Bridge structure self-adapting sampling monitoring method and device adaptive to extreme road environment

Abstract

The application discloses a bridge structure self-adaptive sampling monitoring method and device suitable for an extreme road environment, and belongs to the technical field of bridge structure health monitoring.The core target of the application is to solve the problems of repeated arrangement of existing monitoring equipment, road-bridge data fragmentation, and low self-adaptive sampling regulation precision.Through innovative design of a road-bridge multiplexing integrated sensing node, single-node multi-parameter and full-scene monitoring are realized, and the arrangement cost and construction difficulty are reduced; through construction of a self-adaptive sampling strategy based on fuzzy logic control, a feasible algorithm model is given, continuous dynamic regulation of sampling parameters is realized, monitoring precision and low power consumption are taken into account, deep collaborative linkage of road environment monitoring and bridge structure monitoring is realized, and the safe operation of the bridge structure under the extreme road environment is ensured.

Description

Technical Field

[0001] This invention belongs to the field of bridge structure health monitoring technology, specifically relating to an adaptive sampling monitoring method and device for bridge structures adapted to extreme road environment. Background Technology

[0002] As a core component of transportation infrastructure, the structural safety and durability of highway bridges are directly related to traffic safety. The road environment is complex and changeable. Extreme weather (heavy rain, strong winds, icing), geological disasters (mudslides, slope slips), and special corrosive environments (salt spray) can easily lead to cumulative damage and performance degradation of bridge structures, and even cause safety accidents.

[0003] Currently, most existing bridge structure monitoring technologies employ a fixed-frequency sampling mode. This means that regardless of the stability of the road environment, monitoring equipment collects bridge structural data (such as stress, displacement, vibration, and tilt angle) at a preset fixed frequency. This mode has significant drawbacks: Firstly, in stable road environments, fixed high-frequency sampling generates a large amount of redundant data, increasing data transmission and storage costs, while also consuming a large amount of electricity, resulting in insufficient battery life for the monitoring equipment. This is especially problematic for bridges in remote areas without mains power, leading to extremely high maintenance costs. Secondly, in extreme road environments, fixed low-frequency sampling cannot capture the instantaneous dynamic response of the bridge structure, making it difficult to accurately identify potential structural damage. This can easily result in missed alarms and false alarms, failing to meet the needs of bridge safety monitoring in extreme environments. Summary of the Invention

[0004] In view of this, the purpose of this invention is to provide an adaptive sampling monitoring method and device for bridge structures that adapts to extreme road environments. By constructing an adaptive sampling strategy based on fuzzy logic control, a feasible algorithm model is provided to achieve continuous dynamic adjustment of sampling parameters, balancing monitoring accuracy and low power consumption. This enables deep synergy between road environment monitoring and bridge structure monitoring, ensuring the safe operation of bridge structures under extreme road environments.

[0005] To achieve the above objectives, the present invention provides the following technical solution: An adaptive sampling and monitoring method for bridge structures adapted to extreme road environment includes: Construct the core parameters of the fuzzy logic controller, including input variables, output variables, membership functions of input variables, membership functions of output variables, and fuzzy rule base; The comprehensive risk index for characterizing the health of bridge structures under extreme road conditions, the bridge structure response change rate for reflecting the dynamic change trend of bridge structure response, and the remaining power ratio of monitoring equipment are acquired in real time. These are then substituted into the corresponding input variable membership functions to calculate the membership values ​​of each input variable and determine the fuzzy subset of each input variable. The fuzzy rule base is queried based on the fuzzy subsets of the input variables, and the Mamdani inference method is used to infer the fuzzy subsets of each output variable and their corresponding membership values ​​using the membership values ​​of each fuzzy subset corresponding to the input variables. Among them, the output variables include sampling frequency, sampling resolution, number of monitoring channels, and data upload priority. The discretized centroid method is used to perform defuzzification operations on the fuzzy subsets of the output variables and their corresponding membership values ​​to obtain the precise values ​​of each output variable. These values ​​are then used to generate adaptive sampling commands that are sent to the corresponding monitoring equipment to complete the sampling operation.

[0006] Furthermore, the comprehensive hazard index H, used to characterize the structural health of bridges under extreme road conditions, and the bridge structural response change rate, reflecting the dynamic trend of the bridge structural response, are acquired in real time. ,include: Real-time data collection of road environment parameters and bridge structural health parameters; using the analytic hierarchy process (AHP), initial scores are calculated based on the road environment parameters and bridge structural health parameters, as well as their corresponding weight coefficients. The initial scores include a road environment intensity score and a bridge structural sensitivity score. The overall risk index H is obtained by calculating the initial score using a linear weighted summation method. The bridge structure response change rate is calculated based on the bridge structure sensitivity score at preset time intervals. It is used to reflect the dynamic changing trend of the bridge structure response.

[0007] Furthermore, an adaptive sampling and monitoring method for bridge structures adapted to extreme road environments also includes: obtaining a comprehensive hazard index H and comparing it with an environmental threshold range to determine the level of coordinated working condition between the current road environment and the bridge structure; The environmental threshold range is set as follows: Stable environment: H < 0.3; Mildly extreme environments: 0.3 ≤ H < 0.6; Moderately extreme environments: 0.6 ≤ H < 0.8; Severe extreme environment: H≥0.8.

[0008] Furthermore, the universe of discourse for the comprehensive risk index H is set as follows: This characterizes the overall risk level of the road environment and bridge structure; The universe of discourse for the remaining battery percentage B of the device is set to B=0 indicates that the battery is depleted, and B=1 indicates that the battery is fully charged. Bridge structural response rate of change The domain of discourse is set as The unit is 1 / h, representing the change in the bridge structure sensitivity score per unit time. The larger the value, the more severe the bridge structure response.

[0009] Furthermore, the universe of discourse for the sampling frequency f is set to Via times / minute, corresponding to 1 to 30 times / hour; The universe of discourse of the sampling resolution R is set to R=1 represents the highest resolution, and R=10 represents the lowest resolution; The universe of discourse for the number of monitoring channels N is set to The unit is "one"; The discrete values ​​of the data upload priority P are set as follows: , where 1 represents the lowest priority and 5 represents the highest priority.

[0010] Furthermore, the membership functions of the input variables include the membership functions of input variable H, input variable B, and input variable H. Membership function; The fuzzy subset of the membership function of the input variable H is: These represent no risk, low risk, medium risk, relatively high risk, and high risk, respectively. The membership function expression for the input variable H is: Vertex H=0; Vertex H = 0.2; , Vertex H = 0.325; Vertex H = 0.55; Vertex H = 1.0; The fuzzy subset of the membership function of input variable B is: These represent low battery level, medium battery level, and high battery level, respectively. The membership function expression for input variable B is: LB(B) = B∈[0,0.3], vertex B=0; MB(B) = B∈[0.2,0.7], vertex B=0.45; HB(B)= B∈[0.6,1.0], vertex B=1.0; Input variables The fuzzy subset of the membership function is: These represent slow change, moderate change, and rapid change, respectively. Input variables The membership function expression is: SL(ΔS)= ΔS∈[0,0.03], vertex ; ML(ΔS)= , vertex ; ,vertex .

[0011] Furthermore, the membership functions of the output variables include the membership functions of the output variable f, the output variable R, the output variable N, and the output variable P. The fuzzy subset of the membership function of the output variable f is: These represent extremely low frequency, low frequency, medium frequency, high frequency, and ultra-high frequency, respectively. The membership function expression for the output variable f is: support set ,vertex , ; support set ,vertex 0.167, ; support set ,vertex , ; support set ,vertex 5, ; support set ,vertex , ; The fuzzy subset of the membership function of the output variable R is: These represent high resolution, medium resolution, and low resolution, respectively. The membership function expression for the output variable R is: Vertex R=1; Vertex R=5; Vertex R=10; The fuzzy subset of the membership function of the output variable N is: These represent a few channels, a medium channel, and a full channel, respectively. The membership function expression for the output variable N is: Vertex N=2; Vertex N=6.5; Vertex N=12; The fuzzy subset of the membership function of the output variable P is: These represent low priority, medium priority, and high priority, respectively. The single-point membership function of the output variable P is defined as follows: At P=1 Other P values Deblurring followed by overlay The center of gravity is taken at 1.5. At P=3 Other P values Deblurring followed by overlay Take the center of gravity 3; At P=5 Other P values Deblurring followed by overlay The center of gravity is 4.5.

[0012] An adaptive sampling and monitoring device for bridge structures adapted to extreme road environments is provided to implement an adaptive sampling and monitoring method for bridge structures adapted to extreme road environments. The monitoring device includes: The bridge structure sensing module and the road environment sensing module are used to collect bridge structure health parameters and road environment parameters in real time. The bridge structure perception module and the road environment perception module adopt a distributed perception, edge-side aggregation, and centralized management and control architecture. Different physical quantities of the bridge structure and road environment are separated into independent dedicated sensor devices according to the monitoring type. All perception data are uniformly uploaded to the edge computing gateway to complete aggregation, fusion and decision-making. The energy supply module is used to provide the remaining power percentage of the monitoring equipment. The edge computing gateway is used to uniformly receive data collected by the bridge structure sensing module and the road environment sensing module. After data processing, it uses the built-in environment-structure coupling analysis module to calculate the comprehensive hazard index and the bridge structure response change rate. It is also used to generate adaptive sampling instructions with sampling frequency, sampling resolution, number of monitoring channels, and data upload priority by using the built-in fuzzy logic controller, combined with the remaining power ratio of the equipment, the comprehensive hazard index, and the bridge structure response change rate. These instructions are then sent to the corresponding monitoring equipment to complete the sampling operation. The data transmission module provides data transmission services for the bridge structure sensing module, road environment sensing module, energy supply module, and edge computing gateway.

[0013] Furthermore, the bridge structure perception module is based on the inertial measurement principle of microelectromechanical systems. By collecting triaxial acceleration, angular velocity and magnetic force data, it calculates the three-dimensional attitude and real-time vibration modes of the structure, which serve as health parameters of the bridge structure. The road environment perception module adopts a classified distributed sensor architecture, which is divided into three categories: meteorology, geology and corrosion, into independent sensing devices to achieve flexible deployment in different scenarios, and is used to complete the collection of road environment parameters. The energy supply module adopts a multi-source coupled self-powered architecture, including solar panels, a small wind turbine, a bridge vibration energy harvester, a piezoelectric energy recovery unit, and an energy storage battery; The edge computing gateway is also used to adaptively adjust the energy supply mode of the energy supply module based on light intensity, wind speed and bridge vibration amplitude, as well as device power consumption; The data transmission module adopts LoRa and 5G dual-mode transmission, has a built-in data transmission priority scheduling unit, and works in conjunction with the fuzzy logic controller. It dynamically switches the transmission mode according to the data upload priority output by the fuzzy logic controller.

[0014] Furthermore, the energy supply module is also equipped with a low-power sleep mechanism. The low-power sleep mechanism is as follows: a low-power sleep threshold is preset. When the remaining power of the monitoring device is less than the low-power sleep threshold, the low-power mode is automatically triggered, non-core monitoring channels are shut down, and key parameter acquisition is prioritized.

[0015] The beneficial effects of this invention are as follows: 1. An adaptive sampling strategy based on fuzzy logic control (FLC) is constructed, and a complete, implementable algorithm model is provided (including input and output variables, membership functions, fuzzy rule base, inference and defuzzification algorithms). Compared with existing fixed-frequency sampling or simple level switching sampling, the fuzzy logic control algorithm of this invention can accurately adapt sampling parameters according to the intensity of the road environment, the rate of change of bridge structure response, and equipment power consumption. It can improve the sampling frequency and accuracy in extreme environments, capture the instantaneous dynamic response of the bridge structure, and avoid missed alarms and false alarms. At the same time, it can reduce the sampling frequency and reduce energy consumption in stable environments, completely solving the problems of redundant data, insufficient monitoring accuracy in extreme environments, and short equipment battery life caused by fixed sampling frequency. At the same time, the algorithm has high computational efficiency, adapts to the low power consumption requirements of edge computing gateways, and improves the stability and reliability of control.

[0016] 2. Achieve deep collaborative monitoring of road environment and bridge structure. Through the environment-structure coupling analysis module, achieve deep fusion of data from both. Combined with fuzzy logic control algorithm, realize closed-loop control of "environmental parameter acquisition - comprehensive risk assessment - sampling parameter adjustment - structural monitoring and early warning". The edge computing gateway adopts an event-driven data scheduling mechanism, combined with dual-mode transmission and local storage, to improve data transmission efficiency, ensure the continuity and reliability of monitoring data, reduce redundant data, and improve predictive capabilities.

[0017] 3. It adopts a multi-source coupled self-powered architecture, integrating solar energy, wind energy, bridge vibration energy and piezoelectric energy recovery technology to achieve adaptive energy supply, eliminating the need for wiring and maintenance. It is suitable for bridge monitoring in areas without mains power and in remote road areas, extending the equipment's lifespan. At the same time, the energy supply is linked with the fuzzy logic control strategy to achieve reasonable energy allocation, further reducing energy consumption and ensuring stable operation of the equipment in extreme environments.

[0018] 4. It can achieve accurate monitoring and graded early warning for different types of extreme road environment (heavy rain, strong wind, icing, etc.), quickly identify potential bridge structural damage hazards, provide a scientific basis for bridge maintenance decisions, effectively improve the safety and reliability of bridge structures in extreme road environment, extend the service life of bridges, and adapt to the monitoring needs of different types of bridges (mountainous areas, coastal areas), with strong versatility.

[0019] Other advantages, objectives, and features of the invention will be set forth in the following description and will be apparent to those skilled in the art in some respects, or may be learned by practice of the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0020] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0021] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of a method and device for adaptive sampling and monitoring of bridge structures adapted to extreme road environments, as described in an embodiment of the present invention. Figure 2 This is a schematic diagram of a device module for an adaptive sampling monitoring method and apparatus for bridge structures adapted to extreme road environments, as described in an embodiment of the present invention. Detailed Implementation

[0022] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0023] like Figure 1 , Figure 2 As shown, this invention proposes an adaptive sampling and monitoring device for bridge structures adapted to extreme road environments. Through an innovative design of a road-bridge reusable integrated sensing node, it achieves multi-parameter, full-scenario monitoring of a single node, reducing deployment costs and construction difficulty. Specifically, it adopts a distributed sensing, edge-side aggregation, and centralized control architecture, completely eliminating highly integrated and redundant integrated composite nodes. Different physical quantities of the bridge structure and road environment are separated into independent dedicated sensor devices according to monitoring type. All sensing data is uniformly uploaded to an edge computing gateway for aggregation, fusion, and decision-making. The entire system consists of five parts: a bridge structure sensing subsystem, a road environment sensing subsystem, an edge computing gateway, an energy supply module, and a data transmission module.

[0024] (1) Bridge structure perception module Attitude and Vibration Integrated Sensing Unit: Based on the MEMS (Micro-Electro-Mechanical Systems) inertial measurement principle, it acquires triaxial acceleration, angular velocity, and magnetic force data to calculate three-dimensional attitude (tilt angle) and real-time vibration modes. Tilt Angle Sensing: By integrating the gravitational component of the accelerometer under static / quasi-static conditions and combining it with gyroscope drift compensation, it accurately calculates the pitch / roll angles of bridge piers and supports. Vibration Sensing: Through high-speed sampling of gyroscope and accelerometer data, it captures high-frequency (dynamic strain) and low-frequency (torsional, translational) vibration signals of the bridge to achieve modal analysis. Strain Sensing Unit: Employing fiber optic gratings (FBG) or vibrating wire strain gauges, it independently acquires the strain distribution of key sections.

[0025] (2) Road environment perception module A categorized distributed sensor architecture is adopted, with sensors divided into three main categories: meteorology, geology, and corrosion. This avoids forced integration of different physical quantities and installation locations, enabling flexible deployment based on specific scenarios. The meteorological monitoring group includes rainfall sensors, wind speed and direction sensors, temperature and humidity sensors, and icing thickness sensors, collecting routine meteorological and icing parameters. The geological disaster monitoring group includes slope displacement sensors and debris flow early warning sensors, deployed on slopes and dangerous road sections to monitor geological risks. The corrosion environment monitoring group includes salt spray concentration sensors and corrosion ion concentration sensors, deployed along coastlines or in highly corrosive road sections to collect corrosion parameters.

[0026] (3) Edge computing gateway Data aggregation: Unifies the reception of raw data from all distributed sensors of bridge structure and road environment, and completes verification, filtering, and outlier removal. Coupled calculation: Built-in environment-structure coupling analysis module calculates road environment intensity score Senv, bridge structure sensitivity score Sstruct, comprehensive hazard index H, and structural response change rate ΔSstruct.

[0027] Adaptive Control: An integrated fuzzy logic controller (FLC) combines comprehensive risk assessment, equipment power consumption, and structural change rate to generate control commands such as sampling frequency, resolution, monitoring channels, and upload priority. Command Issuance: Sampling control commands are precisely issued to corresponding sensors, enabling dynamic control based on device, parameter, and operating condition. Data Scheduling: An event-driven mechanism is employed, uploading only critical data and early warning information, significantly reducing transmission and storage costs.

[0028] (4) Energy supply module Employing a multi-source coupled self-powered architecture, combining solar, wind, bridge vibration energy, and piezoelectric energy recovery technologies, the system includes solar panels, a small wind turbine, a bridge vibration energy collector, a piezoelectric energy recovery unit, and a storage battery. The edge computing gateway adaptively adjusts the energy supply mode based on environmental parameters (light intensity, wind speed, bridge vibration amplitude) and device power consumption, prioritizing solar power, activating the wind turbine when wind speed is high, and starting the vibration energy collector and piezoelectric energy recovery unit when bridge vibration is significant, ensuring stable power supply to the device under extreme environments. Simultaneously, the energy supply module works in conjunction with the shared power management unit of the integrated sensing node to achieve reasonable energy allocation and low power consumption control, extending the device's lifespan.

[0029] (5) Data transmission module It adopts LoRa and 5G dual-mode transmission, with a built-in data transmission priority scheduling unit that works in conjunction with a fuzzy logic controller (FLC). The transmission mode is dynamically switched according to the data upload priority P output by the FLC. When the road environment is stable, the LoRa low-power transmission mode is used to transmit only key indicator data. In extreme environments, it automatically switches to 5G high-speed transmission mode to upload all raw data and metadata, while also supporting local data storage to avoid data loss caused by network interruptions in extreme environments.

[0030] The core of the adaptive sampling monitoring method for bridge structures adapted to extreme road environments based on the above-mentioned device lies in achieving coordinated acquisition of "environment-structure" parameters through a road-bridge reusable integrated sensing node. Then, through an adaptive sampling strategy based on fuzzy logic control (FLC) combined with a specific algorithm model, continuous dynamic adjustment of sampling parameters is achieved, rather than the traditional fixed-level switching. The specific steps are as follows: Step 1: Device Initialization Initialize the road environment perception module, bridge structure perception module, edge computing gateway, energy supply module, and data transmission module, focusing on the following settings: Preset the road environment intensity scoring standard and bridge structure sensitivity scoring standard, specifying that the values ​​of Senv and Sstruct are both [0,1]. Senv=0 indicates a stable and risk-free road environment, while Senv=1 indicates an extremely dangerous road environment. Sstruct=0 indicates no sensitive response from the bridge structure, while Sstruct=1 indicates a critical response from the bridge structure. Preset the threshold ranges for the comprehensive hazard index H: H1=0.3 (stable environment threshold), H2=0.6 (mildly extreme environment threshold), and H3=0.8 (severely extreme environment threshold), to assist in determining the working condition level.

[0031] The core parameters for constructing a fuzzy logic controller (FLC) are: to define the input variables, output variables, fuzzy universe of discourse, membership function, and fuzzy rule base of the FLC, and to complete the algorithm initialization.

[0032] Step 2: Construct the core parameters of the fuzzy logic controller (FLC) (1) Input variable definitions (3) Comprehensive risk index H: universe of discourse [0,1], characterizing the comprehensive risk level of the road environment and bridge structure; equipment remaining power ratio B: universe of discourse [0,1], B=0 indicates that the power is exhausted, B=1 indicates that the power is fully charged; bridge structure response change rate ΔSstruct: universe of discourse [0,0.1] (unit: 1 / h), characterizing the change range of the bridge structure sensitivity score per unit time. The larger ΔSstruct is, the more severe the bridge structure response.

[0033] (2) Definition of output variables (4) Sampling frequency f: Universe of discourse [0.017, 30] times / minute (corresponding to 1 time / hour to 30 times / minute); Sampling resolution R: Universe of discourse [1, 10], R=1 indicates the highest resolution (corresponding to the highest precision), R=10 indicates the lowest resolution (corresponding to the lowest precision); Number of monitoring channels N: Universe of discourse [2, 12] (unit: channels); Data upload priority P: discrete values ​​{1, 2, 3, 4, 5}, 1 is the lowest priority, 5 is the highest priority.

[0034] (3) Input variable membership function ① Membership function of input variable H Fuzzy subsets: {NB (no risk), NS (low risk), ZR (medium risk), PS (higher risk), PB (high risk)}; Membership function expression (trigonometric function): μ_NB(H)=max( H∈[0,0.2], vertex H=0; μ_NS(H)=max H∈[0.1,0.3], vertex H=0.2; μ_ZR(H)=max( H∈[0.2,0.45], vertex H=0.325; μ_PS(H)=max H∈[0.4,0.7], vertex H=0.55; μ_PB(H)=max H∈[0.6,1.0], vertex H=1.0.

[0035] ② Membership function of input variable B Fuzzy subset: {LB (low battery), MB (medium battery), HB (high battery)} Membership function expression: μ_LB(B)=max B∈[0,0.3], vertex B=0; μ_MB(B)=max B∈[0.2,0.7], vertex B=0.45; μ_HB(B)=max B∈[0.6,1.0], vertex B=1.0.

[0036] ③ Membership function of input variable ΔSstruct Fuzzy subsets: {SL (slow change), ML (medium change), FL (fast change)} Membership function expression: μ_SL(ΔS)=max ΔS∈[0,0.03], vertex ΔS=0; μ_ML(ΔS)=max ΔS∈[0.02,0.07], vertex ΔS=0.045; μ_FL(ΔS)=max ΔS∈[0.06,0.10], vertex ΔS=0.10.

[0037] (4) Output the membership function of the variable ① Membership function of output variable f (sampling frequency) Domain of discourse: [0.017, 30] times / minute; Fuzzy subset: {VF (Very Low Frequency), LF (Low Frequency), MF (Medium Frequency), HF (High Frequency), VHF (Ultra High Frequency)}; Using a logarithmic scale uniform distribution: μ_VF(f): Support set [0.017, 0.1], vertex f = 0.017 (corresponding to 1 time / hour), μ_VF(f) = max( μ_LF(f): Support set [0.05, 0.5], vertex f=0.167 (corresponding to 10 times / hour), μ_LF(f)=max ); μ_MF(f): Support set [0.3, 3], vertex f=1 (corresponding to 1 time / minute), μ_MF(f)=max ); μ_HF(f): Support set [2, 10], vertex f=5 (corresponding to 5 times / minute), μ_HF(f)=max ); μ_VHF(f): Support set [8, 30], vertex f=30 (corresponding to 30 times / minute), μ_VHF(f)=max( ).

[0038] ② Membership function of output variable R (sampling resolution) Domain of discourse: [1, 10]; Fuzzy subsets: {HR (high resolution / high precision), MR (medium resolution), LR (low resolution)}; Membership function: μ_HR(R) = max , R∈[1,4], vertex R=1 (highest resolution); μ_MR(R) = max R∈[2,8], vertex R=5; μ_LR(R) = max R∈[6,10], vertex R=10 (lowest resolution).

[0039] ③ Membership function of output variable N (number of monitoring channels) Domain of discourse: [2, 12] (domains); Fuzzy subsets: {FN (few channels), MN (medium channels), AN (all channels)}; Membership function: μ_FN(N) = max N∈[2,5], vertex N=2; μ_MN(N) = max N∈[3,10], vertex N =6.5; μ_AN(N) = max N∈[8,12], vertex N=12 ④ Processing method for output variable P (data upload priority) Domain of discourse: {1, 2, 3, 4, 5}; Fuzzy subset: {LP (low priority), MP (medium priority), HP (high priority)}; Singleton membership function definition: μ_LP(P): μ=1 at P=1, and μ=0 at other P values; when covering P∈{1,2} after defuzzification, the centroid is taken as 1.5; μ_MP(P): μ=1 at P=3, and μ=0 at other P values; when covering P∈{2,3,4} after defuzzing, the centroid is taken as 3; μ_HP(P): μ=1 at P=5, and μ=0 at other P values; when covering P∈{4,5} after defuzzification, the centroid is taken as 4.5; Post-processing for deblurring: P_defuzzy = Σ[μ_output(P k )·P k ] / Σ[μ_output(P k )]; P_final = r Rounding to the nearest integer, it maps to {1,2,3,4,5}.

[0040] (5) Fuzzy rule base (full version, 45 rules in total) Rule format: IFH=□ANDB=□ANDΔS=□THENf=□ANDR=□ANDN=□ANDP=□.

[0041] Among them: H∈{NB,NS,ZR,PS,PB}; B∈{LB,MB,HB}; ΔS∈{SL,ML,FL}; f∈{VF,LF,MF,HF,VHF}; R∈{HR,MR,LR}; N∈{FN,MN,AN}; P∈{LP,MP,HP}.

[0042] H=NB (risk-free) low power scenario Rule 1: IF H=NB AND B=LB AND ΔS=SL THEN f=VF, R=LR, N=FN, P=LP; Rule 2: IF H=NB AND B=LB AND ΔS=ML THEN f=VF, R=LR, N=FN, P=LP; Rule 3: IF H=NB AND B=LB AND ΔS=FL THEN f=LF, R=MR, N=FN, P=LP; H=NB (risk-free) medium power scenario Rule 4: IF H=NB AND B=MB AND ΔS=SL THEN f=VF, R=LR, N=FN, P=LP; Rule 5: IF H=NB AND B=MB AND ΔS=ML THEN f=VF, R=LR, N=FN, P=LP; Rule 6: IF H=NB AND B=MB AND ΔS=FL THEN f=LF, R=MR, N=FN, P=LP; H=NB (risk-free) high-power scenario Rule 7: IF H=NB AND B=HB AND ΔS=SL THEN f=VF, R=LR, N=FN, P=LP Rule 8: IF H=NB AND B=HB AND ΔS=ML THEN f=LF, R=LR, N=FN, P=LP Rule 9: IF H=NB AND B=HB AND ΔS=FL THEN f=LF, R=MR, N=FN, P=LP H=NS (Low Risk) Low Power Scenarios Rule 10: IF H=NS AND B=LB AND ΔS=SL THEN f=VF, R=LR, N=FN, P=LP; Rule 11: IF H=NS AND B=LB AND ΔS=ML THEN f=LF, R=MR, N=FN, P=LP; Rule 12: IF H=NS AND B=LB AND ΔS=FL THEN f=LF, R=MR, N=MN, P=MP; H=NS (Low Risk) Medium Power Scenario Rule 13: IF H=NS AND B=MB AND ΔS=SL THEN f=LF, R=LR, N=FN, P=LP; Rule 14: IF H=NS AND B=MB AND ΔS=ML THEN f=LF, R=MR, N=MN, P=MP; Rule 15: IF H=NS AND B=MB AND ΔS=FL THEN f=MF, R=MR, N=MN, P=MP; H=NS (Low Risk) High Power Scenarios Rule 16: IF H=NS AND B=HB AND ΔS=SL THEN f=LF, R=LR, N=FN, P=LP; Rule 17: IF H=NS AND B=HB AND ΔS=ML THEN f=MF, R=MR, N=MN, P=MP; Rule 18: IF H=NS AND B=HB AND ΔS=FL THEN f=MF, R=HR, N=MN, P=MP; H=ZR (Medium Risk) Low Power Scenarios Rule 19: IF H=ZR AND B=LB AND ΔS=SL THEN f=LF, R=MR, N=FN, P=LP; Rule 20: IF H=ZR AND B=LB AND ΔS=ML THEN f=MF, R=MR, N=MN, P=MP; Rule 21: IF H=ZR AND B=LB AND ΔS=FL THEN f=MF, R=HR, N=MN, P=MP; H=ZR (Medium Risk) Medium Electricity Scenario Rule 22: IF H=ZR AND B=MB AND ΔS=SL THEN f=MF, R=MR, N=MN, P=MP; Rule 23: IF H=ZR AND B=MB AND ΔS=ML THEN f=MF, R=HR, N=MN, P=MP; Rule 24: IF H=ZR AND B=MB AND ΔS=FL THEN f=HF, R=HR, N=AN, P=HP; H=ZR (Medium Risk) High Power Scenario Rule 25: IF H=ZR AND B=HB AND ΔS=SL THEN f=MF, R=MR, N=MN, P=MP; Rule 26: IF H=ZR AND B=HB AND ΔS=ML THEN f=HF, R=HR, N=AN, P=HP; Rule 27: IF H=ZR AND B=HB AND ΔS=FL THEN f=HF, R=HR, N=AN, P=HP; H=PS (Higher Risk) Low Power Scenarios Rule 28: IF H=PS AND B=LB AND ΔS=SL THEN f=MF, R=MR, N=MN, P=MP; Rule 29: IF H=PS AND B=LB AND ΔS=ML THEN f=HF, R=MR, N=MN, P=HP; Rule 30: IF H=PS AND B=LB AND ΔS=FL THEN f=HF, R=HR, N=AN, P=HP; H=PS (Higher Risk) Medium Power Scenarios Rule 31: IF H=PS AND B=MB AND ΔS=SL THEN f=HF, R=HR, N=MN, P=HP; Rule 32: IF H=PS AND B=MB AND ΔS=ML THEN f=HF, R=HR, N=AN, P=HP; Rule 33: IF H=PS AND B=MB AND ΔS=FL THEN f=VHF, R=HR, N=AN, P=HP; H=PS (Higher Risk) High-Power Scenarios Rule 34: IF H=PS AND B=HB AND ΔS=SL THEN f=HF, R=HR, N=AN, P=HP Rule 35: IF H=PS AND B=HB AND ΔS=ML THEN f=VHF, R=HR, N=AN, P=HP Rule 36: IF H=PS AND B=HB AND ΔS=FL THEN f=VHF, R=HR, N=AN, P=HP H=PB (High Risk) Low Power Scenarios Rule 37: IF H=PB AND B=LB AND ΔS=SL THEN f=HF, R=MR, N=MN, P=HP; Rule 38: IF H=PB AND B=LB AND ΔS=ML THEN f=VHF, R=HR, N=MN, P=HP; Rule 39: IF H=PB AND B=LB AND ΔS=FL THEN f=VHF, R=HR, N=AN, P=HP; H=PB (High-risk) medium power scenario Rule 40: IF H=PB AND B=MB AND ΔS=SL THEN f=VHF, R=HR, N=AN, P=HP; Rule 41: IF H=PB AND B=MB AND ΔS=ML THEN f=VHF, R=HR, N=AN, P=HP; Rule 42: IF H=PB AND B=MB AND ΔS=FL THEN f=VHF, R=HR, N=AN, P=HP; H=PB (High Risk, High Power Scenarios) Rule 43: IF H=PB AND B=HB AND ΔS=SL THEN f=VHF, R=HR, N=AN, P=HP; Rule 44: IF H=PB AND B=HB AND ΔS=ML THEN f=VHF, R=HR, N=AN, P=HP; Rule 45: IF H=PB AND B=HB AND ΔS=FL THEN f=VHF, R=HR, N=AN, P=HP.

[0043] (6) Fuzzy reasoning and defuzzification algorithms Step 1: Input fuzzification Enter precise values Substitute each membership function into the equation to calculate its membership degree to each fuzzy subset.

[0044] Calculation example: when H=0.66, B=0.8, ΔSstruct=0.05; μ_PS = max( = min(1.73, 0.27) = 0.27; μ_PB = max = 0.15; μ_HB = max = 0.5; μ_ML = max = min(1.2, 0.8) = 0.8; μ_FL = 0.

[0045] The actual activation rule is a combination of H=PS / PB, B=HB, and ΔS=ML.

[0046] Step 2: Fuzzy Reasoning (Mamdani Method) Using the Mamdani inference method, the activation strength of a rule is determined by the minimum membership degree of each antecedent: αᵢ = min(μ_Hᵢ(H), μ_Bᵢ(B), μ_ΔSᵢ(ΔS)); Where αᵢ is the activation strength of the i-th rule.

[0047] Taking the aforementioned input as an example, the key activation rule is: Rule 34 : α 34 = min(0.27, 0.5, 0.8) = 0.27 → f=HF, R=HR, N=AN, P=HP Rule 35 (H=PS, B=HB, ΔS=ML): According to the rule base, rule 35 is: IF H=PS AND B=HB AND ΔS=ML THEN f=VHF, R=HR, N=AN, P=HP; when ΔS=0.05, μ_FL=0. In this example, ΔS=ML (μ_ML=0.8), therefore rule 35 is actually activated, α 35 =min(0.27,0.5,0.8)=0.27 → f=VHF, R=HR, N=AN, P=HP The membership function of the output fuzzy set is obtained by taking the minimum value: μ_outputⱼ(x) = maxᵢ[min(αᵢ, μ_consequentᵢⱼ(x))]; Where μ_consequentᵢⱼ(x) is the consequent membership function of the i-th rule on the j-th output variable.

[0048] Step 3: Defuzzification (Center of Gravity Discretization, COG) For the output variable (f, R, N) in a continuous universe of discourse: x_defuzzy = Σ k [μ_output(x k ) · x k · Δx] / Σ k [μ_output(x k ) · Δx]; in, x k For discrete sampling points within the universe of discourse (it is recommended to take 5-9 sampling points for each fuzzy subset, evenly distributed); Δx is the sampling interval; μ_output(x k ) represents the output membership degree after aggregation.

[0049] Sampling point setting recommendations: f: Take 20-30 points in the [0.017,30] logarithmically uniformly distributed range; R: Take 10 points in [1,10] with a linear uniform distribution (step size 1). N: Take 11 points in [2,12] with a linear uniform distribution (step size 1, integer points).

[0050] For discrete output variable P: P_defuzzy = Σ k [μ_output(P k ) · P k ] / Σ k [μ_output(P k )]; P_final = r (Rounded to the nearest integer); If P_final exceeds {1,2,3,4,5}, then it is truncated to the boundary value (<1 takes 1, >5 takes 5).

[0051] Calculation example (taking sampling frequency f as an example): Assuming the aggregated output fuzzy set has non-zero membership at VF, LF, MF, and HF, the sampling points are: f1=0.017 f2=0.1 f3=0.5 f4=1 f5=5 Then f_defuzzy = (0.1×0.017 + 0.3×0.1 + 0.6×0.5 + 0.2×1 + 0.05×5) / ≈ 0.63 times / minute.

[0052] Step 3: Collaborative collection of road environment and structural parameters The road environment perception module and bridge structure perception module are synchronously activated through integrated perception nodes. Each integrated perception node's environmental perception unit and bridge structure monitoring unit (including a MEMS inertial measurement subunit) work synchronously, collecting real-time road environment parameters (meteorological parameters, geological disaster parameters, and corrosion environment parameters) and bridge structure health parameters. The collected data undergoes preliminary preprocessing (outlier removal and smoothing filtering) via the node's built-in micro-edge processor, and is then synchronously transmitted to the edge computing gateway via a wireless self-organizing network. This enables bidirectional interaction and synchronous acquisition of road-bridge data, providing data support for comprehensive hazard index calculation and fuzzy logic control. During this process, the integrated perception node can implement some low-power control based on the initial instructions from the edge computing gateway, reducing ineffective energy consumption.

[0053] Step 4: Calculation of Comprehensive Risk Index (1) Calculation of Senv score for road environment intensity The analytic hierarchy process (AHP) is used to perform a weighted summation of the environmental parameters: Senv = Σ(wᵢ·xᵢ); Where wᵢ represents the weight of the i-th environmental parameter (e.g., rainfall weight 0.3, wind speed weight 0.25, slope displacement weight 0.2, salt spray concentration weight 0.15, and icing thickness weight 0.1).

[0054] xᵢ is the normalized value of the i-th environmental parameter: xᵢ = (actual value - minimum value) / (maximum value - minimum value), ensuring that xᵢ∈[0,1].

[0055] (2) Calculation of bridge structural sensitivity score Sstruct The analytic hierarchy process is also used: Sstruct =Σ(kⱼ · yⱼ); Among them, \(k_j\) is the weight of the \(j\)th structural parameter (such as the stress weight of 0.3, the vibration frequency weight of 0.25, the displacement weight of 0.2, the pier inclination weight of 0.15, and the bearing settlement weight of 0.1); \(y_j\) is the normalized value of the \(j\)th structural parameter: \(y_j=(actual value - safety value) / (critical value - safety value)\), ensuring that \(y_j\in[0,1]\).

[0056] (3) Calculation of the comprehensive hazard index \(H\) The linear weighted summation formula is adopted: \(H = \alpha S_{env}+\beta\cdot S_{struct}\); Among them, \(\alpha\) and \(\beta\) are weight coefficients, which are dynamically adjusted according to the bridge type (such as \(\alpha = 0.6\), \(\beta = 0.4\) for mountain bridges, and \(\alpha = 0.5\), \(\beta = 0.5\) for coastal bridges); ensure that \(H\in[0,1]\).

[0057] At the same time, calculate the change rate of the bridge structure response: \(\Delta S_{struct}=|S_{struct}(t)-S_{struct}|\) / \(\Delta t\); | / \(\Delta t\); Among them, \(\Delta t\) is the time interval (preset to 1 hour), which is used to reflect the dynamic change trend of the bridge structure response.

[0058] (4) Judgment of the working condition level According to the value of \(H\) and combined with the preset threshold, judge the collaborative working condition level of the current road area environment and the bridge structure: Steady environment: \(H < H_1 = 0.3\); Mild extreme environment: \(H_1\leq H < H_2 = 0.6\); Moderate extreme environment: \(H_2\leq H < H_3 = 0.8\); Severe extreme environment: \(H\geq H_3 = 0.8\).

[0059] This level judgment is used to assist in understanding the FLC regulation result, and the actual sampling parameters are dynamically calculated and determined by the fuzzy logic controller.

[0060] Step 5: Fuzzy logic adaptive sampling parameter adjustment (core step) The edge computing gateway realizes the dynamic regulation of sampling parameters through the built-in fuzzy logic controller (FLC) according to the following process: The first step: Fuzzification of input variables Substitute the calculated \(H\), \(\Delta S_{struct}\), and the remaining battery percentage \(B\) of the device collected in real time into the modified membership function, convert them into corresponding fuzzy sets, and obtain the membership values of each input variable.

[0061] The second step: Fuzzy inference Based on the fuzzy set of the input variables, the complete fuzzy rule base (45 rules) is queried. Using the Mamdani inference method, the fuzzy set and membership values ​​of the output variables (f, R, N, P) are obtained through "AND" (min) and "OR" (max) logical operations.

[0062] Step 3: Deblurring The discretized centroid method is used to perform defuzzification operations on the fuzzy set of output variables to obtain the precise values ​​of sampling frequency f, sampling resolution R, number of monitoring channels N, and data upload priority P, and to generate adaptive sampling instructions.

[0063] Step 4: Issuance and Execution of Commands The edge computing gateway sends sampling instructions to each integrated sensing node, and the node adjusts its own sampling parameters according to the instructions to achieve precise control.

[0064] Typical control parameter ranges (reference values) for each operating condition level: Operating condition level, sampling frequency f (times / minute), sampling resolution R (1=highest), number of monitoring channels N (number), data priority P (1-5); Stable environment (H<0.3), 0.017-0.083 (1-5 times / hour); 7-10 (low resolution); 2-3 (core channel); 1-2 (low priority); Mild extreme (0.3≤H<0.6), 0.167-0.5 (10-30 times / hour); 4-7 (medium resolution); 4-6 (normal channels); 2-3 (medium priority); Moderately extreme (0.6≤H<0.8), 1-5 (1-5 times / minute); 2-4 (higher resolution); 7-10 (most channels); 4 (higher priority); Severe extreme (H≥0.8), 10-30 (10-30 times / minute); 1-2 (high resolution); 11-12 (full channel); 5 (highest priority); Note: The table above shows the parameter range reference values ​​under typical operating conditions. The actual parameters are determined by FLC through continuous calculation based on the precise values ​​of H, B, and ΔSstruct, rather than by simple level switching. R is the sampling resolution (corrected), R=1 indicates the highest resolution (highest precision), and R=10 indicates the lowest resolution (lowest precision).

[0065] Step 6: Special Operating Condition Adjustment When the device battery level B < 0.2 (low battery), regardless of the value of H, FLC will automatically reduce the sampling frequency f (by 30%-50%) and the number of monitoring channels N (retaining only 2-3 core channels). The MEMS inertial measurement subunit will only retain the core calibration function to ensure continuous operation of the device and avoid monitoring interruption due to battery depletion.

[0066] This mechanism is linked to the FLC's power input B: when B∈[0,0.2), μ_LB(B) approaches 1, automatically activating low power-related rules (rules 1-3, 10-12, 19-21, 28-30, 37-39); when B∈[0.2,0.3), μ_LB(B) gradually decreases, and μ_MB(B) gradually increases, achieving a smooth transition and avoiding drastic changes in sampling parameters (hysteresis effect) caused by power fluctuations near the threshold.

[0067] Step 7: Data Processing and Early Warning The edge computing gateway performs local preprocessing on the collected bridge structural parameters, and analyzes the health status of the bridge structure by combining the comprehensive hazard index H and road environment parameters. If the structural parameters are detected to exceed the preset threshold, a graded warning (mild, moderate, severe) is immediately triggered, and the abnormal data, warning information and corresponding road environment parameters are uploaded to the cloud platform. If the comprehensive hazard index H drops below the stable environment threshold, the FLC automatically adjusts the sampling parameters to a low-power mode to achieve "on-demand monitoring".

[0068] Step 8: Adaptive Energy Supply from Multi-Source Coupled Sources The energy supply module adaptively adjusts the power supply mode based on road environment parameters (sunlight intensity, wind speed, bridge vibration amplitude) and equipment power consumption. When there is sufficient sunlight, solar power is used first; when the wind speed is high, the wind turbine is activated; when the bridge vibration is significant, the vibration energy collector and piezoelectric energy recovery unit are activated, and the energy storage battery stores redundant electrical energy in real time. At the same time, the energy supply module is linked with the fuzzy logic controller to dynamically adjust the energy distribution ratio according to the sampling parameters output by the FLC, prioritizing the power supply of the core monitoring channels. When the battery power is lower than the preset threshold, the non-core monitoring channels are automatically shut down to ensure the continuous and stable operation of the equipment in extreme environments.

[0069] Step 9: Low-power sleep mechanism A preset low-power sleep threshold is set. When the device power B < 0.2, the low-power mode is automatically triggered, non-core monitoring channels are shut down, and key parameter acquisition is prioritized.

[0070] This mechanism is linked to the FLC's power input B, and achieves this through the smooth transition characteristics of the fuzzy membership function: When B∈[0,0.2), μ_LB(B) approaches 1, activating the low-charge-related rule; When B∈[0.2,0.3), μ_LB(B) gradually decreases and μ_MB(B) gradually increases, achieving a smooth transition; To avoid drastic changes in sampling parameters (hysteresis effect) caused by fluctuations in power level near the threshold.

[0071] The present invention will be further described in detail below with reference to specific embodiments. The following embodiments are used to illustrate the present invention, but are not limited to the scope of protection of the present invention. They mainly demonstrate the practical application of road-bridge reusable integrated sensing nodes and fuzzy logic control algorithms: Example: Applicable to mountainous highway bridges (prone to extreme environments such as heavy rain and mudslides). Four integrated road-bridge sensing nodes were deployed on the slopes of both sides of the mountain highway bridge. Each node simultaneously collects road environment parameters such as rainfall, slope displacement, wind speed, and temperature, as well as structural parameters such as bridge vibration, stress, and pier tilt angle. The bridge structure monitoring unit has a built-in MEMS inertial measurement subunit to simultaneously assist in calibrating vibration and tilt angle data, thereby improving the accuracy of data acquisition. The edge computing gateway is deployed in the control box at the end of the bridge, with preset weighting coefficients α=0.6 and β=0.4, and comprehensive hazard index thresholds H1=0.3, H2=0.6, and H3=0.8.

[0072] Fuzzy Logic Controller (FLC) Parameter Settings (Revised): The input variable H has a domain of discourse of [0,1] and a fuzzy subset of {NB, NS, ZR, PS, PB}, where the support set of ZR is corrected to [0.2,0.45] (vertex H=0.325), and PS remains [0.4,0.7] (vertex H=0.55). The two form a reasonable overlap, eliminating the coverage blind spot at the original H=0.4. B has a domain of discourse [0,1] and a fuzzy subset {LB, MB, HB}. ΔSstruct has a universe of discourse [0, 0.1] and fuzzy subsets {SL, ML, FL}. The output variable has a universe of discourse of f[0.017,30] (log-scale distribution), a universe of discourse of R[1,10] (sampling resolution, corrected), a universe of discourse of N[2,12], and a universe of discourse of P{1,2,3,4,5} (Singleton processing). The membership function uses trigonometric functions (P uses Singleton), the fuzzy rule base contains 45 complete rules, the inference uses the Mamdani inference method, and the defuzzification uses the discretization centroid method.

[0073] Scenario 1: Moderately extreme environment (heavy rain + strong wind) Input conditions: Rainfall 60mm / 24h, wind speed 9, slope displacement 1.2mm / d → S_env=0.7; The bridge's vibration frequency is abnormal, and the stress is too high → Sstruct=0.6; With weighting coefficients α=0.6 and β=0.4, H=0.6×0.7+0.4×0.6=0.66; Device power B = 0.8 (high power); The structural response change rate ΔSstruct = 0.05 (1 / h, moderate change).

[0074] Fuzzy computation (corrected): μ_PS =max =min(1.73,0.27)=0.27; μ_PB =max =0.15; μ_HB =max =0.5; μ_ML =max =min(1.2,0.8)=0.8; μ_FL =0.

[0075] Key rules for activation: Rule 34 : α=min(0.27,0.5,0.8)=0.27 → f=HF,R=HR,N=AN,P=HP; Rule 35 (H=PS, B=HB, ΔS=ML): α 35 =min(0.27,0.5,0.8)=0.27 → f=VHF,R=HR,N=AN,P=HP (Equal activation strength as in rule 34, both contribute to the output fuzzy set); Rule... (Other combinations are activated based on non-zero membership).

[0076] Defuzzy results (discretized COG): Sampling frequency f ≈ 8 times / minute (high frequency); Sampling resolution R≈1.5 (high resolution, close to the highest precision); The number of monitoring channels N ≈ 11 (close to all channels); Data priority P=5 (highest priority).

[0077] Regulatory actions: Automatically switch to 5G high-speed transmission, enable bridge vibration mode recognition, and use high-frequency assisted calibration of the MEMS inertial measurement subunit.

[0078] Scenario 2: Stable environment (normal weather) Input conditions: Rainfall 8mm / 24h, wind speed 5, slope displacement 0.3mm / d → S_env=0.15; Bridge structural parameters are normal → S_struct=0.1; H = 0.6 × 0.15 + 0.4 × 0.1 = 0.13; Device power B = 0.9 (high power); The structural response rate of change ΔS_struct = 0.01 (1 / h, slow change).

[0079] Fuzzy computation: μ_NB =max =0.35; μ_NS =max =min(0.3,1.7)=0.3; μ_HB =max =0.75; μ_SL =max =0.67; μ_ML =0.

[0080] Key rules for activation: Rule 7 : α=min(0.35,0.75,0.67)=0.35 → f=VF,R=LR,N=FN,P=LP; Rule 8 : α=min(0.3,0.75,0.67)=0.3 → f=LF,R=LR,N=FN,P=LP.

[0081] Defuzzing results: The sampling frequency f ≈ 0.05 times / minute (approximately 3 times / hour, extremely low frequency); Sampling resolution R≈9 (low resolution); The number of monitoring channels is approximately N = 2 (core channels); Data priority P=1 (low priority).

[0082] Regulatory actions: Entering a low-power sleep mode, using LoRa low-power transmission, and intermittently waking up the MEMS inertial measurement subunit to extend battery life.

[0083] In summary, this invention provides an adaptive sampling monitoring device and method for bridge structures adapted to extreme road environments. It is applicable to highway bridges and municipal bridges in extreme road environments such as rainstorms, strong winds, icing, mudslides, salt spray, and freeze-thaw cycles. It is especially suitable for bridges in remote road areas without mains power or stable network coverage, enabling dynamic, accurate, and low-power monitoring and early warning of the health status of bridge structures.

[0084] Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made to it in form and detail without departing from the scope defined by the claims of the present invention.

Claims

1. An adaptive sampling and monitoring method for bridge structures adapted to extreme road environment, characterized in that, include: Construct the core parameters of the fuzzy logic controller, including input variables, output variables, membership functions of input variables, membership functions of output variables, and fuzzy rule base; The comprehensive risk index for characterizing the health of bridge structures under extreme road conditions, the bridge structure response change rate for reflecting the dynamic change trend of bridge structure response, and the remaining power ratio of monitoring equipment are acquired in real time. These are then substituted into the corresponding input variable membership functions to calculate the membership values ​​of each input variable and determine the fuzzy subset of each input variable. The fuzzy rule base is queried based on the fuzzy subsets of the input variables, and the Mamdani inference method is used to infer the fuzzy subsets of each output variable and their corresponding membership values ​​using the membership values ​​of each fuzzy subset corresponding to the input variables. Among them, the output variables include sampling frequency, sampling resolution, number of monitoring channels, and data upload priority. The discretized centroid method is used to perform defuzzification operations on the fuzzy subsets of the output variables and their corresponding membership values ​​to obtain the precise values ​​of each output variable. These values ​​are then used to generate adaptive sampling commands that are sent to the corresponding monitoring equipment to complete the sampling operation.

2. The adaptive sampling and monitoring method for bridge structures adapted to extreme road environments according to claim 1, characterized in that, Real-time acquisition of the comprehensive hazard index H, used to characterize the structural health of bridges under extreme road conditions, and the bridge structural response change rate, reflecting the dynamic trend of bridge structural response. ,include: Real-time road environment parameters and bridge structural health parameters are collected. Using the analytic hierarchy process (AHP), initial scores are calculated based on the road environment parameters and bridge structural health parameters, as well as their corresponding weight coefficients. The initial scores include a road environment intensity score and a bridge structural sensitivity score. The overall risk index H is obtained by calculating the initial score using a linear weighted summation method. The bridge structure response change rate is calculated based on the bridge structure sensitivity score at preset time intervals. It is used to reflect the dynamic changing trend of the bridge structure response.

3. The adaptive sampling and monitoring method for bridge structures adapted to extreme road environments according to claim 2, characterized in that, Also includes: The comprehensive hazard index H is obtained and compared with the environmental threshold range to determine the current working condition level of the road environment and bridge structure. The environmental threshold range is set as follows: Stable environment: H < 0.3; Mildly extreme environments: 0.3 ≤ H < 0.6; Moderately extreme environments: 0.6 ≤ H < 0.8; Severe extreme environment: H≥0.

8.

4. The adaptive sampling and monitoring method for bridge structures adapted to extreme road environments according to claim 1, characterized in that, The universe of discourse for the comprehensive hazard index H is set as follows: This characterizes the overall risk level of the road environment and bridge structure; The universe of discourse for the remaining power percentage B of the device is set as follows: B=0 indicates that the battery is depleted, and B=1 indicates that the battery is fully charged. The bridge structure response change rate The domain of discourse is set as The unit is 1 / h, representing the change in the bridge structure sensitivity score per unit time. The larger the value, the more severe the bridge structure response.

5. The adaptive sampling and monitoring method for bridge structures adapted to extreme road environments according to claim 1, characterized in that, The universe of discourse of the sampling frequency f is set to Via times / minute, corresponding to 1 to 30 times / hour; The universe of discourse of the sampling resolution R is set to R=1 represents the highest resolution, and R=10 represents the lowest resolution; The universe of discourse for the number of monitoring channels N is set as follows: The unit is "one"; The discrete values ​​of the data upload priority P are set as follows: , where 1 represents the lowest priority and 5 represents the highest priority.

6. The adaptive sampling and monitoring method for bridge structures adapted to extreme road environments according to claim 1, characterized in that, The input variable membership functions include the membership functions of input variable H, input variable B, and input variable H. Membership function; The fuzzy subset of the membership function of the input variable H is: These represent no risk, low risk, medium risk, relatively high risk, and high risk, respectively. The membership function expression for the input variable H is: Vertex H=0; Vertex H = 0.2; ， Vertex H = 0.325; Vertex H = 0.55; Vertex H = 1.0; The fuzzy subset of the membership function of the input variable B is: These represent low battery level, medium battery level, and high battery level, respectively. The membership function expression for the input variable B is: LB(B) = B∈[0,0.3], vertex B=0; MB(B) = B∈[0.2,0.7], vertex B=0.45; HB(B)= B∈[0.6,1.0], vertex B=1.0; The input variables The fuzzy subset of the membership function is: These represent slow change, moderate change, and rapid change, respectively. The input variables The membership function expression is: SL(ΔS)= ΔS∈[0,0.03], vertex ; ML(ΔS)= , vertex ; ,vertex .

7. The adaptive sampling and monitoring method for bridge structures adapted to extreme road environments according to claim 1, characterized in that, The membership functions of the output variables include the membership functions of the output variable f, the output variable R, the output variable N, and the output variable P. The fuzzy subset of the membership function of the output variable f is: These represent extremely low frequency, low frequency, medium frequency, high frequency, and ultra-high frequency, respectively. The membership function expression for the output variable f is: support set ,vertex , ; support set ,vertex 0.167, ; support set ,vertex , ; support set ,vertex 5, ; support set ,vertex , ; The fuzzy subset of the membership function of the output variable R is: These represent high resolution, medium resolution, and low resolution, respectively. The membership function expression for the output variable R is: Vertex R=1; Vertex R=5; Vertex R=10; The fuzzy subset of the membership function of the output variable N is: These represent a few channels, a medium channel, and a full channel, respectively. The membership function expression for the output variable N is: Vertex N=2; Vertex N=6.5; Vertex N=12; The fuzzy subset of the membership function of the output variable P is: These represent low priority, medium priority, and high priority, respectively. The single-point membership function of the output variable P is defined as follows: At P=1 Other P values Deblurring followed by overlay The center of gravity is taken at 1.

5. At P=3 Other P values Deblurring followed by overlay Take the center of gravity 3; At P=5 Other P values Deblurring followed by overlay The center of gravity is 4.

5.

8. An adaptive sampling and monitoring device for bridge structures adapted to extreme road environments, characterized in that, For performing the monitoring method as described in any one of claims 1-7, the monitoring device comprises: The bridge structure sensing module and the road environment sensing module are used to collect bridge structure health parameters and road environment parameters in real time. The bridge structure sensing module and the road environment sensing module adopt a distributed sensing, edge-side aggregation, and centralized control architecture. Different physical quantities of the bridge structure and road environment are divided into independent dedicated sensor devices according to the monitoring type. All sensing data are uniformly uploaded to the edge computing gateway to complete aggregation, fusion and decision-making. The energy supply module is used to provide the remaining power percentage of the monitoring equipment. The edge computing gateway is used to uniformly receive data collected by the bridge structure sensing module and the road environment sensing module. After data processing, it uses the built-in environment-structure coupling analysis module to calculate the comprehensive hazard index and the bridge structure response change rate. It is also used to generate adaptive sampling instructions with sampling frequency, sampling resolution, number of monitoring channels, and data upload priority by using the built-in fuzzy logic controller, combined with the remaining power ratio of the equipment, the comprehensive hazard index, and the bridge structure response change rate. These instructions are then sent to the corresponding monitoring equipment to complete the sampling operation. The data transmission module provides data transmission services for the bridge structure sensing module, road environment sensing module, energy supply module, and edge computing gateway.

9. The adaptive sampling and monitoring device for bridge structures adapted to extreme road environments according to claim 8, characterized in that: The bridge structure sensing module is based on the inertial measurement principle of microelectromechanical systems. By collecting triaxial acceleration, angular velocity and magnetic force data, it calculates the three-dimensional attitude and real-time vibration modes of the structure, which serve as health parameters of the bridge structure. The road environment perception module adopts a classified distributed sensor architecture, which is divided into three major categories: meteorology, geology and corrosion, into independent sensing devices to achieve flexible deployment in different scenarios, and is used to complete the collection of road environment parameters. The energy supply module adopts a multi-source coupled self-powered architecture, including a solar panel, a small wind turbine, a bridge vibration energy harvester, a piezoelectric energy recovery unit, and an energy storage battery. The edge computing gateway is also used to adaptively adjust the energy supply mode of the energy supply module according to light intensity, wind speed and bridge vibration amplitude, as well as device power consumption; The data transmission module adopts LoRa and 5G dual-mode transmission, has a built-in data transmission priority scheduling unit, is linked with the fuzzy logic controller, and dynamically switches the transmission mode according to the data upload priority output by the fuzzy logic controller.

10. The adaptive sampling and monitoring device for bridge structures adapted to extreme road environments according to claim 8, characterized in that, The energy supply module is also equipped with a low-power sleep mechanism, which is as follows: a low-power sleep threshold is preset, and when the remaining power of the monitoring device is less than the low-power sleep threshold, the low-power mode is automatically triggered, non-core monitoring channels are shut down, and key parameter acquisition is prioritized.

Images