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

Abstract

The invention discloses an online stress monitoring device and method based on Modbus-RTU stress sensing and digital communication, and the method comprises the steps: collecting original stress data and a multi-dimensional environment parameter sequence through stress sensing nodes, constructing an exclusive environment gene map, generating a dynamic compensation model, and carrying out the environment coupling decoupling and self-adaptive correction of the original data; a periodic self-diagnosis process is executed in the monitoring process, signal impedance characteristics and output signal stability are analyzed, and an equipment state fault code is generated and actively uploaded through a Modbus-RTU protocol when abnormity occurs; triggering a cooperative monitoring mechanism based on the corrected stress measurement value, controlling adjacent nodes to enter a high-frequency sampling mode and generating a stress propagation dynamic cloud; meanwhile, a node performance baseline is established, and a gain self-regeneration process is started when performance degradation is detected. Accurate compensation and dynamic correction of stress data are realized, and the environmental adaptability, the operation reliability and the fault early warning capability of a monitoring system are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of structural health monitoring technology, specifically to an online stress monitoring device and method based on Modbus-RTU stress sensing and digital communication. Background Technology

[0002] Stress monitoring is a crucial tool for structural safety assessment and health management, especially in critical facilities such as petrochemical plants, bridges, and large equipment manufacturing plants, where real-time and accurate monitoring of structural stress is of paramount safety importance. Traditional stress monitoring methods primarily rely on sensors such as resistance strain gauges, which convert minute resistance changes caused by strain into voltage signals via a Wheatstone bridge. These signals are then conditioned and converted from analog to digital before being output as stress data. However, these methods have significant limitations in practical applications: firstly, strain gauge output signals are susceptible to interference from multi-dimensional environmental factors such as temperature and time, leading to decreased measurement accuracy. Furthermore, conventional compensation methods are mostly static or empirical corrections, making it difficult to adapt to dynamic changes under complex operating conditions. Secondly, existing monitoring systems often focus on data acquisition and transmission, lacking effective diagnostic and performance maintenance mechanisms for the sensors themselves. Over long-term operation, issues such as loose leads and component aging can easily lead to monitoring failures or data distortion. In addition, when abnormal stress is detected, the system often cannot coordinate with adjacent nodes for collaborative analysis, making it difficult to capture the dynamic characteristics of stress propagation and limiting the timeliness of early warning and decision-making. Summary of the Invention

[0003] In view of the above problems, the present invention provides an online stress monitoring device and method based on Modbus-RTU stress sensing and digital communication. By constructing a dynamic compensation model and a node self-diagnosis mechanism, it realizes accurate perception and reliable diagnosis of stress data, and solves the problems of insufficient stress monitoring accuracy and low system reliability in complex environments.

[0004] To achieve the above objectives, in a first aspect, this application provides an online stress monitoring method based on Modbus-RTU stress sensing and digital communication, comprising:

[0005] Raw stress data and corresponding multidimensional environmental parameter sequences are continuously collected through stress sensing nodes. The multidimensional environmental parameter sequences include at least temperature parameters and time parameters.

[0006] Based on multidimensional environmental parameter sequences and raw stress data, a unique environmental gene map for each stress sensing node is constructed through dynamic regression analysis, generating a dynamic compensation model with spatiotemporal adaptability.

[0007] The raw stress data collected in real time is input into the dynamic compensation model for environmental coupling decoupling and adaptive correction to obtain the corrected stress measurement value.

[0008] During the monitoring process, the self-diagnosis process of the stress sensing node is periodically executed. By analyzing the signal impedance characteristics and output signal stability, the lead connection status and the health of the sensing element of the stress sensing node are determined.

[0009] When the self-diagnosis result is abnormal, a device status fault code is generated based on the diagnosis result and actively uploaded to the master station via the Modbus-RTU protocol.

[0010] Threshold judgment is made based on the corrected stress measurement value. When the detected stress value exceeds the preset stress threshold, the collaborative monitoring mechanism is triggered to control the adjacent stress sensing nodes to enter the high-frequency sampling mode and generate a dynamic cloud map of stress propagation.

[0011] Establish a node performance baseline, continuously monitor the performance degradation trend, and when the sensitivity or signal-to-noise ratio is detected to be lower than the adaptive threshold, initiate the gain self-regeneration process and fine-tune the signal conditioning parameters to restore measurement accuracy.

[0012] Based on the corrected stress measurements, equipment status fault codes, stress propagation dynamic cloud map, and gain self-regeneration results, a comprehensive stress monitoring report is generated and output.

[0013] In some embodiments, based on multidimensional environmental parameter sequences and raw stress data, a unique environmental gene map for each stress sensing node is constructed through dynamic regression analysis to generate a dynamic compensation model with spatiotemporal adaptability, including:

[0014] Within the preset initial learning period, the synchronous change data of the stress zero-point drift of each stress sensing node and the multi-dimensional environmental parameter sequence are collected.

[0015] A mapping model between multidimensional environmental parameter sequences and stress zero-point drift was established using a multivariate nonlinear regression algorithm.

[0016] The coefficient matrix of the mapping relationship model is stored as the environmental gene map of the corresponding stress sensing node;

[0017] During operation, the coefficient matrix of the environmental gene map is dynamically updated based on the real-time collected multidimensional environmental parameter sequences;

[0018] Based on the updated environmental gene map, a dynamic compensation model with spatiotemporal adaptability is generated.

[0019] In some embodiments, a mapping model between a multidimensional environmental parameter sequence and the zero-point stress drift is established using a multivariate nonlinear regression algorithm, including:

[0020] Data preprocessing is performed on the multidimensional environmental parameter sequence and the zero-point drift of stress.

[0021] Based on the preprocessed multidimensional environmental parameter sequence and the zero-point stress drift, a nonlinear regression equation is constructed with environmental parameters as independent variables and the zero-point stress drift as dependent variables. The nonlinear regression equation includes a quadratic term for the temperature parameter and a cross-coupling term between different environmental parameters.

[0022] The parameter estimates of the nonlinear regression equation are obtained by using the recursive least squares algorithm, thus obtaining the initial mapping relationship model;

[0023] The goodness of fit of the initial mapping relationship model is verified by residual analysis. If the preset accuracy requirement is not met, a regularization term is introduced to optimize the model parameters.

[0024] The optimized model parameters are constructed into a coefficient matrix to form a mapping relationship model.

[0025] In some embodiments, the raw stress data acquired in real time is input into a dynamic compensation model for environmental coupling decoupling and adaptive correction to obtain corrected stress measurements, including:

[0026] The real-time collected multidimensional environmental parameter sequences are input into the environmental gene map, and the stress drift compensation amount under the current environmental conditions is calculated through the mapping relationship model in the environmental gene map.

[0027] An environmental coupling matrix is ​​constructed based on the stress drift compensation amount. The environmental coupling matrix characterizes the coupling strength and direction of each environmental parameter to the stress measurement.

[0028] Tensor operations are performed between the environmental coupling matrix and the real-time acquired raw stress data, and singular value decomposition is used to decouple environmental factors from real stress.

[0029] An adaptive weighted fusion algorithm is used to reconstruct the decoupled multi-dimensional stress components to obtain preliminary stress estimates after environmental interference is eliminated.

[0030] The initial stress estimate is optimized by time-series processing using the Kalman filter algorithm to eliminate the influence of random noise and measurement fluctuations, resulting in optimized stress data.

[0031] The optimized stress data is restored to its original dimensions and converted to engineering units to generate corrected stress measurement values.

[0032] The corrected stress measurement value is stored in the holding register specified by the Modbus-RTU protocol.

[0033] In some embodiments, during the monitoring process, a self-diagnostic procedure for the stress sensing node is periodically executed. By analyzing the signal impedance characteristics and output signal stability, the lead connection status and sensing element health of the stress sensing node are determined, including:

[0034] Within a preset monitoring period, a characteristic excitation signal is applied to the signal path of the stress sensing node, and the corresponding response signal is collected.

[0035] Based on the amplitude and phase relationship between the characteristic excitation signal and the response signal, the real-time impedance characteristic value of the signal path is calculated;

[0036] The real-time impedance characteristic value is compared with the preset impedance reference range. If it exceeds the impedance reference range, the lead connection status is determined to be abnormal.

[0037] The output signal sequence of the stress sensing node is monitored synchronously, and the time-frequency domain characteristics of the output signal are analyzed by wavelet transform.

[0038] High-frequency noise energy and low-frequency drift amplitude of the output signal are extracted as health assessment indicators.

[0039] The health assessment index is compared with the preset stability threshold. If the stability threshold is exceeded, the health of the sensing element is determined to be abnormal.

[0040] When an abnormal lead connection status or sensor health status is detected, the abnormality type and severity are recorded, and the device status fault code generation process is triggered.

[0041] In some embodiments, a threshold judgment is performed based on the corrected stress measurement value. When the detected stress value exceeds a preset stress threshold, a collaborative monitoring mechanism is triggered to control adjacent stress sensing nodes to enter a high-frequency sampling mode and generate a dynamic stress propagation cloud map, including:

[0042] When the corrected stress measurement value of any stress sensing node exceeds the preset stress threshold, the stress sensing node is marked as an abnormal source node.

[0043] Based on the node topology, multiple associated stress sensing nodes adjacent to the anomaly source node are identified and denoted as associated nodes.

[0044] The Modbus-RTU protocol is used to send high-frequency sampling commands to associated nodes, controlling them to enter the preset high-frequency sampling mode.

[0045] Synchronously collect stress data sequences of anomaly source nodes and associated nodes within the same time period;

[0046] The collected stress data sequence is time-aligned and spatially interpolated to construct the spatiotemporal distribution matrix of the stress field;

[0047] Based on the spatiotemporal distribution matrix, a continuous dynamic cloud map of stress propagation is generated using the Kriging interpolation algorithm.

[0048] By analyzing the stress gradient variation characteristics in the dynamic stress propagation cloud map, the propagation path and influence range of stress anomalies are identified, and the identification results are obtained.

[0049] The dynamic cloud map of stress propagation and the identification results are stored.

[0050] In some embodiments, based on the spatiotemporal distribution matrix, a continuous dynamic contour map of stress propagation is generated using the Kriging interpolation algorithm, including:

[0051] Based on the spatial coordinates and stress measurements of each stress sensing node in the spatiotemporal distribution matrix, spatial autocorrelation parameters are calculated.

[0052] A semi-variogram model is constructed based on spatial autocorrelation parameters to determine the weight coefficients of the Kriging interpolation algorithm;

[0053] By using weighting coefficients, the stress values ​​in the unsampled area are estimated optimally and unbiasedly to generate a continuous stress field covering the monitoring area.

[0054] By fusing continuous stress field data with time-dimensional data, a dynamic cloud map of stress propagation is obtained;

[0055] Analyzing the stress gradient variation characteristics in the dynamic stress propagation cloud map, the propagation path and influence range of stress anomalies are identified, and the identification results are obtained, including:

[0056] Calculate the stress gradient vector at each spatial point in the dynamic stress propagation cloud map, and construct the stress gradient field;

[0057] Based on the stress gradient field, the direction of the most significant stress change is traced to determine the propagation path of stress anomalies;

[0058] The region growing algorithm is used to identify connected regions where stress values ​​exceed a preset threshold, thereby determining the range of stress anomalies.

[0059] By combining the propagation path of stress anomalies with the range of their influence, identification results are generated that include the location of the anomaly source, the direction of propagation, and the degree of influence.

[0060] In some embodiments, a node performance baseline is established, and the performance degradation trend is continuously monitored. When the sensitivity or signal-to-noise ratio is detected to be lower than an adaptive threshold, a gain self-regeneration process is initiated to fine-tune the signal conditioning parameters to restore measurement accuracy, including:

[0061] During the initial calibration phase of the stress sensing node, the initial sensitivity, initial signal-to-noise ratio, and initial zero-point stability parameters of the stress sensing node are recorded to establish a node performance baseline.

[0062] During operation, the current sensitivity and signal-to-noise ratio are periodically calculated and compared with the node performance baseline, including:

[0063] The long-term decay trend of the sensitivity and signal-to-noise ratio of stress sensing nodes is identified based on the sliding window, and the degree of performance degradation is obtained.

[0064] When the current sensitivity or current signal-to-noise ratio is detected to be consistently below an adaptive threshold dynamically calculated based on historical data, a gain self-regeneration process is triggered, including:

[0065] Calculate the gain compensation coefficient required for the signal conditioning circuit based on the degree of performance degradation;

[0066] Gain compensation is achieved by fine-tuning the signal conditioning parameters using a digital potentiometer or programmable amplifier.

[0067] Verify whether the current sensitivity and current signal-to-noise ratio after compensation have recovered to the range of the adaptive threshold, and update the node performance baseline.

[0068] In some embodiments, the degree of performance degradation is obtained by identifying the long-term degradation trend of the sensitivity and signal-to-noise ratio of the stress sensing node based on a sliding window, including:

[0069] Within a sliding window of a preset time length, continuous sensitivity sequences and signal-to-noise ratio sequences of stress sensing nodes are collected.

[0070] Trend decomposition was performed on the sensitivity sequence and signal-to-noise ratio sequence to extract their long-term variation components.

[0071] Based on the long-term variation components, the unit time attenuation rate of sensitivity and signal-to-noise ratio is calculated respectively.

[0072] The decay rate per unit time is compared with a preset decay rate threshold to determine whether there is a significant decay trend;

[0073] When there is a significant decay trend, the degree of performance decay is quantitatively calculated based on the relative deviation between the current value and the initial value.

[0074] By combining the degree of sensitivity attenuation and the degree of signal-to-noise ratio attenuation, the final performance attenuation is obtained through a weighted fusion algorithm.

[0075] In a second aspect, the present invention also provides an online stress monitoring device based on Modbus-RTU stress sensing and digital communication, applicable to the method described in the first aspect. The device includes a data acquisition module, an environmental gene map construction module, a stress correction module, a self-diagnosis module, a fault reporting module, a collaborative monitoring module, a performance maintenance module, and a report generation module. The data acquisition module is used to continuously acquire raw stress data and corresponding multidimensional environmental parameter sequences through stress sensing nodes. The environmental gene map construction module is used to construct an environmental gene map specific to each stress sensing node based on the multidimensional environmental parameter sequences and raw stress data through dynamic regression analysis, generating a dynamic compensation model with spatiotemporal adaptability. The stress correction module is used to input the real-time acquired raw stress data into the dynamic compensation model for environmental coupling decoupling and adaptive correction to obtain the corrected stress measurement value. The self-diagnosis module is used to periodically perform self-diagnosis of the stress sensing nodes during the monitoring process. The process involves analyzing signal impedance characteristics and output signal stability to determine the lead connection status and sensing element health of stress sensing nodes. A fault reporting module generates equipment status fault codes based on the self-diagnosis results when abnormal results are found, and actively uploads them to the master station via the Modbus-RTU protocol. A collaborative monitoring module performs threshold judgment based on corrected stress measurements; when a stress value exceeds a preset threshold, a collaborative monitoring mechanism is triggered, controlling adjacent stress sensing nodes to enter high-frequency sampling mode and generating a dynamic stress propagation cloud map. A performance maintenance module establishes a node performance baseline and continuously monitors performance degradation trends; when sensitivity or signal-to-noise ratio falls below an adaptive threshold, a gain self-regeneration process is initiated to fine-tune signal conditioning parameters to restore measurement accuracy. A report generation module generates and outputs a comprehensive stress monitoring report based on corrected stress measurements, equipment status fault codes, dynamic stress propagation cloud maps, and gain self-regeneration results.

[0076] Unlike existing technologies, the above-mentioned technical solution provides an online stress monitoring device and method based on Modbus-RTU stress sensing and digital communication. It collects raw stress data and multi-dimensional environmental parameter sequences through stress sensing nodes, constructs a dedicated environmental gene map, and generates a dynamic compensation model. The raw data undergoes environmental coupling decoupling and adaptive correction. During monitoring, a periodic self-diagnosis process is executed, analyzing signal impedance characteristics and output signal stability. In case of anomalies, a device status fault code is generated and actively uploaded via the Modbus-RTU protocol. Based on the corrected stress measurement values, a collaborative monitoring mechanism is triggered, controlling adjacent nodes to enter high-frequency sampling mode and generating a dynamic stress propagation cloud. Simultaneously, a node performance baseline is established, and a gain self-regeneration process is initiated when performance degradation is detected. This invention achieves accurate compensation and dynamic correction of stress data, significantly improving the environmental adaptability, operational reliability, and fault early warning capability of the monitoring system.

[0077] The above description of the invention is merely an overview of the technical solution of this application. In order to enable those skilled in the art to better understand the technical solution of this application and to implement it based on the description and drawings, and to make the above-mentioned objectives and other objectives, features and advantages of this application easier to understand, the following description is provided in conjunction with the specific embodiments and drawings of this application. Attached Figure Description

[0078] The accompanying drawings are only used to illustrate the principles, implementation methods, applications, features, and effects of specific embodiments of the present invention and other related contents, and should not be considered as limitations on this application.

[0079] In the accompanying drawings of the instruction manual:

[0080] Figure 1 This is a schematic diagram illustrating steps S101 to S108 of the online stress monitoring method described in the specific implementation embodiment;

[0081] Figure 2 This is a schematic diagram illustrating steps S201 to S205 of the online stress monitoring method described in the specific implementation embodiment;

[0082] Figure 3 This is a schematic diagram of the online stress monitoring device described in a specific embodiment.

[0083] The reference numerals used in the above figures are explained as follows:

[0084] 1. Online stress monitoring device;

[0085] 11. Data acquisition module;

[0086] 12. Environmental Genome Mapping Module;

[0087] 13. Stress correction module;

[0088] 14. Self-diagnosis module;

[0089] 15. Fault reporting module;

[0090] 16. Collaborative monitoring module;

[0091] 17. Performance Maintenance Module;

[0092] 18. Report generation module. Detailed Implementation

[0093] To illustrate the possible application scenarios, technical principles, implementable specific solutions, and achievable objectives and effects of this application in detail, the following description, in conjunction with the listed specific embodiments and accompanying drawings, provides a detailed explanation. The embodiments described herein are merely illustrative of the technical solutions of this application and are therefore intended to limit the scope of protection of this application.

[0094] In this document, the term "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The term "embodiment" appearing in various places throughout the specification does not necessarily refer to the same embodiment, nor does it specifically limit its independence or connection with other embodiments. In principle, in this application, as long as there are no technical contradictions or conflicts, the technical features mentioned in each embodiment can be combined in any way to form corresponding implementable technical solutions.

[0095] Unless otherwise defined, the technical terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the use of related terms herein is merely for the purpose of describing particular embodiments and is not intended to limit this application.

[0096] In the description of this application, the term "and / or" is used to describe the logical relationship between objects, indicating that three relationships can exist. For example, A and / or B means: A exists, B exists, and A and B exist simultaneously. Additionally, the character " / " in this document generally indicates that the preceding and following objects have an "or" logical relationship.

[0097] In this application, terms such as “first” and “second” are used only to distinguish one entity or operation from another, and do not necessarily require or imply any actual quantity, hierarchy or order relationship between these entities or operations.

[0098] Without further limitations, the use of terms such as “comprising,” “including,” “having,” or other similar open-ended expressions in this application is intended to cover non-exclusive inclusion, which does not exclude the presence of additional elements in a process, method, or product that includes the stated elements, such that a process, method, or product that includes a list of elements may include not only those defined elements but also other elements not expressly listed, or elements inherent to such a process, method, or product.

[0099] As understood in the Examination Guidelines, in this application, expressions such as "greater than," "less than," and "exceeding" are understood to exclude the stated number; expressions such as "above," "below," and "within" are understood to include the stated number. Furthermore, in the description of the embodiments in this application, "multiple" means two or more (including two), and similar expressions related to "multiple" are also understood in this way, such as "multiple groups" and "multiple times," unless otherwise explicitly specified.

[0100] In the description of the embodiments of this application, the space-related expressions used, such as "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "vertical," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential," indicate the orientation or positional relationship based on the orientation or positional relationship shown in the specific embodiments or drawings. They are only for the purpose of describing the specific embodiments of this application or for the reader's understanding, and do not indicate or imply that the device or component referred to must have a specific position, a specific orientation, or be constructed or operated in a specific orientation. Therefore, they should not be construed as limitations on the embodiments of this application.

[0101] The processor described in the embodiments of this application can be implemented by hardware, firmware, software, or a combination thereof. It can be a circuit, one or more of an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field-programmable gate array (FPGA), a central processing unit (CPU), a controller, a microcontroller, or a microprocessor. It also includes other physical, biological, or chemical structures that can implement the same or equivalent functions as the processors listed above, such as biological neurons, quantum computing units, DNA computing units, etc., so that the processor can execute some or all of the steps in the computer program or method involved in the various embodiments of this application, or any combination of the steps mentioned therein.

[0102] The computer program involved in the embodiments can be stored in a computer device readable storage medium, which includes, but is not limited to, disks, magnetic tapes, magnetic cards, floppy disks, flash memory, optical disks, optical cards, read-only memory (ROM), random access memory (RAM), erasable programmable ROM (EPROM), and electrically erasable programmable ROM (EEPROM), etc., and also includes other biological, physical, or chemical structures that can achieve the same or equivalent functions as the storage media listed above, such as DNA, RNA, proteins, and other units with information storage capabilities. In specific embodiments, the storage medium involved can be one of the above-mentioned media types, or a combination of the above-mentioned media types. In different embodiments, the computer program involved in the embodiments can be centrally stored in a single medium, or distributed and stored in multiple media. The memory containing the computer device readable storage medium can be non-volatile memory or random access memory. These computer device readable storage media can be built into the device, or can be connected to the device involved in the embodiments as an external device or part of an external device. In some embodiments, the memory having a computer device readable storage medium is deployed locally; in other embodiments, the memory may be deployed remotely from the processor, for example, as a network-attached memory accessed via RF circuitry or an external port and a communication network, wherein the communication network may be the Internet, one or more intranets, a local area network (LAN), a wide area network (WLAN), a storage area network (SAN), or a suitable combination thereof, as long as computer device access to the memory is enabled. Furthermore, the computer program involved in the embodiments may be stored in plaintext / ciphertext form, or it may be designed as training data, integrated and recombined through model training and implicitly stored in the parameter states of a deep neural network or other machine learning model.

[0103] Please see Figure 1 In a first aspect, this embodiment provides an online stress monitoring method based on Modbus-RTU stress sensing and digital communication, comprising:

[0104] S101. Continuously collect raw stress data and corresponding multidimensional environmental parameter sequences through stress sensing nodes. The multidimensional environmental parameter sequences include at least temperature parameters and time parameters.

[0105] S102. Based on multidimensional environmental parameter sequences and original stress data, an environmental gene map specific to each stress sensing node is constructed through dynamic regression analysis to generate a dynamic compensation model with spatiotemporal adaptability.

[0106] S103. Input the real-time acquired raw stress data into the dynamic compensation model for environmental coupling decoupling and adaptive correction to obtain the corrected stress measurement value.

[0107] S104. During the monitoring process, the self-diagnosis process of the stress sensing node is periodically executed. By analyzing the signal impedance characteristics and output signal stability, the lead connection status and the health of the sensing element of the stress sensing node are determined.

[0108] S105. When the self-diagnosis result is abnormal, generate a device status fault code based on the diagnosis result and actively upload it to the master station via Modbus-RTU protocol.

[0109] S106. Based on the corrected stress measurement value, a threshold judgment is made. When the stress value is detected to exceed the preset stress threshold, the collaborative monitoring mechanism is triggered to control the adjacent stress sensing nodes to enter the high-frequency sampling mode and generate a stress propagation dynamic cloud map.

[0110] S107. Establish a node performance baseline and continuously monitor the performance degradation trend. When the sensitivity or signal-to-noise ratio is detected to be lower than the adaptive threshold, start the gain self-regeneration process and fine-tune the signal conditioning parameters to restore measurement accuracy.

[0111] S108. Based on the corrected stress measurement values, equipment status fault codes, stress propagation dynamic cloud map, and gain self-regeneration results, generate and output a comprehensive stress monitoring report.

[0112] In step S101, the stress sensing node is a sensing device based on the resistance strain principle, which converts the resistance change caused by strain into a voltage signal through a Wheatstone bridge. The raw stress data represents the initial measurement value without environmental compensation and can be obtained through an analog-to-digital conversion module. The multidimensional environmental parameter sequence includes at least temperature and time parameters, where the temperature parameter is acquired through an integrated temperature sensor, and the time parameter records the timestamp of the data acquisition. This step establishes a data foundation for subsequent environmental compensation by synchronously acquiring stress and environmental data.

[0113] In step S102, the environmental gene map refers to a mathematical model characterizing the dynamic relationship between environmental parameters and stress measurements, constructed through dynamic regression analysis. The dynamic regression analysis employs a recursive least squares algorithm to establish a nonlinear mapping relationship between temperature parameters, time parameters, and stress zero-point drift. The spatiotemporally adaptive dynamic compensation model can automatically adjust compensation parameters according to environmental changes; its spatiotemporal adaptability is reflected in its ability to synchronously compensate for seasonal temperature changes and long-term drift.

[0114] In step S103, environmental coupling decoupling separates the influence of environmental factors on stress measurement through matrix operations, and adaptive correction is a process of real-time correction of the original stress data based on a dynamic compensation model. This step applies the compensation coefficients in the environmental gene map to the real-time data, eliminates environmental interference through weighted fusion and filtering, and finally outputs the corrected stress measurement value to the holding register of the Modbus-RTU protocol.

[0115] In step S104, the self-diagnostic process is achieved by analyzing signal impedance characteristics and output signal stability. Signal impedance characteristics reflect changes in lead connection resistance and are calculated by applying a characteristic excitation signal and measuring the response signal. Output signal stability is assessed by monitoring signal fluctuation amplitude and noise level. Abnormal lead connection status is manifested as an impedance value exceeding a preset range, and abnormal sensor health is manifested as signal drift or increased noise.

[0116] In step S105, the device status fault code is a standardized fault identifier generated based on the self-diagnosis results and is actively uploaded through the exception register specified by the Modbus-RTU protocol. This active upload mechanism ensures that fault information is transmitted to the master station in a timely manner, which, unlike the traditional polling method, improves system response speed.

[0117] In step S106, the preset stress threshold is set based on structural safety standards, and the collaborative monitoring mechanism automatically activates adjacent nodes when an anomaly is detected. The high-frequency sampling mode increases the sampling frequency to several times that of the conventional mode, and the stress propagation dynamic cloud map is generated through a spatial interpolation algorithm, intuitively displaying the spatial distribution characteristics of stress anomalies.

[0118] In step S107, the node performance baseline records key parameters such as sensitivity and signal-to-noise ratio during the initial calibration phase. Performance degradation trends are identified using a sliding window algorithm, and the adaptive threshold is dynamically calculated based on historical data. The gain self-regeneration process recovers performance degradation caused by component aging by adjusting the amplification factor of the signal conditioning circuit.

[0119] In step S108, the integrated stress monitoring report integrates corrected stress data, equipment status information, stress distribution characteristics, and performance maintenance records, and outputs them to the monitoring system via the Modbus-RTU protocol.

[0120] This embodiment achieves accurate stress data measurement and reliable system operation through environmental adaptive compensation, intelligent self-diagnosis, and collaborative monitoring mechanisms. The dynamic compensation model effectively overcomes environmental interference, the self-diagnosis function promptly detects equipment anomalies, and the collaborative monitoring mechanism enhances anomaly identification capabilities. Overall, the solution significantly improves the accuracy and reliability of the stress monitoring system.

[0121] Please see Figure 2 In some embodiments, based on multidimensional environmental parameter sequences and raw stress data, a unique environmental gene map for each stress sensing node is constructed through dynamic regression analysis, generating a dynamic compensation model with spatiotemporal adaptability, including:

[0122] S201. Within the preset initial learning period, collect synchronous change data of the stress zero-point drift of each stress sensing node and the multi-dimensional environmental parameter sequence.

[0123] S202. A mapping relationship model between multidimensional environmental parameter sequences and stress zero-point drift is established using a multivariate nonlinear regression algorithm.

[0124] S203. Store the coefficient matrix of the mapping relationship model as the environmental gene map of the corresponding stress sensing node;

[0125] S204. During operation, the coefficient matrix of the environmental gene map is dynamically updated based on the real-time collected multidimensional environmental parameter sequences.

[0126] S205. Based on the updated environmental gene map, a dynamic compensation model with spatiotemporal adaptability is generated.

[0127] In step S201, the preset initial learning period is a specific time period after the system is put into operation, typically set to 24 to 72 hours, used to establish a baseline relationship between environmental parameters and stress measurements. The stress zero-point drift characterizes the offset of the stress measurement value under no-load conditions, obtained by recording the change of the zero-point stress value over time. The multidimensional environmental parameter sequence includes at least temperature and time parameters, where the temperature parameter is acquired through an integrated temperature sensor, and the time parameter records the timestamp of the data acquisition. Synchronous data changes require that the environmental parameters and stress data have the same time reference; timestamp alignment ensures data correlation. This step, by systematically collecting the correspondence between environmental and stress data, lays the data foundation for establishing an accurate compensation model.

[0128] In step S202, a multivariate nonlinear regression algorithm is used to establish the complex mapping relationship between environmental parameters and stress drift. By constructing a nonlinear equation containing a quadratic temperature term and an interaction term with environmental parameters, it accurately describes the influence of temperature gradient changes and parameter coupling effects on stress measurement. Specifically, the collected data is first standardized to eliminate dimensional differences; then, the regression coefficients are solved using the recursive least squares method, and the goodness of fit of the model is verified through residual analysis; when the fitting accuracy is insufficient, a regularization term is introduced to optimize parameter estimation. This algorithm can effectively capture the nonlinear characteristics between environmental factors and stress drift, providing an accurate mathematical model for subsequent compensation.

[0129] In step S203, the coefficient matrix of the mapping model stores the quantitative relationship between environmental parameters and stress drift, including the weighting coefficients and interaction term coefficients of each environmental parameter. The environmental gene map, serving as the carrier of this coefficient matrix, records the unique characteristics of a specific stress sensing node in its environment. The coefficient matrix is ​​stored in non-volatile memory to ensure data integrity after a system restart. Each stress sensing node possesses an independent environmental gene map, reflecting a comprehensive consideration of individual device differences and environmental specificity.

[0130] In step S204, the dynamic update process is implemented through an online learning algorithm, continuously optimizing the coefficient matrix based on real-time collected environmental data. The update strategy employs a sliding window mechanism, retaining recent data while gradually discarding historical data, ensuring the model maintains both stability and adaptability. The update frequency is adaptively adjusted according to the rate of environmental change; the update frequency is increased when drastic fluctuations in environmental parameters are detected and decreased when the environment is stable. This process enables the environmental genomic map to track long-term environmental change trends, maintaining the timeliness of the compensation model.

[0131] In step S205, the spatiotemporally adaptive dynamic compensation model generates compensation amounts by inputting real-time environmental parameters into the updated environmental gene map. Spatiotemporal adaptability is reflected in the model's ability to simultaneously adapt to temporal features such as seasonal temperature changes and diurnal temperature variations, as well as spatial features such as environmental differences at different installation locations. The model calculates the expected drift amount under current environmental conditions using the coefficient matrix in the environmental gene map, and then compensates and corrects the original stress measurements.

[0132] This embodiment achieves personalized and adaptive environmental compensation for stress measurement by establishing a dedicated environmental gene map and a dynamic update mechanism. A multivariate nonlinear regression algorithm accurately characterizes the complex influence of environmental factors on stress measurement, the dynamic update mechanism ensures the compensation model continuously adapts to environmental changes, and the spatiotemporal adaptive design ensures the solution maintains good performance in different seasons and installation locations. This data-driven compensation method significantly improves the measurement accuracy and reliability of the stress monitoring system in complex environments.

[0133] In some embodiments, a mapping model between a multidimensional environmental parameter sequence and the zero-point stress drift is established using a multivariate nonlinear regression algorithm, including:

[0134] Data preprocessing is performed on the multidimensional environmental parameter sequence and the zero-point drift of stress.

[0135] Based on the preprocessed multidimensional environmental parameter sequence and the zero-point stress drift, a nonlinear regression equation is constructed with environmental parameters as independent variables and the zero-point stress drift as dependent variables. The nonlinear regression equation includes a quadratic term for the temperature parameter and a cross-coupling term between different environmental parameters.

[0136] The parameter estimates of the nonlinear regression equation are obtained by using the recursive least squares algorithm, thus obtaining the initial mapping relationship model;

[0137] The goodness of fit of the initial mapping relationship model is verified by residual analysis. If the preset accuracy requirement is not met, a regularization term is introduced to optimize the model parameters.

[0138] The optimized model parameters are constructed into a coefficient matrix to form a mapping relationship model.

[0139] In this embodiment, the data preprocessing process includes outlier removal, data normalization, and time series alignment. Outlier removal employs a three-standard-deviation principle to exclude measurement data that significantly deviates from the normal range. Data normalization converts environmental parameters of different dimensions to the same order of magnitude, eliminating the impact of differences in parameter magnitudes on modeling. Time series alignment uses linear interpolation to ensure that environmental parameters and stress drift are matched under the same time reference. This step guarantees the quality and consistency of the modeling data, laying the foundation for accurately establishing mapping relationships.

[0140] The nonlinear regression equation is expressed by formula (1) as follows:

[0141] ;

[0142] In formula (1), Indicates the amount of zero-point stress shift. For the first One environmental parameter, For the first One environmental parameter, For constant terms, For the first The coefficient of the first term of each environmental parameter, For the first The environmental parameter and the first The coefficients of the quadratic or cross-term of each environmental parameter. The term represents the random error. The quadratic term for the temperature parameter reflects the nonlinear effect of temperature change, while the cross-coupling term describes the interaction between different environmental parameters. This equation structure can comprehensively characterize the combined influence of environmental parameters on stress drift.

[0143] The recursive least squares algorithm solves for regression coefficients iteratively. Each time new observation data is added, updates are only needed based on the estimation results from the previous time step, avoiding redundant calculations of historical data. The algorithm includes prediction and correction steps. The prediction step calculates the predicted values ​​for new data based on the current parameter estimates, while the correction step adjusts the parameter estimates according to the prediction error. This algorithm is computationally efficient and suitable for online modeling and real-time update requirements.

[0144] Residual analysis quantitatively evaluates the goodness of fit of the initial mapping model by calculating the coefficient of determination (COD) and root mean square (RMS) error. The COD measures the model's ability to explain variations in observed data; a value closer to 1 indicates a better fit. The RMS error quantifies the average deviation between the model's predicted values ​​and the actual observed values. Preset accuracy requirements need to be set according to the specific engineering application scenario, typically requiring a COD higher than 0.85 and an RMS error less than 1% of the sensor's full scale. If the model's evaluation results fail to simultaneously meet these two preset accuracy requirements, a regularization optimization process is triggered.

[0145] When the model's goodness of fit fails to meet the preset accuracy requirement, it indicates that the model may be overfitting, meaning it has excessively adapted to the noise in the training data rather than the underlying patterns. In this case, a regularization term is introduced to optimize the model parameters. This optimization process is achieved by adding a penalty term related to the magnitude of the model parameters to the original loss function. Mathematically, this is achieved by adding a term to the least squares objective function. ,in, Specifically refers to the first parameter in the model parameter vector to be solved. One portion, Let be the total dimension of the model parameter vector to be solved. is the regularization coefficient, whose value is determined through cross-validation and is used to balance model fit and complexity. This optimization method constrains the growth of parameter values, thereby improving the model's generalization ability to new data.

[0146] The optimized model parameters are constructed into a coefficient matrix according to the order of environmental parameters. The rows of the matrix correspond to different environmental parameters and their combinations, while the columns store the corresponding regression coefficient values. This matrix, as a core component of the environmental gene map, fully records the quantitative relationship between environmental parameters and stress drift, providing a calculation basis for dynamic compensation.

[0147] This embodiment establishes an accurate and reliable environmental impact mapping model through a systematic modeling process and rigorous quality control. Data preprocessing ensures the quality of modeling data, nonlinear regression equations fully describe the environmental impact mechanism, recursive least squares ensures computational efficiency, and regularization optimization improves the model's generalization ability. The final mapping relationship model provides a theoretical foundation and technical support for accurate environmental compensation.

[0148] In some embodiments, the raw stress data acquired in real time is input into a dynamic compensation model for environmental coupling decoupling and adaptive correction to obtain corrected stress measurements, including:

[0149] The real-time collected multidimensional environmental parameter sequences are input into the environmental gene map, and the stress drift compensation amount under the current environmental conditions is calculated through the mapping relationship model in the environmental gene map.

[0150] An environmental coupling matrix is ​​constructed based on the stress drift compensation amount. The environmental coupling matrix characterizes the coupling strength and direction of each environmental parameter to the stress measurement.

[0151] Tensor operations are performed between the environmental coupling matrix and the real-time acquired raw stress data, and singular value decomposition is used to decouple environmental factors from real stress.

[0152] An adaptive weighted fusion algorithm is used to reconstruct the decoupled multi-dimensional stress components to obtain preliminary stress estimates after environmental interference is eliminated.

[0153] The initial stress estimate is optimized by time-series processing using the Kalman filter algorithm to eliminate the influence of random noise and measurement fluctuations, resulting in optimized stress data.

[0154] The optimized stress data is restored to its original dimensions and converted to engineering units to generate corrected stress measurement values.

[0155] The corrected stress measurement value is stored in the holding register specified by the Modbus-RTU protocol.

[0156] In this embodiment, the stress drift compensation amount refers to the offset of the stress measurement value caused by the current environmental factors, calculated through the mapping relationship model in the environmental gene map. This calculation process substitutes the real-time collected multi-dimensional environmental parameter sequences, such as temperature and time, into the established regression equation, directly outputting the corresponding compensation amount value. This achieves a quantitative assessment of the environmental impact and provides accurate input data for subsequent compensation.

[0157] The environmental coupling matrix is ​​a square matrix whose dimension equals the number of environmental parameters. The values ​​of the matrix elements are determined by the stress drift compensation and the contribution weights of each environmental parameter. The diagonal elements of the environmental coupling matrix represent the direct coupling strength of a single environmental parameter to stress measurement, while the off-diagonal elements represent the direction and strength of cross-coupling between different environmental parameters. By constructing this matrix, the dispersed environmental influences are integrated into a unified mathematical expression.

[0158] Tensor operations organize the environmental coupling matrix and the original stress data into a high-order tensor structure, where the environmental coupling matrix serves as the transformation kernel and the original stress data as the input vector. This tensor is then decomposed using singular value decomposition (SVD) to obtain a set of singular values ​​and corresponding singular vectors. The magnitude of the singular values ​​characterizes the strength of the environmental coupling effect, while the direction of the singular vectors indicates the interaction mode between environmental factors and the actual stress. By retaining the components corresponding to the main singular values ​​and discarding the minor components, effective separation of environmental noise from the actual stress signal is achieved.

[0159] The adaptive weighted fusion algorithm dynamically assigns weights to each stress component based on its signal-to-noise ratio (SNR) and stability. Components with higher SNR are given larger weights, while those with greater fluctuations are given smaller weights. The weighting coefficients are obtained by calculating and normalizing the inverse variance of each component in real time. This algorithm uses the weighted reconstructed output as a preliminary stress estimate, preserving the effective signal while suppressing random interference.

[0160] The Kalman filter algorithm uses the initial stress estimate as the observation input and performs an optimal estimate of the true stress value through the state equation and the observation equation. The algorithm consists of two recursive steps: prediction and update. The prediction step predicts the current state based on the estimate from the previous time step; the update step corrects the prediction using the current observation and calculates the latest estimation error covariance. This process effectively smooths out random fluctuations in the measurement data and improves the temporal stability of the stress data.

[0161] Dimensional Restoration and Engineering Unit Conversion: Based on the sensor calibration coefficients, the filtered dimensionless data is converted into physically meaningful stress values. The conversion relationship is achieved through linear transformation. The corrected stress measurement values ​​are finally written into the holding register specified by the Modbus-RTU protocol. The specific storage address follows the standard Modbus register mapping specification, facilitating direct reading by the host computer system.

[0162] This embodiment achieves precise decoupling through environmental coupling matrix construction and singular value decomposition, and combines adaptive fusion and Kalman filtering to complete data optimization, forming a complete environmental compensation and data correction process. This scheme effectively overcomes the shortcomings of insufficient accuracy in traditional compensation methods, significantly improves the accuracy of stress measurement in complex environments, and ensures seamless integration with industrial control systems through standardized output.

[0163] In some embodiments, during the monitoring process, a self-diagnostic procedure for the stress sensing node is periodically executed. By analyzing the signal impedance characteristics and output signal stability, the lead connection status and sensing element health of the stress sensing node are determined, including:

[0164] Within a preset monitoring period, a characteristic excitation signal is applied to the signal path of the stress sensing node, and the corresponding response signal is collected.

[0165] Based on the amplitude and phase relationship between the characteristic excitation signal and the response signal, the real-time impedance characteristic value of the signal path is calculated;

[0166] The real-time impedance characteristic value is compared with the preset impedance reference range. If it exceeds the impedance reference range, the lead connection status is determined to be abnormal.

[0167] The output signal sequence of the stress sensing node is monitored synchronously, and the time-frequency domain characteristics of the output signal are analyzed by wavelet transform.

[0168] High-frequency noise energy and low-frequency drift amplitude of the output signal are extracted as health assessment indicators.

[0169] The health assessment index is compared with the preset stability threshold. If the stability threshold is exceeded, the health of the sensing element is determined to be abnormal.

[0170] When an abnormal lead connection status or sensor health status is detected, the abnormality type and severity are recorded, and the device status fault code generation process is triggered.

[0171] In this embodiment, the preset monitoring period is set according to system reliability requirements, typically between 1 and 10 minutes. The characteristic excitation signal is an AC signal with a specific frequency and amplitude, applied to the sensor bridge through a signal conditioning circuit. The response signal is obtained after analog-to-digital conversion, and its amplitude ratio and phase difference with the excitation signal are used to calculate the impedance characteristic value. This step obtains the state parameters of the sensor path through active excitation, providing a basis for connection status diagnosis.

[0172] The real-time impedance characteristic value is obtained through complex number operations. Its real part reflects the line resistance characteristics, while its imaginary part reflects the capacitive and inductive reactance characteristics. The impedance reference range is determined based on the sensor's initial calibration value and the theoretical line impedance, and is typically set to within ±10% of the nominal value. When the measured impedance exceeds this range, it indicates an abnormal condition such as poor contact, breakage, or short circuit in the lead wire.

[0173] The output signal sequence refers to the continuous measurement values ​​output by the stress sensor under normal operating conditions. Wavelet transform employs a multi-resolution analysis method to decompose the signal into different frequency bands. High-frequency noise energy is obtained by calculating the sum of squares of detail coefficients, reflecting the electromagnetic compatibility performance of the sensor and circuit; low-frequency drift amplitude is calculated by the rate of change of the approximation coefficients, characterizing the long-term stability of the sensor.

[0174] The stability threshold is dynamically adjusted based on the sensor's accuracy level and historical operating data. A high-frequency noise energy threshold prevents signal quality degradation, while a low-frequency drift amplitude threshold controls long-term stability. When any indicator exceeds the threshold, it indicates that the sensing element has aged or been damaged. The anomaly type is determined based on the specific indicator exceeding the limit, and the severity is recorded according to the level of the exceedance, providing an accurate basis for subsequent maintenance.

[0175] This embodiment achieves a comprehensive diagnosis of the connection status and component health of stress sensing nodes through impedance analysis and signal stability monitoring. The active excitation method improves the reliability of status detection, wavelet transform ensures the accuracy of signal feature extraction, and the dual-threshold judgment mechanism ensures the precision of fault determination, providing effective technical support for system maintenance.

[0176] In some embodiments, a threshold judgment is performed based on the corrected stress measurement value. When the detected stress value exceeds a preset stress threshold, a collaborative monitoring mechanism is triggered to control adjacent stress sensing nodes to enter a high-frequency sampling mode and generate a dynamic stress propagation cloud map, including:

[0177] When the corrected stress measurement value of any stress sensing node exceeds the preset stress threshold, the stress sensing node is marked as an abnormal source node.

[0178] Based on the node topology, multiple associated stress sensing nodes adjacent to the anomaly source node are identified and denoted as associated nodes.

[0179] The Modbus-RTU protocol is used to send high-frequency sampling commands to associated nodes, controlling them to enter the preset high-frequency sampling mode.

[0180] Synchronously collect stress data sequences of anomaly source nodes and associated nodes within the same time period;

[0181] The collected stress data sequence is time-aligned and spatially interpolated to construct the spatiotemporal distribution matrix of the stress field;

[0182] Based on the spatiotemporal distribution matrix, a continuous dynamic cloud map of stress propagation is generated using the Kriging interpolation algorithm.

[0183] By analyzing the stress gradient variation characteristics in the dynamic stress propagation cloud map, the propagation path and influence range of stress anomalies are identified, and the identification results are obtained.

[0184] The dynamic cloud map of stress propagation and the identification results are stored.

[0185] In this embodiment, the preset stress threshold is pre-set based on the material properties and safety specifications of the monitored structure. When the corrected stress measurement value of any node exceeds this threshold, the system automatically marks that node as an abnormal source node. This marking process triggers the activation of the collaborative monitoring mechanism, providing target location for subsequent refined monitoring.

[0186] The node topology records the spatial location and connectivity of all sensor nodes in the monitoring network. Based on this topology, the system automatically identifies multiple nodes that are physically adjacent to the anomaly source node and defines them as associated nodes, ensuring that the scope of collaborative monitoring can effectively cover the area where stress anomalies may propagate.

[0187] High-frequency sampling commands are encapsulated and transmitted using function code 15 or 16 of the Modbus-RTU protocol, explicitly specifying the sampling frequency boost factor and duration. Upon receiving the command, the associated node adjusts its analog-to-digital converter's sampling rate to a preset high-frequency mode. This mode typically uses a sampling frequency five to ten times higher than the normal mode to capture transient details of stress propagation.

[0188] Temporal alignment uses interpolation to unify the stress data sequences collected from each node to the same timestamp. Spatial interpolation, on the other hand, meshes the discrete measurement points based on the spatial coordinates of the nodes. Through these processes, a distribution matrix that fully reflects the continuous changes of the stress field in the spatiotemporal dimensions is constructed.

[0189] The Kriging interpolation algorithm, based on the principle of spatial autocorrelation, determines the weight contribution of each known data point to the unknown point by calculating the variogram. This algorithm utilizes data from the spatiotemporal distribution matrix to generate a continuous stress field estimate covering the entire monitoring area, and displays the dynamic process of stress propagation in the form of a visual cloud map.

[0190] Stress gradient variation characteristics are obtained by calculating the stress change rate vector at each spatial point in the cloud map. By tracing the direction field of the gradient vector, the main propagation path of stress anomalies can be identified. Simultaneously, a region growing algorithm is used to identify connected regions where stress values ​​exceed the warning threshold, thereby precisely defining the scope of stress anomaly influence.

[0191] This embodiment achieves refined capture and visualization of stress anomaly events through triggered collaborative monitoring and spatial interpolation analysis. The dynamic cloud map generation mechanism transforms discrete nodal measurement data into continuous field distribution information, and gradient analysis and region identification accurately reveal the propagation patterns of anomalies, providing intuitive and reliable spatiotemporal data support for structural safety early warning and diagnosis.

[0192] In some embodiments, based on the spatiotemporal distribution matrix, a continuous dynamic contour map of stress propagation is generated using the Kriging interpolation algorithm, including:

[0193] Based on the spatial coordinates and stress measurements of each stress sensing node in the spatiotemporal distribution matrix, spatial autocorrelation parameters are calculated.

[0194] A semi-variogram model is constructed based on spatial autocorrelation parameters to determine the weight coefficients of the Kriging interpolation algorithm;

[0195] By using weighting coefficients, the stress values ​​in the unsampled area are estimated optimally and unbiasedly to generate a continuous stress field covering the monitoring area.

[0196] By fusing continuous stress field data with time-dimensional data, a dynamic cloud map of stress propagation is obtained;

[0197] Analyzing the stress gradient variation characteristics in the dynamic stress propagation cloud map, the propagation path and influence range of stress anomalies are identified, and the identification results are obtained, including:

[0198] Calculate the stress gradient vector at each spatial point in the dynamic stress propagation cloud map, and construct the stress gradient field;

[0199] Based on the stress gradient field, the direction of the most significant stress change is traced to determine the propagation path of stress anomalies;

[0200] The region growing algorithm is used to identify connected regions where stress values ​​exceed a preset threshold, thereby determining the range of stress anomalies.

[0201] By combining the propagation path of stress anomalies with the range of their influence, identification results are generated that include the location of the anomaly source, the direction of propagation, and the degree of influence.

[0202] In this embodiment, the spatial autocorrelation parameter is used to quantify the spatial dependence of stress measurements, and is obtained by calculating the variogram values ​​between all node pairs. The semi-variogram model describes the variation of spatial autocorrelation with distance, and is typically fitted using a spherical or exponential model. Based on the fitted model, the weighting coefficients required for Kriging interpolation can be calculated. These coefficients ensure that the interpolation results satisfy both the unbiasedness condition and have the minimum estimation variance.

[0203] Using predetermined weighting coefficients, stress values ​​are estimated for locations within the monitoring area where no sensors are deployed. This estimation process is achieved through a weighted linear combination of known nodes, ultimately generating a spatially continuous stress field covering the entire monitoring area, overcoming the limitations of discrete point measurements. The continuous stress field is then overlaid with the corresponding timestamp sequence to form a dynamic stress propagation cloud map. This cloud map, with time as the fourth dimension, visually demonstrates the evolution of the stress field over time, providing a data foundation for analyzing dynamic propagation behavior.

[0204] The stress gradient vector is obtained by calculating the rate of change of the stress field along the coordinate direction at each spatial point. Its magnitude reflects the drasticness of stress change, and its direction indicates the direction of the fastest change. The stress gradient field constructed based on the gradient vector clearly reveals the trend of stress transmission in space.

[0205] The propagation path of stress anomalies is determined by tracing the mainstream direction of vectors in the stress gradient field. Starting from the anomaly source node, path integration along the gradient direction can outline the main diffusion channels of stress anomalies and identify the most susceptible propagation paths in the structure.

[0206] The region growth algorithm uses stress points exceeding a preset threshold as seed points. By iteratively merging adjacent high-stress regions, it gradually expands to form a complete stress anomaly influence range, accurately defining the dangerous areas of stress concentration in the structure.

[0207] The final identification result integrates the precise location of the anomaly source, the geometric description of the propagation path, and the quantitative assessment of the impact range, forming a structured diagnostic report that includes spatial location, directional information, and severity.

[0208] This embodiment transforms discrete nodal measurements into continuous field distribution information through spatial interpolation and gradient analysis techniques, achieving a precise description and visualization of the stress anomaly propagation process. Kriging interpolation ensures the accuracy of spatial estimation, gradient tracking reveals the propagation mechanism, and region growth quantifies the impact range, providing a comprehensive spatiotemporal analysis basis for structural health diagnosis and safety early warning.

[0209] In some embodiments, a node performance baseline is established, and the performance degradation trend is continuously monitored. When the sensitivity or signal-to-noise ratio is detected to be lower than an adaptive threshold, a gain self-regeneration process is initiated to fine-tune the signal conditioning parameters to restore measurement accuracy, including:

[0210] During the initial calibration phase of the stress sensing node, the initial sensitivity, initial signal-to-noise ratio, and initial zero-point stability parameters of the stress sensing node are recorded to establish a node performance baseline.

[0211] During operation, the current sensitivity and signal-to-noise ratio are periodically calculated and compared with the node performance baseline, including:

[0212] The long-term decay trend of the sensitivity and signal-to-noise ratio of stress sensing nodes is identified based on the sliding window, and the degree of performance degradation is obtained.

[0213] When the current sensitivity or current signal-to-noise ratio is detected to be consistently below an adaptive threshold dynamically calculated based on historical data, a gain self-regeneration process is triggered, including:

[0214] Calculate the gain compensation coefficient required for the signal conditioning circuit based on the degree of performance degradation;

[0215] Gain compensation is achieved by fine-tuning the signal conditioning parameters using a digital potentiometer or programmable amplifier.

[0216] Verify whether the current sensitivity and current signal-to-noise ratio after compensation have recovered to the range of the adaptive threshold, and update the node performance baseline.

[0217] In this embodiment, the node performance baseline is a performance benchmark established during the initial calibration phase of the sensor. It is constructed by recording initial sensitivity, initial signal-to-noise ratio, and initial zero-point stability parameters. These parameters characterize the sensor's response capability to input signals, signal quality level, and output stability under no-load conditions, respectively, and together constitute a complete index system for evaluating the sensor's performance status.

[0218] The current sensitivity and signal-to-noise ratio (SNR) are obtained through periodic sampling. Sensitivity is obtained by applying a standard load and calculating the rate of change of the output signal; the SNR is determined by analyzing the ratio of signal power to noise power. The comparative analysis process calculates the difference between the current measured value and the performance baseline reference value, providing data support for performance degradation assessment.

[0219] The sliding window mechanism uses a data series of fixed time length for trend analysis. Linear regression is performed on the sensitivity and signal-to-noise ratio sequences within the window, and the performance degradation rate per unit time is obtained by calculating the regression slope. This degradation rate is compared with a preset degradation rate threshold; if the threshold is exceeded, a significant degradation trend is identified, and the degree of performance degradation is quantified based on the relative deviation between the current value and the initial value.

[0220] The adaptive threshold is dynamically adjusted based on the statistical characteristics of historical operating data, typically set as a specific percentage of the performance baseline. When the current sensitivity or signal-to-noise ratio remains below this threshold for multiple sampling periods, the system automatically triggers a gain self-regeneration process. This dynamic threshold mechanism can adapt to the needs of different operating environments and aging stages.

[0221] The gain compensation coefficient is calculated and determined based on the degree of performance attenuation, and its value is directly proportional to the degree of attenuation. Precise fine-tuning of the signal conditioning parameters is achieved by adjusting the bridge excitation voltage using a digital potentiometer or by adjusting the signal amplification factor using a programmable amplifier. After compensation, the sensitivity and signal-to-noise ratio are remeasured to verify whether they have returned to the normal range, and the performance baseline data is updated accordingly.

[0222] This embodiment achieves autonomous maintenance of sensor performance by establishing a performance baseline and dynamic threshold monitoring. Sliding window trend recognition ensures early detection of performance degradation, and a gain self-regeneration mechanism enables automatic recovery of measurement accuracy, forming a complete closed-loop performance control system that significantly improves the long-term reliability and measurement accuracy of the monitoring system.

[0223] In some embodiments, the degree of performance degradation is obtained by identifying the long-term degradation trend of the sensitivity and signal-to-noise ratio of the stress sensing node based on a sliding window, including:

[0224] Within a sliding window of a preset time length, continuous sensitivity sequences and signal-to-noise ratio sequences of stress sensing nodes are collected.

[0225] Trend decomposition was performed on the sensitivity sequence and signal-to-noise ratio sequence to extract their long-term variation components.

[0226] Based on the long-term variation components, the unit time attenuation rate of sensitivity and signal-to-noise ratio is calculated respectively.

[0227] The decay rate per unit time is compared with a preset decay rate threshold to determine whether there is a significant decay trend;

[0228] When there is a significant decay trend, the degree of performance decay is quantitatively calculated based on the relative deviation between the current value and the initial value.

[0229] By combining the degree of sensitivity attenuation and the degree of signal-to-noise ratio attenuation, the final performance attenuation is obtained through a weighted fusion algorithm.

[0230] In this embodiment, the sliding window of a preset time length is set according to the sensor aging characteristics, typically covering continuous monitoring data from several hours to several days. The sensitivity sequence and signal-to-noise ratio sequence are obtained through periodic measurements, recording the temporal changes in sensor response characteristics and signal quality, respectively.

[0231] Trend decomposition employs moving average or exponential smoothing methods to separate long-term variation components from the original sequence. This component eliminates interference from short-term fluctuations and random noise, accurately reflecting the inherent degradation law of sensor performance and providing a stable basis for quantitative evaluation.

[0232] The decay rate per unit time is obtained by calculating the slope of the linear regression of the long-term variation component, and its value directly characterizes the rate of decline of the performance parameter. The preset decay rate threshold is set based on the sensor's lifespan index. When the calculated value exceeds this threshold, it indicates that the performance decay has exceeded the normal range, and compensation measures need to be initiated.

[0233] The degree of performance degradation is quantified as a percentage of the relative deviation between the current measurement and the initial reference value. Sensitivity degradation reflects the extent of the decrease in sensor responsiveness, while signal-to-noise ratio degradation reflects the degree of signal quality deterioration; together, they constitute a complete dimension of performance evaluation.

[0234] The weighted fusion algorithm assigns weights based on the degree to which sensitivity and signal-to-noise ratio (SNR) affect measurement accuracy. Sensitivity typically has a higher weight because it directly impacts measurement accuracy; the SNR weight is secondary, primarily affecting signal stability. The overall performance degradation is obtained through weighted summation, providing a precise basis for adjusting gain compensation.

[0235] This embodiment achieves precise quantification of sensor performance degradation through sliding window trend analysis and multi-parameter fusion evaluation. Extraction of long-term variation components effectively eliminates short-term interference, and the dual-parameter weighted evaluation comprehensively reflects the performance status, providing a reliable decision-making basis for gain self-regeneration and ensuring the long-term stability of the monitoring system.

[0236] Please see Figure 3In a second aspect, this embodiment also provides an online stress monitoring device 1 based on Modbus-RTU stress sensing and digital communication, applicable to the method described in the first aspect. The device includes a data acquisition module 11, an environmental gene map construction module 12, a stress correction module 13, a self-diagnosis module 14, a fault reporting module 15, a collaborative monitoring module 16, a performance maintenance module 17, and a report generation module 18. The data acquisition module 11 is used to continuously acquire raw stress data and corresponding multidimensional environmental parameter sequences through stress sensing nodes. The environmental gene map construction module 12 is used to construct an environmental gene map specific to each stress sensing node based on the multidimensional environmental parameter sequences and raw stress data through dynamic regression analysis, generating a dynamic compensation model with spatiotemporal adaptability. The stress correction module 13 is used to input the real-time acquired raw stress data into the dynamic compensation model for environmental coupling decoupling and adaptive correction, obtaining the corrected stress measurement value. The self-diagnosis module 14 is used to periodically execute during the monitoring process. The stress sensing node's self-diagnosis process analyzes signal impedance characteristics and output signal stability to determine the lead connection status and sensing element health. The fault reporting module 15 generates a device status fault code based on the self-diagnosis results when abnormal results are found, and actively uploads it to the master station via the Modbus-RTU protocol. The collaborative monitoring module 16 performs threshold judgment based on corrected stress measurements. When a stress value exceeds a preset threshold, a collaborative monitoring mechanism is triggered, controlling adjacent stress sensing nodes to enter high-frequency sampling mode and generating a dynamic stress propagation cloud map. The performance maintenance module 17 establishes a node performance baseline and continuously monitors performance degradation trends. When sensitivity or signal-to-noise ratio falls below an adaptive threshold, a gain self-regeneration process is initiated to fine-tune signal conditioning parameters to restore measurement accuracy. The report generation module 18 generates and outputs a comprehensive stress monitoring report based on corrected stress measurements, device status fault codes, dynamic stress propagation cloud maps, and gain self-regeneration results.

[0237] In this embodiment, the various modules of the device are implemented collaboratively through hardware circuits and embedded software. The data acquisition module 11 includes a Wheatstone bridge and an analog-to-digital converter, used to convert strain gauge resistance changes into digital signals. The environmental gene mapping module 12 uses a microprocessor to execute a dynamic regression algorithm to establish a mapping relationship between environmental parameters and stress drift. The stress correction module 13 uses a digital signal processor to perform environmental coupling and decoupling operations. The self-diagnosis module 14 integrates impedance analysis circuitry and a signal analysis unit for monitoring the health status of the sensors.

[0238] The fault reporting module 15 implements the Modbus-RTU protocol stack through an RS485 interface chip, supporting proactive reporting of abnormal states. The collaborative monitoring module 16 includes a network communication unit and a spatial interpolation processor, enabling multi-node collaborative sampling and cloud map generation. The performance maintenance module 17 incorporates a digital potentiometer and a programmable amplifier, maintaining measurement accuracy by automatically adjusting signal conditioning parameters. The report generation module 18 integrates the output data from all modules to generate a standardized monitoring report.

[0239] This device achieves accurate acquisition, intelligent compensation, and reliable diagnosis of stress data through the coordinated operation of its various functional modules. After the data acquisition module 11 acquires the raw signal, the environmental gene map construction module 12 establishes a compensation model, the stress correction module 13 eliminates environmental interference, the self-diagnosis module 14 monitors the equipment status in real time, the fault reporting module 15 transmits abnormal information in a timely manner, the collaborative monitoring module 16 captures stress propagation characteristics, the performance maintenance module 17 ensures long-term measurement accuracy, and finally, the report generation module 18 outputs comprehensive monitoring results, forming a complete intelligent stress monitoring solution.

[0240] In practical applications, the core sensing unit of the above technical solution is based on the working principle of a resistance strain gauge. The strain gauge is firmly adhered to the surface of the structure being measured using adhesive. When the structure is subjected to stress and deformation, the strain gauge expands and contracts, generating a change in resistance. This change in resistance is converted into a voltage signal via a Wheatstone bridge, amplified by a signal conditioning circuit, and then converted into a digital signal by an analog-to-digital converter. A quarter-bridge configuration is preferred, with the strain gauge serving as one arm of the bridge. Matching compensation resistors eliminate temperature effects, ensuring basic measurement accuracy.

[0241] In practical implementation, the installation of stress sensing nodes must follow a standardized procedure: First, mark the positioning lines on the test piece surface, then grind the area to be fitted to remove oxides and increase roughness, followed by cleaning with alcohol to ensure surface cleanliness. Ceramic adhesive is used as the bonding agent, and a thin-layer application and smoothing process is employed to achieve a firm bond to the strain gauge. The curing process can be selected based on the operating conditions, choosing either room temperature curing for 24 hours or heat curing. The latter, involving holding at 80-150℃ for 2 hours followed by slow cooling, yields more stable bonding performance.

[0242] The hardware implementation of the device is based on a dedicated signal conditioning module, which integrates calibration, filtering, and zeroing functions, supporting a measurement accuracy of ±0.03% and an internal resolution of 10 grams. The module uses an RS485 interface to implement the Modbus-RTU communication protocol, supporting a network of up to 247 nodes by mapping stress measurements, status flags, and other parameters through holding registers. In actual deployment, the sensor nodes are connected via a five-pin interface, with a 6-12V DC power supply and A+ / B- differential signal transmission for communication, ensuring stable communication within a 1000-meter range.

[0243] For system maintenance, a dedicated production configuration tool is provided to support parameter settings and firmware upgrades. This tool communicates with the module via a USB-to-serial chip and supports online program updates. When module performance degradation is detected, gain self-regeneration can be achieved by adjusting the parameters of the digital potentiometer or programmable amplifier, ensuring the stability of measurement accuracy during long-term operation.

[0244] By adopting the above technical solutions, this invention differs from existing technologies and possesses the following beneficial effects: By constructing a unique environmental gene map for stress sensing nodes and generating a dynamic compensation model with spatiotemporal adaptability, it effectively overcomes the coupling interference of multi-dimensional environmental factors such as temperature and time on stress measurement, significantly improving measurement accuracy and data reliability under complex working conditions. The system monitors the lead connection status and sensing element health in real time through a periodic self-diagnosis process, combined with an active reporting mechanism to ensure timely transmission of fault information, enhancing the operational stability and maintainability of the monitoring system. The collaborative monitoring mechanism triggered when an anomaly is detected achieves precise visualization of the stress propagation path and impact range through high-frequency sampling and dynamic cloud map generation technology, providing an intuitive basis for structural safety early warning. Simultaneously, the gain self-regeneration process based on the node performance baseline can automatically compensate for sensor performance degradation, maintaining long-term measurement accuracy and solving the problem of accuracy decline caused by component aging in traditional monitoring systems. The final comprehensive stress monitoring report integrates environmental compensation data, equipment status information, and anomaly propagation characteristics, forming a complete intelligent monitoring closed loop, comprehensively improving the accuracy, reliability, and intelligence level of stress monitoring.

[0245] Finally, it should be noted that although the above embodiments have been described in the text and drawings of this application, this should not limit the scope of patent protection of this application. Any technical solutions that are based on the essential concept of this application and utilize the content described in the text and drawings of this application, resulting in equivalent structural or procedural substitutions or modifications, as well as the direct or indirect application of the technical solutions of the above embodiments to other related technical fields, are all included within the scope of patent protection of this application.

Claims

1. An online stress monitoring method based on Modbus-RTU stress sensing and digital communication, characterized in that, include: The raw stress data and the corresponding multidimensional environmental parameter sequence are continuously collected through stress sensing nodes. The multidimensional environmental parameter sequence includes at least temperature parameters and time parameters. Based on the multidimensional environmental parameter sequence and the original stress data, a unique environmental gene map for each stress sensing node is constructed through dynamic regression analysis, generating a dynamic compensation model with spatiotemporal adaptability. The raw stress data collected in real time is input into the dynamic compensation model for environmental coupling decoupling and adaptive correction to obtain the corrected stress measurement value. During the monitoring process, the self-diagnosis process of the stress sensing node is periodically executed. By analyzing the signal impedance characteristics and output signal stability, the lead connection status and the health of the sensing element of the stress sensing node are determined. When the self-diagnosis result is abnormal, a device status fault code is generated based on the diagnosis result and actively uploaded to the master station via the Modbus-RTU protocol. Based on the corrected stress measurement value, a threshold judgment is made. When the stress value is detected to exceed the preset stress threshold, a collaborative monitoring mechanism is triggered to control the adjacent stress sensing nodes to enter the high-frequency sampling mode and generate a dynamic cloud map of stress propagation. Establish a node performance baseline, continuously monitor the performance degradation trend, and when the sensitivity or signal-to-noise ratio is detected to be lower than the adaptive threshold, initiate the gain self-regeneration process and fine-tune the signal conditioning parameters to restore measurement accuracy. Based on the corrected stress measurement values, equipment status fault codes, stress propagation dynamic cloud map, and gain self-regeneration results, a comprehensive stress monitoring report is generated and output.

2. The online stress monitoring method based on Modbus-RTU stress sensing and digital communication according to claim 1, characterized in that, Based on the aforementioned multidimensional environmental parameter sequence and original stress data, a unique environmental gene map for each stress sensing node is constructed through dynamic regression analysis, generating a dynamic compensation model with spatiotemporal adaptability, including: Within the preset initial learning period, the stress zero-point drift of each stress sensing node and the synchronous change data of the multidimensional environmental parameter sequence are collected. A mapping model between the multidimensional environmental parameter sequence and the zero-point drift of stress is established by a multivariate nonlinear regression algorithm. The coefficient matrix of the mapping relationship model is stored as an environmental gene map of the corresponding stress sensing node; During operation, the coefficient matrix of the environmental gene map is dynamically updated based on the real-time collected multidimensional environmental parameter sequences. Based on the updated environmental gene map, a dynamic compensation model with spatiotemporal adaptability is generated.

3. The online stress monitoring method based on Modbus-RTU stress sensing and digital communication according to claim 2, characterized in that, A mapping model between the multidimensional environmental parameter sequence and the zero-point stress drift is established using a multivariate nonlinear regression algorithm, including: Data preprocessing is performed on the multidimensional environmental parameter sequence and the stress zero-point drift. Based on the preprocessed multidimensional environmental parameter sequence and the stress zero-point drift, a nonlinear regression equation is constructed with environmental parameters as independent variables and stress zero-point drift as dependent variables. The nonlinear regression equation includes a quadratic term for the temperature parameter and a cross-coupling term between different environmental parameters. The parameter estimates of the nonlinear regression equation are obtained by using the recursive least squares algorithm, thus obtaining the initial mapping relationship model; The goodness of fit of the initial mapping relationship model is verified by residual analysis. If the preset accuracy requirement is not met, a regularization term is introduced to optimize the model parameters. The optimized model parameters are constructed into a coefficient matrix to form the mapping relationship model.

4. The online stress monitoring method based on Modbus-RTU stress sensing and digital communication according to claim 1, characterized in that, The raw stress data acquired in real time is input into the dynamic compensation model for environmental coupling decoupling and adaptive correction to obtain the corrected stress measurement values, including: The multidimensional environmental parameter sequence collected in real time is input into the environmental gene map, and the stress drift compensation amount under the current environmental conditions is calculated through the mapping relationship model in the environmental gene map. An environmental coupling matrix is ​​constructed based on the stress drift compensation amount. The environmental coupling matrix characterizes the coupling strength and direction of each environmental parameter to the stress measurement. The environmental coupling matrix is ​​subjected to tensor operations with the real-time acquired raw stress data, and the environmental factors and the real stress are decoupled and separated through singular value decomposition. An adaptive weighted fusion algorithm is used to reconstruct the decoupled multi-dimensional stress components to obtain preliminary stress estimates after environmental interference is eliminated. The preliminary stress estimate is optimized by using a Kalman filter algorithm to eliminate the influence of random noise and measurement fluctuations, resulting in optimized stress data. The optimized stress data is restored to its original dimensions and converted to engineering units to generate corrected stress measurement values. The corrected stress measurement value is stored in the holding register specified by the Modbus-RTU protocol.

5. The online stress monitoring method based on Modbus-RTU stress sensing and digital communication according to claim 1, characterized in that, During monitoring, a self-diagnostic process for the stress sensing nodes is periodically executed. By analyzing signal impedance characteristics and output signal stability, the lead connection status and sensing element health of the stress sensing nodes are determined, including: Within a preset monitoring period, a characteristic excitation signal is applied to the signal path of the stress sensing node, and the corresponding response signal is collected. Based on the amplitude and phase relationship between the characteristic excitation signal and the response signal, the real-time impedance characteristic value of the signal path is calculated; The real-time impedance characteristic value is compared with a preset impedance reference range. If it exceeds the impedance reference range, the lead connection status is determined to be abnormal. The output signal sequence of the stress sensing node is monitored synchronously, and the time-frequency domain characteristics of the output signal are analyzed by wavelet transform. High-frequency noise energy and low-frequency drift amplitude of the output signal are extracted as health assessment indicators. The health assessment index is compared with a preset stability threshold. If the health threshold is exceeded, the sensor element is determined to be abnormal. When an abnormal lead connection status or sensor health status is detected, the abnormality type and severity are recorded, and the device status fault code generation process is triggered.

6. The online stress monitoring method based on Modbus-RTU stress sensing and digital communication according to claim 1, characterized in that, Based on the corrected stress measurement value, a threshold judgment is performed. When the detected stress value exceeds the preset stress threshold, a collaborative monitoring mechanism is triggered to control adjacent stress sensing nodes to enter a high-frequency sampling mode, generating a dynamic stress propagation cloud map, including: When the corrected stress measurement value of any stress sensing node exceeds the preset stress threshold, the stress sensing node is marked as an abnormal source node. Based on the node topology, multiple associated stress sensing nodes adjacent to the anomaly source node are identified and denoted as associated nodes. The Modbus-RTU protocol is used to send high-frequency sampling commands to the associated nodes, controlling them to enter a preset high-frequency sampling mode; The stress data sequences of the anomaly source node and associated node within the same time period are collected simultaneously. The collected stress data sequence is time-aligned and spatially interpolated to construct the spatiotemporal distribution matrix of the stress field; Based on the aforementioned spatiotemporal distribution matrix, a continuous dynamic cloud map of stress propagation is generated using the Kriging interpolation algorithm. By analyzing the stress gradient change characteristics in the stress propagation dynamic cloud map, the propagation path and influence range of stress anomalies are identified, and the identification results are obtained. The stress propagation dynamic cloud map and the identification results are stored.

7. The online stress monitoring method based on Modbus-RTU stress sensing and digital communication according to claim 6, characterized in that, Based on the aforementioned spatiotemporal distribution matrix, a continuous dynamic contour map of stress propagation is generated using the Kriging interpolation algorithm, including: Based on the spatial coordinates and stress measurement values ​​of each stress sensing node in the spatiotemporal distribution matrix, spatial autocorrelation parameters are calculated. A semi-variogram model is constructed based on the spatial autocorrelation parameters to determine the weight coefficients of the Kriging interpolation algorithm; The stress values ​​in the unsampled area are estimated optimally using the weighting coefficients to generate a continuous stress field covering the monitoring area. The continuous stress field is fused with time-dimensional data to obtain the stress propagation dynamic cloud map; Analyzing the stress gradient change characteristics in the stress propagation dynamic cloud map, the propagation path and influence range of stress anomalies are identified, and the identification results are obtained, including: Calculate the stress gradient vector at each spatial point in the stress propagation dynamic cloud map, and construct the stress gradient field; Based on the stress gradient field, the direction of the most significant stress change is traced to determine the stress anomaly propagation path; The region growing algorithm is used to identify connected regions where stress values ​​exceed a preset threshold, thereby determining the range of stress anomalies. By combining the propagation path of stress anomalies with the range of their influence, identification results are generated that include the location of the anomaly source, the direction of propagation, and the degree of influence.

8. The online stress monitoring method based on Modbus-RTU stress sensing and digital communication according to claim 1, characterized in that, Establish a node performance baseline, continuously monitor performance degradation trends, and when sensitivity or signal-to-noise ratio is detected to be below the adaptive threshold, initiate a gain self-regeneration process to fine-tune signal conditioning parameters to restore measurement accuracy, including: During the initial calibration phase of the stress sensing node, the initial sensitivity, initial signal-to-noise ratio, and initial zero-point stability parameters of the stress sensing node are recorded to establish a node performance baseline. During operation, the current sensitivity and current signal-to-noise ratio are periodically calculated and compared with the node performance baseline, including: The long-term decay trend of the sensitivity and signal-to-noise ratio of stress sensing nodes is identified based on the sliding window, and the degree of performance degradation is obtained. When the current sensitivity or current signal-to-noise ratio is detected to be consistently below an adaptive threshold dynamically calculated based on historical data, a gain self-regeneration process is triggered, including: Calculate the gain compensation coefficient required for the signal conditioning circuit based on the degree of performance degradation; Gain compensation is achieved by fine-tuning the signal conditioning parameters using a digital potentiometer or programmable amplifier. Verify whether the current sensitivity and current signal-to-noise ratio after compensation have recovered to the range of the adaptive threshold, and update the node performance baseline.

9. The online stress monitoring method based on Modbus-RTU stress sensing and digital communication according to claim 8, characterized in that, Based on the long-term degradation trend of the sensitivity and signal-to-noise ratio of stress sensing nodes identified by the sliding window, the degree of performance degradation is obtained, including: Within a sliding window of a preset time length, continuous sensitivity sequences and signal-to-noise ratio sequences of stress sensing nodes are collected. The sensitivity sequence and signal-to-noise ratio sequence are respectively subjected to trend decomposition to extract their long-term variation components; Based on the long-term variation components, the unit time attenuation rate of sensitivity and signal-to-noise ratio is calculated respectively. The decay rate per unit time is compared with a preset decay rate threshold to determine whether there is a significant decay trend; When there is a significant decay trend, the degree of performance decay is quantitatively calculated based on the relative deviation between the current value and the initial value. By combining the degree of sensitivity attenuation and the degree of signal-to-noise ratio attenuation, the final performance attenuation is obtained through a weighted fusion algorithm.

10. An online stress monitoring device based on Modbus-RTU stress sensing and digital communication, characterized in that, The apparatus applicable to the method of any one of claims 1 to 9 comprises: The data acquisition module is used to continuously acquire raw stress data and corresponding multidimensional environmental parameter sequences through stress sensing nodes; The environmental gene map construction module is used to construct an environmental gene map specific to each stress sensing node based on the multidimensional environmental parameter sequence and the original stress data through dynamic regression analysis, and generate a dynamic compensation model with spatiotemporal adaptability. The stress correction module is used to input the real-time acquired raw stress data into the dynamic compensation model for environmental coupling decoupling and adaptive correction, so as to obtain the corrected stress measurement value. The self-diagnostic module is used to periodically execute the self-diagnostic process of the stress sensing node during the monitoring process. By analyzing the signal impedance characteristics and output signal stability, it can determine the lead connection status and the health of the sensing element of the stress sensing node. The fault reporting module is used to generate a device status fault code based on the self-diagnosis result when the self-diagnosis result is abnormal, and actively upload it to the master station via the Modbus-RTU protocol; The collaborative monitoring module is used to make a threshold judgment based on the corrected stress measurement value. When the stress value is detected to exceed the preset stress threshold, the collaborative monitoring mechanism is triggered to control the adjacent stress sensing nodes to enter the high-frequency sampling mode and generate a dynamic cloud map of stress propagation. The performance maintenance module is used to establish a node performance baseline, continuously monitor the performance degradation trend, and when the sensitivity or signal-to-noise ratio is detected to be lower than the adaptive threshold, the gain self-regeneration process is initiated to fine-tune the signal conditioning parameters to restore measurement accuracy. The report generation module is used to generate and output a comprehensive stress monitoring report based on the corrected stress measurement values, equipment status fault codes, stress propagation dynamic cloud map, and gain self-regeneration results.

Images