Search

CN-121996901-A - Online stress monitoring device and method based on Modbus-RTU stress sensing and digital communication

CN121996901ACN 121996901 ACN121996901 ACN 121996901ACN-121996901-A

Abstract

The invention discloses an online stress monitoring device and method based on Modbus-RTU stress sensing and digital communication, which are characterized in that original stress data and multidimensional environment parameter sequences are collected through stress sensing nodes, an exclusive environment genetic map is built, a dynamic compensation model is generated, environment coupling decoupling and self-adaptive correction are carried out on the original data, a periodic self-diagnosis process is executed in a monitoring process, signal impedance characteristics and output signal stability are analyzed, equipment state fault codes are generated and actively uploaded through a Modbus-RTU protocol when abnormal conditions occur, a collaborative monitoring mechanism is triggered based on corrected stress measurement values, adjacent nodes are controlled to enter a high-frequency sampling mode and generate stress propagation dynamic clouds, a node performance base line is established at the same time, and a gain self-regeneration process is started when performance attenuation is detected. The invention realizes the accurate compensation and dynamic correction of stress data, and remarkably improves the environmental adaptability, the operation reliability and the fault early warning capability of the monitoring system.

Inventors

  • Chi Shuihua
  • LIU GUANGLEI

Assignees

  • 福建赢鱼网络科技有限公司

Dates

Publication Date
20260508
Application Date
20260210

Claims (10)

  1. 1. An on-line stress monitoring method based on Modbus-RTU stress sensing and digital communication is characterized by comprising the following steps: Continuously acquiring original stress data and a corresponding multidimensional environment parameter sequence through a stress sensing node, wherein the multidimensional environment parameter sequence at least comprises a temperature parameter and a time parameter; Constructing an exclusive environment genetic map of each stress sensing node through dynamic regression analysis based on the multidimensional environment parameter sequence and the original stress data, and generating a dynamic compensation model with space-time adaptability; inputting the original stress data acquired in real time into the dynamic compensation model for environmental coupling decoupling and self-adaptive correction to obtain corrected stress measurement values; In the monitoring process, a self-diagnosis flow of the stress sensing node is periodically executed, and the lead connection state of the stress sensing node and the health degree of the sensing element are judged by analyzing the signal impedance characteristics and the stability of the output signal; When the self-diagnosis result is abnormal, generating a device state fault code based on the diagnosis result, and actively uploading the device state fault code to a master station through a Modbus-RTU protocol; judging a threshold value based on the corrected stress measurement value, triggering a collaborative monitoring mechanism when the stress value is detected to exceed a preset stress threshold value, controlling adjacent stress sensing nodes to enter a high-frequency sampling mode, and generating a stress propagation dynamic cloud image; Establishing a node performance base line, continuously monitoring performance attenuation trend, starting a gain self-regeneration flow when the sensitivity or the signal-to-noise ratio is detected to be lower than the self-adaptive threshold, and finely adjusting signal conditioning parameters to restore measurement accuracy; and generating and outputting a comprehensive stress monitoring report based on the corrected stress measurement value, the equipment state fault code, the stress propagation dynamic cloud picture and the gain self-regeneration result.
  2. 2. The on-line stress monitoring method based on Modbus-RTU stress sensing and digital communication according to claim 1, wherein the constructing an environment genetic map dedicated to each stress sensing node by dynamic regression analysis based on the multi-dimensional environment parameter sequence and the original stress data, generating a dynamic compensation model with space-time adaptability comprises: In a preset initial learning period, collecting synchronous change data of the stress zero drift amount of each stress sensing node and the multidimensional environment parameter sequence; Establishing a mapping relation model between the multidimensional environment parameter sequence and the stress zero drift amount by a multivariate nonlinear regression algorithm; storing the coefficient matrix of the mapping relation model as an environment genetic map corresponding to the stress sensing node; in the running process, dynamically updating the coefficient matrix of the environment genetic map according to the multidimensional environment parameter sequence acquired in real time; based on the updated environmental genetic map, a dynamic compensation model with space-time adaptability is generated.
  3. 3. The on-line stress monitoring method based on Modbus-RTU stress sensing and digital communication according to claim 2, wherein the establishing a mapping relation model between the multi-dimensional environment parameter sequence and the stress zero drift amount by a multi-element nonlinear regression algorithm comprises: carrying out data preprocessing on the multidimensional environment parameter sequence and the stress zero drift amount; constructing a nonlinear regression equation which takes an environment parameter as an independent variable and the stress zero drift amount as a dependent variable according to the preprocessed multidimensional environment parameter sequence and the stress zero drift amount, wherein the nonlinear regression equation comprises a quadratic term of a temperature parameter and a cross coupling term among different environment parameters; Solving a parameter estimation value of the nonlinear regression equation by adopting a recursive least square algorithm to obtain an initial mapping relation model; Verifying the fitting goodness of the initial mapping relation model through residual analysis, and introducing regularization term optimization model parameters if the fitting goodness of the initial mapping relation model does not meet the preset precision requirement; and constructing the optimized model parameters into coefficient matrixes to form the mapping relation model.
  4. 4. The on-line stress monitoring method based on Modbus-RTU stress sensing and digital communication according to claim 1, wherein inputting the raw stress data collected in real time into the dynamic compensation model for performing environment coupling decoupling and adaptive correction to obtain a corrected stress measurement value comprises: Inputting the multidimensional environmental parameter sequence acquired in real time into the environmental genetic map, and calculating to obtain the stress drift compensation quantity under the current environmental condition through a mapping relation model in the environmental genetic map; Constructing an environment coupling matrix based on the stress drift compensation quantity, wherein the environment coupling matrix represents the coupling strength and the acting direction of each environment parameter on stress measurement; Performing tensor operation on the environment coupling matrix and original stress data acquired in real time, and decoupling and separating environment factors and real stress through singular value decomposition; Reconstructing the decoupled multidimensional stress component by adopting a self-adaptive weighted fusion algorithm to obtain a preliminary stress estimated value after the environmental interference is eliminated; Performing time sequence optimization processing on the preliminary stress estimation value through a Kalman filtering algorithm, eliminating the influence of random noise and measurement fluctuation, and obtaining optimized stress data; performing dimension reduction and engineering unit conversion on the optimized stress data to generate a corrected stress measurement value; And storing the corrected stress measurement value into a holding register specified by the Modbus-RTU protocol.
  5. 5. The on-line stress monitoring method based on Modbus-RTU stress sensing and digital communication according to claim 1, wherein in the monitoring process, a self-diagnosis process of the stress sensing node is periodically executed, and the judging of the lead connection state of the stress sensing node and the health of the sensing element by analyzing the signal impedance characteristic and the stability of the output signal comprises: applying a characteristic excitation signal to a signal path of the stress sensing node in a preset monitoring period, and collecting a corresponding response signal; Calculating a real-time impedance characteristic value of a signal path based on the amplitude phase relation of the characteristic excitation signal and the response signal; Comparing the real-time impedance characteristic value with a preset impedance reference range, and judging that the lead connection state is abnormal if the real-time impedance characteristic value exceeds the impedance reference range; synchronously monitoring an output signal sequence of the stress sensing node, and analyzing the time-frequency domain characteristics of the output signal through wavelet transformation; extracting high-frequency noise energy and low-frequency drift amplitude of an output signal as health evaluation indexes; Comparing the health degree evaluation index with a preset stability threshold, and judging that the health degree of the sensing element is abnormal if the health degree evaluation index exceeds the stability threshold; when abnormality of the lead connection state or abnormality of the health degree of the sensing element is detected, recording the abnormality type and severity degree, and triggering the equipment state fault code generation flow.
  6. 6. The Modbus-RTU stress sensing and digital communication based on-line stress monitoring method of claim 1, wherein performing threshold judgment based on the corrected stress measurement value, when detecting that the stress value exceeds a preset stress threshold, triggering a cooperative monitoring mechanism, controlling adjacent stress sensing nodes to enter a high-frequency sampling mode, and generating a stress propagation dynamic cloud image comprises: When the corrected stress measured value of any stress sensing node exceeds a preset stress threshold value, marking the stress sensing node as an abnormal source node; determining a plurality of associated stress sensing nodes adjacent to the abnormal source node based on the node topological relation, and marking the stress sensing nodes as associated nodes; sending a high-frequency sampling instruction to the associated node through a Modbus-RTU protocol, and controlling the associated node to enter a preset high-frequency sampling mode; synchronously collecting stress data sequences of the abnormal source node and the associated node in the same period; Performing time alignment and spatial interpolation on the acquired stress data sequence to construct a space-time distribution matrix of the stress field; generating a continuous stress propagation dynamic cloud image through a Kriging interpolation algorithm based on the space-time distribution matrix; Analyzing the stress gradient change characteristics in the stress propagation dynamic cloud image, and identifying the stress abnormal propagation path and the influence range to obtain an identification result; and storing the stress propagation dynamic cloud image and the identification result.
  7. 7. The Modbus-RTU stress sensing and digital communication based on-line stress monitoring method of claim 6, wherein generating a continuous dynamic cloud of stress propagation by a kriging interpolation algorithm based on the spatio-temporal distribution matrix comprises: Calculating a spatial autocorrelation parameter based on the spatial coordinates and stress measurement values of each stress sensing node in the spatial-temporal distribution matrix; constructing a semi-variation function model according to the space autocorrelation parameters, and determining a weight coefficient of a Kriging interpolation algorithm; carrying out optimal unbiased estimation on stress values of the non-sampled area by using the weight coefficient, and generating a continuous stress field covering the monitoring area; Fusing the continuous stress field and time dimension data to obtain the stress propagation dynamic cloud image; analyzing the stress gradient change characteristics in the stress propagation dynamic cloud image, and identifying the stress abnormal propagation path and the influence range to obtain an identification result, wherein the method comprises the following steps of: calculating stress gradient vectors of all spatial points in the stress propagation dynamic cloud picture, and constructing a stress gradient field; tracking the direction in which the stress change is most remarkable based on the stress gradient field, and determining a stress abnormal propagation path; Identifying a communication region with a stress value exceeding a preset threshold value through a region growing algorithm, and determining a stress abnormality influence range; and combining the stress anomaly propagation path and the stress anomaly influence range to generate a recognition result comprising the position, the propagation direction and the influence degree of the anomaly source.
  8. 8. The Modbus-RTU stress sensing and digital communication based on-line stress monitoring method of claim 1, wherein establishing a node performance baseline, continuously monitoring performance decay trend, starting a gain self-regeneration process when sensitivity or signal-to-noise ratio is detected to be lower than an adaptive threshold, and fine tuning signal conditioning parameters to restore measurement accuracy, comprises: in an initial calibration stage of the stress sensing node, recording initial sensitivity, initial signal-to-noise ratio and initial zero point stability parameters of the stress sensing node, and constructing a node performance base line; In the operation process, the current sensitivity and the current signal-to-noise ratio are calculated periodically, and compared and analyzed with the node performance base line, and the method comprises the following steps: Identifying the sensitivity and the long-term attenuation trend of the signal to noise ratio of the stress sensing node based on the sliding window to obtain the performance attenuation degree; When the current sensitivity or the current signal-to-noise ratio is detected to be continuously lower than the adaptive threshold value dynamically calculated based on the historical data, triggering the gain self-regeneration flow, comprising: calculating a gain compensation coefficient required by the signal conditioning circuit according to the performance attenuation degree; fine-tuning signal conditioning parameters through a digital potentiometer or a programmable amplifier, and implementing gain compensation; And verifying whether the compensated current sensitivity and the current signal-to-noise ratio are restored to be within the range of the self-adaptive threshold value, and updating the node performance baseline.
  9. 9. The Modbus-RTU stress sensing and digital communication based on-line stress monitoring method of claim 8, wherein identifying long-term attenuation trends of sensitivity and signal-to-noise ratio of stress sensing nodes based on sliding windows, obtaining performance attenuation degree, comprises: Collecting a continuous sensitivity sequence and a continuous signal-to-noise ratio sequence of the stress sensing node in a sliding window with a preset time length; trend decomposition is carried out on the sensitivity sequence and the signal to noise ratio sequence respectively, extracting a long-term variation component thereof; Calculating unit time attenuation rates of sensitivity and signal-to-noise ratio based on the long-term variation components; Comparing the attenuation rate in unit time with a preset attenuation rate threshold value, and judging whether a significant attenuation trend exists or not; When a significant attenuation trend exists, quantitatively calculating the performance attenuation degree according to the relative deviation of the current value and the initial value; combining the sensitivity attenuation degree and the signal-to-noise ratio attenuation degree, and obtaining the final performance attenuation degree through a weighted fusion algorithm.
  10. 10. An on-line stress monitoring device based on Modbus-RTU stress sensing and digital communication, characterized in that it is adapted to the method according to any one of claims 1 to 9, said device comprising: The data acquisition module is used for continuously acquiring original stress data and a corresponding multidimensional environment parameter sequence through the stress sensing node; The environment genetic map construction module is used for constructing an exclusive environment genetic map of each stress sensing node through dynamic regression analysis based on the multidimensional environment parameter sequence and the original stress data, and generating a dynamic compensation model with space-time adaptability; the stress correction module is used for inputting the original stress data acquired in real time into the dynamic compensation model to perform environment coupling decoupling and self-adaptive correction to obtain a corrected stress measured value; The self-diagnosis module is used for periodically executing a self-diagnosis flow of the stress sensing node in the monitoring process, and judging the lead connection state of the stress sensing node and the health degree of the sensing element by analyzing the signal impedance characteristics and the stability of the output signal; The fault reporting module is used for generating a device state fault code based on the diagnosis result when the self-diagnosis result is abnormal, and actively uploading the device state fault code to the master station through a Modbus-RTU protocol; The collaborative monitoring module is used for judging a threshold value based on the corrected stress measurement value, triggering a collaborative monitoring mechanism when the stress value is detected to exceed a preset stress threshold value, controlling adjacent stress sensing nodes to enter a high-frequency sampling mode, and generating a stress propagation dynamic cloud image; the performance maintenance module is used for establishing a node performance baseline, continuously monitoring performance attenuation trend, starting a gain self-regeneration flow when detecting that the sensitivity or the signal-to-noise ratio is lower than the self-adaptive threshold, and finely adjusting signal conditioning parameters to restore measurement accuracy; And the report generation module is used for generating and outputting a comprehensive stress monitoring report based on the corrected stress measurement value, the equipment state fault code, the stress propagation dynamic cloud picture and the gain self-regeneration result.

Description

Online stress monitoring device and method based on Modbus-RTU stress sensing and digital communication Technical Field The invention relates to the technical field of structural health monitoring, in particular to an on-line stress monitoring device and method based on Modbus-RTU stress sensing and digital communication. Background Stress monitoring is an important means for engineering structure safety assessment and health management, and has important safety significance for real-time and accurate monitoring of structural stress especially in key facilities such as petrochemical industry, bridges, large-scale equipment and the like. The traditional stress monitoring method mainly relies on sensors such as resistance strain gages, micro resistance changes caused by strain are converted into voltage signals through a Wheatstone bridge, and then stress data are output after signal conditioning and analog-to-digital conversion. However, the method has obvious limitations in practical application, on one hand, strain gauge output signals are easily interfered by multidimensional environmental factors such as temperature, time and the like, so that measurement accuracy is reduced, most of conventional compensation means are static or empirical correction and are difficult to adapt to dynamic changes under complex working conditions, on the other hand, the conventional monitoring system is mainly focused on data acquisition and transmission, and lacks an effective diagnosis and performance maintenance mechanism for the state of the sensor, so that monitoring failure or data distortion is easily caused by problems such as lead loosening, component aging and the like in long-term operation. In addition, when abnormal stress is monitored, the system often cannot link adjacent nodes to carry out collaborative analysis, so that dynamic characteristics of stress propagation are difficult to capture, and timeliness of early warning and decision making is limited. Disclosure of Invention In view of the above problems, the invention provides an on-line stress monitoring device and method based on Modbus-RTU stress sensing and digital communication, which can realize accurate sensing and reliable diagnosis of stress data by constructing a dynamic compensation model and a node self-diagnosis mechanism, and solve the problems of insufficient stress monitoring precision and low system reliability in a complex environment. To achieve the above object, in a first aspect, the present application provides an on-line stress monitoring method based on Modbus-RTU stress sensing and digital communication, including: Continuously acquiring original stress data and a corresponding multidimensional environment parameter sequence through a stress sensing node, wherein the multidimensional environment parameter sequence at least comprises a temperature parameter and a time parameter; Constructing an exclusive environment genetic map of each stress sensing node through dynamic regression analysis based on the multidimensional environment parameter sequence and the original stress data, and generating a dynamic compensation model with space-time adaptability; inputting the original stress data acquired in real time into a dynamic compensation model for environmental coupling decoupling and self-adaptive correction to obtain corrected stress measurement values; In the monitoring process, a self-diagnosis flow of the stress sensing node is periodically executed, and the lead connection state of the stress sensing node and the health degree of the sensing element are judged by analyzing the signal impedance characteristics and the stability of the output signal; When the self-diagnosis result is abnormal, generating a device state fault code based on the diagnosis result, and actively uploading the device state fault code to a master station through a Modbus-RTU protocol; judging a threshold value based on the corrected stress measurement value, triggering a collaborative monitoring mechanism when the stress value is detected to exceed a preset stress threshold value, controlling adjacent stress sensing nodes to enter a high-frequency sampling mode, and generating a stress propagation dynamic cloud image; Establishing a node performance base line, continuously monitoring performance attenuation trend, starting a gain self-regeneration flow when the sensitivity or the signal-to-noise ratio is detected to be lower than the self-adaptive threshold, and finely adjusting signal conditioning parameters to restore measurement accuracy; And generating and outputting a comprehensive stress monitoring report based on the corrected stress measurement value, the equipment state fault code, the stress propagation dynamic cloud picture and the gain self-regeneration result. In some embodiments, based on the multidimensional environmental parameter sequence and the original stress data, constructing an exclusive environmental genetic map of each stress se