Lightning arrester resistor disc degradation degree monitoring system and method in extremely cold environment

Abstract

The invention relates to the technical field of electrical monitoring and intelligent monitoring, in particular to a system and a method for monitoring the deterioration degree of a resistor disc of a lightning arrester in an extremely cold environment, which are used for acquiring the size, the depth and the expansion trend of a crack in real time and generating holographic model data of the crack through a multi-band optical irradiation and optical interference measurement technology. Thirdly, calculating stress distribution and energy transfer characteristics of the crack based on the holographic model data of the crack, optimizing a crack propagation path in combination with a reinforcement learning method, and generating crack degradation trend data; through combination of multilayer time sequence analysis and crack distribution influence factors, the long-term degradation trend of the lightning arrester resistor disc is predicted, measurement parameters of the optical monitoring unit are adaptively adjusted, and the calculation weight of the crack modeling and trend prediction unit is optimized. According to the invention, high-precision monitoring of crack propagation in an extremely cold environment can be realized, a monitoring strategy is optimized according to real-time data, and the response precision and adaptive capacity of the system are improved.

Description

Technical Field

[0001] The present invention relates to the technical fields of electrical monitoring technology and intelligent monitoring technology, and particularly relates to a monitoring system and method for the deterioration degree of arrester resistor chips in an extremely cold environment. Background Art

[0002] An arrester is a key device in the power system, and its main function is to guide excessive current to the ground during lightning or overvoltage events, thereby protecting electrical equipment from damage. The deterioration degree of the arrester resistor chip directly affects its working efficiency and safety. Especially in an extremely cold environment, temperature changes will exacerbate the physical deterioration of the material, thereby affecting the performance of the arrester. Therefore, timely and effectively monitoring and evaluating the deterioration degree of the resistor chip is a key measure to ensure the long-term stable operation of the equipment and avoid equipment failures. In the prior art, traditional methods for evaluating the life of arrester resistor chips mostly rely on regular tests of electrical performance, such as parameters like leakage current density and power loss. However, these methods have poor applicability in an extremely cold environment and cannot accurately reflect the influence of temperature changes on the performance of the resistor chip, resulting in errors in the life evaluation results. In addition, most of the prior art ignores the dynamic expansion of cracks in the resistor chip and the subsequent structural changes, and cannot comprehensively predict the influence of crack expansion on the overall performance of the resistor chip.

[0003] The prior art (Chinese invention patent, publication number: CN118795247A, title: A method for evaluating the life of an arrester resistor chip) mainly focuses on evaluating the life of an arrester resistor chip by simulating lightning strikes or electrical performance tests (such as leakage current and power loss). For example, some methods obtain loss coefficients by simulating environmental changes (such as temperature and humidity) and lightning strike effects, and predict the life of the resistor chip based on these coefficients. However, these methods have the following defects:

[0004] The prior art fails to monitor the expansion process of cracks in the resistor chip in real time and cannot accurately capture the influence of crack expansion on the performance of the resistor chip; the influence of temperature changes on material performance is non-linear, and the prior art often cannot dynamically adapt to the changes in crack expansion in an extremely cold environment, resulting in insufficient accuracy of life evaluation; traditional methods rely on static models and preset parameters and lack an adaptive optimization function based on real-time monitoring data, resulting in a weak response ability of the monitoring system to sudden environmental changes. Summary of the Invention

[0005] In view of the many problems existing in the above-mentioned prior art, the present invention provides a monitoring system and method for the deterioration degree of lightning arrester resistor chips in extremely cold environments. The present invention realizes high-precision monitoring of crack propagation through optical resonance technology and crack modeling and trend prediction methods, and optimizes the crack propagation path through reinforcement learning. By dynamically adjusting the crack propagation rate and direction, combining multi-layer time series analysis and crack distribution influence factors, crack deterioration trend data is generated, and the monitoring strategy is further optimized to improve the system response ability. The present invention provides strong support for the long-term stable operation of lightning arrester resistor chips.

[0006] A monitoring system for the deterioration degree of lightning arrester resistor chips in extremely cold environments, the system comprising:

[0007] An optical monitoring unit, configured to perform multi-band optical irradiation on the lightning arrester resistor chip, and based on optical interference, measure the optical resonance characteristics of the crack area, obtain the size, depth and propagation trend of the crack, and generate crack holographic model data;

[0008] A crack modeling and trend prediction unit, configured to receive the crack holographic model data, calculate the stress distribution, energy transfer characteristics and propagation driving force of the crack, construct a crack propagation dynamics model, and based on the crack evolution path optimization method, combine reinforcement learning to dynamically adjust the crack propagation rate and direction, and generate crack deterioration trend data.

[0009] A deterioration evaluation and optimization unit, configured to receive the crack deterioration trend data, combine the multi-layer time series analysis and the crack distribution influence factor calculation model, predict the long-term deterioration trend of the lightning arrester resistor chip, generate deterioration evaluation data, and adaptively adjust the measurement parameters of the optical monitoring unit based on the deterioration evaluation data, and optimize the calculation weights of the crack modeling and trend prediction unit.

[0010] Preferably, the optical monitoring unit comprises:

[0011] A multi-band laser irradiation device, configured to emit multi-wavelength lasers to the surface of the lightning arrester resistor chip, and receive the optical signal reflected from the crack area through an optical reflection device;

[0012] An optical signal processing module, configured to convert the received optical signal into optical resonance data, and obtain the size, depth and propagation trend of the crack by comparing and analyzing the spectral changes in the crack area, and generate crack holographic model data.

[0013] Preferably, the crack modeling and trend prediction unit comprises:

[0014] A stress field calculation module, configured to receive the crack holographic model data, and calculate the stress distribution in the crack area based on the crack holographic model data, and analyze the interaction between the external stress and the internal stress of the material received by the crack;

[0015] An energy transfer analysis module for analyzing the energy transfer characteristics within the crack region, calculating the driving force for crack propagation, and generating crack energy data;

[0016] A crack propagation model construction module for constructing a crack propagation dynamics model, simulating the crack propagation process, and generating crack stress data and crack energy data based on the stress distribution and energy data.

[0017] Preferably, the crack modeling and trend prediction unit further includes:

[0018] A crack propagation rate calculation module for calculating the crack propagation rate based on the crack stress data and crack energy data, and inferring the time, rate, and direction of crack propagation;

[0019] A crack path optimization module for inferring the future crack propagation path based on the crack propagation rate using the finite element method and the multi-point iteration method, and generating crack propagation prediction data.

[0020] Preferably, the crack modeling and trend prediction unit further includes:

[0021] A reinforcement learning optimization module for dynamically adjusting the crack propagation rate and direction based on the crack propagation data and crack path optimization data, adaptively optimizing the crack evolution path through a reinforcement learning model, and generating crack deterioration trend data.

[0022] Preferably, the deterioration evaluation and optimization unit includes:

[0023] A time series analysis module for receiving the crack deterioration trend data, predicting the long-term deterioration trend of the arrester resistor chip through multiple sampling and trend analysis methods, and generating long-term deterioration trend data;

[0024] A crack distribution influence factor calculation module for analyzing the spatial distribution of cracks and their influence on the performance of the resistor chip based on the crack distribution factor and historical monitoring data, and generating first deterioration evaluation data.

[0025] Preferably, the deterioration evaluation and optimization unit further includes:

[0026] An adaptive adjustment module for dynamically adjusting the measurement parameters of the optical monitoring unit according to the first deterioration evaluation data, and optimizing the calculation weights of the crack modeling and trend prediction unit in combination with the monitoring feedback signal to improve the accuracy of the crack prediction model.

[0027] Preferably, the deterioration evaluation and optimization unit further includes:

[0028] The monitoring data fusion optimization module is used to fuse historical data and real-time data, train an intelligent monitoring model based on the fused data, and optimize the response ability of the monitoring system through an adaptive algorithm to generate a final adaptive monitoring strategy.

[0029] Preferably, the system further includes:

[0030] The data storage unit is used to store crack holographic model data, crack deterioration trend data, deterioration assessment data, and adaptive monitoring strategies for subsequent analysis and system optimization.

[0031] A method for monitoring the deterioration degree of lightning arrester resistor chips in an extremely cold environment, which is used to execute the monitoring system for the deterioration degree of lightning arrester resistor chips in the extremely cold environment. The steps of the present invention include:

[0032] Perform multi-band optical irradiation on the lightning arrester resistor chip through an optical quantum resonance sensor, and based on optical interference measurement of the optical resonance characteristics of the crack area, obtain the size, depth, and expansion trend of the crack, and generate crack holographic model data;

[0033] Receive the crack holographic model data, calculate the stress distribution, energy transfer characteristics, and expansion driving force of the crack, and construct a crack expansion dynamics model; based on the crack evolution path optimization method, combine reinforcement learning to dynamically adjust the crack expansion rate and direction, and generate crack deterioration trend data;

[0034] Receive the crack deterioration trend data, combine multi-layer time series analysis and the crack distribution influence factor calculation model to predict the long-term deterioration trend of the lightning arrester resistor chip, and generate deterioration assessment data; adaptively adjust the measurement parameters of the optical monitoring unit based on the deterioration assessment data, and optimize the calculation weights of the crack modeling and trend prediction unit.

[0035] Compared with the prior art, the advantages and beneficial effects of the present invention are as follows:

[0036] Through multi-band optical irradiation technology and optical interference measurement, high-precision detection of cracks is achieved, providing reliable data support for subsequent deterioration prediction. At the same time, through reinforcement learning and adaptive algorithms, the crack expansion path and the response ability of the monitoring system are dynamically optimized, making the prediction of the crack expansion trend more accurate. In an extremely cold environment, this technical means can effectively overcome the adverse effects of temperature on traditional monitoring methods, improving the adaptability and accuracy of the monitoring system.

[0037] The advantages of the present invention not only lie in the real-time monitoring of crack expansion, but also include dynamic adjustment and adaptive optimization based on real-time data, ensuring that the system can flexibly respond to temperature changes and environmental fluctuations. These innovative measures make the application of the present invention in an extremely cold environment have high accuracy and reliability. Description of the Drawings

[0038] Figure 1 This is the structural block diagram of the system of the present invention;

[0039] Figure 2 This is the schematic diagram of the interaction between the data flow and the optimization process in the present invention;

[0040] Figure 3 This is the process schematic diagram of the method of the present invention. Detailed implementation manners

[0041] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present disclosure. In the following detailed description, for the sake of explanation, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the present disclosure.

[0042] As Figure 1 - Figure 2 shown, a monitoring system for the deterioration degree of arrester resistor chips in an extremely cold environment, the system includes:

[0043] An optical monitoring unit, configured to perform multi-band optical irradiation on the arrester resistor chip, and based on the optical resonance characteristics of the crack area measured by optical interference, obtain the size, depth and expansion trend of the crack, and generate crack holographic model data.

[0044] In the present invention, the core function of the optical monitoring unit is to monitor the crack state of the arrester resistor chip in real time through multi-band optical irradiation, especially in an extremely cold environment, to ensure efficient and stable measurement of the crack size, depth and expansion trend. This unit uses optical interference measurement technology, irradiates the crack area with multi-wavelength lasers, and combines optical resonance characteristic measurement to accurately capture the minute changes of the crack and generate crack holographic model data in real time. This technology not only breaks through the influence of the extremely cold environment on traditional monitoring methods, but also provides more accurate crack detection results.

[0045] In an extremely cold environment, the cracks and material changes of the arrester resistor chip are usually minute and difficult to detect. In order to capture these subtle changes, the optical monitoring unit adopts multi-band optical irradiation technology. By combining irradiation in the ultraviolet, visible light and near-infrared light bands, lasers with different wavelengths have different penetration depths and scattering characteristics, and can adapt to cracks with different depths and properties.

[0046] Ultraviolet light (200nm - 400nm) has a relatively short wavelength and can be used to detect the shallowest minute damages on the crack surface, especially sensitive in the detection of surface stress changes and oxide layers of the crack.

[0047] Visible light (400nm - 700nm) is applicable to areas with relatively shallow crack depths and can effectively scan for cracks in the shallow and middle layers of the material.

[0048] Near-infrared light (700nm - 1400nm) is suitable for detecting deeper cracks, can penetrate the surface layer of the material, and provide information on the size and propagation of deep cracks.

[0049] The laser source of each wavelength is precisely adjusted to irradiate the cracks by automatically selecting the appropriate wavelength and power, ensuring that all detection requirements for different types of cracks can be covered. The optical signal after laser irradiation will be reflected back, collected by an optical reflection device, and transmitted to the optical signal processing module for analysis.

[0050] Optical resonance technology obtains the size, depth, and propagation trend of cracks through the interference effect. In the present invention, the optical resonance signal is a direct reflection of the crack position and morphology. Using optical interferometry, an interference fringe pattern can be formed in the crack area to reflect the minute changes of the crack. The propagation path of light is closely related to the surface morphology, depth of the crack, and changes in the internal stress of the crack. Using this method, the minute expansion and stress distribution of cracks can be detected in extremely cold environments.

[0051] Specifically, the optical monitoring unit irradiates the crack surface of the arrester resistor with a laser light source, exciting the propagation of light waves in the crack and the surrounding material. The refraction, reflection, scattering, etc. of light waves in the crack area will change the propagation characteristics of light waves, and then these changes can be accurately quantified through interferometry. The reflected optical signal can extract the depth, size, and propagation trend of the crack after interference analysis, and generate crack holographic model data.

[0052] In the optical monitoring unit, the crack holographic model data is a comprehensive description of the crack area. Combining the laser reflection spectral signals of different wavelengths, the three-dimensional structure of the crack is reconstructed through optical coherence tomography (OCT). This process can provide detailed information on the spatial distribution, depth distribution, etc. of the crack. Different from the traditional single-dimensional crack detection method, the holographic model data provides a multi-dimensional and all-round analysis of the crack, providing accurate basic data for subsequent crack propagation prediction and deterioration assessment.

[0053] The generation of crack holographic model data is based on a complex optical analysis process. First, the multi-band optical signal is converted into spectral information, and then the Fourier Transform technology is applied to convert the spectral information into time-domain data, thereby obtaining detailed structural information of the crack area. During this process, the phase change and spectral change of light directly reflect the morphological characteristics of the crack and the stress state of the crack.

[0054] This data generation method not only improves the sensitivity and accuracy of crack detection, but also can operate continuously and stably in extremely cold environments, overcoming the interference of environmental temperature changes on the monitoring results and ensuring the reliability and efficiency of the monitoring results.

[0055] Preferably, the optical monitoring unit includes:

[0056] A multi-band laser irradiation device for emitting multi-wavelength laser to the surface of the arrester resistor chip and receiving the optical signal reflected from the crack area through an optical reflection device;

[0057] An optical signal processing module for converting the received optical signal into optical resonance data, and obtaining the size, depth and expansion trend of the crack by comparing and analyzing the spectral changes in the crack area, and generating crack holographic model data.

[0058] In extremely cold environments, the monitoring of the deterioration degree of the arrester resistor chip relies on the optical monitoring unit, which monitors the morphology, depth and expansion trend of cracks in real time through multi-band optical irradiation and optical interferometry. Traditional electrical parameter monitoring is vulnerable to the influence of the external environment in extremely cold environments, resulting in unstable measurement results. In contrast, optical monitoring technology can provide more stable and accurate crack monitoring. Especially in low-temperature and harsh climate conditions, it can ensure efficient operation for a long time.

[0059] The optical monitoring unit uses multi-band laser irradiation technology. By irradiating the crack area with lasers of different wavelengths, cracks of different depths and sizes can be effectively detected. Each band of laser has different penetration depths and scattering characteristics, so the appropriate wavelength can be selected according to the type of crack and environmental conditions. These reflected optical signals are measured by optical interferometry, and by analyzing the phase change of the light wave, the size, depth and expansion trend of the crack are obtained.

[0060] Specifically, the laser irradiation will interact with the surface and internal structure of the crack, and the optical wave signals generated by this interaction will be received and analyzed by the optical monitoring unit. The distribution and stress state of cracks in the flake material will affect the propagation of light waves, and the phase, frequency and amplitude of the optical signals will change. By analyzing these changes, the system can accurately calculate the depth and expansion of the crack.

[0061] One of the key technologies of the optical monitoring unit is optical interferometry, that is, using the interference phenomenon to convert the phase change of the reflected light wave into crack characteristics. This process depends on the wavelength of light and the characteristics of the light source. Usually, a single-mode fiber laser is used to emit light waves of a certain frequency, and the crack information is obtained through the phase change caused by the reflection and refraction of the light wave in the crack area. The received reflected optical signal enters the interferometer, and the depth, length and expansion direction of the crack are determined through phase difference analysis.

[0062] The advantage of interferometry lies in its high resolution. Even in extremely cold environments, when the changes in cracks are very small, the interferometry technology can still capture these tiny changes, thus providing extremely accurate crack data. The optical monitoring unit can not only accurately measure the surface characteristics of cracks, but also measure the depth of cracks according to different light wavelengths. By integrating optical signals in multiple frequency bands, the system can obtain all-round information about cracks and create a three-dimensional holographic model of the cracks for subsequent expansion prediction and deterioration assessment.

[0063] In an embodiment, in actual operation, the optical monitoring unit will use the following specific method for crack detection:

[0064] Multi-band laser irradiation: The system selects ultraviolet light (200 - 400 nm), visible light (400 - 700 nm), and near-infrared light (700 - 1400 nm) to irradiate the crack area. Suppose a laser with a wavelength of ultraviolet light (355 nm) is selected. It has a high resolution and can effectively capture the initial surface stress changes and micro-cracks of the cracks. Near-infrared light (800 nm) is suitable for deep detection of cracks, which can penetrate the material surface layer and analyze the crack depth.

[0065] After the laser irradiates the crack area, the surface and internal structures of the crack will cause refraction, reflection, or scattering of light waves. The optical reflection device that receives the optical signal captures these reflected signals and transmits them to the optical signal processing module. Suppose the phase change of the reflected light is Δφ, and there is a clear mathematical relationship between Δφ and the crack depth d and the expansion area A.

[0066] For example, the relationship between the phase change of the optical signal and the crack depth and expansion area can be expressed by the following formula:

[0067] Δφ = 4πd / λ · cos(θ)

[0068] Where, Δφ is the phase change of the reflected light, d is the crack depth, λ is the laser wavelength, and θ is the angle between the light ray and the crack surface.

[0069] Through this formula, the optical monitoring unit can calculate the crack depth and expansion trend based on the phase difference of the reflected light and generate corresponding crack holographic model data.

[0070] Combined with multi - band optical signals, the system can map the spectral changes of the reflected light into a three - dimensional model of the crack. Assuming the system uses optical coherence tomography (OCT) to perform three - dimensional imaging on the crack area, through multiple laser scans and data acquisition, holographic data of the crack is obtained. For example, by performing interference measurements on the reflected light of two wavelengths, the system can obtain information on the length (L), width (W), and depth (D) of the crack, and the finally generated crack holographic model is a three - dimensional grid model, where each point represents a specific position of the crack.

[0071] By combining the holographic model data with time - series data, the system can deduce the propagation trend of the crack. During the monitoring process, the propagation speed (V) of the crack and the changes in crack morphology will be recorded and analyzed. Assuming that in each measurement, the propagation speed of the crack can be calculated by the formula:

[0072]

[0073] where V is the crack propagation speed, L1 and L2 are the lengths of the crack at two measurements respectively, and t1 and t2 are the measurement time points.

[0074] Based on these calculations, the system can predict the propagation trend of the crack, generate prediction data, and provide early warnings for future crack propagation.

[0075] The crack modeling and trend prediction unit is used to receive crack holographic model data, calculate the stress distribution, energy transfer characteristics, and propagation driving force of the crack, construct a crack propagation dynamics model, and based on the crack evolution path optimization method, combine reinforcement learning to dynamically adjust the crack propagation rate and direction, and generate crack deterioration trend data.

[0076] In the present invention, the core task of the crack modeling and trend prediction unit is to calculate the stress distribution and energy transfer characteristics of the crack by receiving crack holographic model data, and combine the crack propagation dynamics model to predict the propagation trend of the crack. The design of this unit can not only predict the propagation behavior of the crack, but also optimize the propagation rate and direction of the crack, so as to provide accurate data for subsequent deterioration assessment and repair strategies.

[0077] The driving force for crack propagation is usually determined by the stress distribution and energy transfer characteristics of the material. In extremely cold environments, these factors may change due to factors such as temperature changes and stress accumulation. Therefore, it is necessary to accurately simulate these changes in order to timely detect potential risks of crack propagation.

[0078] The theoretical basis of this unit is the stress field theory and the energy release rate (G). The stress field theory mainly calculates the driving force for crack propagation through the stress distribution in the crack area; while the energy release rate is the energy released during crack propagation, which can help predict the crack propagation ability.

[0079] The dynamic calculation of crack propagation calculates the stress state in the crack area through finite element analysis (FEM), combines the energy release characteristics of the crack, and establishes a dynamic model for crack propagation. Through this modeling, the system can calculate the rate and direction of crack propagation at the crack tip and in the surrounding area, so as to predict the future development trend of the crack.

[0080] The stress distribution of the crack is one of the key factors in crack propagation. The system first obtains the morphological information of the crack by receiving the crack holographic model data from the optical monitoring unit. On this basis, the finite element method (FEM) is used to simulate the stress field around the crack. By calculating the stress intensity factor (SIF) in different regions, the driving force for crack propagation is obtained.

[0081] The energy transfer characteristics of the crack are also one of the important factors in expansion prediction. The system calculates the elastic energy and plastic energy during crack propagation through the energy release rate calculation, and judges the possibility of crack propagation. Specifically, the energy release rate G is the standard for calculating crack propagation, indicating the energy released per unit crack extension.

[0082] The formula for the energy release rate is:

[0083]

[0084] where G is the energy release rate, U is the total energy during crack propagation, and A is the area of crack propagation.

[0085] Based on the calculated value of G, the system can infer whether the crack will further propagate and provide the speed and direction of propagation.

[0086] The propagation of the crack is not just a simple stress transfer problem. It is also affected by the energy density at the crack tip and the material toughness. In the present invention, the crack propagation dynamic model performs dynamic calculations through finite element analysis by integrating stress distribution, energy transfer characteristics, and material toughness. This model infers the crack propagation trend based on the mechanical properties of the material, the morphology of the crack, and the changes in external stress.

[0087] For example, when the crack growth rate (V) is relatively fast, it indicates that the crack is being pushed by a relatively large external load or internal stress. During the crack growth process, the system predicts the crack growth direction and adjusts the future crack growth path according to the crack path optimization model. This process is quantified by the following calculation formula: This formula can help the system determine the crack growth speed and predict the crack growth time period.

[0088] During the crack growth process, due to the complexity of the crack morphology and growth path, the system needs to use the crack evolution path optimization method to further adjust the crack growth trend. In the present invention, reinforcement learning technology is adopted, and Q-learning or Deep Q-Learning is used to adaptively optimize the crack growth path.

[0089] The application of reinforcement learning in this process mainly evaluates the advantages and disadvantages of each growth path through Q-values and continuously adjusts according to environmental changes. The system calculates the Q-value function to evaluate the optimal path of crack growth in the current state, thereby optimizing the crack growth rate and direction.

[0090] The basic steps of reinforcement learning are as follows: Initialize the initial state of crack growth (such as crack length, depth, and stress state). Calculate the Q-value of each state and select the growth path according to the current crack state. Evaluate the Q-value of the new state after crack growth and adjust the path selection according to the feedback. Through multiple iterations, optimize the crack growth path and rate. In this way, the system can dynamically adapt to the crack growth process, avoid the limitations of the preset path, and make the prediction of crack growth more accurate and flexible.

[0091] Preferably, the crack modeling and trend prediction unit includes:

[0092] A stress field calculation module, configured to receive crack holographic model data, calculate the stress distribution in the crack region based on the crack holographic model data, and analyze the interaction between the external stress and the internal stress of the material acting on the crack;

[0093] An energy transfer analysis module, configured to analyze the energy transfer characteristics in the crack region, calculate the driving force for crack growth, and generate crack energy data;

[0094] A crack growth model construction module, configured to construct a crack growth dynamics model, simulate the crack growth process, and generate crack stress data and crack energy data based on the stress distribution and energy data.

[0095] The main function of the crack modeling and trend prediction unit is to analyze the crack propagation process through the stress field calculation module and the energy transfer analysis module. By combining the stress distribution and energy transfer characteristics of the crack, a dynamic model of crack propagation is constructed to accurately predict the crack propagation trend. This process is crucial for monitoring the deterioration degree of the arrester resistor chip in extremely cold environments, especially for helping the system to track the dynamic evolution of the crack in real time and providing accurate basis for the repair decision of the arrester.

[0096] The core technologies of the present invention include the stress distribution analysis of cracks, the energy transfer characteristic analysis, and the establishment of a crack propagation model. The crack propagation is affected by various factors such as external stress, material properties, and crack path. Therefore, it is necessary to comprehensively evaluate the crack propagation trend and driving force through multi-dimensional data analysis and calculation. In extremely cold environments, these calculations can provide high-precision predictions of the crack propagation process, especially for identifying the subtle changes in crack propagation under the influence of low-temperature environments on materials.

[0097] The main driving force for crack propagation comes from the stress distribution within the crack region. In the crack modeling and trend prediction unit, the stress field calculation module is responsible for calculating the stress distribution in the crack region based on the crack holographic model data. The stress distribution of the crack is formed by the combined action of external loads and internal stresses of the material. To accurately simulate these stress states, the system uses the finite element method (FEM) to perform stress field simulation calculations.

[0098] Through the crack holographic model data, this module can accurately capture the morphology, size of the crack region, and the stress changes of the surrounding materials. Within the crack region, the material is affected by external loading forces, resulting in stress concentration and a large stress field at the crack tip. The magnitude of the stress directly affects the crack propagation rate. Therefore, accurately calculating the stress distribution in the crack region is the key to predicting the crack propagation trend.

[0099] Specifically, the system discretizes the material in the crack region through finite element mesh generation, converting the stress field near the crack into a numerical calculation problem. The stress at each mesh node can be solved by the following formula:

[0100]

[0101] where σ is the stress, F is the acting force, and A is the area.

[0102] The calculated stress field data provides a strong basis for crack propagation, enabling accurate simulation of the stress state of the crack and the driving force for crack propagation.

[0103] The propagation of cracks is not only driven by external stresses but also affected by the energy transfer within the material. In the energy transfer analysis module, the system analyzes the energy distribution and energy transfer characteristics within the crack region, calculates the driving force for crack propagation, and generates crack energy data. The core of energy transfer analysis is the Energy Release Rate (G), which reflects the energy released by the crack during the crack propagation process.

[0104] The calculation of the energy release rate is closely related to the stress state of the crack. Through the stress intensity factor (K) of the crack and the energy release rate formula, the driving force for crack propagation can be calculated. The system gradually evaluates the energy required for the crack according to different stages of crack propagation:

[0105]

[0106] where G is the energy release rate, U is the energy released during crack propagation, and A is the area of crack propagation.

[0107] In this formula, the energy release rate G represents the energy change per unit area during crack propagation, and the crack propagation direction and speed can be predicted through this data.

[0108] In the dynamic modeling of crack propagation, the crack propagation model construction module is responsible for establishing a dynamic model of crack propagation through stress distribution and energy data. This module simulates various dynamic processes of the crack from the initial state to the propagation process, and finally predicts the trend of crack propagation. By calculating the dynamic behavior of crack propagation through the finite element method, the system can simulate the propagation path of the crack under different loading conditions.

[0109] The crack propagation direction and rate are affected by multiple factors, including the toughness of the material, the initial shape of the crack, external loads, and the internal stress field, etc. Therefore, the system combines the stress data and energy data of the crack and uses the crack propagation dynamic model to simulate the crack propagation process. This model considers factors such as crack stress concentration, energy release rate, and material plasticity, and provides more accurate predictions during the simulation of crack propagation.

[0110] The dynamic formula for crack propagation usually combines the calculation using the stress intensity factor and the energy release rate. The system combines the stress and energy data of the crack and derives the crack propagation rate and direction through the following formula:

[0111]

[0112] where V is the crack propagation rate, K I is the stress intensity factor of the crack, E is the elastic modulus of the material, and G is the energy release rate.

[0113] Through this dynamic calculation, the system can continuously adjust the crack propagation path and perform adaptive optimization according to changes in the external environment.

[0114] Embodiment. In practical applications, assume that a crack appears on the surface of the arrester resistor disc, and the optical monitoring unit collects holographic model data of the crack. The crack modeling and trend prediction unit first calculates the stress distribution in the crack area through the stress field calculation module and simulates the stress intensity factor at the crack tip using the finite element method. Subsequently, the energy transfer analysis module calculates the energy release rate during crack propagation to obtain the driving force for crack propagation. Finally, the crack propagation model construction module combines this data to deduce the crack propagation path and rate, providing a scientific basis for subsequent degradation assessment.

[0115] Preferably, the crack modeling and trend prediction unit further includes:

[0116] A crack propagation rate calculation module for calculating the crack propagation rate based on crack stress data and crack energy data, and estimating the time, rate, and direction of crack propagation;

[0117] A crack path optimization module for estimating the future crack propagation path based on the crack propagation rate using the finite element method and the multi-point iteration method, and generating crack propagation prediction data.

[0118] The crack propagation rate is a key parameter during crack propagation, which directly affects the degradation degree and service life of the arrester resistor disc. The core task of the crack propagation rate calculation module is to accurately calculate the crack propagation rate based on crack stress data and crack energy data, and estimate the time, rate, and direction of crack propagation. In extremely cold environments, due to the influence of temperature changes on material properties, the behavior of crack propagation is more complex, so accurately predicting the crack propagation rate is crucial.

[0119] The crack propagation rate is closely related to the external stress applied to the crack and the energy release characteristics of the material. When a crack propagates, the stress intensity factor in the crack area and the energy release rate of the material determine the crack propagation rate. In the present invention, the calculation of the crack propagation rate depends on the combination of stress data and energy data. Specifically, the crack propagation rate V can be calculated by the following formula:

[0120]

[0121] where V is the crack propagation rate, K I is the stress intensity factor of the crack, E is the elastic modulus of the material, and G is the energy release rate.

[0122] This formula reflects the relationship between the crack growth rate and the stress and energy release in the crack region. The crack growth rate is positively correlated with the elastic modulus E of the material and the stress intensity factor K of the crack I At the same time, the energy release rate G also plays a key role in crack propagation. In extremely cold environments, as the temperature decreases, the elastic modulus of the material and the stress concentration effect of the crack may change. Therefore, the system needs to dynamically calculate these parameters and conduct comprehensive analysis.

[0123] By obtaining the stress data and energy data in the crack region in real time, the calculation module can calculate the crack growth rate, and combine this data with the crack propagation direction to further estimate the time of crack propagation.

[0124] Example: Assume that in an actual environment, the crack length L of the arrester resistor chip is 2 mm, and the stress intensity factor K at the crack tip I is calculated as The elastic modulus E of the material is 200 GPa, and the energy release rate G during crack propagation is 0.05 J / m 2 . According to the above formula, the crack growth rate V can be calculated as follows:

[0125]

[0126] Therefore, the crack growth rate is 2×10 -6 m / s, and based on this value, the crack propagation behavior within a certain time can be further estimated.

[0127] The crack path optimization module is used to estimate the future crack propagation path through the finite element method and the multi-point iteration method, and generate crack propagation prediction data. The optimization of the crack path not only considers stress concentration and the crack growth rate, but also needs to comprehensively consider the material properties of the crack, the crack morphology, and the action of external stress. Since the crack propagation process may be affected by multiple factors, the crack propagation path is often irregular and difficult to predict. Especially in extremely cold environments, the crack propagation behavior may change significantly. Therefore, the optimization of the crack path plays an important role in improving the accuracy of degradation prediction.

[0128] The theoretical basis of the crack path optimization module is the finite element method (FEM) and the multi-point iteration method. The finite element method is mainly used for stress field simulation during crack propagation. Combining the stress intensity factor and energy release rate during crack propagation, it generates the crack propagation path. The multi-point iteration method optimizes the prediction of the future crack propagation direction and rate by repeatedly calculating the crack propagation paths at different time nodes.

[0129] During the crack propagation process, the crack path is not only affected by the stress at the crack tip but may also be influenced by material heterogeneity, stress direction, and external loads. Therefore, the system calculates the crack propagation path for each point through the multi-point iteration method and simulates the crack propagation under different stress fields based on the finite element method to finally obtain the optimal crack propagation path.

[0130] The multi-point iteration method mainly conducts multiple calculations at different nodes of the crack, gradually deduces the crack propagation behavior, and adjusts the crack propagation direction according to the results of each calculation. The calculation of each crack path node is based on the stress distribution and energy data of the crack to ensure that the predicted path has sufficient accuracy and robustness.

[0131] Example: Assume that the initial position of the crack is P0, and the crack path optimization module obtains the optimal crack propagation path through multiple iterative calculations. First, the system calculates the stress state of crack propagation to obtain the initial crack propagation path P1, and then makes dynamic adjustments according to the elastic properties of the material and external loads to obtain a new crack propagation path P2, and so on. After multiple iterations, the crack propagation path will tend to be stable, thus providing an accurate prediction for future propagation trends.

[0132] The system optimizes the crack path through the following iterative formula:

[0133]

[0134] where P n+1 is the crack path at the (n + 1)-th step, ΔP is the increment of the crack path, Δσ is the stress change during crack propagation, and E is the elastic modulus of the material.

[0135] Through this method, the prediction of the crack path can be continuously optimized with each iteration and provide more accurate predictions at each stage of crack propagation.

[0136] By combining the crack propagation rate calculation module and the crack path optimization module, the system can provide accurate crack propagation prediction data. In an extremely cold environment, the crack propagation process of the arrester resistor chip may be affected by factors such as temperature fluctuations and external loads, resulting in changes in the crack propagation path. Therefore, the path optimization method based on the finite element method and the multi-point iteration method can be dynamically adjusted according to different states of the crack, thereby improving the accuracy of crack propagation prediction.

[0137] Preferably, the crack modeling and trend prediction unit further includes:

[0138] The reinforcement learning optimization module is used to dynamically adjust the crack propagation rate and direction based on crack propagation data and crack path optimization data, adaptively optimize the crack evolution path through a reinforcement learning model, and generate crack deterioration trend data.

[0139] The reinforcement learning optimization module is a key component in the crack modeling and trend prediction unit. It is used to dynamically adjust the crack propagation rate and direction based on crack propagation data and crack path optimization data. Through the reinforcement learning model, it adaptively optimizes the crack evolution path and generates crack deterioration trend data, providing decision-making support for subsequent deterioration assessment.

[0140] In extremely cold environments, the crack propagation of arrester resistor chips is affected by factors such as temperature fluctuations and changes in material physical properties. Therefore, the dynamic evolution of crack propagation is extremely complex, and traditional crack propagation prediction methods often cannot fully cope with the challenges brought by extremely cold environments. Reinforcement learning enables the prediction of crack evolution to be more flexible and accurate under various environmental conditions by adaptively adjusting the crack propagation path.

[0141] The application of reinforcement learning in this process is to model the state and behavior of the crack propagation process. The system adjusts the next action strategy according to the feedback of each crack state. The reinforcement learning model can learn the optimal crack propagation path based on crack propagation data and crack path optimization data, and uses a reward function to strengthen the selection of the best behavior during crack propagation, thereby obtaining a more accurate crack propagation prediction.

[0142] The implementation of the reinforcement learning optimization module depends on the Q-learning or deep Q-learning algorithm. The basic principle of reinforcement learning is that the system learns how to select the optimal action strategy according to the current state through interaction with the environment. The crack propagation process can be regarded as a Markov decision process (MDP), where each state represents the current state of the crack, the action represents the possible direction of crack propagation, and the reward function is used to evaluate the effect of crack propagation in the current state.

[0143] In this module, the crack propagation process is modeled as a dynamic system, where each state S represents the current state of the crack. The state of the crack includes information such as the crack position, length, depth, stress distribution, and energy release rate. The action A represents the selection of the crack propagation path, that is, the direction in which the crack propagates.

[0144] The reward function R(S,A) is used to measure the quality of the crack propagation path. In the present invention, the design of the reward function not only considers the speed of crack propagation, but also needs to consider the impact of crack propagation on the varistor resistor chips, such as the degree of damage to the material caused by the crack and the impact of the propagation on the device performance. Assuming that the goal of the system is to minimize the negative impact of crack propagation, that is, to minimize stress concentration and damage caused by cracks as much as possible during crack propagation, the reward function can be designed in the following way:

[0145] E(S,A) = -α·crack damage value + β·propagation rate

[0146] Among them, the crack damage value can be evaluated by the energy release rate of the material and the propagation depth of the crack, indicating the degree of damage to the material caused by the crack. The propagation rate represents the speed of crack propagation, and generally, the faster the crack propagation, the higher the risk. α and β are weight factors that control the balance between the crack propagation rate and damage.

[0147] In reinforcement learning, the Q-value is used to represent the expected value of future rewards that can be obtained by performing a certain action in a certain state. The update of the Q-value adopts the Bellman equation:

[0148]

[0149] Among them, Q(S t ,A t ) is the Q-value of the current state S t and action A t , R(S t ,A t ) is the reward for the current state and action, γ is the discount factor that controls the impact of future rewards, is the maximum Q-value of all possible actions in the next state S t+1 .

[0150] By iteratively updating the Q-value, the system can learn the optimal crack propagation path after multiple interactions and gradually optimize the rate and direction of crack propagation.

[0151] For complex crack propagation problems, traditional Q-learning may not be able to effectively handle a large state space. At this time, deep learning techniques (especially deep neural networks) can be used to approximate the Q-value function, that is, to approximate the Q-value through a neural network model and train through a deep Q-network (DQN). The deep Q-network can handle complex and high-dimensional state spaces, making the optimization of the crack propagation path more accurate and efficient.

[0152] Embodiment. Assume that in practical applications, cracks appear on the surface of the arrester resistor chip, and the optical monitoring unit generates holographic model data of the cracks through laser interference measurement. The crack modeling and trend prediction unit first obtains the stress distribution of the cracks based on the stress field calculation module, and then uses the energy transfer analysis module to calculate the driving force for crack propagation. Next, the reinforcement learning optimization module adaptively optimizes the crack propagation path based on the crack propagation data using the Q-learning algorithm, and adjusts the rate and direction of crack propagation through the reward function. Through multiple iterations of optimization, the system finally generates the deterioration trend data of the cracks, providing high-precision predictions for subsequent deterioration assessments.

[0153] The deterioration assessment and optimization unit is used to receive the crack deterioration trend data, combine the multi-layer time series analysis and the crack distribution influence factor calculation model to predict the long-term deterioration trend of the arrester resistor chip, generate the deterioration assessment data, and adaptively adjust the measurement parameters of the optical monitoring unit based on the deterioration assessment data to optimize the calculation weights of the crack modeling and trend prediction unit.

[0154] The deterioration assessment and optimization unit is one of the key modules in the present invention. By receiving the crack deterioration trend data and combining the multi-layer time series analysis and the crack distribution influence factor calculation model, it accurately predicts the long-term deterioration trend of the arrester resistor chip. The design of this unit can not only evaluate the current state of crack deterioration, but also provide reasonable repair and monitoring strategies based on the long-term deterioration trend to ensure the reliability and long-term safety of the arrester in extremely cold environments.

[0155] The propagation and deterioration of cracks are affected by various factors, including temperature changes, external loads, material fatigue, etc. In extremely cold environments, the mechanical properties of materials and the crack propagation laws may change significantly. Therefore, traditional deterioration assessment methods may not accurately reflect the actual development of cracks in extremely cold environments. For this reason, the present invention proposes a deterioration assessment method based on multi-layer time series analysis and crack distribution influence factor calculation models.

[0156] The core of multi-layer time series analysis lies in the ability to capture the dynamic changes during the crack deterioration process through continuously updated data in the time window, and then effectively predict the future deterioration trend. In the present invention, the propagation and deterioration of cracks is a complex dynamic process, affected by both external factors (such as environmental temperature, load changes, etc.) and internal factors (such as material aging, fatigue, etc.). Therefore, the system uses a multi-level time series model to analyze the crack propagation behavior in different time periods.

[0157] The key steps of multi-layer time series analysis include:

[0158] Data acquisition and preprocessing: Crack propagation data is collected by the monitoring unit and preprocessed, including denoising, smoothing, and normalization.

[0159] Time series modeling: The ARIMA (AutoRegressive Integrated Moving Average) model is used to model the crack propagation process to obtain the trend of crack propagation in different time periods.

[0160] Long short-term memory network (LSTM): To solve the problem of long-term dependencies in time series, an LSTM neural network is used to model the crack propagation data. When dealing with time series data, LSTM can better retain the dependencies over long time periods, thus more accurately predicting the long-term deterioration trend of cracks.

[0161] In extremely cold environments, the propagation of cracks is not only affected by the stress concentration at the crack tip but also closely related to factors such as environmental temperature and the microstructure of the material. Therefore, a crack distribution influence factor calculation model is introduced into the system to consider the propagation characteristics of cracks at different positions. The main role of the crack distribution influence factor model is to quantify the crack propagation behavior in different regions and combine it with factors such as the aging characteristics of the material and external stress to provide a more accurate prediction of the deterioration trend.

[0162] The key parameters of this model include:

[0163] Stress intensity factor (SIF): The stress intensity factor at the crack tip is the main driving force for crack propagation. By calculating the stress intensity factor, the crack propagation rate can be predicted.

[0164] Temperature sensitivity factor (TSF): In extremely cold environments, the impact of temperature changes on materials is crucial. The brittleness of materials at low temperatures may lead to accelerated crack propagation. By introducing the temperature sensitivity factor, the impact of low temperature on crack propagation can be accurately predicted.

[0165] Material fatigue factor (MFF): After the material undergoes multiple stress cycles, fatigue damage will accelerate crack propagation. By calculating the fatigue factor, the deterioration trend of cracks under long-term working loads can be evaluated.

[0166] Combining these influence factors, the crack distribution influence factor calculation model can provide a more refined prediction of the deterioration trend. By simulating the crack propagation at different positions, the system can generate accurate crack deterioration trend data.

[0167] During the crack propagation process, it is necessary not only to evaluate the current state of the crack, but also to predict the future crack propagation in order to provide data support for equipment maintenance and management. The degradation evaluation and optimization unit further optimizes the crack propagation prediction model by combining crack propagation rate data with crack path optimization data, providing an adaptive adjustment function.

[0168] Specifically, based on the prediction results, the system will adaptively adjust the measurement parameters of the optical monitoring unit and optimize the calculation weights of the crack modeling and trend prediction unit. For example, when the crack propagation rate reaches a certain critical value, the system will automatically adjust the measurement accuracy of the optical monitoring unit to further improve the accuracy of crack monitoring.

[0169] Preferably, the degradation evaluation and optimization unit includes:

[0170] A time series analysis module, which is used to receive crack degradation trend data, predict the long-term degradation trend of the arrester resistor chip through multiple sampling and trend analysis methods, and generate long-term degradation trend data;

[0171] A crack distribution influence factor calculation module, which is used to analyze the spatial distribution of cracks and its influence on the performance of the resistor chip based on the crack distribution factor and historical monitoring data, and generate the first degradation evaluation data.

[0172] The degradation evaluation and optimization unit is mainly used to predict the long-term degradation trend of the arrester resistor chip based on the crack degradation trend data through the time series analysis module and the crack distribution influence factor calculation module, and evaluate the overall performance of the resistor chip by combining the influence of the spatial distribution of cracks. Especially in extremely cold environments, the influence of temperature changes, material fatigue and external stress on crack propagation is relatively complex. Therefore, how to accurately predict the long-term evolution trend of cracks and the influence of the spatial distribution of cracks on the performance of resistor chips is the core of evaluating the reliability and life of arresters.

[0173] The time series analysis module is a key module in this unit, which is responsible for predicting the long-term degradation trend of cracks according to the crack degradation trend data using multiple sampling and trend analysis methods, and generating long-term degradation trend data. The core task of this module is to capture the law of crack propagation through changes in the time dimension, so as to provide decision support for long-term monitoring and maintenance.

[0174] Time series analysis can extract the long-term law of crack propagation through multiple sampling and analysis of crack degradation trend data. During the crack propagation process, especially in extremely cold environments, the crack propagation is not only related to the stress state, but also closely related to the environmental temperature and material fatigue. Therefore, through the time series method, the long-term trend of crack propagation can be modeled to predict the possible development trend of cracks in the future.

[0175] In the present invention, crack propagation data generally includes the length, depth, propagation rate, etc. of cracks. After multiple sampling analyses and combined with a time series prediction model, it can effectively avoid the prediction errors caused by ignoring environmental changes in traditional methods. The time series analysis method mainly uses the autoregressive integrated moving average (ARIMA) model and the long short-term memory (LSTM) network to model the time dependence of crack propagation.

[0176] ARIMA model: This model predicts the future development of cracks through autoregressive analysis of time series and combined with historical data. The present invention can identify seasonal changes, trend changes, and periodic fluctuations in crack propagation.

[0177] LSTM network: In the case of a large amount of data with complex time series dependencies, the LSTM network can capture the long-term trends and short-term fluctuations in crack propagation through long short-term memory processing of crack propagation data.

[0178] By combining these two methods, the time series analysis module can predict the future behavior of crack propagation, generate long-term deterioration trend data, and provide a basis for subsequent equipment maintenance and repair.

[0179] Embodiment: Assume that during the initial propagation of a crack, the crack length data at each time point is recorded as L(t). The time series analysis module processes this data to generate prediction data for future times t+1 and t+2. For example, using the ARIMA model analysis, the crack propagation length in the next three months is obtained as:

[0180] L(t+1) = φ1L(t) + φ2L(t - 1) + ∈

[0181] where φ1 and φ2 are the parameters of the ARIMA model, and ∈ is the error term.

[0182] This result is used to predict the future crack propagation and provide data support for equipment management based on the long-term trend.

[0183] The propagation of cracks inside the material is affected not only by local stress but also by the spatial distribution of cracks, the interaction at the crack tip, and the overall structure of the material. The crack distribution influence factor calculation module aims to analyze the spatial distribution of cracks in the material and its influence on the overall performance of the resistor chip. By combining the crack distribution factor and historical monitoring data, this module provides a comprehensive assessment of the influence of cracks on the performance of the arrester resistor chip, and then generates the first deterioration assessment data.

[0184] The spatial distribution of cracks has an important impact on the mechanical properties of materials. In lightning arrester resistor chips, the distribution of cracks may lead to stress concentration, which in turn accelerates crack propagation and affects the overall performance of the material. The crack distribution factor calculation module analyzes historical monitoring data, considers the distribution characteristics of cracks at different positions, and analyzes the interaction between cracks and the material.

[0185] The main tasks of this module are:

[0186] Crack distribution factor calculation: Calculate the crack distribution factor according to the position of cracks in the resistor chip. This factor reflects the influence of cracks at different positions on the material properties, especially the influence of crack concentration areas on the performance of the resistor chip.

[0187] Stress concentration analysis: Combine finite element analysis (FEM) to evaluate the stress concentration effect in the crack area and determine the regional material weaknesses that may be caused by crack propagation.

[0188] Integration of historical monitoring data: By combining past crack monitoring data, calculate the distribution changes of cracks at different time points and evaluate the long-term impact of cracks on the performance of the resistor chip.

[0189] Example: The crack distribution factor α can be calculated by the following formula:

[0190]

[0191] where n is the number of crack monitors, and the crack stress concentration factor O i Calculated by the finite element method, it represents the stress concentration effect of cracks at the i-th position. Through the calculated crack distribution factor, the expansion effect of cracks in different regions can be evaluated, providing important spatial data support for predicting the crack deterioration trend.

[0192] Through the collaborative work of the time series analysis module and the crack distribution impact factor calculation module, the deterioration assessment and optimization unit can provide a more accurate long-term deterioration trend prediction for lightning arrester resistor chips. Especially in extremely cold environments, the influence of temperature changes on crack propagation is very significant, and it is difficult for traditional methods to comprehensively consider these influencing factors. By combining the above two methods, the system can accurately capture the dynamic changes of crack propagation and provide more powerful data support for the maintenance and management of equipment.

[0193] Specifically, the time series analysis module can effectively capture the long-term trends and periodic fluctuations in the crack propagation process, while the crack distribution impact factor calculation module can accurately evaluate the influence of the spatial distribution of cracks on the performance of the resistor chip, providing a decision-making basis for subsequent repair and optimization.

[0194] Preferably, the deterioration assessment and optimization unit further includes:

[0195] An adaptive adjustment module, which is used to dynamically adjust the measurement parameters of the optical monitoring unit according to the first deterioration evaluation data, and optimize the calculation weights of the crack modeling and trend prediction unit in combination with the monitoring feedback signal, so as to improve the accuracy of the crack prediction model.

[0196] The adaptive adjustment module is one of the core components in the deterioration evaluation and optimization unit. It aims to dynamically adjust the measurement parameters of the optical monitoring unit according to the first deterioration evaluation data, and optimize the calculation weights of the crack modeling and trend prediction unit in combination with the monitoring feedback signal, thereby improving the accuracy of the crack prediction model. This module mainly optimizes through the real-time feedback of monitoring data to achieve accurate prediction and effective monitoring of crack propagation. In extremely cold environments, the stress and crack propagation behavior of materials may fluctuate violently due to temperature changes. Therefore, the real-time optimization function of the adaptive adjustment module is crucial.

[0197] In extremely cold environments, the crack propagation process of the arrester resistor chip is not only affected by the stress at the crack tip, but may also be jointly affected by multiple factors such as temperature and material aging. Therefore, the adaptive adjustment mechanism based on real-time data can help the system adjust the monitoring and prediction models according to environmental changes, improving the prediction accuracy and monitoring accuracy.

[0198] The basic principle of adaptive adjustment is to evaluate the system performance according to the first deterioration evaluation data (the crack propagation trend data generated by the crack propagation prediction model), and dynamically adjust the parameters of the monitoring system through the action of the feedback signal. This adjustment is not limited to the measurement parameters of the optical monitoring unit, but also includes the calculation weights of the crack modeling and trend prediction unit.

[0199] Specifically, the first deterioration evaluation data provides the state information and prediction results of crack deterioration. Based on these data, the system can identify possible errors or deviations in the current monitoring process, and then make adjustments. For example, if the crack propagation rate is high, the system will automatically increase the sampling frequency of the optical monitoring unit to obtain more accurate crack propagation data. At the same time, based on the monitoring feedback signal, the system can optimize the weight configuration in the crack modeling and trend prediction unit, making the prediction of the future expansion trend of the model more accurate.

[0200] Embodiment. Assume that in practical applications, the cracks of the arrester resistor chip are monitored by an optical monitoring unit, and the system collects the data of crack propagation in real time. According to these data, the first deterioration evaluation data is generated and fed back to the adaptive adjustment module. For example, assume that the optical monitoring unit collects a relatively fast crack propagation rate and a non-linear change at the initial measurement. The system will make adjustments through the following steps:

[0201] Adjust the parameters of the optical monitoring unit based on the first deterioration assessment data: By evaluating the crack growth rate, the system finds that the crack growth speed has significantly accelerated and the direction of crack growth has changed greatly. Therefore, the sampling frequency and optical resolution of the optical monitoring unit will be adjusted according to the crack growth characteristics. The system may increase the sampling frequency so that the optical monitoring unit can provide more detailed crack growth information at shorter time intervals. In addition, the detection spectral range of the monitoring unit may also be optimized according to the change in crack morphology to capture more details.

[0202] Optimize the calculation weights of the crack modeling and trend prediction unit in combination with the feedback signal: The system will optimize the calculation weights in the crack modeling and trend prediction unit through the feedback signal according to the adjustment results of the optical monitoring unit. If the crack growth pattern becomes more complex, the system can adjust the weights of different data sources (such as stress data, crack morphology data, etc.) in the crack modeling unit, making the model pay more attention to the key parameters related to the crack growth trend. For example, if the change in crack depth has a greater impact on the growth rate, the system will increase the weight of the crack depth data in the model.

[0203] During the optimization process, the system dynamically adjusts the parameter weights of the model according to the feedback signal, making the crack growth prediction more accurate. The goal of optimization is to maximize the accuracy and adaptability of the model in predicting crack growth, especially in the extremely cold environment with rapid changes.

[0204] Adaptive adjustment calculation expression: Assume that the crack growth prediction model used in the system is based on the energy release rate G and the stress intensity factor K I of the calculation formula. The adaptive adjustment module can adjust the weight parameters in the calculation formula of the crack growth rate according to the first deterioration assessment data. Taking the following formula as an example:

[0205]

[0206] In the above formula, n is the weight factor dynamically adjusted according to the feedback signal, K I is the crack stress intensity factor, G is the energy release rate, and E is the elastic modulus of the material. The system dynamically adjusts the value of n according to the crack growth trend data to ensure the more accurate calculation of the crack growth rate V.

[0207] Through the dynamic adjustment of the adaptive adjustment module, the system can effectively cope with the uncertainty of crack growth behavior in the extremely cold environment. In the extremely cold environment, the crack growth of the arrester resistor chip is not only affected by external stress, but also closely related to factors such as temperature and material aging. The adaptive adjustment module adjusts the measurement accuracy of the monitoring unit and optimizes the calculation weights of the crack prediction model according to the real-time data of crack growth and the first deterioration assessment data, so as to ensure the accuracy and reliability of crack growth prediction.

[0208] Preferably, the deterioration evaluation and optimization unit further includes:

[0209] A monitoring data fusion and optimization module, configured to fuse historical data and real-time data, train an intelligent monitoring model based on the fused data, and optimize the response ability of the monitoring system through an adaptive algorithm to generate a final adaptive monitoring strategy.

[0210] The monitoring data fusion and optimization module is one of the important components of the present invention. Its core function is to fuse historical monitoring data and real-time monitoring data, train an intelligent monitoring model using the fused data, and optimize the response ability of the monitoring system through an adaptive algorithm to finally generate an adaptive monitoring strategy. In extremely cold environments, temperature fluctuations can have a significant impact on the mechanical properties of materials, and the crack propagation behavior may be more complex. Therefore, traditional monitoring methods may not be fully adapted to the crack propagation characteristics under extremely cold conditions. The design purpose of the monitoring data fusion and optimization module is precisely to address this challenge. By integrating data from different sources and optimizing the monitoring response ability, the monitoring system can still maintain high accuracy and response speed in extreme environments.

[0211] Monitoring data fusion refers to combining historical data and real-time data to form a complete data set for subsequent training of the intelligent monitoring model. In the present invention, historical data usually includes information such as past crack propagation data, monitoring results, and environmental conditions, while real-time data comes from the current crack monitoring process, including real-time changing data such as temperature, crack length, propagation rate, and stress.

[0212] Fusion of historical data and real-time data: Through a data fusion algorithm, the historical monitoring data and real-time data are effectively combined. Historical data provides long-term observation results of crack propagation and the aging characteristics of materials, while real-time data can reflect the current state of crack changes. By fusing these two types of data, more comprehensive information can be obtained to ensure that the monitoring system responds to crack propagation more timely and accurately.

[0213] The core algorithms used in the data fusion process include:

[0214] Kalman filtering algorithm: mainly used for smoothing and denoising real-time data. During the processing, the sensor noise is modeled so that the system can more accurately obtain the real-time dynamics of crack propagation.

[0215] Weighted average method: By weighting the historical data and real-time data, the weighted average of the fused data can be achieved, enabling the model to better adapt to changes at different monitoring stages.

[0216] Deep learning: Combine deep learning methods to further learn and optimize the fused data. Especially for non-linear relationships and complex crack propagation patterns, deep learning methods can effectively improve the prediction ability of the model.

[0217] The intelligent monitoring model is mainly trained by using the fused data to provide support for the long-term monitoring and management of the system. During the training process of the intelligent monitoring model, the system uses different data sets (such as crack size, depth, propagation rate, etc.) for training, and adopts common methods such as deep neural network (DNN) and convolutional neural network (CNN). Through these technologies, features are extracted from a large amount of historical data to achieve accurate prediction of crack propagation.

[0218] During the training process, the system continuously optimizes the loss function to improve the prediction accuracy of the model. For example, the loss function L can be expressed as follows:

[0219]

[0220] where y i is the real crack propagation data, is the crack propagation data predicted by the model, and N is the number of data samples.

[0221] Adaptive algorithm to optimize the response ability of the monitoring system: The core of the adaptive algorithm is to dynamically adjust the parameters of the monitoring system according to the monitoring feedback signal. These adjustments include parameters such as the sampling frequency and measurement accuracy of the optical monitoring unit, as well as the model calculation weights in the crack modeling and trend prediction unit. The adaptive adjustment algorithm continuously optimizes the monitoring system through real-time feedback results to ensure that the system can respond according to the crack propagation trend.

[0222] During this process, the feedback signal contains information such as crack changes and crack propagation rate that occur during the current monitoring process. The adaptive algorithm uses these feedback signals to adjust the reaction ability of the system, so that the accuracy of crack monitoring always remains at a relatively high level.

[0223] For example, assume that the crack propagation speed suddenly increases at a certain moment. The system will increase the monitoring sampling frequency through the adaptive algorithm, thereby improving the accuracy of crack data collection. In addition, combined with the adjustment of the feedback signal, the intelligent monitoring model can optimize its calculation weights and improve the prediction accuracy of the crack propagation trend.

[0224] Embodiment. Assume that during the application process, the optical monitoring unit provides real-time data on cracks, including information such as the length, propagation rate, and crack depth of the cracks. The system fuses the real-time data with historical data, uses the Kalman filtering algorithm to denoise and smooth the real-time data, and at the same time uses the weighted average method to fuse the historical data and the real-time data. The intelligent monitoring model is trained through a deep learning algorithm, and the fused data is input into this model to predict the crack propagation trend.

[0225] If the intelligent monitoring model predicts that the crack propagation rate is about to reach the critical danger value, the system dynamically adjusts the sampling frequency and accuracy of the optical monitoring unit through the adaptive adjustment module. For example, at the critical moment of crack propagation, the sampling frequency of the optical monitoring unit may be increased from the original 1 Hz to 5 Hz to accurately capture every detail of the crack change. At the same time, based on the monitoring feedback signal, the system will also optimize the calculation weights in the crack prediction model, so as to significantly improve the accuracy of crack propagation prediction.

[0226] Preferably, the system further includes:

[0227] A data storage unit for storing crack holographic model data, crack deterioration trend data, deterioration assessment data, and adaptive monitoring strategies for subsequent analysis and system optimization.

[0228] The data storage unit not only needs to support the storage of large-scale data, but also needs to be able to perform data reading and writing operations efficiently, so as to quickly respond and adjust the monitoring strategy during the actual operation process. This unit must have a high degree of reliability and real-time performance to ensure that the system can continue to operate in extremely cold environments and still maintain high-efficiency data processing capabilities when the data volume is increasing.

[0229] Storage of crack holographic model data: Crack holographic model data is the crack information collected by the optical monitoring unit, including data such as the size, depth, and propagation trend of the cracks. The storage of these data requires high precision and must ensure long-term access stability. To achieve this, the data storage unit adopts high-performance storage technologies such as solid-state drives (SSDs) or distributed storage systems to ensure the persistence and reliability of the data in extremely cold environments.

[0230] During storage, the crack holographic model data is preprocessed through compression and encryption algorithms to reduce the storage space and ensure the security of the data. When needed, the system can quickly retrieve these data for subsequent analysis and prediction.

[0231] Storage of crack deterioration trend data: The crack deterioration trend data is generated by the crack modeling and trend prediction unit, including the crack growth rate, growth direction, and dynamic data of crack evolution. Since the crack deterioration trend data contains a large amount of time-series data, an efficient time-series database is required for storage. This data is usually stored using compression technology and index optimization is performed to enable rapid retrieval and trend prediction in subsequent analyses.

[0232] During the storage process, the crack deterioration trend data is optimized by combining the historical evolution trend of the crack and real-time monitoring data through database indexing and data compression. This can improve the data reading efficiency and ensure the integrity and consistency of the crack deterioration trend data.

[0233] Storage of deterioration assessment data: The deterioration assessment data includes long-term deterioration prediction results generated based on the crack deterioration trend data and other analysis data. These data play a decisive role in the adjustment of the monitoring strategy and equipment maintenance. The storage of this data requires high precision and high reliability to ensure that data is not lost in extremely cold environments.

[0234] The deterioration assessment data usually involves the calculation results of multiple variables, so each set of data needs to be stored and marked separately. The system organizes and manages the data through a relational database or a NoSQL database, enabling independent data records to be generated for each assessment according to different conditions.

[0235] Storage of adaptive monitoring strategies: The adaptive monitoring strategies are optimized strategies obtained by adjusting the monitoring parameters based on the real-time data of crack growth. These strategies directly affect the response ability and prediction accuracy of the monitoring system, so their storage method needs to ensure high efficiency and flexibility. To ensure that the monitoring system can respond to changes in a timely manner, the strategy data must be quickly accessible and updatable.

[0236] When storing the adaptive monitoring strategies, the system uses a fast cache technology so that the strategy data can be quickly read and applied when needed. At the same time, the system supports incremental updates and version control to ensure that the strategy data is not lost during the continuous update process.

[0237] As Figure 3 shown, a method for monitoring the deterioration degree of arrester resistor chips in an extremely cold environment, which is used to implement the system for monitoring the deterioration degree of arrester resistor chips in an extremely cold environment. The steps of the present invention include:

[0238] Perform multi - band optical irradiation on the arrester resistor chip through an optical quantum resonance sensor, and based on optical interference measurement of the optical resonance characteristics of the crack area, obtain the size, depth, and expansion trend of the crack, and generate crack holographic model data; use the optical quantum resonance sensor to perform multi - band optical irradiation on the arrester resistor chip, and capture the reflection signal of the crack area. Based on optical interference technology, analyze the optical resonance characteristics of the crack area, obtain the size, depth, and expansion trend of the crack, and generate crack holographic model data. This step ensures the accurate acquisition of information such as crack morphology and size through high - precision optical imaging and interference measurement of the crack, providing basic data for subsequent analysis.

[0239] Receive the crack holographic model data, calculate the stress distribution, energy transfer characteristics, and expansion driving force of the crack, and construct a crack expansion dynamics model; based on the crack evolution path optimization method, combine reinforcement learning to dynamically adjust the crack expansion rate and direction, and generate crack deterioration trend data; use the crack holographic model data to calculate the stress distribution and energy transfer characteristics of the crack area, and analyze the crack expansion driving force. Based on the above data, construct a crack expansion dynamics model to simulate the crack expansion behavior under different stress and temperature environments. Adopt the crack evolution path optimization method, combine reinforcement learning to dynamically adjust the crack expansion rate and direction, and generate crack deterioration trend data. The reinforcement learning algorithm optimizes the crack expansion prediction through a feedback mechanism, enhancing the adaptability and accuracy of the model.

[0240] Receive the crack deterioration trend data, combine the multi - layer time series analysis and the crack distribution influence factor calculation model to predict the long - term deterioration trend of the arrester resistor chip, and generate deterioration evaluation data; based on the deterioration evaluation data, adaptively adjust the measurement parameters of the optical monitoring unit, and optimize the calculation weights of the crack modeling and trend prediction unit. Receive the crack deterioration trend data, combine the multi - layer time series analysis method to analyze the long - term evolution law of the crack. Use the crack distribution influence factor calculation model to evaluate the influence of crack distribution on the performance of the resistor chip, and predict the long - term deterioration trend. Generate deterioration evaluation data to provide an evaluation of the overall performance of the resistor chip, providing a basis for equipment maintenance and repair decisions.

[0241] Based on the deterioration evaluation data, adaptively adjust the measurement parameters of the optical monitoring unit to optimize the response accuracy of the monitoring system. By optimizing the calculation weights of the crack modeling and trend prediction unit, improve the accuracy of the crack expansion prediction model, making the system more adaptable to the complex changes in crack expansion in extremely cold environments.

[0242] The above are only embodiments of the present application and are not used to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the scope of the claims of the present application.

Claims

1. A system for monitoring the degradation degree of lightning arrester resistors in extremely cold environments, characterized in that: The system includes: Optical monitoring unit, used to collect crack information of arrester resistors; A crack modeling and trend prediction unit, used to calculate the stress distribution, energy transfer characteristics and expansion driving force of the crack according to the crack information, build a crack expansion model and generate crack degradation trend data; The degradation assessment and optimization unit is used to assess the long-term degradation degree of the resistor and generate degradation assessment data based on the crack degradation trend data in combination with multi-layer time series analysis and crack distribution influencing factor calculation model.

2. The system for monitoring the degradation degree of arrester resistors in extremely cold environments according to claim 1 is characterized in that: The optical monitoring unit comprises: A multi-band laser irradiation device, used to emit multi-wavelength lasers to the surface of the arrester resistor and collect optical signals reflected or interfered from the crack area; The optical signal processing module is used to convert the optical signal into crack information including crack size, depth and initial expansion characteristics.

3. The system for monitoring the degradation degree of arrester resistors in extremely cold environments according to claim 1 is characterized in that: The crack modeling and trend prediction unit comprises: The stress field calculation module is used to analyze the stress distribution in the crack area and obtain the stress intensity information of the crack; Energy transfer analysis module, used to analyze the energy release and transfer characteristics within the crack and determine the driving force of crack extension; The crack growth model building module is used to build a crack growth model by combining stress field and energy data, and output crack degradation trend data.

4. The system for monitoring the degradation degree of arrester resistors in an extremely cold environment according to claim 1 is characterized in that: The crack modeling and trend prediction unit further comprises: A crack growth rate and path calculation module, used to calculate the crack growth rate and growth path according to the crack growth model; The reinforcement learning optimization module is used to dynamically adjust the crack expansion direction and speed based on the expansion rate and path data, and to adaptively optimize the crack evolution path.

5. The system for monitoring the degradation degree of arrester resistors in extremely cold environments according to claim 1 is characterized in that: The degradation assessment and optimization unit comprises: Multi-layer time series analysis module, used to perform multi-period and multi-resolution trend decomposition and modeling of crack degradation trend data, and combined with historical monitoring data to predict the long-term degradation trend of resistors in extreme cold environments; The crack distribution influence factor calculation module is used to quantify the distribution characteristics of cracks in different areas of the resistor and their impact on the overall performance based on the crack distribution influence factor calculation model, and generate degradation assessment data.

6. The system for monitoring the degradation degree of arrester resistors in extremely cold environments according to claim 5 is characterized in that: The degradation assessment and optimization unit makes a comprehensive judgment on the long-term degradation degree of the resistor according to the results of the multi-layer time series analysis and crack distribution influence factor calculation module, and outputs degradation assessment data for further monitoring or maintenance decision-making.

7. The system for monitoring the degradation degree of arrester resistors in extremely cold environments according to claim 5 or 6, characterized in that: The degradation assessment and optimization unit further includes an adaptive adjustment module for adjusting the measurement parameters of the optical monitoring unit in real time according to the degradation assessment data and the monitoring feedback signal, and optimizing the calculation weights of the crack modeling and trend prediction unit.

8. The system for monitoring the degradation degree of arrester resistors in extremely cold environments according to claim 7 is characterized in that: The degradation assessment and optimization unit also includes a monitoring data fusion optimization module, which is used to fuse historical data with real-time data, optimize the response capability of the monitoring system through an adaptive algorithm, and output a final adaptive monitoring strategy.

9. The system for monitoring the degradation degree of arrester resistors in an extremely cold environment according to claim 1 is characterized in that: It also includes a data storage unit for storing crack data collected by the optical monitoring unit, extended model data output by the crack modeling and trend prediction unit, and degradation assessment data and adaptive monitoring strategies generated by the degradation assessment and optimization unit.

10. A method for monitoring the degradation degree of a lightning arrester resistor in an extremely cold environment, used to implement the system for monitoring the degradation degree of a lightning arrester resistor in an extremely cold environment as claimed in any one of claims 1 to 9, characterized in that: The method comprises the following steps: Perform multi-band optical irradiation on the arrester resistor through an optical monitoring unit, collect optical signals reflected or interfered from the crack area, and convert the optical signals into crack information including crack size, depth and initial extension characteristics; Inputting the crack information into a crack modeling and trend prediction unit, calculating the stress distribution, energy transfer characteristics and expansion driving force of the crack, and building a crack expansion model based on the crack information; generating crack degradation trend data through the model; Inputting the crack degradation trend data and historical monitoring data into a multi-layer time series analysis model to perform trend decomposition and prediction on crack extension behaviors at different time scales and resolutions; According to the calculation model of crack distribution influencing factors, the distribution characteristics of cracks inside the resistor and their impact on the overall performance of the resistor are evaluated; Comprehensively judge the output results of multi-layer time series analysis and the calculation results of crack distribution influencing factors to obtain the long-term degradation degree of the resistor; output degradation assessment data for subsequent monitoring adjustment or maintenance decision-making; According to the degradation assessment data, the measurement parameters of the optical monitoring unit and the calculation weights of the crack modeling and trend prediction unit are dynamically adjusted.

Images