1 Introduction

With the increasing disturbance of human engineering activities to the geological environment, the frequency of landslide disasters is getting higher and higher, resulting in great economic losses and casualties (Huang 2007). The causes of landslide disasters mainly include earthquake, rainfall, human factors and so on. The occurrence of landslide disaster has the characteristics of strong suddenness, great harm and high difficulty in treatment. If the deformation evolution behaviour of landslide can’t be accurately predicted in time, it will cause immeasurable loss (Lin et al. 2009; Xu et al. 2008).

Aiming at the problem of landslide early warning and prediction, relevant experts have done a lot of research on this (Wang et al. 2023; Ma et al. 2017; Aristizábal et al. 2016). A new data-driven method for spatiotemporal prediction of landslides at the regional scale has been proposed, which integrates landslide susceptibility maps using the RF algorithm and probabilistic hydrological and meteorological thresholds (Abraham et al. 2023). Moretto et al. (2021) confirmed that satellite InSAR can be successfully applied to the early monitoring of landslides, and emphasized its limitations. Meng et al. (2023) proposed a dynamic hybrid prediction model of landslide displacement based on empirical mode decomposition (EMD), gated recurrent unit (GRU) and error correction, which provides an important basis for the prediction and prevention of stepped reservoir landslide. Miao et al. (2017) drew an improved Saito curve based on the experience of Saito method, and verified that the improved Saito method is reliable and economical for landslide prediction through field monitoring data. Aggarwal et al. (2020) applied different landslide prediction techniques to the same set of data for analysis, and concluded that the dynamic neural network model is the best model for time series landslide prediction. Compared with other techniques, the DNN model has the smallest error rate. Soranzo et al. (2023) combined machine learning with strength reduction method, which greatly increased the computational efficiency.

Because the surface displacement, deep displacement, groundwater level, pore water pressure, rainfall, soil moisture content and other parameters are only necessary conditions for the occurrence of landslide geological disasters, rather than sufficient conditions. Through the landslide warning of these parameters, the warning signal can’t be issued in time. Based on this, He (2016), He et al. (2017) proposed a two-body catastrophe mechanics model based on interface Newton force measurement. And through the independent research and development of Newton force change remote monitoring and early warning system for on-site monitoring, it is confirmed that the scientific phenomenon of ' Newton force suddenly drops and landslide occurs '. Tao et al. (2022, 2019) obtained the applicability of NPR anchor cable in landslide early warning through research. The early warning mode and level of landslide Newton force disaster are put forward, and the integrated control technology of landslide reinforcement-monitoring-early warning is formed, which successfully solves the scientific problem of landslide short-term prediction.

With the development of optical fiber sensing technology (Ivanov et al. 2021), many scholars (Feng et al. 2017; Liu et al. 2022; Minardo et al. 2021; Wu et al. 2017; Xu and Yin 2016) began to apply it to the monitoring of slope engineering, and achieved some good results. Zhu et al. (2015) developed an innovative quasi-distributed optical fiber sensing network based on fiber Bragg grating (FBG) technology to monitor the strain distribution of soil nails, the underground displacement of slope and the internal strain of soil. Combined with laser displacement sensor to measure surface displacement, it is proved that optical fiber monitoring can effectively identify the evolution stage of slope. Zhang et al. (2018) Calculating the correlation between landslide displacement and axial strain applied to sensing optical fibers based on distributed optical fibers. And some interferometric fiber optic sensors have also been used to test the ability to identify unstable precursors in a reduced model of rainfall induced landslides composed of granular materials (Longoni et al. 2022). Based on probability migration, Jiang (2021) used 15 fiber Bragg gratings to effectively predict landslide geological disasters. Yin et al. (2023) used optical fibers to monitor cracks in the canopy area of the landslide, providing early warning for landslide disasters. Shi et al. (2004), Shi et al. made a systematic and creative contribution to the theory, technology and application of distributed optical fiber sensing in rock and soil mass, and pointed out the future development direction and frontier research topics of optical fiber sensing technology for rock and soil mass catastrophe.

At present, the application of optical fiber sensing technology in slope engineering and Newton force monitoring technology have gradually matured. In view of the fact that the Newton force monitoring system cannot effectively locate the potential sliding surface in real time and the optical fiber sensing technology cannot reinforce the slope. In this paper, on the basis of the above research, through the three-dimensional physical model test, the fiber grating and NPR anchor cable are physically combined, and the Newtonian force monitoring curve is combined and compared with the fiber strain results. The advantages and characteristics of NPR-OFST landslide disaster joint sensing and early warning method are discussed, so as to provide a more effective comprehensive monitoring and control method for landslide disaster in field engineering practice.

2 Fundamentals

2.1 FBG sensing technology

Fiber grating is a special preparation method to establish the grating structure in a cylindrical optical waveguide structure, that is, the core or cladding structure. The strain acting on the fiber grating will cause the grating period to change. Pulling or squeezing the fiber grating will cause deformation in the grating period. The working principle of the fiber grating is shown in Fig. 1. The light wave entering the fiber will propagate in a specific direction. The wavelength that satisfies the fiber Bragg condition will reflect, and other wavelengths will continue to propagate through the fiber grating. When some physical quantities of the measured object change and strain the fiber grating, the central wavelength of the reflected light changes accordingly, and the reflected light is received by the receiver. Finally, it is possible to measure the changes in physical quantities of the measured object (Zhang et al. 2016).

Fig. 1
figure 1

Working principle of the FBG sensing technology

Full size image

The expression of the fiber grating Bragg wavelength is:

$$\lambda_{B} = 2n_{eff} \Lambda$$
(1)

In the formula, \(\lambda_{B}\) is the central wavelength of the fiber grating, i.e., Bragg wavelength, \(n_{eff}\) is the effective refractive index of the fiber core, and \(\Lambda\) is the grating period. When the broadband light source enters the Brgg grating, only the light wave satisfying Eq. (1) will be reflected back, and other wavelengths will be transmitted out.

When the grating is subjected to external force to produce strain or temperature change, its effective refractive index and grating period will change accordingly (Zhu et al. 2020). The central wavelength of the reflected light, strain and temperature satisfy the following relationship:

$$\frac{{\Delta \lambda_{B} }}{{\lambda_{B} }} = (1 - \rho_{e} )\Delta \varepsilon + (\alpha + \zeta )\Delta T$$
(2)

In the formula, \(\lambda_{B}\) is the original Bragg wavelength without strain and at 0 °C; \(\Delta \lambda_{B}\) is the wavelength shift caused by the strain and temperature changes; \((1 - \rho_{e} )\Delta \varepsilon\) denotes the shift of the central wavelength caused by the axial strain, where \(\rho_{e}\) is the effective photoelastic coefficient; \((\alpha + \zeta )\Delta T\) is the shift of the central wavelength caused by temperature change, where \(\alpha\) and \(\zeta\) are the thermal expansion coefficient and the thermo-optic coefficient, respectively. The experimental environment in this study is a constant temperature environment, and the data collection time is short. Therefore, it is reasonable to assume that the offset of the central wavelength is independent of the temperature change (Wang et al. 2017). Therefore, \(\Delta \lambda_{B}\) in the expression (2) is only related to the strain increment \(\Delta \varepsilon\), and the expression (2) can be transformed into:

$$\Delta \varepsilon = \Delta \lambda_{B} /K_{\varepsilon }$$
(3)

In the formula, \(\Delta \varepsilon\) is the strain increment; \(\Delta \lambda_{B}\) is the wavelength drift; \(K_{\varepsilon }\) is the strain sensitivity coefficient. The Bragg wavelength shift is converted to the strain value by this formula, where the strain sensitivity coefficient \(K_{\varepsilon } = 1.2\) (Zhang and Chen 2017).

2.2 Newton force monitoring principle and NPR anchor cable

He et al. (2021) pointed out that the landslide from gestation to disaster is a complex nonlinear process, and its Newton force changes through the gradual deformation stage, the sudden change stage and the failure movement stage. The landslide is a gradual evolution process from local damage to the whole block instability. It is a process of accumulation and release of deformation energy. According to Newton 's second law, if an object wants to move, it needs to produce acceleration, and force is the cause of acceleration, that is, acceleration occurs only after a strong change. The generation of force is a necessary and sufficient condition for the movement of objects. This force is the resultant force F (Newton force). The essence of landslide geological disaster is the relative movement of sliding body and sliding bed along the potential sliding surface. The variation of Newtonian force can characterize the true motion state of landslides. The change of Newton force can characterize the real motion state of landslide. According to the indoor landslide Newton force monitoring physical model experiment, the landslide Newton force change curve is obtained as shown in Fig. 2.

Fig. 2
figure 2

The schematic diagram of Newton 's force change law of landslide

Full size image

The occurrence of landslide disaster is mainly due to the development of the disaster-causing rock mass from small deformation to nonlinear large deformation, which in turn leads to the occurrence of landslide disaster. The existing supporting materials for landslide prevention and control engineering are mainly made of materials with Poisson 's ratio effect. This kind of material cannot meet the conditions of large deformation, and is prone to necking fracture, which can’t effectively prevent and control the large deformation disaster of rock mass. In order to deal with the problem of large deformation of rock mass, He et al. (2016, 2014) developed a constant resistance large deformation anchor cable with negative Poisson 's ratio effect, and applied it to slope reinforcement and landslide monitoring and early warning. The NPR anchor cable is mainly composed of constant resistance casing, constant resistance conical body and steel strand, which has extraordinary mechanical properties such as high constant resistance, large deformation, energy absorption and impact resistance. In the process of large deformation of rock mass, the constant resistance body slips in the constant resistance casing. Because the diameter of the large end of the constant resistance body is slightly larger than the inner diameter of the casing, the casing expands radially while being stretched, resulting in NPR effect. It can monitor the change of Newtonian force throughout the entire process of landslides and achieve early warning of landslide disaster. The reinforcement mechanism is shown in Fig. 3.

Fig. 3
figure 3

NPR anchor cable slope reinforcement mechanism

Full size image

3 Slope physical model test

This experiment investigates the feasibility and superiority of joint monitoring of fiber Bragg grating and NPR anchor cable. A three-dimensional physical model test under 1 g condition was carried out in the State Key Laboratory of Deep Geotechnical and Underground Engineering, China University of Mining and Technology (Beijing). In the experiment, the NPR anchor cable and the fiber Bragg grating are physically combined to form an anchor cable sensor. Establish an integrated reinforcement monitoring system to achieve continuous spatiotemporal monitoring of the entire life cycle of the slope.

3.1 Model box and model material

This paper mainly studies the prediction and location of the potential sliding surface position, and therefore adopts manual loading landslide induction method. The size of the model box is 1.5 m × 2 m × 1.8 m (Fig. 4). The main body of the model box is built by a rigid frame, and the two sides are tempered glass, which is convenient for observing the change process of the model. A pre-tightening force application device is set on the back of the model box. The pre-tightening force can be applied to the anchor cable by twisting the connecting rod, and then fixed by tightening the nut, so as to apply the pre-tightening force to the NPR anchor cable used in the test.

Fig. 4
figure 4

Model test box pictures

Full size image

The simulation object of this model test is the location and failure of the potential sliding surface of the slope under the action of external force, which is mainly affected by shear deformation. Therefore, the main mechanical parameters of the indoor model are determined as density, compressive strength, elastic modulus, cohesion and internal friction angle. The model material selected river sand, gypsum powder and water ratio of 8:1:1 to make the landslide body. The selected materials are common geotechnical materials. The mixing of gypsum and water can not only shorten the time of model making, but also adjust the mechanical parameters of landslide in a wide range. The final parameters are shown in Table 1.

Table 1 Physical and mechanical parameters of landslide body
Full size table

3.2 Anchor cable sensor and monitoring equipment

In this experiment, NPR anchor cable was simulated by solder wire with a diameter of 2 mm. Solder wire has similar mechanical properties to NPR anchor cable, which can meet the mechanical conditions required for NPR anchor cable in laboratory tests. The static tensile test of solder wire was carried out by ZQ-990A static tensile machine (as shown in Fig. 5a). The obtained tensile curve is shown in Fig. 5b. It can be seen from the curve that the solder wire with a diameter of 2 mm approximately has ideal elastic–plastic characteristics during static tension. Therefore, it is reasonable to use solder wire with a diameter of 2 mm to simulate NPR anchor cable in this test.

Fig. 5
figure 5

Study on tensile properties of solder wire: a Static tensile machine and static tensile test; b Static tensile curve of 2 mm diameter solder wire

Full size image

The monitoring equipment used in this paper includes fiber grating strain monitoring system and Newton force monitoring system. The FBG strain monitoring system is composed of FBG demodulator, FBG strain sensor and signal transmission cable. The Newton force monitoring system consists of NPR anchor cable, PVC pipe, anchor cable axial force sensor, acquisition instrument, tray and anchor pier (Fig. 6). The anchor pier is simulated by a disc with an outer diameter of 50 mm and an inner diameter of 5 mm. The tray is used to connect the anchor cable and the sensor to ensure the uniform force of the anchor cable to the sensor. The inner diameter of anchor cable axial force sensor is 6 mm. The optical fiber sensor adopts the high transmission tight envelope grating strain sensing optical cable of Suzhou Nanzhi Sensing Technology Co., Ltd. The NPR anchor cable, PVC pipe, FBG strain sensor, anchor cable axial force sensor, tray and anchor pier are combined to form an anchor cable sensor to reinforce and monitor the slope body.

Fig. 6
figure 6

Test model material

Full size image

3.3 Monitoring principle and scheme

Compared with other materials, PVC pipes have lower elastic modulus and can maintain deformation suitable for soil. And PVC pipe can protect and fix the direction of solder wire. Therefore, PVC tube is selected as the carrier of fiber grating. The outer diameter of PVC pipe is 6 mm and the inner diameter is 3 mm. The fiber grating optical cable is fixed on the PVC pipe with epoxy resin glue. The epoxy resin glue has a protective effect on the optical fiber at the same time. The stress characteristics of the PVC pipe in the soil are shown in Fig. 7. Because the anchor cable sensor is fixed on the model box and anchored on the surface of the slope, both ends are fixed. After soil loading, the upper side of the PVC pipe above the sliding surface is in the extrusion state, and the lower side is in the tensile state. The upper side of the PVC pipe below the sliding surface is in the tensile state, and the lower side is in the extrusion state. Therefore, the FBG measuring points on both sides of the sliding surface will show different response characteristics, and the position of the sliding surface can be judged by the different responses of the measuring points.

Fig. 7
figure 7

Force characteristics of anchor cable sensor

Full size image

The anchor cable sensor is anchored in the slope body, as shown in Fig. 8a, and a total of 8 anchor cable sensors are anchored. Fiber Bragg grating strain monitoring points and anchor cable axial force monitoring points are shown in Fig. 8b. A total of 16 optical fiber measuring points are set up, five optical fiber measuring points are set up for each of the third layer anchor cables, and three optical fiber measuring points are set up for each of the second layer anchor cables. Anchor cable axial force monitoring points are set on each anchor cable sensor, and a total of 8 anchor cable axial force monitoring points are set.

Fig. 8
figure 8

Anchor cable sensor and monitoring layout diagram: a Anchor cable sensor structure diagram and anchoring position; b Monitoring point layout diagram

Full size image

3.4 Model building process and test procedure

The bottom size of the test slope model is length × width = 1.5 m × 1.5 m, the height is 1.3 m, and the inclination angle is about 45°. The model is compacted by 100 mm layered compaction. The inclination angle of the anchor cable sensor is 15°, the horizontal spacing is 50 cm, and the vertical spacing is 20 cm. The anchor cable sensor is embedded in the soil and anchored. The model reserves the sliding surface by laying mica sheet. The loading of the model is carried out by manual stacking. Place an acrylic plate with a width of 20 cm, a length of 1.5 m and a thickness of 30 mm was placed above the model, and a stacking load with an initial load of 100 kg was applied to the acrylic plate. The complete test process is shown in Fig. 9.

Fig. 9
figure 9

Complete process of test

Full size image

4 Test results and analysis

4.1 Strain monitoring result of FBG sensing array

The fiber grating measuring points on both sides of the reserved sliding surface are selected for analysis, and whether it is reasonable to use the position of the sliding surface as the interface between the compression zone and the tensile zone of the PVC pipe is discussed. The fiber strain time history curve obtained by the test is shown in Fig. 10. According to the arrangement of optical fiber strain monitoring points, it can be seen that the measuring points 1,4,5,9,12 and 13 are all located in the soil below the preset sliding surface. Each measuring point is in a tensile state, and the measured strain values are all tensile strain. The measuring points 3, 7, 8, 11, 15 and 16 are all located in the soil body above the preset sliding surface. Each measuring point is in a compressive state, and the measured strain values are compressive strains. The test results show that it is accurate to take the interface between the compression zone and the tensile zone of PVC pipe as the location of the sliding surface.

Fig. 10
figure 10

Fiber strain time history curve: a Strain value of anchor cable sensor 2; b Strain value of anchor cable sensor 6; c Strain value of anchor cable sensor 3; d Strain value of anchor cable sensor 7

Full size image

When the load on the slope increases gradually, the slope has a tendency to slide along the preset sliding surface, and the strain accumulation of the slope is mainly concentrated near the preset sliding surface. And the load on the slope is transmitted from top to bottom, and the internal deformation of the upper part of the slope is higher than that of the lower part. Reflected in the PVC pipe, it should be shown that the closer to the preset sliding surface, the greater strain accumulation will be produced, and the influence of the upper PVC pipe should be greater than that of the lower. Through data analysis, it can be seen that during the monitoring time, the maximum strain values measured at measuring points 1,3,9,11 reached about 3000με, the maximum strain values measured at measuring points 5,7,13,15 reached more than 2000με, and the maximum strain values measured at measuring points 4,8,12,16 did not reach 2000με. It can be seen from the arrangement of optical fiber strain measuring points that the positions of measuring points 1,3,5,7,9,11,13 and 15 are closer to the preset sliding surface, and the positions of measuring points 1, 3, 9 and 11 are located above the vertical direction of the slope. According to the strain distribution characteristics of each monitoring point, the measuring point near the preset sliding surface has a larger strain monitoring value. And the monitoring point above the slope has a larger strain monitoring value than the monitoring point below the slope. It can be seen that PVC pipe and soil have the ability to maintain deformation coordination. It is feasible to analyze the instability state of slope by arranging measuring points on PVC pipe.

4.2 Slip surface monitoring result analysis

Monitoring the sliding surface can locate the potential sliding surface in time and capture the key information of the force and displacement of the sliding surface. In the whole loading process, the position of the sliding surface can be located to effectively analyze the stability of the slope in time. The strain distribution curve fitted by Bezier spline using typical time point strain data is shown in Fig. 11. According to the distribution curve, the strain distribution law of each anchor cable sensor is basically the same, and the curve form conforms to the deflection fitting curve when the PVC pipe is stressed. It can be concluded that the PVC pipe can be divided into tension zone and compression zone with 30 cm from the end of the pipe as the dividing point, and the closer to the sliding surface, the greater the strain value. According to the distribution characteristics of the fitting curve and the theoretical analysis in Fig. 7, the sliding position of about 30 cm is determined, which is basically consistent with the preset sliding surface position.

Fig. 11
figure 11

Each fiber optic sensor monitors the strain-position curve: a Strain distribution of anchor cable sensor 3; b Strain distribution of anchor cable sensor 7

Full size image

In this experiment, when designing the optical fiber measuring point, the monitoring points are also arranged at the preset sliding surface position to explore the corresponding law of the optical fiber strain at the sliding surface. Because the strain value measured at the sliding surface is disordered, the characteristic points are selected for fitting to obtain the optical fiber strain monitoring curve at the sliding surface, as shown in Fig. 12. It can be seen from the optical fiber strain monitoring curve that after loading for a certain time, the strain value at the preset sliding surface increases rapidly and is much larger than the optical fiber strain value at the non-sliding surface. After further loading, the optical fiber measuring point at the sliding surface fails, that is, the data cannot be continuously collected. Through analysis, it is determined that this phenomenon is due to the excessive loss of the optical fiber measuring point, which leads to the failure of the optical fiber measuring point. However, at this time, other measuring points can still carry out data monitoring, and by comparison, it can be seen that the failure time of the measuring points lower the slope is earlier than that of the above measuring points. It is concluded that there will be a large strain accumulation at the potential sliding surface, and the strain accumulation closer to the force position is greater than that farther away, and the optical fiber will fail when it is subjected to a large loss. That is, in the field application, not only to protect the optical fiber from fracture and bending, but also to consider the measurement range and loss degree of the optical fiber, so that the optical fiber can realize the whole life cycle monitoring.

Fig. 12
figure 12

Strain monitoring curve at sliding surface: a Measuring point 2 optical fiber strain value; b Measuring point 10 optical fiber strain value; c Measuring point 6 optical fiber strain value; d Measuring point 14 optical fiber strain value

Full size image

4.3 Comparative analysis of Newton force monitoring and strain results

According to Newton 's law, force is the reason for changing the state of an object. The generation of force will cause the acceleration of the object, which will lead to the change of the state of the object, that is, the force and the acceleration are closely related. In this paper, the optical fiber strain acceleration value (The following is collectively referred to as the acceleration value) is roughly simulated by the way of twice differential of optical fiber strain to time. The obtained acceleration value (expression is as follows) is fitted to obtain the corresponding curve, and the axial force monitoring curve of the anchor cable is compared with the fitting curve. As shown in Fig. 13, the selected optical fiber measuring points are close to the sliding surface to make the results more obvious. In this paper, through the direct measurement of the axial force of NPR anchor cable, the unmeasurable Newton force is indirectly obtained, so as to realize the measurement of Newton force.

$$\varepsilon_{t} = \frac{{L(t) - L_{0} }}{{L_{0} }}$$
(4)
$$\varepsilon{\prime} (t) = \frac{d\varepsilon }{{dt}} = \frac{1}{{L_{0} }}\frac{dL(t)}{{dt}} = \frac{v(t)}{{L_{0} }}$$
(5)
$$\varepsilon^{^{\prime\prime}} (t) = \frac{{d\varepsilon{\prime} }}{dt} = \frac{1}{{L_{0} }}\frac{dv(t)}{{dt}} = \frac{a(t)}{{L_{0} }}$$
(6)

where \(\varepsilon_{t}\) is the strain value, \(L(t)\) is the length of time t, \(L_{0}\) is the original length, \(v_{t}\) is the velocity of the ends away from each other, and \(a_{t}\) is the acceleration of the ends away from each other.

Fig. 13
figure 13

Joint monitoring curve of anchor cable axial force and optical fiber strain: a Joint detection curve of optical fiber sensor 3; b Joint detection curve of optical fiber sensor 7; c Joint detection curve of optical fiber sensor 6; d Joint detection curve of optical fiber sensor 2

Full size image

According to the joint detection curve, it can be seen that the axial force of the anchor cable increases gradually after the external load is applied. At this time, the stable state inside the slope body begins to change, and plastic deformation occurs in the slope body. There are cracks on the surface of the slope body, and strain accumulation occurs near the sliding surface. According to the acceleration curve, the strain value near the sliding surface is in a state of continuous accumulation. When the load increases, the acceleration value shows an increasing trend. At this time, the strain accumulation rate increases and gradually turns into a steady state. The acceleration value decreases and the strain accumulation rate tends to be gentle. When the loading time reaches 50 s, the axial force curve of No.6 and No.2 anchor cable sensors appears a sudden drop phenomenon. And after the axial force decreases, the acceleration value of the corresponding optical fiber measuring point shows an upward state. At this time, the acceleration values of the corresponding optical fiber measuring points of the No.3 and No.7 anchor cable sensors also show an upward state, but the axial force of the anchor cable does not show a sudden drop. At this time, the cracks in the upper sliding surface of the slope are penetrated, and the sliding surface is initially generated but not completely penetrated. After the loading time reaches 150 s, the axial force curves of each anchor cable sensor drop suddenly one after another. After the sudden drop of the axial force, the acceleration curves of the corresponding optical fiber measuring points all rise. At this time, the sliding surface is completed, the complete sliding surface is formed inside the slope body, and the landslide body is in a critical sliding state.

According to the analysis of the results of the joint monitoring curve, the change of force is the cause of acceleration. After the sudden drop of Newton 's force, it is followed by an increase in acceleration value, but there is a time difference in this process. It shows that the change of force is ahead of the change of acceleration, and more ahead of the change of displacement. The accuracy of using Newton force as the main parameter of landslide monitoring and early warning is verified. Through the joint detection method, timely warning of landslide disasters and timely positioning of potential sliding surfaces can be achieved. Before the Newtonian force issued a warning of imminent sliding, the characteristics, location, volume and damage degree of the potential sliding surface were analyzed by optical fiber strain monitoring and corresponding measures were taken. It provides a new idea for the treatment of landslide disaster, which is of great significance.

5 Conclusions

In this paper, the whole life cycle characteristics of NPR anchor cable anchored slope under external load are studied. The evolution process of model slope stability and deformation is simulated by applying load on top of slope. The fiber grating strain sensor is physically combined with the NPR anchor cable to form an anchor cable sensing device. A monitoring-reinforcement integrated system is formed to monitor the whole life cycle and analyze the stability of the slope. By processing the strain monitoring results of fiber optic gratings and comparing with the results of Newton force monitoring, the results are compared and analyzed. The rationality of using Newton force as the main parameter of landslide monitoring and early warning is verified, and the instability deformation and mechanical characteristics of slope are analyzed. According to this experimental study, the following conclusions are drawn:

  1. (1)

    The reliability and accuracy of the slope stability of the optical fiber sensor network monitoring model are verified. According to the monitoring results, it is concluded that the fiber grating measuring point will fail when it is subjected to large deformation. Therefore, in the field construction, the monitoring range of optical fiber should be fully considered, and appropriate protection measures should be taken to prevent the failure of optical fiber due to fracture, bending and excessive loss.

  2. (2)

    The feasibility of using PVC tube as the medium for the combination of fiber grating and NPR anchor cable is verified. The formation of anchor cable sensing device can make full use of the high sensitivity of fiber grating strain sensor, and make full use of the ability of PVC pipe to maintain soil deformation coordination. Through the anchor cable sensing device, the whole life cycle monitoring of the slope can be realized.

  3. (3)

    The internal displacement of slope rock mass can be measured by using fiber grating anchor cable sensing device. By comparing the strain data measured by fiber grating with the monitoring data of Newton force, it is concluded that the change of Newton force is ahead of the change of acceleration, and the change of force is the reason for the change of object state. The mechanical law of “Newton force sudden drops, immediately catastrophe happens” is further verified.

  4. (4)

    Through the test, the anchor cable sensing device can form a good reinforcement effect on the slope body. It can timely warn the occurrence of landslide disaster and locate the potential failure surface in time. The comprehensive monitoring and treatment of slope instability are realized. The NPR-OFST joint sensing and early warning method is a further optimization of the Newton force monitoring system.