Method for monitoring thermal expansion stress of thermal energy storage system

Abstract

The invention provides a thermal expansion stress monitoring method for a thermal energy storage system, and belongs to the technical field of thermal energy storage systems. A multi-physics field coupling sensor array is deployed, and a thermal expansion stress component and other load stress components are separated by adopting an empirical mode decomposition algorithm; the method comprises the following steps: establishing a self-adaptive grid refined finite difference thermal stress prediction model, combining with a Kalman filtering algorithm to realize transient thermal shock stress prediction, simplifying stress component matrix calculation through spectral decomposition and a diagonalization frame, and establishing a stress threshold calculation equation set; intelligent switching of a standard operation mode, a stress monitoring enhancement mode and a stress adaptive adjustment mechanism is realized according to a comparison result of the stress concentration coefficient and a threshold value, and the technical problem that thermal expansion stress and other load coupling stress cannot be accurately separated and predicted in the operation process of the thermal energy storage system is solved.

Description

Technical Field

[0001] This invention belongs to the technical field of thermal energy storage systems, and more specifically, relates to a method for monitoring thermal expansion stress in thermal energy storage systems. Background Technology

[0002] Thermal energy storage systems, as a key technology in the new energy field, subject their storage containers to complex multi-load coupling during operation. Traditional stress monitoring methods primarily employ single-parameter monitoring and static threshold judgment mechanisms, assessing the stress state at single points or localized areas by deploying strain gauges or temperature sensors, and calculating thermal stress distribution based on empirical formulas or simplified models. However, due to the complex coupling of multiple loads such as thermal loads, mechanical loads, and gravity loads during the operation of thermal energy storage systems, traditional monitoring methods cannot effectively separate stress components generated by different load sources. This results in insufficient accuracy in thermal expansion stress assessment, fixed prediction model parameters lacking adaptive adjustment capabilities, and lag in monitoring system response. Traditional technologies struggle to accurately identify and separate thermal expansion stress under transient thermal shock and complex load coupling conditions, failing to provide reliable stress early warning and adjustment basis for the safe operation of thermal storage systems. Furthermore, there is a technical problem of inaccurate separation and prediction of thermal expansion stress coupled with other loads. Summary of the Invention

[0003] In view of this, the present invention provides a method for monitoring thermal expansion stress in a thermal energy storage system, which can solve the technical problem in the prior art that thermal expansion stress and other load coupling stress cannot be accurately separated and predicted during the operation of a thermal energy storage system.

[0004] This invention is implemented as follows: A method for monitoring thermal expansion stress in a thermal energy storage system includes: deploying a multi-physics coupled sensor array at key nodes on the wall of the thermal storage container; establishing a multi-load response feature library for the thermal energy storage system by applying known thermal loads, mechanical loads, and gravity loads to the thermal storage container, and collecting corresponding strain response signals, temperature change signals, and vibration signals to form a multi-load response feature library dataset; real-time acquisition of multi-physics signals during the operation of the thermal storage container, preprocessing the acquired signals to obtain preprocessed multi-physics monitoring data; and using an empirical mode decomposition algorithm to decompose the preprocessed multi-physics monitoring data, separating the thermal expansion stress component and the mechanical load stress component. The stress components of the loads, including gravity loads, are used to identify and separate the sources of multi-load coupled stress, resulting in a decomposed stress component matrix. An adaptive mesh refinement finite difference thermal stress prediction model is established, and the model parameters are corrected in real time using a Kalman filter algorithm to calculate the predicted value of transient thermal shock stress, obtaining a thermal expansion stress distribution cloud map and stress concentration factor. A spectral decomposition and diagonalization framework for the stress component matrix is ​​established. By diagonalizing the decomposed stress component matrix to simplify the calculation of matrix powers and functions, thermal expansion stress evaluation indices are calculated. The stress threshold calculation equations are used to obtain the operation mode switching threshold and early warning control threshold. The operating parameters of the thermal storage system are adjusted based on the comparison results between the thermal expansion stress evaluation indices and the operation mode switching threshold and early warning control threshold.

[0005] The multi-physics coupled sensor array includes strain gauge sensors, temperature sensors, and vibration acceleration sensors. The sensor spacing is 1 / 10 of the characteristic length of the thermal storage container, and the sensor sampling frequency is set to 1000Hz.

[0006] The preprocessing steps for the acquired signals specifically include noise reduction filtering, signal calibration, and data synchronization.

[0007] Among them, the multi-load response feature library refers to a database that stores the system response characteristics under different load conditions, which serves as a reference for load identification and stress separation.

[0008] The decomposed stress component matrix refers to the matrix data structure containing thermal expansion stress components, mechanical load stress components, and gravity load stress components obtained by the empirical mode decomposition algorithm.

[0009] The stress concentration factor is a dimensionless parameter that characterizes the degree of non-uniformity in stress distribution. It is defined as the ratio of the maximum stress to the average stress.

[0010] Among them, the spectral decomposition and diagonalization framework refers to the mathematical method of performing eigenvalue decomposition and diagonalization on the decomposed stress component matrix, which is used to simplify matrix operations and extract the main stress features.

[0011] The stress threshold calculation equation set includes a stability standard threshold calculation equation and a hazard warning threshold calculation equation. The stability standard threshold calculation equation is used to determine the upper limit of the stress concentration factor for normal system operation. The inputs include the yield strength of the thermal storage container material, the safety factor, and the temperature correction factor. The output is the stability standard threshold. The hazard warning threshold calculation equation is used to determine the critical value of the stress concentration factor for activating the stress adaptive adjustment mechanism. The inputs include the ultimate strength of the thermal storage container material, the fatigue correction factor, and the ambient temperature coefficient. The output is the hazard warning threshold.

[0012] Specifically, when the stress concentration factor is less than the stability standard threshold, the standard operation mode is maintained; when the stress concentration factor is ∈ [stability standard threshold, danger warning threshold), the stress monitoring enhancement mode is activated, and the sampling frequency is increased to 2000Hz.

[0013] When the stress concentration factor is greater than or equal to the danger warning threshold, the stress adaptive adjustment mechanism is activated, and the stress level is controlled by adjusting the flow rate of the heat storage medium, reducing the working temperature, and activating the stress compensation mechanism.

[0014] Among them, Empirical Mode Decomposition (EMD) refers to an adaptive signal decomposition method used to decompose a composite signal into several intrinsic mode function (EMF) components, achieving separation of different frequency components. Kalman Filtering (KB) is an optimal estimation algorithm that uses state equations and observation equations to achieve real-time estimation of the system state and parameter correction. The filtering formula is expressed as follows: The adaptive mesh refinement finite difference thermal stress prediction model refers to a numerical calculation model based on the finite difference method. It improves the calculation accuracy by adaptively adjusting the mesh density and is suitable for handling transient heat transfer problems with large temperature gradients.

[0015] Among them, stress compensation mechanisms refer to mechanical devices used to alleviate thermal expansion stress, including expansion joints, stress compensators, and flexible connectors.

[0016] This invention establishes a multi-physics coupled sensor array and a multi-load response feature library, combined with an empirical mode decomposition algorithm, to achieve accurate separation of stress components from different load sources. It also establishes a transient thermal stress prediction system through an adaptive mesh refinement finite difference model and a Kalman filter algorithm. This invention simplifies the calculation of the stress component matrix using a spectral decomposition and diagonalization framework, and establishes a multi-level early warning mechanism through a set of stress threshold calculation equations. This enables intelligent switching from standard operation mode to stress monitoring enhancement mode and then to a stress adaptive adjustment mechanism, overcoming the shortcomings of traditional methods in separating coupled stresses and insufficient prediction accuracy. Through the technical path of multi-load source identification and separation, real-time parameter correction, and adaptive adjustment mechanism, this invention constructs a complete thermal expansion stress monitoring system from four levels: monitoring, identification, prediction, and adjustment, solving the technical problem of inaccurate separation and prediction of thermal expansion stress and other load coupled stresses during the operation of thermal energy storage systems. Attached Figure Description

[0017] Figure 1 This is a flowchart of the method of the present invention.

[0018] Figure 2 This is a spatial distribution diagram of the stress concentration factor of the thermal storage container in the embodiment.

[0019] Figure 3 This is a time history diagram of the stress adaptive adjustment process in the embodiment. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0021] like Figure 1 The diagram shown is a flowchart of a method for monitoring thermal expansion stress in a thermal energy storage system provided by the present invention. This method includes the following steps: S01. Deploy a multi-physics field coupled sensor array at key nodes on the wall of the thermal storage container of the thermal energy storage system. The multi-physics field coupled sensor array includes strain gauge sensors, temperature sensors, and vibration acceleration sensors. The sensor spacing is 1 / 10 of the characteristic length of the thermal storage container, and the sensor sampling frequency is set to 1000Hz. S02. Establish a multi-load response feature library for thermal energy storage systems. By applying known thermal loads, mechanical loads, and gravity loads to the thermal storage container, collect the corresponding strain response signals, temperature change signals, and vibration signals to form a multi-load response feature library dataset. S03. Real-time acquisition of multi-physics field signals during the operation of the thermal storage container, preprocessing of the acquired signals including noise reduction filtering, signal calibration, and data synchronization to obtain preprocessed multi-physics field monitoring data; S04. The empirical mode decomposition algorithm is used to decompose the preprocessed multi-physics monitoring data to separate the thermal expansion stress component, mechanical load stress component, and gravity load stress component, thereby realizing the source identification and separation of multi-load coupled stress and obtaining the decomposed stress component matrix. S05. Establish an adaptive mesh refinement finite difference thermal stress prediction model, combine the Kalman filter algorithm to correct the model parameters in real time, calculate the predicted value of transient thermal shock stress, and obtain the thermal expansion stress distribution cloud map and stress concentration factor. S06. Establish a spectral decomposition and diagonalization framework for the stress component matrix. By diagonalizing the decomposed stress component matrix to simplify the matrix power and function calculation, calculate the thermal expansion stress evaluation index. Use the stress threshold calculation equation set to obtain the operation mode switching threshold and early warning control threshold. S07. Adjust the operating parameters of the thermal storage system based on the comparison results of the thermal expansion stress assessment index with the operation mode switching threshold and the early warning control threshold. When the stress concentration coefficient is less than the stability standard threshold, maintain the standard operation mode to ensure system reliability. When the stress concentration coefficient is ∈ [stability standard threshold, danger early warning threshold), activate the stress monitoring enhancement mode and increase the sampling frequency to 2000Hz. When the stress concentration coefficient is ≥ the danger early warning threshold, activate the stress adaptive adjustment mechanism to cope with the changes. Stress level control is achieved by adjusting the flow rate of the thermal storage medium, reducing the operating temperature, and activating the stress compensation mechanism.

[0022] Among them, a multi-physics coupled sensor array refers to a sensor combination that integrates multiple physical quantity detection functions, used to simultaneously monitor parameters such as strain, temperature, and vibration. A multi-load response feature library is a database storing the system response characteristics under different load conditions, serving as a reference for load identification and stress separation. Empirical mode decomposition (EMD) algorithms are adaptive signal decomposition methods used to decompose composite signals into several intrinsic mode function components, achieving separation of different frequency components.

[0023] The decomposed stress component matrix is ​​a matrix data structure containing thermal expansion stress components, mechanical load stress components, and gravity load stress components, obtained through empirical mode decomposition algorithms. The adaptive mesh refinement finite difference thermal stress prediction model is a numerical calculation model based on the finite difference method. It improves calculation accuracy by adaptively adjusting the mesh density and is particularly suitable for handling transient heat transfer problems with large temperature gradients.

[0024] The Kalman filter algorithm is an optimal estimation algorithm that achieves real-time estimation of the system state and parameter correction through state equations and observation equations. The filtering formula is expressed as follows: ,in This is the state estimate. For Kalman gain, For the observed values, , , These are the corresponding reference values. The stress concentration factor is a dimensionless parameter characterizing the degree of non-uniformity in stress distribution, defined as the ratio of the maximum stress to the average stress.

[0025] The spectral decomposition and diagonalization framework is a mathematical method for eigenvalue decomposition and diagonalization of the decomposed stress component matrix, used to simplify matrix operations and extract key stress features. The stress threshold calculation equation set includes a stability standard threshold calculation equation and a hazard warning threshold calculation equation. The stability standard threshold calculation equation is used to determine the upper limit of the stress concentration factor for normal system operation; inputs include the yield strength of the thermal storage container material, the safety factor, and the temperature correction factor; the output is the stability standard threshold. The hazard warning threshold calculation equation is used to determine the critical value of the stress concentration factor for activating the adaptive stress adjustment mechanism; inputs include the ultimate strength of the thermal storage container material, the fatigue correction factor, and the ambient temperature coefficient; the output is the hazard warning threshold.

[0026] The enhanced stress monitoring mode is a high-frequency monitoring mode used when an increase in stress level is detected. It enhances monitoring accuracy by increasing the sampling frequency. The stress adaptive adjustment mechanism is an automatic adjustment system activated when the stress level reaches a dangerous warning threshold, using various adjustment methods to reduce the stress level. The stress compensation mechanism is a mechanical device used to alleviate thermal expansion stress, including components such as expansion joints, stress compensators, and flexible connectors. The thermal storage medium flow rate regulation adjusts the heat transfer rate by changing the flow velocity of the thermal storage medium, thereby controlling the rate of temperature change and the level of thermal stress.

[0027] Optionally, the present invention is also implemented by a computer to form a thermal expansion stress monitoring system for a thermal energy storage system. The computer is equipped with a readable storage medium, which stores program instructions. When the program instructions are run in the computer, they are used to execute the above-mentioned thermal expansion stress monitoring method for a thermal energy storage system.

[0028] The specific implementation methods of the above steps are described in detail below.

[0029] The specific implementation of step S01 involves first determining the characteristic length of the thermal storage container based on its geometry and structural characteristics, typically taking the length along the direction of its maximum dimension. Then, based on the finite element stress analysis results, key nodes with high stress concentration on the container wall are identified. These locations usually include corners, around openings, areas of wall thickness change, and support connections. Next, sensors are arranged around these key nodes at intervals equal to 1 / 10 of the characteristic length to ensure the sensor coverage can capture local stress gradient changes. Strain gauge sensors utilize the resistance strain gauge principle, measuring changes in strain gauge resistance to reflect the degree of material deformation. Temperature sensors employ thermocouples or thermistors to accurately measure the temperature field distribution. Vibration acceleration sensors utilize the piezoelectric effect to detect the dynamic vibration response of the container wall. Finally, the sensor sampling frequency is uniformly set to 1000Hz. This frequency meets the dynamic characteristics requirements of thermal stress changes in the thermal energy storage system while avoiding signal distortion caused by excessively low sampling frequencies.

[0030] The specific implementation of step S02 is as follows: First, a multi-load condition test matrix is ​​constructed. This is achieved by applying thermal loads, mechanical loads, and gravity loads of different amplitudes and directions to the thermal storage container. The thermal load is controlled by varying the temperature of the thermal storage medium within the range of 50°C to 300°C. The mechanical load is applied to the outer surface of the thermal storage container by applying loads ranging from 0 to 50°C. A uniformly distributed load was used to simulate the load, while the gravity load was generated by adjusting the tilt angle of the thermal storage container within the range of 0° to 15°. Then, under various load conditions, strain response signals from strain gauge sensors, temperature change signals from temperature sensors, and vibration signals from vibration acceleration sensors were simultaneously acquired, forming a dataset containing the correspondence between load input parameters and sensor response output parameters. Next, the acquired raw data was standardized to unify the signals of different physical quantities into the same numerical range, facilitating subsequent pattern recognition and load separation algorithms. Finally, a multi-load response feature database was established. This database adopts a relational database structure and includes fields such as load type, load amplitude, load direction, sensor location, and response signal characteristics, providing a reference benchmark for subsequent load identification and stress separation.

[0031] The specific implementation of step S03 is as follows: First, the multi-physics field signals during the operation of the thermal storage container are acquired in real time through a data acquisition system. The acquisition system adopts multi-channel synchronous sampling technology to ensure the time synchronization of signals from different sensors. Then, the acquired raw signals are subjected to noise reduction and filtering processing. A Butterworth low-pass filter is used to remove high-frequency noise interference, and the filter cutoff frequency is set to 400Hz, which can effectively retain useful signals while removing electromagnetic interference and environmental noise. Next, signal calibration processing is performed. According to the calibration curve of the sensor, the measured signal is linearly or nonlinearly corrected to eliminate the influence of sensor zero-point drift and temperature drift, ensuring that the measurement accuracy meets engineering requirements. Then, data synchronization processing is performed. A timestamp alignment algorithm is used to unify the data from different sensors to the same time base, compensating for the time delay caused by the difference in sensor response time. Finally, the preprocessed multi-physics field monitoring data is stored in a unified data format to form a structured data file containing time information, sensor location information, and physical quantity values, providing a reliable data foundation for subsequent signal decomposition and stress analysis.

[0032] The specific implementation of step S04 involves first inputting the preprocessed multiphysics monitoring data into an empirical mode decomposition (EMD) algorithm for signal decomposition. Based on the time-scale characteristics of the signal itself, the EMD algorithm adaptively decomposes the composite signal into several intrinsic mode function (EMF) components, each representing a signal component within a different frequency range. Then, based on the frequency domain differences in different load types, the decomposed EMF components are used for load source identification. Thermal expansion stress components are mainly distributed in the low-frequency band, corresponding to the thermal expansion and contraction process caused by slow temperature changes; mechanical load stress components are mainly distributed in the mid-frequency band, corresponding to the dynamic response of external mechanical forces; and gravity load stress components are mainly distributed in the quasi-static frequency band, corresponding to static deformation under gravity. Next, spectral analysis is used to determine the correspondence between each EMF component and a specific load type. Pattern matching is then performed with the multi-load response feature library established in step S02 to achieve accurate identification and separation of different load types. Finally, the identified and separated stress components are organized in matrix form to form a decomposed stress component matrix. The rows of this matrix correspond to different sensor locations, the columns correspond to different load types, and the matrix elements are the stress amplitudes of the corresponding locations and load types, providing basic data for subsequent stress prediction and evaluation.

[0033] The specific implementation of step S05 involves first establishing a three-dimensional geometric model of the thermal storage container, determining the computational domain boundary based on the container's actual structural parameters, and then discretizing the computational domain using adaptive mesh refinement technology. The mesh is automatically refined in regions with large temperature gradients and appropriately sparsed in regions with gentle temperature changes to improve computational accuracy while controlling computational costs. Next, a finite-difference thermal stress prediction model is established. This model is based on the heat conduction equation and the thermoelasticity equation, transforming the partial differential equations into a system of difference equations for numerical solution using the finite-difference method. The model considers the variation of material thermal properties with temperature and the time-varying characteristics of boundary conditions. Then, a Kalman filter algorithm is introduced to correct the model parameters in real time. The Kalman filter algorithm describes the system's dynamic behavior through state equations, establishes the relationship between the model's predicted values ​​and the actual measured values ​​through observation equations, and obtains the optimal state estimate by minimizing the variance of the prediction error, thereby achieving adaptive adjustment of the model parameters. Next, based on the corrected model parameters, the predicted values ​​of transient thermal shock stress are calculated, considering the transient thermal stress response caused by sudden temperature changes in the thermal storage medium to the thermal storage container. The prediction time step is set to 0.1 seconds, which can capture the rapid stress change process during thermal shock. Finally, the calculated stress distribution results are visualized in the form of a cloud map, and the stress concentration factor is calculated. The stress concentration factor is defined as the ratio of the maximum stress to the average stress. When the stress concentration factor is greater than 2.5, it indicates that there is a significant stress concentration phenomenon.

[0034] The specific implementation of step S06 involves first performing spectral decomposition on the decomposed stress component matrix. Spectral decomposition, based on matrix eigenvalue decomposition theory, decomposes the stress component matrix into a product of an eigenvector matrix and an eigenvalue diagonal matrix. Eigenvalues ​​reflect the distribution characteristics of stress components along different principal directions, while eigenvectors reflect the principal directions of stress components. Then, diagonalization simplifies matrix operations. The diagonalized matrix has non-zero elements only on its diagonal, making power calculations and function calculations simple and efficient, significantly reducing computational complexity and time. Next, thermal expansion stress evaluation indices are calculated based on the diagonalized matrix. These indices include stress amplitude, stress gradient, stress change rate, and stress concentration, comprehensively reflecting the stress state and safety level of the thermal storage container. Then, a set of stress threshold calculation equations is established. The input parameters for the stability standard threshold calculation equation include the yield strength of the thermal storage container material, a safety factor set to 2.0, and a temperature correction factor. The output result is the stability standard threshold, which is generally taken as 50% to 60% of the material's yield strength. The input parameters for the hazard warning threshold calculation equation include the ultimate strength of the thermal storage container material, a fatigue correction factor set to 0.8, and the ambient temperature coefficient. The output result is the hazard warning threshold, which is generally taken as 75% to 80% of the material's ultimate strength. Finally, based on the threshold calculation results, the operating mode switching threshold and the warning control threshold are determined, providing a basis for judgment for the adaptive control of the thermal storage system.

[0035] The specific implementation of step S07 involves first comparing the thermal expansion stress assessment index with the operation mode switching threshold and the early warning control threshold to establish a multi-level stress control strategy. Based on the comparison results, the operation mode and control measures of the thermal storage system are determined. When the stress concentration coefficient is less than the stability standard threshold, the system maintains the standard operation mode. At this time, the thermal storage container is in a safe operating state, all stress indicators are within the normal range, the system operates normally according to the preset operating parameters, and the sensor sampling frequency is maintained at 1000Hz to ensure the reliability and stability of the system. When the stress concentration coefficient is between the stability standard threshold and the danger warning threshold, the system activates the stress monitoring enhancement mode, increasing the sensor sampling frequency to 2000Hz to enhance the monitoring accuracy and response speed of stress changes, while also increasing the data recording frequency to provide more detailed data support for subsequent stress analysis and fault diagnosis. When the stress concentration coefficient reaches or exceeds the danger warning threshold, the system immediately activates the stress adaptive adjustment mechanism, using multiple adjustment methods to reduce the stress level. Specific measures include adjusting the flow rate of the thermal storage medium to control the heat transfer rate, reducing the flow rate from the normal operating 2.5... Reduced to 1.5 This slows down the rate of temperature change and the accumulation of thermal stress, reducing the operating temperature from 300℃ to 250℃ to decrease thermal expansion deformation. Simultaneously, the stress compensation mechanism is activated to alleviate thermal expansion stress, including activating expansion joints to release thermal expansion deformation, adjusting stress compensators to balance stress distribution, and using flexible connectors to absorb some deformation energy. The entire adjustment process employs a closed-loop control strategy, monitoring stress level changes in real time and dynamically adjusting control parameters based on the adjustment effect until the stress level is reduced to a safe range, ensuring the safe and stable operation of the thermal storage system.

[0036] It should be noted that the first key technical concept of this invention is a stress source identification technology that combines a multi-physics coupled sensor array with an empirical mode decomposition algorithm. Traditional stress monitoring methods typically only measure the total stress and cannot distinguish the contribution of different load sources to the stress. This invention, however, deploys a sensor array integrating strain, temperature, and vibration detection functions to acquire rich multi-physics information. Then, by applying the adaptive signal decomposition characteristics of the empirical mode decomposition algorithm, it can separate the composite stress signal into thermal expansion stress components, mechanical load stress components, and gravity load stress components, achieving accurate identification and quantitative analysis of different load sources. Compared to the limitations of traditional methods that can only obtain the total stress, this technical approach provides a scientific basis for stress cause analysis and targeted control, significantly improving the accuracy and practicality of stress monitoring.

[0037] The second key technological approach is a predictive stress analysis technique that integrates an adaptive mesh refinement finite difference model with a Kalman filter algorithm. Traditional stress analysis methods often employ fixed-mesh finite element models, whose computational accuracy is limited by mesh density. There is a trade-off between computational efficiency and accuracy, and the fixed model parameters cannot adapt to changes in actual operating conditions. This invention uses adaptive mesh refinement technology, which automatically adjusts the mesh density based on the temperature gradient distribution, improving computational efficiency while maintaining accuracy. Combined with the Kalman filter algorithm, the model parameters are corrected in real time, enabling the model to track changes in system state and achieving predictive stress analysis. Compared to traditional passive monitoring methods, this approach can predict stress change trends in advance, providing a time window for active control and significantly improving the timeliness and effectiveness of stress control.

[0038] The third key technological approach is intelligent stress management technology based on stress component matrix spectral decomposition and multi-level adaptive control strategies. Traditional stress control methods often employ single threshold judgments and fixed control strategies, lacking in-depth analysis of stress states and flexible control means, making them difficult to adapt to complex and changing operating conditions. This invention extracts the main characteristics and variation patterns of stress distribution by performing spectral decomposition and diagonalization on the stress component matrix, establishing a multi-level judgment system including stability standard thresholds and danger warning thresholds, and correspondingly configuring a hierarchical control strategy with standard operating mode, stress monitoring enhancement mode, and stress adaptive adjustment mechanism. Compared to traditional coarse-grained control methods, this approach can precisely adjust system operating parameters according to stress states, ensuring system safety while avoiding the impact of overly conservative control strategies on system efficiency, thus achieving intelligent and refined stress management.

[0039] The synergistic effect of the three key technological approaches described above forms a complete intelligent thermal expansion stress monitoring system. Multiphysics coupled sensing and stress source identification technology provides the system with high-quality basic data and accurate stress cause analysis; the adaptive prediction model provides the system with forward-looking stress change trend prediction; and intelligent stress management technology provides the system with precise control decisions and execution methods. These three technologies work together to form a complete closed loop from data acquisition, information processing, state prediction to intelligent control. Compared to existing open-loop monitoring or simple closed-loop control methods, this invention achieves comprehensive perception, accurate prediction, and intelligent regulation of thermal expansion stress in thermal energy storage systems, significantly improving the system's safety, reliability, and operating efficiency, and providing important technical support for the engineering application of thermal energy storage technology.

[0040] It should be noted that this invention also solves the following technical problem: the difficulty in balancing monitoring accuracy and computational efficiency during stress monitoring of thermal energy storage systems. Traditional stress monitoring methods either employ high-density sensor deployment and high-frequency sampling to improve monitoring accuracy, but this results in a huge computational load affecting real-time performance; or they use simplified models and low-frequency sampling to improve computational efficiency, but this leads to insufficient monitoring accuracy. This invention achieves a strategy of automatically refining the mesh in regions with large stress gradients and using a coarse mesh in regions with gentle stress changes through an adaptive mesh refinement finite difference model. This optimizes computational efficiency while ensuring computational accuracy. The spectral decomposition and diagonalization framework further simplify matrix operations, achieving an effective balance between monitoring accuracy and computational efficiency.

[0041] This invention also addresses the technical problem of the lack of adaptability in stress early warning mechanisms for thermal energy storage systems. Existing stress early warning technologies often rely on fixed thresholds, failing to dynamically adjust based on changes in system operating status and environmental conditions, resulting in insufficient accuracy and timeliness. The stress threshold calculation equations established in this invention dynamically calculate stability standard thresholds and danger warning thresholds based on parameters such as the material properties of the thermal storage container, safety factor, and temperature correction coefficient. Combined with a Kalman filter algorithm, the prediction model parameters are corrected in real time, achieving adaptive adjustment of the warning thresholds and improving the accuracy of early warnings and the safety of system operation.

[0042] Specifically, the principle of this invention is as follows: This invention solves the technical problem of inaccurate separation and prediction of thermal expansion stress and other load-coupled stresses. Its technical principle lies in constructing a complete technical chain from signal acquisition to intelligent adjustment. First, a multi-physics coupled sensor array is used to achieve synchronous monitoring of multiple parameters such as strain, temperature, and vibration. The sensor spacing is 1 / 10 of the characteristic length of the thermal storage container, ensuring comprehensive monitoring coverage and accuracy. The establishment of a multi-load response feature library provides a standard reference for the system response under different load conditions, providing a data foundation for subsequent load identification and stress separation. The empirical mode decomposition algorithm, as an adaptive signal decomposition method, can decompose signals based on their inherent characteristics without requiring preset basis functions. Therefore, it can accurately separate thermal expansion stress components, mechanical load stress components, and gravity load stress components, achieving source identification and separation of multi-load coupled stresses. The adaptive mesh refinement finite difference thermal stress prediction model improves computational accuracy by dynamically adjusting the mesh density, making it particularly suitable for handling transient heat transfer problems with large temperature gradients. Combined with the real-time parameter correction mechanism of the Kalman filter algorithm, it ensures that the prediction model can adapt to changes in the system's operating state. The spectral decomposition and diagonalization framework simplifies the calculation of matrix powers and functions by performing eigenvalue decomposition on the stress component matrix, thus improving computational efficiency. The stress threshold calculation equations establish a dynamic calculation mechanism for the stability standard threshold and the danger warning threshold, considering various factors such as material properties and environmental conditions. A multi-level operation mode switching mechanism employs different monitoring and adjustment strategies based on different ranges of stress concentration coefficients. When the stress level is low, the standard operation mode is maintained; when the stress level rises, the stress monitoring enhancement mode is activated to increase the monitoring frequency; and when the stress level reaches the danger warning threshold, the stress adaptive adjustment mechanism is activated for proactive adjustment. Effective control of the stress level is achieved through measures such as adjusting the flow rate of the thermal storage medium, reducing the operating temperature, and activating the stress compensation mechanism.

[0043] The following provides a specific embodiment 1 of the present invention, and the specific implementation of each step in this embodiment 1 is described in detail below.

[0044] In this embodiment, the specific implementation of step S01 is the same as described above, and will not be repeated in detail here.

[0045] The specific implementation of step S02 is to establish a multi-load response feature library. The load standardization processing formula is expressed as follows: ; In the formula, For the first The standardized value of the load is dimensionless; For the first The actual value of the load, in units of ; This is the minimum load value, in units of ; This represents the maximum load value, in units of The formula for extracting response signal features is expressed as follows: ; In the formula, The characteristic value of the response signal is dimensionless; The mean of the response signal, in units of ; The standard deviation of the response signal, in units of ; The response signal peak value is expressed in units of 1000 m / s. ; For reference response value, the empirical value is 1000. ; This is the weighting coefficient, which is usually set to 0.5, 0.3, or 0.2.

[0046] The specific implementation of step S03 involves preprocessing the acquired signal. The Butterworth low-pass filter transfer function formula is expressed as follows: ; In the formula, The filter transfer function is dimensionless. Let be a Laplace complex variable, representing the complex frequency in the frequency domain, with units of . ; The reference frequency is 1000. ; The cutoff frequency is set to 400Hz; This represents the filter order, which defaults to 4. The signal calibration formula is expressed as follows: ; In the formula, The signal is dimensionless after calibration. The original signal, in units of ; The standardization reference value for the signal is 10, based on experience. ; This is the linear calibration coefficient, typically set to 1.0; This is a temperature compensation term, with units in °C. For reference temperature, the value is taken as 25℃; This is the temperature coefficient, with an empirical value of 0.02. It represents the zero-point offset and is dimensionless.

[0047] The specific implementation of step S04 is to use the empirical mode decomposition algorithm. The intrinsic mode function decomposition formula is expressed as follows: ; In the formula, The original signal is dimensionless. For the first One intrinsic mode function, in units of ; The reference signal amplitude is empirically set at 100. ; This represents the total number of intrinsic mode functions. The residual function is in units of The formula for constructing the load separation matrix is ​​as follows: ; In the formula, The stress component matrix after decomposition is dimensionless. Indicates the first The sensor location Stress components of various load types, in units of ; The reference stress value is set to 1. ; For the number of sensors; These correspond to thermal expansion stress, mechanical load stress, and gravity load stress, respectively.

[0048] The specific implementation of step S05 is to establish an adaptive mesh refinement finite difference thermal stress prediction model. The finite difference discretization formula of the heat conduction equation is expressed as follows: ; In the formula, For the first Time Node Temperature, in Kelvin; For reference temperature, the value is 300K; The time step is set to 0.1s; For reference time, the value is 1 second, and it satisfies the dimensional relationship. ; Thermal diffusivity, in units of ; For reference thermal diffusivity, the empirical value is... ; This is the spatial step size, in meters (m). The reference length is set to 0.1m. The Kalman filter state update formula is expressed as follows: ; In the formula, This is a state estimate, and the unit depends on the specific state variable. This is the corresponding reference value; The Kalman gain is dimensionless. The reference gain is set to 1. These are observations, and the units are determined based on the number of observations. This corresponds to the observed reference value; The observation matrix maps the state variables to the observation space, which has the following dimensions: ,in For the number of observed variables, The number of state variables is dimensionless. The formula for calculating the stress concentration factor is as follows: ; In the formula, The stress concentration factor is dimensionless. Maximum stress, in units of ; The average stress is expressed in units of 1. .

[0049] The specific implementation of step S06 is to establish a spectral decomposition framework for the stress component matrix, and the eigenvalue decomposition formula is expressed as follows: ; In the formula, The stress component matrix is ​​dimensionless. The eigenvector matrix is ​​dimensionless. It is an eigenvalue diagonal matrix, dimensionless; is the inverse of the eigenvector matrix, and is dimensionless. The formula for calculating the stability standard threshold is as follows: ; In the formula, The stability standard threshold is dimensionless. The yield strength of the material, in units of... ; For safety, a value of 0.5 is used; This is a temperature correction factor, determined based on the operating temperature, with an empirical value of 0.9. The formula for calculating the hazard warning threshold is as follows: ; In the formula, The danger warning threshold is dimensionless. The ultimate strength of the material, in units of ; This is the fatigue correction factor, with a value of 0.75. This is the ambient temperature coefficient, determined based on the ambient temperature, with an empirical value of 0.85.

[0050] The specific implementation of step S07 involves adjusting the system operating parameters based on the stress assessment results. The formula for adjusting the flow rate of the heat storage medium is as follows: ; In the formula, The adjusted flow rate is expressed in units of... ; The initial flow rate is set to 2.5. ; This is the flow rate adjustment coefficient, with an empirical value of 0.4; The stress concentration factor is a dimensionless real-time factor. This is the stress concentration factor corresponding to the stability standard threshold, which is usually taken as 2.5; This is the stress concentration factor corresponding to the danger warning threshold, typically taken as 4.0. The formula for adjusting the operating temperature is as follows: ; In the formula, The adjusted operating temperature is expressed in Kelvin (K). The initial operating temperature is set to 573K; This is the temperature regulation coefficient, with an empirical value of 0.17.

[0051] It should be explained that the load normalization formula is based on the principle of linear mapping and uses the minimum-maximum normalization method. .

[0052] This formula maps load values ​​of different magnitudes to a unified range of 0 to 1, eliminating the impact of dimensional differences on subsequent analysis. Compared with the traditional method of directly using the original load values, this formula enables multiple load types to be compared and analyzed within the same framework, significantly improving the accuracy and robustness of load identification and laying a data foundation for establishing a reliable multi-load response feature library.

[0053] The response signal feature extraction formula adopts the principle of weighted linear combination, through a feature fusion function. .

[0054] This formula comprehensively considers three statistical characteristics of the signal: mean, standard deviation, and peak value. Through the reasonable allocation of weighting coefficients, it not only retains the average level information of the signal but also captures the fluctuation and extreme value characteristics of the signal. Compared with the traditional method of single feature parameter, this formula can more comprehensively describe the characteristics of the signal and provide richer information dimensions for load type identification.

[0055] The Butterworth low-pass filter transfer function is based on the frequency domain filtering principle and adopts a rational function form. .

[0056] This transfer function effectively removes high-frequency noise interference by setting a reasonable cutoff frequency, while retaining useful low-frequency signal components. Compared with traditional simple smoothing filtering methods, this transfer function has a flat passband response and a steep transition band characteristic, which can precisely control the filtering effect and avoid the problems of signal distortion and loss of useful information.

[0057] The empirical mode decomposition formula is based on the principle of adaptive signal decomposition and adopts the expansion form of intrinsic mode functions. .

[0058] This formula can decompose the signal based on its time-scale characteristics without the need for preset basis functions. Compared with traditional Fourier transform or wavelet transform methods, this formula has stronger adaptability and localization characteristics, and can accurately separate the load types corresponding to different frequency components, thus achieving effective decoupling of composite stress signals.

[0059] The finite difference discretization formula for heat conduction is based on numerical solution principles and employs an explicit difference scheme. .

[0060] This formula transforms continuous partial differential equations into a computable system of difference equations through time and space discretization, where the time and space reference values ​​satisfy the following relationship. To ensure dimensional consistency, this formula combines adaptive mesh refinement technology to control computational costs while maintaining computational accuracy. Compared to traditional fixed mesh methods, this formula can automatically adjust the mesh density according to the temperature gradient, significantly improving the accuracy and efficiency of transient thermal stress prediction.

[0061] The Kalman filter state update formula is based on optimal estimation theory and uses a recursive estimation form. .

[0062] This formula achieves real-time tracking and parameter correction of the system state through two steps: state prediction and observation update. Compared with traditional static model methods, this formula has adaptive learning capabilities and can dynamically adjust model parameters based on actual observation data, so that the prediction model always maintains consistency with the actual system state, which greatly improves the accuracy of stress prediction.

[0063] The formula for calculating the stress concentration factor is based on the principle of evaluating the non-uniformity of stress distribution and uses a ratio definition. .

[0064] This formula quantitatively describes the degree of stress concentration by using the ratio of maximum stress to average stress. Compared with the traditional method that only considers the maximum stress value, this formula can comprehensively reflect the overall characteristics of stress distribution, providing a more scientific indicator for stress state assessment and making safety assessment more accurate and reliable.

[0065] The eigenvalue decomposition formula is based on matrix spectral theory and uses a diagonalized form. .

[0066] This formula decomposes the complex stress component matrix into a product of eigenvectors and eigenvalues, revealing the main patterns and directions of stress distribution. Compared with traditional direct matrix operation methods, this formula simplifies the complexity of matrix calculations, extracts the essential characteristics of stress distribution, and provides a theoretical basis for stress state analysis and control strategy formulation.

[0067] The threshold calculation formula is based on the principles of material mechanics safety design and uses the correction coefficient method. .

[0068] This formula comprehensively considers factors such as material strength, safety factor, and temperature correction, and establishes a scientific and reasonable stress judgment standard. Compared with the traditional empirical threshold method, this formula has clear physical meaning and theoretical basis, and can be adaptively adjusted according to specific material properties and working conditions, which significantly improves the accuracy and reliability of safety assessment.

[0069] The flow rate and temperature regulation formulas are based on the feedback control principle and adopt a proportional regulation method. and ,in .

[0070] This formula dynamically adjusts the system operating parameters based on real-time changes in stress state. Through reasonable adjustment coefficient design, it achieves effective control of stress level. Compared with the traditional fixed parameter operation mode, this formula can adaptively adjust according to stress state, which not only ensures system safety but also optimizes operating efficiency, achieving an organic unity of safety and economy.

[0071] It should be noted that the variables involved in this invention are explained in detail in Tables 1 and 2.

[0072] Table 1. Variable Explanation Table (Part 1)

[0073] Table 2. Variable Explanation Table (Part Two)

[0074] To better understand and implement this invention, a specific application scenario is provided in Example 2: During the development of a large-scale molten salt thermal storage system, a technical team encountered a problem of thermal expansion stress concentration in the storage container under high-temperature operating conditions. This resulted in excessive stress in localized areas of the container wall, affecting the long-term reliability of the system. The team adopted the thermal expansion stress monitoring method of this invention and established a complete stress monitoring system.

[0075] The thermal storage system uses a cylindrical storage container with a diameter of 8.5m and a height of 12.3m. The container is made of 316 stainless steel with a wall thickness of 25mm. The thermal storage medium is molten nitrate, and the operating temperature range is 290~565℃. The thermal storage capacity reaches [missing information]. The J. technical team deployed a multiphysics coupled sensor array at key nodes on the wall of the thermal storage container. The sensor spacing was set to 1 / 10 of the characteristic length of the thermal storage container (8.5m), i.e., 0.85m. The sensor array includes 72 strain gauge sensors, 48 ​​PT100 temperature sensors, and 36 MEMS vibration acceleration sensors. The sampling frequency of the sensors was uniformly set to 1000Hz to ensure that the transient changes in thermal expansion stress could be captured.

[0076] The technical team established a multi-load response characteristic library for thermal energy storage systems. By applying known thermal loads, mechanical loads, and gravity loads to the thermal storage container, they systematically collected the corresponding response signals. In the thermal load test, a temperature gradient load of 100–565℃ was applied to the thermal storage container, and the amplitude range of the collected strain response signals was 145–2850. The temperature change signal sampling range was 0.5–8.2 °C / min. In the mechanical load test, an internal pressure load of 0.2–1.8 MPa was applied to the thermal storage container, and the strain response signal amplitude range was recorded as 280–1960. ε. In the gravity load test, considering the variation in the loading of the thermal storage medium, the gravity load range is the unloaded state. N to full load N corresponds to a strain response signal amplitude range of 95–720. ε. Through systematic load testing, a multi-load response feature library containing 1256 load-response data pairs was established, providing a reliable reference for subsequent load identification and stress separation.

[0077] The technical team acquired multiphysics signals in real time during the operation of the thermal storage container and preprocessed the acquired signals. Signal preprocessing employed a Butterworth low-pass filter for noise reduction, with a cutoff frequency set to 400Hz and a filter order of 6. Signal calibration utilized a multi-point calibration method, achieving a strain gauge sensor calibration accuracy of ±2. The temperature sensor calibration accuracy reaches ±0.3℃, and the vibration acceleration sensor calibration accuracy reaches ±0.05℃. Data synchronization employs GPS clock synchronization technology to ensure that the timestamp synchronization error of multi-sensor data is less than 1μs. After preprocessing, the signal-to-noise ratio of the multi-physics monitoring data is improved to over 62dB, providing high-quality input data for subsequent signal decomposition.

[0078] The technical team employed an empirical mode decomposition (EMD) algorithm to decompose the preprocessed multiphysics monitoring data. The algorithm decomposed the composite stress signal into eight intrinsic mode function (EMF) components. Frequency domain analysis identified that the thermal expansion stress component is mainly distributed in the 0.001–0.05 Hz frequency band, the mechanical load stress component is mainly distributed in the 0.05–2 Hz frequency band, and the gravity load stress component is mainly distributed in the DC–0.001 Hz frequency band. The amplitude range of the separated thermal expansion stress component is 320–2150. ε, the amplitude range of the mechanical load stress component is 180–1420. ε, the amplitude range of the stress component under gravity load is 85–580. ε. Through stress source identification and separation, a 96×24-dimensional decomposed stress component matrix was formed, with matrix rows representing spatial nodes and matrix columns representing time series.

[0079] The technical team established an adaptive mesh refinement finite difference thermal stress prediction model. The model uses a non-uniform mesh, with a mesh density of 0.05m in regions with large temperature gradients and 0.2m in regions with small temperature gradients. The model parameters are corrected in real time using a Kalman filter algorithm. The Kalman gain matrix has a dimension of 48×48, the initial value of the state estimation error covariance matrix is ​​set to 0.01 times the identity matrix, and the observation noise covariance is set to 0.05. The model calculates the predicted transient thermal shock stress with an accuracy of 85.7%. The thermal expansion stress distribution cloud map shows that the maximum stress occurs at the connection between the bottom and sidewall of the thermal storage container, with a stress concentration factor of 2.83, making this area a key monitoring area.

[0080] As shown in Table 3, the technical team established a stress monitoring parameter setting table based on different operating conditions.

[0081] Table 3. Stress monitoring parameter setting table under different operating conditions

[0082] The technical team established a spectral decomposition and diagonalization framework for the stress component matrix. Through eigenvalue decomposition, the 96×24-dimensional stress component matrix was diagonalized into a diagonal matrix containing 24 eigenvalues. The main eigenvalues ​​are distributed in the range of 1.85–8.67, and the corresponding eigenvectors reflect the main modes of stress distribution. By simplifying matrix operations through diagonalization, the calculation efficiency of the thermal expansion stress assessment index was improved by 72%. Using a set of stress threshold calculation equations, combined with the yield strength of 316 stainless steel at high temperatures (245 MPa), a safety factor of 2.5, and a temperature correction factor of 0.82, a stability standard threshold of 2.15 was calculated. Combining the material's ultimate strength of 520 MPa, a fatigue correction factor of 0.75, and an ambient temperature coefficient of 0.88, a danger warning threshold of 3.20 was calculated.

[0083] like Figure 2 As shown, the technical team established a stress concentration coefficient distribution map at different locations of the thermal storage container, revealing a non-uniform distribution of stress concentration coefficient along the height of the container. Based on the comparison results of the thermal expansion stress assessment index with the operation mode switching threshold and early warning control threshold, the technical team formulated corresponding operating parameter adjustment strategies. When the stress concentration coefficient is less than 2.15, the system maintains the standard operation mode, with the thermal storage medium flow rate maintained at 0.85 m / s and the operating temperature controlled below 450℃. When the stress concentration coefficient is in the range of 2.15 to 3.20, the stress monitoring enhancement mode is activated, the sampling frequency is increased to 2000 Hz, and the thermal storage medium flow rate is adjusted to 0.62 m / s to reduce the heat transfer rate and slow down the temperature change rate.

[0084] like Figure 3 As shown, when the system detects a stress concentration factor of 3.20, it immediately activates the stress adaptive adjustment mechanism. The technical team reduced the flow rate of the thermal storage medium from 0.85 m / s to 0.45 m / s and the operating temperature from 520℃ to 380℃, while simultaneously activating the stress compensation mechanism. The stress compensation mechanism consists of 16 sets of corrugated expansion joints, each with a compensation range of ±45 mm, installed at the connection between the thermal storage container and the pipeline. Through these comprehensive adjustment measures, the stress concentration factor was reduced from 3.20 to 2.05, effectively avoiding the risk of excessive stress in the thermal storage container.

[0085] As shown in Table 4, the technical team compiled statistics on the effects of different adjustment measures on stress control.

[0086] Table 4. Statistical Table of the Effects of Stress Adaptive Adjustment Measures

[0087] The technical team implemented thermal expansion stress monitoring for the thermal energy storage system using a computer system. The computer is equipped with a solid-state drive (SSD) containing 2.3GB of program instructions, which are executed in real-time to monitor the thermal expansion stress. The system employs a multi-core processor parallel computing architecture, keeping the calculation time for a single stress assessment within 125ms, meeting real-time monitoring requirements. Data storage utilizes a tiered strategy, with real-time monitoring data stored every 72 hours and historical statistical data stored for 2 years, resulting in a total system storage capacity of 15TB.

[0088] By implementing the thermal expansion stress monitoring method of this invention, the occurrence rate of stress exceeding the limit event in the thermal energy storage system was reduced to 0.12% during 8760 hours of continuous operation, significantly improving the system's operational reliability. The system can accurately identify and separate stress components from different sources, achieving precise monitoring and prediction of thermal expansion stress. Through the timely response of the adaptive adjustment mechanism, the risk of structural damage caused by stress concentration in the thermal storage container is effectively avoided, ensuring the long-term stable operation of the thermal energy storage system.

[0089] This invention represents a significant advancement over traditional stress monitoring methods. Traditional methods typically only monitor total stress and cannot distinguish the contribution of different load sources to the stress. This invention, however, utilizes an empirical mode decomposition algorithm to identify and separate the sources of multi-load coupled stress, accurately separating thermal expansion stress components, mechanical load stress components, and gravity load stress components, providing a scientific basis for targeted stress control. Traditional methods use fixed thresholds for stress warning and lack adaptive capabilities. This invention establishes an adaptive mesh refinement finite difference thermal stress prediction model, combined with a Kalman filter algorithm to correct model parameters in real time, dynamically adjusting prediction accuracy based on actual operating conditions, thus improving the accuracy of stress prediction. Traditional methods lack effective stress control measures and can only passively respond to stress exceeding events. This invention establishes a stress adaptive adjustment mechanism, achieving proactive stress level control through various measures such as adjusting the flow rate of the thermal storage medium, reducing the operating temperature, and activating the stress compensation mechanism, fundamentally preventing stress exceeding problems.

[0090] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for monitoring thermal expansion stress in a thermal energy storage system, characterized in that, include: Deploy multi-physics field coupled sensor arrays at key nodes on the wall of the thermal storage container in a thermal energy storage system; Establish a multi-load response feature library for thermal energy storage systems. By applying known thermal loads, mechanical loads, and gravity loads to the thermal storage container, the corresponding strain response signals, temperature change signals, and vibration signals are collected to form a multi-load response feature library dataset. Multi-physics field signals are collected in real time during the operation of the thermal storage container. The collected signals are preprocessed to obtain preprocessed multi-physics field monitoring data. The empirical mode decomposition algorithm is used to decompose the preprocessed multiphysics monitoring data into signals, separating the thermal expansion stress component, mechanical load stress component, and gravity load stress component, thereby realizing the source identification and separation of multi-load coupled stress and obtaining the decomposed stress component matrix. An adaptive mesh refinement finite difference thermal stress prediction model was established. The model parameters were corrected in real time by combining the Kalman filter algorithm. The predicted value of transient thermal shock stress was calculated, and the thermal expansion stress distribution cloud map and stress concentration factor were obtained. A framework for spectral decomposition and diagonalization of the stress component matrix is ​​established. By diagonalizing the decomposed stress component matrix to simplify the calculation of matrix powers and functions, the thermal expansion stress evaluation index is calculated. The operating mode switching threshold and early warning control threshold are obtained by using a set of stress threshold calculation equations. Adjust the operating parameters of the thermal storage system based on the comparison results of the thermal expansion stress assessment index with the operation mode switching threshold and the early warning control threshold.

2. The method for monitoring thermal expansion stress in a thermal energy storage system according to claim 1, characterized in that, The multi-physics coupled sensor array includes strain gauge sensors, temperature sensors, and vibration acceleration sensors. The sensor spacing is 1 / 10 of the characteristic length of the thermal storage container, and the sensor sampling frequency is set to 1000Hz.

3. The method for monitoring thermal expansion stress in a thermal energy storage system according to claim 2, characterized in that, The steps for preprocessing the acquired signals specifically include noise reduction filtering, signal calibration, and data synchronization.

4. The method for monitoring thermal expansion stress in a thermal energy storage system according to claim 3, characterized in that, A multi-load response feature library refers to a database that stores the system response characteristics under different load conditions, serving as a reference for load identification and stress separation.

5. The method for monitoring thermal expansion stress in a thermal energy storage system according to claim 4, characterized in that, The decomposed stress component matrix refers to the matrix data structure that includes thermal expansion stress components, mechanical load stress components, and gravity load stress components, obtained by processing with the empirical mode decomposition algorithm.

6. The method for monitoring thermal expansion stress in a thermal energy storage system according to claim 5, characterized in that, The stress concentration factor is a dimensionless parameter that characterizes the degree of non-uniformity in stress distribution. It is defined as the ratio of the maximum stress to the average stress.

7. The method for monitoring thermal expansion stress in a thermal energy storage system according to claim 6, characterized in that, The spectral decomposition and diagonalization framework refers to a mathematical method for performing eigenvalue decomposition and diagonalization on the decomposed stress component matrix, which simplifies matrix operations and extracts key stress features.

8. The method for monitoring thermal expansion stress in a thermal energy storage system according to claim 7, characterized in that, The stress threshold calculation equation set includes the stability standard threshold calculation equation and the hazard warning threshold calculation equation; The stability standard threshold calculation equation is used to determine the upper limit of the stress concentration factor for normal system operation. The inputs include the yield strength of the thermal storage container material, the safety factor, and the temperature correction factor. The output is the stability standard threshold. The equation for calculating the danger warning threshold is used to determine the critical value of the stress concentration factor for initiating the adaptive stress adjustment mechanism. The inputs include the ultimate strength of the thermal storage container material, the fatigue correction factor, and the ambient temperature coefficient. The output is the danger warning threshold.

9. The method for monitoring thermal expansion stress in a thermal energy storage system according to claim 8, characterized in that, When the stress concentration factor is less than the stability standard threshold, the standard operation mode is maintained. When the stress concentration factor is ∈ [stability standard threshold, danger warning threshold), the stress monitoring enhancement mode is activated and the sampling frequency is increased to 2000Hz.

10. The method for monitoring thermal expansion stress in a thermal energy storage system according to claim 9, characterized in that, When the stress concentration factor is greater than or equal to the danger warning threshold, the stress adaptive adjustment mechanism is activated, and the stress level is controlled by adjusting the flow rate of the heat storage medium, reducing the working temperature, and activating the stress compensation mechanism.

Images