Slope site safety monitoring method and system

Abstract

The invention relates to the technical field of electric digital data processing, and provides a side slope field safety monitoring method and system, and the method comprises the following steps: L1, obtaining n paths of monitoring levels from n monitoring electrodes, and enabling the monitoring electrodes to be installed at n point positions of a side slope and a side slope support; l2, respectively carrying out FFT conversion on the n paths of monitoring levels to obtain n frequency lists; l3, inputting the n frequency lists into a preset probability graph model to calculate a risk probability value p; l4, judging whether the risk probability value p exceeds a preset upper limit value x or not, if so, entering the next step, and if not, returning to the step L1; and L5, sending a warning signal to an upper system. According to the method, a probabilistic graph model calculation mode is adopted after multi-path parallel FFT transformation, only two chips are used for core calculation, both the calculation amount and the data transmission amount are in an extremely low level, and meanwhile high early warning accuracy can be ensured.

Description

Technical Field

[0001] This invention provides a method and system for on-site safety monitoring of slopes, relating to the field of electrical digital data processing technology, and specifically to the field of force analysis or optimization technology. Background Technology

[0002] In terms of slope stability monitoring, for soil slopes or rock slopes, the main focus is on monitoring their slippage. For example, Chinese Patent Application No. CN202410454123.6 discloses a method, system and early warning method for predicting slope slippage. In this type of scheme, the electrodes used for monitoring mainly reflect the resistivity of the soil. After obtaining the resistance value of the electrodes, the system performs time-series analysis and calculation to achieve early warning.

[0003] However, rock slopes differ significantly in that their strata contain little or no soil. The main risk comes from damage caused by vibration between rocks. Therefore, safety monitoring of rock slopes primarily involves monitoring and analyzing vibration. Vibration electrodes (i.e., vibration electrode probes, typically metal wires with a diameter of 10–200 micrometers, commonly made of platinum alloy or stainless steel, controlled by a piezoelectric vibrator or piezoelectric ceramic connected to the tail end to measure potential changes in the area and convert them into current signals) can be used to detect vibration. The data generated by the vibration electrodes is the potential change produced by the electrodes, which manifests as a circuit signal of fluctuating levels. Both its form and underlying physical meaning are fundamentally different from the electrodes used to monitor soil conditions in existing technologies for monitoring soil or rock-soil slopes. The signals from vibration electrodes cannot be processed using existing methods for handling resistance changes, nor can they be analyzed using existing resistance-based monitoring and analysis methods.

[0004] On the other hand, rock slopes are often located in remote areas with poor communication and a lack of local power generation (it is difficult to deploy wind power equipment nearby, and there is no condition for large-scale deployment of photovoltaic power generation equipment). Therefore, it is difficult to use large-scale data transmission or high-performance edge computing solutions for monitoring rock slopes. The signal processing and analysis of vibration electrodes must minimize the amount of computation (to ensure low power consumption), while at the same time pursuing the highest possible early warning accuracy and the least amount of remote data transmission. Existing technologies are difficult to meet such stringent requirements. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a method and system for on-site slope safety monitoring. This method and system is based on multi-channel parallel FFT transformation followed by probabilistic graphical model calculation. Only two chips are used for core computation, resulting in extremely low computational and data transmission volumes while ensuring a high accuracy rate for early warning.

[0006] The present invention is achieved through the following technical solutions.

[0007] The present invention provides a method for on-site safety monitoring of slopes, comprising the following steps: L1. Obtain n monitoring levels from n monitoring electrodes, which are installed at n points on the slope and its support. L2. Perform FFT transformation on each of the n monitoring levels to obtain a list of n frequencies; L3. Input the list of n frequencies into the preset probability graph model to calculate the risk probability value p; L4. Determine whether the risk probability value p exceeds the preset upper limit value x. If it does, proceed to the next step; otherwise, return to step L1. L5. Send a warning signal to the upper-level system.

[0008] The value n ranges from 10 to 30.

[0009] After step L3, the risk probability value p is further filtered.

[0010] The filtering process employs the Kalman filtering method.

[0011] The probabilistic graphical model employs a Markov network.

[0012] The preset upper limit value x ranges from 0.7 to 0.95.

[0013] The present invention also provides a slope on-site safety monitoring system, comprising: An on-site FPGA is used to implement n independent FFT transformation circuits. The input of each FFT transformation circuit is connected to a monitoring electrode, and the outputs of the n FFT transformation circuits are all connected to the on-site main control to execute steps L1 and L2 in the slope on-site safety monitoring method as described in claim 1. The on-site main control unit implements the probabilistic graphical model and uses the data from the on-site FPGA as the input of the probabilistic graphical model to execute steps L3 and L4 in the slope on-site safety monitoring method as described in claim 1. The signal transmitting unit executes step L5 in the slope field safety monitoring method as described in claim 1.

[0014] The on-site main control also implements a Kalman filter to perform filtering processing on the risk probability value p in the slope on-site safety monitoring method as described in claim 3.

[0015] The field controller uses a data stream approach, directly forwarding the data from the field FPGA as a data stream input to the probabilistic graphical model.

[0016] The outputs of the n-channel FFT transformation circuit are connected to one bus interface of the field master controller via bus communication.

[0017] The beneficial effects of this invention are as follows: based on the multi-path parallel FFT transformation and the use of probabilistic graphical model for calculation, only two core operation chips are used, so both the amount of computation and the amount of data transmission are kept at a very low level, while ensuring a high early warning accuracy. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating at least one embodiment of the present invention; Figure 2 This is a schematic diagram of connection and data stream transmission in at least one embodiment of the present invention. Detailed Implementation

[0019] The technical solution of the present invention is further described below, but the scope of protection is not limited to what is described.

[0020] The first embodiment of the present invention relates to, for example Figure 1 The method for on-site safety monitoring of slopes, as shown, includes the following steps: L1. Obtain n monitoring levels from n monitoring electrodes, which are installed at n points on the slope and its support. L2. Perform FFT transformation on each of the n monitoring levels (generally, take the monitoring level values ​​within 1 second and combine them in time sequence before performing FFT transformation) to obtain n frequency lists, such as L1{a11,a12,……,a1m}, L2{a21,a22,……,a2m},……,Ln{an1,an2,……,anm}; L3. Input the list of n frequencies into the preset probability graph model to calculate the risk probability value p; L4. Determine whether the risk probability value p exceeds the preset upper limit value x. If it does, proceed to the next step; otherwise, return to step L1. L5. Send a warning signal to the upper-level system.

[0021] In this embodiment, an undirected graphical model is preferred for the probabilistic graphical model. When an undirected graphical model is used, a globally fully connected Markov random field can be used for training and parameter tuning with historical data collected from other slope monitoring electrodes. This approach is simple, straightforward, and effective, and can basically meet the monitoring needs. However, from an accuracy perspective, the results of the undirected graphical model are less accurate, and the directed graphical model performs better. In this case, targeted adjustments are needed based on the specific slope engineering site conditions and the installation points of the n monitoring electrodes. Moreover, when using a directed graphical model, historical data collected from other slope monitoring electrodes cannot be used for training and parameter tuning. Developers need to be stationed on-site to adjust the model parameters for actual conditions. Therefore, although the accuracy is higher, it is generally not recommended due to its high cost.

[0022] Therefore, the essence of this implementation method is to directly perform FFT transformation on the multi-channel monitoring levels of multiple monitoring electrodes (generally implemented using FPGA), take multiple spectra and use a probabilistic graphical model to perform brute-force solution (generally implemented using MCU), which makes it easy to implement edge computing on the field control terminal, and the early warning accuracy is greatly improved compared with the single-value judgment method.

[0023] The second embodiment of the present invention is largely equivalent to the first embodiment, mainly in the further optimization of details, and the value of n ranges from 10 to 30.

[0024] Furthermore, after step L3, the risk probability value p is also filtered.

[0025] Furthermore, the filtering process employs the Kalman filtering method.

[0026] Furthermore, the probabilistic graphical model employs a Markov network.

[0027] Furthermore, the preset upper limit value x ranges from 0.7 to 0.95.

[0028] The third embodiment of the present invention relates to, for example Figure 2 The slope field safety monitoring system shown includes: An on-site FPGA is used to implement n independent FFT transformation circuits. The input of each FFT transformation circuit is connected to a monitoring electrode, and the outputs of the n FFT transformation circuits are all connected to the on-site main control to execute steps L1 and L2 in the slope on-site safety monitoring method as described in claim 1. The on-site main control unit implements the probabilistic graphical model and uses the data from the on-site FPGA as the input of the probabilistic graphical model to execute steps L3 and L4 in the slope on-site safety monitoring method as described in claim 1. The signal transmitting unit executes step L5 in the slope field safety monitoring method as described in claim 1.

[0029] The fourth embodiment of the present invention is largely equivalent to the third embodiment, mainly in the further optimization of details. The on-site main control also implements a Kalman filter to perform filtering processing on the risk probability value p in the slope on-site safety monitoring method as described in claim 3.

[0030] Furthermore, the field controller uses a data stream approach, directly forwarding the data from the field FPGA as a data stream input to the probabilistic graphical model.

[0031] Furthermore, the outputs of the n-channel FFT transformation circuit are connected to one bus interface of the field controller via bus communication.

[0032] Those skilled in the art will understand that the above embodiments are specific examples of implementing the present invention, and in practical applications, various changes in form and detail may be made without departing from the spirit and scope of the present invention.

Claims

1. A method for on-site safety monitoring of slopes, characterized in that, Includes the following steps: L1. Obtain n monitoring levels from n monitoring electrodes, which are installed at n points on the slope and its support. L2. Perform FFT transformation on each of the n monitoring levels to obtain a list of n frequencies; L3. Input the list of n frequencies into the preset probability graph model to calculate the risk probability value p; L4. Determine whether the risk probability value p exceeds the preset upper limit value x. If it does, proceed to the next step; otherwise, return to step L1. L5. Send a warning signal to the upper-level system.

2. The slope on-site safety monitoring method as described in claim 1, characterized in that, The value n ranges from 10 to 30.

3. The slope on-site safety monitoring method as described in claim 1, characterized in that, After step L3, the risk probability value p is further filtered.

4. The slope on-site safety monitoring method as described in claim 1, characterized in that, The filtering process employs the Kalman filtering method.

5. The slope on-site safety monitoring method as described in claim 1, characterized in that, The probabilistic graphical model employs a Markov network.

6. The slope on-site safety monitoring method as described in claim 1, characterized in that, The preset upper limit value x ranges from 0.7 to 0.

95.

7. A slope on-site safety monitoring system, characterized in that, include: An on-site FPGA is used to implement n independent FFT transformation circuits. The input of each FFT transformation circuit is connected to a monitoring electrode, and the outputs of the n FFT transformation circuits are all connected to the on-site main control to execute steps L1 and L2 in the slope on-site safety monitoring method as described in claim 1. The on-site main control unit implements the probabilistic graphical model and uses the data from the on-site FPGA as the input of the probabilistic graphical model to execute steps L3 and L4 in the slope on-site safety monitoring method as described in claim 1. The signal transmitting unit executes step L5 in the slope field safety monitoring method as described in claim 1.

8. The slope on-site safety monitoring system as described in claim 7, characterized in that, The on-site main control also implements a Kalman filter to perform filtering processing on the risk probability value p in the slope on-site safety monitoring method as described in claim 3.

9. The slope on-site safety monitoring system as described in claim 7, characterized in that, The field controller uses a data stream approach, directly forwarding the data from the field FPGA as a data stream input to the probabilistic graphical model.

10. The slope on-site safety monitoring system as described in claim 7, characterized in that, The outputs of the n-channel FFT transformation circuit are connected to one bus interface of the field master controller via bus communication.

Images