarm prediction point, is processed to obtain the monitoring period, and the unit area is monitored in real time during the monitoring period; When conducting real-time monitoring of a unit area, if the data monitored by any sensor installed in the unit area reaches the warning threshold, a warning signal is generated and sent to the slope safety supervision center.

[0015] The technical solution of the embodiment of the present invention is: by analyzing the predicted time point of the slope abnormality alarm occurring in the unit area, the occurrence type of the slope abnormality alarm occurring in the unit area is identified and analyzed, and there are uniform occurrence and uneven occurrence. Secondly, under the condition of uneven occurrence, it is analyzed whether there is a trend of concentrated occurrence. The present invention is beneficial to the optimization of the strategy of slope abnormality monitoring and early warning by predicting and analyzing the occurrence of slope abnormality alarm in the unit area. Example 3

[0016] like Figure 2 As shown, an active warning full coverage slope safety monitoring and warning system according to an embodiment of the present invention includes: Sensor configuration module: divide the slope area into unit areas according to the monitoring range of the sensor, obtain the monitoring information of the unit area in multiple historical periods through the slope monitoring report, process and analyze the monitoring information, and set the appropriate sensor group for the unit area according to the analysis results; Prediction and analysis module: Process and analyze the predicted time point of slope abnormality alarm in the unit area caused by various inducing factors corresponding to the sensor groups installed in the unit area, and determine the prediction type of slope abnormality alarm in the unit area according to the analysis results, where the prediction type includes uniform occurrence type and non-uniform occurrence type, and the non-uniform occurrence type includes concentrated type and non-concentrated type; Monitoring module: Determine the monitoring method of the unit area according to the predicted type of slope abnormality alarm occurring in the unit area.

[0017] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A full-coverage slope safety monitoring and early warning method with active early warning, characterized in that: include: Step 1: Divide the slope area into unit areas according to the monitoring range of the sensor, obtain the monitoring information of the unit areas in multiple historical periods through the slope monitoring report, process and analyze the monitoring information, and set an adaptive sensor group for the unit area according to the analysis results; Step 2: Process and analyze the predicted time points of the slope abnormality alarm in the unit area caused by various inducing factors corresponding to the sensor groups set in the unit area, and determine the predicted type of the slope abnormality alarm in the unit area according to the analysis results, wherein the prediction types include uniform occurrence type and non-uniform occurrence type, and the non-uniform occurrence type includes concentrated type and non-concentrated type; Step 3: Determine the monitoring method for the unit area according to the predicted type of slope abnormality alarm occurring in the unit area.

2. The active early warning full coverage slope safety monitoring and early warning method according to claim 1 is characterized by: The process of dividing the slope area into unit areas according to the monitoring range of the sensor is as follows: among all sensors, the monitoring range of each sensor is compared, the minimum monitoring range of the sensor is extracted, and the slope area is divided into a number of unit areas of equal area according to the minimum monitoring range of the sensor. The division restriction conditions are: the minimum monitoring range of the sensor is greater than or equal to the area range of the unit area and the monitoring range of the sensor is sufficient to cover the unit area.

3. The active early warning full coverage slope safety monitoring and early warning method according to claim 1 is characterized by: The monitoring information includes the inducing factors when the slope abnormality alarm occurs in the unit area; The process of setting the sensor group adapted to the unit area is as follows: Based on any one of the predisposing factors in any unit area; The number of times the same inducing factor causes abnormal slope alarms in the unit area is counted, and the ratio is processed with the total number of abnormal slope alarms in the unit area to obtain the inducing number ratio, which is marked as YF; Through the slope monitoring report, the time point when the slope abnormality alarm occurs in the unit area due to the same inducing factor is obtained, and the predicted time point when the slope abnormality alarm occurs in the unit area due to the same inducing factor is determined by processing and analysis; The predicted time point of slope abnormality alarm in the unit area caused by the same inducing factor is combined with the current cycle for comparison and analysis to obtain the alarm prediction frequency ratio CS and the alarm deviation value PL; By formula: The alarm prediction value BJY is obtained, where s1, s2 and s3 are all preset proportional coefficients, s1 is 1.14, s2 is 1.26, and s3 is 1.47; If the alarm prediction value is greater than or equal to the alarm prediction threshold, a sensor corresponding to the same inducing factor is set in the unit area, and the operation is repeated. After the processing and analysis of various inducing factors in each unit area are completed, the sensors set in each unit area are integrated into a sensor group, and the adaptation setting of the sensor group in each unit area is completed.

4. The active early warning full coverage slope safety monitoring and early warning method according to claim 3 is characterized by: The process of determining the predicted time point when the same inducing factor causes the occurrence of slope abnormality alarm in the unit area is as follows: The time point when the same inducing factor causes the abnormal slope alarm in the unit area is marked as the inducing time point, and the inducing time points are integrated to obtain the inducing time series; The induced time series is processed and analyzed using the rolling average method to determine whether the induced time series is stable; If the induced time series is stable, the interval between every two adjacent induced time points in the induced time series is obtained, and the sum is taken to obtain the average to obtain the predicted interval time. The induced time series is expanded according to the predicted interval time to obtain the expanded induced time series. In the expanded induced time series, the expanded induced time point is the predicted time point at which the same induced factor causes the slope abnormality alarm in the unit area. If the induced time series is unstable, the random forest method is used to expand the induced time series. In the expanded induced time series, the expanded induced time point is the predicted time point when the same inducing factor causes the slope abnormality alarm in the unit area.

5. The active early warning full coverage slope safety monitoring and early warning method according to claim 4 is characterized by: The judgment process of whether the induced time series is stable is: S1, acquisition of evoked time series; S2, set the rolling window, the rolling window is a historical period; S3, as the rolling window rolls, the mean of the data in the rolling window at each roll, i.e., the rolling average value, is calculated; S4, integrating all the rolling averages obtained to obtain a rolling average sequence; S5, extracting every two adjacent rolling averages in the rolling average sequence, integrating them into a rolling adjacent group, and calculating the change rate of the rolling adjacent group; S6, comparing the obtained change rate with a preset change rate; if the change rate is less than the preset change rate, marking the scrolling adjacent group as a normal scrolling adjacent group; Count the number of normal scrolling adjacent groups and compare it with the preset number threshold; If the number of normal rolling adjacent groups is greater than or equal to the preset number threshold, it means that the induced time series is stable; If the number of normal scrolling adjacent groups is less than a preset number threshold, it indicates that time series instability is induced.

6. The active early warning full coverage slope safety monitoring and early warning method according to claim 4 is characterized by: The alarm prediction frequency ratio is obtained as follows: The predicted time point of slope abnormality alarm in the unit area caused by the same inducing factor is compared with the current cycle, and the number of predicted time points of slope abnormality alarm in the unit area caused by the same inducing factor in the current cycle is counted, and the ratio is processed with the highest number of slope abnormality alarms in the unit area in the historical cycle to obtain the alarm prediction number ratio, which is marked as CS; The alarm deviation value PL is obtained in the following manner: The predicted time deviation is calculated based on the predicted time point when the slope abnormality alarm occurs in the unit area due to the same inducing factor and the mid-term time point of the current cycle. The mean of all predicted time deviations is summed and averaged to obtain the mean of the predicted time deviation. The mean of the predicted time deviation is ratioed with the duration corresponding to the current cycle to obtain the alarm deviation value, which is marked as PL.

7. The active early warning full coverage slope safety monitoring and early warning method according to claim 1 is characterized by: The determination process of the uniform appearance type and the non-uniform appearance type is as follows: Mark the predicted time point at which the slope abnormality alarm occurs in the unit area due to various inducing factors corresponding to the sensor group set in the unit area as the alarm prediction point; In the current cycle, the current cycle is divided into a number of continuous sub-unit cycles of equal length, the number of alarm prediction points in the current cycle is counted and the ratio is processed with the number of divided sub-unit cycles to obtain the average value of the alarm prediction points; Based on any subunit cycle; The number of alarm prediction points in the sub-unit cycle is counted and compared with the mean value of the alarm prediction points. If the number of alarm prediction points in the sub-unit cycle is not equal to the mean value of the alarm prediction points, the sub-unit cycle is marked as an unequal sub-cycle; if the number of alarm prediction points in the sub-unit cycle is equal to the mean value of the alarm prediction points, the sub-unit cycle is marked as an equal sub-cycle. The equal sub-periods and unequal sub-periods are processed and analyzed to obtain the variance value FC of the equal period ratio XD and the quantity deviation group. For Xining data processing, the formula is: The predicted average value JY is obtained, where z1 and z2 are both preset proportional coefficients, z1 is 1.416, and z2 is 1.013; If the predicted uniform value JY is greater than or equal to the predicted uniform threshold, it means that the prediction type of slope abnormality alarm in the unit area is uniform occurrence type; If the predicted uniform value JY is less than the predicted uniform threshold, it means that the predicted type of slope abnormality alarm occurring in the unit area is non-uniform occurrence type.

8. The active early warning full coverage slope safety monitoring and early warning method according to claim 7 is characterized by: The equal period ratio XD and the variance value FC of the number deviation group are obtained by counting the number of equal sub-periods and performing ratio processing with the number of sub-unit periods to obtain the equal period ratio, which is marked as XD; Based on unequal sub-periods, the number of alarm prediction points in each unequal sub-period is subtracted from the mean of the alarm prediction points to obtain the quantity deviation of the alarm prediction points. The quantity deviations of all alarm prediction points are integrated into a quantity deviation group, and the variance of the quantity deviation group is calculated to obtain the variance value FC of the quantity deviation group.

9. The active early warning full coverage slope safety monitoring and early warning method according to claim 5 is characterized by: The determination process of the centralized type and the decentralized type is as follows: In the current cycle, the alarm prediction points in the current cycle are sorted in chronological order, the difference between the first-ranked alarm prediction point and the start time point of the current cycle is processed, and the absolute value of the difference is taken to obtain the start time deviation; the difference between the alarm prediction point at the bottom of the ranking and the end time point of the current cycle is processed, and the absolute value of the difference is taken to obtain the end time deviation, the start time deviation and the end time deviation are summed to obtain the total time deviation, the total time deviation is processed by ratio with the time corresponding to the current cycle to obtain the time deviation ratio, and marked as SP; Based on the alarm prediction point, the interval time between every two adjacent alarm prediction points is obtained, and the sum is taken to get the average of the interval time, and the average of the interval time is processed by ratio processing with the time corresponding to the current cycle to obtain the interval performance value, which is marked as JG; By formula: The concentrated performance value JZc is obtained, where a1 and a2 are both preset proportional coefficients, a1 is 1.314, and a2 is 1.286; If the concentration performance value is greater than or equal to the concentration performance threshold, it means that the prediction type of slope abnormality alarm in the unit area is concentrated; If the concentration performance value is less than the concentration performance threshold, it means that the prediction type of slope abnormality alarm occurring in the unit area is non-concentrated.

10. The active early warning full coverage slope safety monitoring and early warning method according to claim 1 is characterized by: The process of determining the monitoring method of the unit area is as follows: If the prediction type of slope abnormality alarm in the unit area is uniform occurrence, the unit area is monitored in real time during the current cycle; If the prediction type of slope abnormality alarm in the unit area is non-centralized, the unit area is monitored in real time during the current cycle; If the prediction type of slope abnormality alarm in the unit area is concentrated, then in the current cycle, the difference between the first ranked alarm prediction point and the last ranked alarm prediction point is processed to obtain the monitoring period, and the unit area is monitored in real time during the monitoring period.

Images