Post insulator crack damage detection and early warning method, device and system

Abstract

The invention relates to the field of power equipment health monitoring, and particularly discloses a post insulator crack damage detection and early warning method, device and system.The post insulator crack damage detection and early warning method comprises the steps that loaded mechanical characteristics of a post insulator are analyzed on the basis of stress wave signals, and a crack dynamic survival model is constructed according to the loaded mechanical characteristics to predict existence of cracks; determining an optimal phase space reconstruction parameter and drawing a recurrence plot for the post insulator predicted to have the crack, determining a multi-scale damage characteristic according to a change trend of the recurrence plot, and quantifying a crack damage degree based on the multi-scale damage characteristic; a damage process prediction model is constructed through the crack damage degree to calculate a damage evolution factor, the crack damage evolution trend of the post insulator with cracks is predicted, a health threshold value and a failure threshold value are set for double-layer early warning, and early accurate diagnosis and trend pre-judgment of the crack damage of the post insulator can be achieved. And the monitoring efficiency and the early warning reliability are improved.

Description

Technical Field

[0001] This invention relates to the field of power equipment health monitoring, and more specifically, to a method, device, and system for detecting and warning of crack damage in post insulators. Background Technology

[0002] In power systems, post insulators serve as critical support and insulation components, and their health directly impacts the safe operation of the power network. With the continuous expansion of power system capacity, post insulators operate under high voltage, heavy loads, and complex climatic environments for extended periods, making them highly susceptible to multiple factors such as mechanical shock, thermal expansion and contraction, arc burns, and environmental corrosion. This can lead to internal or surface damage such as cracks, glaze peeling, and discharge erosion, potentially causing insulation breakdown, equipment short circuits, or even large-scale power outages.

[0003] Currently, crack damage monitoring mainly relies on manual inspections or image recognition-assisted systems. Manual inspections depend on regular on-site checks by professionals, which not only have long inspection cycles and limited coverage, but are also limited by the working environment and human subjective judgment, resulting in a high risk of missed detections and misjudgments. Furthermore, operating near high-voltage equipment poses significant safety hazards. Existing automatic monitoring technologies still have shortcomings in real-time performance, accuracy, and adaptability to complex operating conditions. To address these shortcomings, current solutions rely on periodic manual inspections or automated monitoring methods using fixed thresholds to assess the health status of post insulators. However, these solutions neglect the dynamic evolution characteristics of cracks during actual service, such as the temporal information of crack initiation, propagation, and connection stages, and lack a sensitive sensing mechanism for minute crack propagation trends.

[0004] Therefore, it is necessary to provide a method, device, and system for detecting and warning of crack damage in post insulators to solve the above-mentioned technical problems. In order to solve the above problems, a technical solution is provided. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art, the present invention provides a method, device, and system for detecting and warning of crack damage in post insulators. This addresses the problems of low efficiency in existing manual inspections and high false alarm rates caused by the lack of accurate assessment of the dynamic survival state of cracks in automated monitoring methods, which are unable to effectively warn of the evolution trend of crack damage.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A method for detecting and warning of crack damage in post insulators includes the following steps: By deploying a voltage sensor array on the post insulator to acquire stress wave signals, analyzing the load-bearing mechanical characteristics of the post insulator based on the stress wave signals, and constructing a dynamic crack survival model based on the load-bearing mechanical characteristics to predict the existence of cracks; For post insulators predicted to have cracks, the optimal phase space reconstruction parameters are determined based on stress wave signals, and a recursion diagram is drawn. Multi-scale damage characteristics are determined based on the changing trend of the recursion diagram, and the degree of crack damage is quantified based on the multi-scale damage characteristics. A damage history prediction model is constructed by constructing a damage history prediction model based on the degree of crack damage to calculate the damage evolution factor, predict the crack damage evolution trend of the post insulator with cracks, and set a health threshold and a failure threshold for dual-layer early warning. The process of predicting the crack damage evolution trend of cracked post insulators includes: The first signal library of post insulators with cracks is obtained, and the first stress wave signal of post insulators with cracks at the same monitoring point and at the same time point is classified into the third signal library. A first stress variation curve is plotted using a first signal library, and a first stress value is extracted based on the first stress variation curve; a third stress variation curve is plotted using a third signal library, and a third stress value is extracted based on the third stress variation curve. A damage history prediction model is established based on the degree of crack damage, the first signal library, and the third signal library to calculate the damage evolution factor and predict the crack damage evolution trend of the post insulator with cracks.

[0007] As a further aspect of the present invention, based on the deployment of a voltage sensor array on the post insulator, specifically, by uniformly arranging piezoelectric sensors from top to bottom on the post insulator as monitoring points to acquire the first stress wave signal and construct a first signal library, and statistically analyzing the first stress wave signals of all post insulators at the same monitoring point at the same time to construct a second signal library, and preprocessing the first signal library and the second signal library respectively.

[0008] As a further aspect of the present invention, the mechanical characteristics of the post insulator under load are analyzed based on stress wave signals. These mechanical characteristics include average stress intensity and local stress concentration intensity. The average stress intensity includes a first average stress intensity and a second average stress intensity, and the local stress concentration intensity includes a first local stress concentration intensity and a second local stress concentration intensity. The specific process is as follows: The first stress variation curve is plotted using the first signal library, and the second stress variation curve is plotted using the second signal library. The stress value is then extracted based on the stress variation curve. The first average stress intensity is calculated based on the second signal library of the post insulator at the same time; the second average stress intensity is calculated based on the first signal library of the post insulator respectively; the first local stress concentration intensity of each post insulator is calculated based on the stress value and the first average stress intensity; and the second local stress concentration intensity of each post insulator is calculated based on the stress value and the second average stress intensity.

[0009] As a further aspect of the present invention, a dynamic crack survival model is constructed based on the mechanical characteristics of the load to predict the existence of cracks. The specific process includes: Based on the pre-calculation of the weight function at several reference crack lengths Sensor response matrix under several far-field stress conditions , The first crack tip stress is determined based on the reference crack length. With sensor response matrix Establish weight function mapping as simulation samples In the formula: Let be the sensor hotspot vector at time t. The stress at the first crack tip at time t; The second crack tip stress is obtained by correcting the output of the weight function mapping model through linear regression. A dynamic crack survival model is built based on the dynamic stress concentration factor at the crack tip and the stress at the second crack tip, and the dynamic crack survival factor is calculated. Extract the crack dynamic survival factor and compare it with a preset threshold. If the crack dynamic growth factor is greater than the preset threshold, the prediction output indicates that a crack exists; if the crack dynamic growth factor is less than the preset threshold, the prediction output indicates that no crack exists.

[0010] As a further aspect of the present invention, multi-scale damage characteristics are determined based on the changing trend of the recursion graph, and the degree of crack damage is quantified based on the multi-scale damage characteristics. The specific process includes: calculating the recursion rate, entropy, and average diagonal length based on the recursion graph as multi-scale damage characteristics, and establishing a crack damage degree assessment model based on the multi-scale damage characteristics to calculate the degree of crack damage.

[0011] As a further aspect of the present invention, a dual-layer early warning system is implemented by setting a health threshold and a failure threshold. The specific process includes: obtaining historical damage evolution factors of the post insulator cracks as threshold classification samples; determining the health threshold and failure threshold based on the threshold classification samples; and comparing the damage evolution factors with the health threshold and failure threshold respectively to determine the damage state of the post insulator. Specifically: If the damage evolution factor is greater than the health threshold, the post insulator can still be used normally; If the damage evolution factor is less than or equal to the health threshold and greater than or equal to the failure threshold, the post insulator can still be used, but the detection frequency needs to be increased. At this time, a level one warning is triggered. If the damage evolution factor is less than the failure threshold, the post insulator cannot continue to be used, and a level two warning is triggered.

[0012] As a further aspect of the present invention, the health threshold is determined by taking the mean value of the historical damage evolution factors of post insulators that have no abnormalities during use but have cracks. The failure threshold is determined by taking the mean value of the historical damage evolution factors of post insulators that have abnormalities and cracks during use.

[0013] A device for detecting and warning of crack damage in a post insulator, applied to the aforementioned method for detecting and warning of crack damage in a post insulator, the device comprising: a voltage sensor, a crack prediction unit, a crack damage quantification unit, and a damage evolution prediction unit; Voltage sensors are used to acquire stress wave signals; The crack prediction unit is used to analyze the mechanical characteristics of the post insulator under load based on stress wave signals, and to construct a dynamic crack survival model based on the mechanical characteristics under load to predict the existence of cracks. The crack damage quantification unit is used to determine the optimal phase space reconstruction parameters based on the first stress wave signal for the post insulator predicted to have cracks, and to draw a recursion graph. Based on the changing trend of the recursion graph, the multi-scale damage characteristics are determined, and the degree of crack damage is quantified based on the multi-scale damage characteristics. The damage evolution prediction unit is used to construct a damage history prediction model based on the degree of crack damage, calculate the damage evolution factor, predict the damage evolution trend of cracked post insulators, and set health threshold and failure threshold for dual-layer early warning.

[0014] A crack damage detection and early warning system for post insulators includes: Memory, used to store programs; A processor is used to execute the program stored in the memory. When the program is executed, the processor is used to execute the above-described method for detecting and warning of crack damage in post insulators.

[0015] The technical effects and advantages of this invention, which provides a method, device, and system for detecting and warning of crack damage in post insulators, are as follows: This invention analyzes the load-bearing mechanical characteristics of post insulators based on stress wave signals, constructs a dynamic crack survival model based on these characteristics to predict the presence of cracks, determines the optimal phase space reconstruction parameters and draws a recursion graph for predicted cracks in post insulators, determines multi-scale damage characteristics based on the changing trends of the recursion graph, and quantifies the degree of crack damage based on these multi-scale damage characteristics, constructs a damage history prediction model based on the degree of crack damage to calculate damage evolution factors, predicts the crack damage evolution trend of post insulators with cracks, and sets health thresholds and failure thresholds for dual-layer early warning. This enables early and accurate diagnosis and trend prediction of crack damage in post insulators, improving monitoring efficiency and the reliability of early warning.

[0016] This invention utilizes a voltage-type sensor array to acquire stress wave signals in real time, enabling precise perception of the load state of post insulators and improving the real-time and comprehensiveness of monitoring. Based on the load-bearing mechanical characteristics extracted from the stress wave signals, a dynamic crack survival model is constructed, which can scientifically and accurately predict the existence of cracks, avoiding false alarms and missed detections caused by relying solely on surface detection. For post insulators predicted to have cracks, phase space reconstruction and recursive graph analysis methods are used to mine the multi-scale dynamic characteristics of the stress wave signals, achieving a quantitative assessment of the crack damage degree and enhancing the ability to understand the internal evolution process of cracks, effectively improving the accuracy and stability of crack detection. A damage history prediction model based on the crack damage degree dynamically reflects the crack evolution trend by calculating damage evolution factors. Combined with set health and failure thresholds, a dual-layer early warning mechanism is implemented, ensuring the safe operation of equipment and optimizing maintenance resource allocation, avoiding blind replacement or delayed repairs. Attached Figure Description

[0017] Figure 1 A flowchart illustrating a method for detecting and warning of crack damage in post insulators provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of a device for detecting and warning of crack damage in a post insulator, provided in an embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of this invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described technical solutions are only a part of this invention, and not all of it. All other technical solutions obtained by those skilled in the art based on the technical solutions of this invention without inventive effort are within the scope of protection of this invention.

[0019] like Figure 1 The diagram shown is a flowchart of a method for detecting and warning of crack damage in post insulators provided by an embodiment of the present invention. Figure 1 The execution entity of the method shown can be a software and / or hardware device. The execution entity of this application can include, but is not limited to, at least one of the following: user equipment, network equipment, etc. User equipment can include, but is not limited to, computers, smartphones, personal digital assistants (PDAs), and the aforementioned electronic devices. Network equipment can include, but is not limited to, a single network server, a server group consisting of multiple network servers, or a cloud based on cloud computing consisting of a large number of computers or network servers. Cloud computing is a type of distributed computing, consisting of a super virtual computer composed of a group of loosely coupled computers. This embodiment does not limit this. Steps S1 to S3 are detailed as follows: S1. By deploying a voltage sensor array on the post insulator to obtain stress wave signals, the load mechanical characteristics of the post insulator are analyzed based on the stress wave signals, and a dynamic crack survival model is constructed based on the load mechanical characteristics to predict the existence of cracks. In this embodiment, a post insulator in operation at a 220kV substation is used as the research object. A voltage sensor array is uniformly arranged axially on the surface of the post insulator to collect stress wave signals in real time during operation and under external excitation. The sensors are reliably in contact with the surface of the post insulator through conductive fasteners and are calibrated to ensure consistent signal response sensitivity. This achieves full-process online monitoring from stress wave signal acquisition and load-bearing mechanical characteristic analysis to crack dynamic prediction, providing a highly timely and accurate technical means for the health assessment of post insulators.

[0020] S2. For post insulators predicted to have cracks, determine the optimal phase space reconstruction parameters based on stress wave signals and draw a recursion diagram. Determine multi-scale damage characteristics based on the changing trend of the recursion diagram and quantify the degree of crack damage based on the multi-scale damage characteristics. In this embodiment, a certain type of post insulator in operation is used as the research object. High-sensitivity stress wave sensors deployed at key locations on the surface and inside the insulator are used to acquire the stress wave signals required for crack damage monitoring. The sampling frequency of the sensor is set to 1MHz to ensure that the high-frequency dynamic characteristics during crack initiation and propagation can be captured. The acquired raw stress wave signals are bandpass filtered and DC drift removed by a pre-signal conditioning module to suppress power frequency and low-frequency environmental noise interference, while retaining the high-frequency components generated by stress release at the crack tip.

[0021] For the pre-processed stress wave signal, the optimal delay time parameter is first determined using the average mutual information method, and the optimal embedding dimension is determined using the pseudo-neighbor method to obtain the best phase space reconstruction effect. Using the optimal phase space reconstruction parameters, the single-channel time series is mapped to a 3D phase space trajectory, and a recursive matrix is ​​constructed. During the construction of the recursive matrix, the threshold is adaptively determined based on the quantiles of the Euclidean distance distribution between points in the phase space, ensuring that the recursion rate is within a reasonable range to guarantee the sensitivity of the recursive structure to changes in system dynamics.

[0022] A recursive graph is generated based on the recursive matrix, and the structure of the recursive graph within a sliding window at different time points is dynamically monitored. With crack initiation, propagation, and interactions, significant changes occur in the recursive graph, such as the distribution of diagonal length, dense recursive blocks, and isolated point distribution. Multi-scale damage characteristics, including recursion rate, entropy, and average diagonal length, are calculated using recursive quantitative analysis at multiple time scales, such as the original scale, 2x, and 4x coarse-grained scales.

[0023] To quantify the degree of crack damage, multi-scale damage features are weighted and fused to obtain the crack damage degree. The feature weights are determined by fitting historical calibration data to ensure the best correlation between the crack damage degree and the actual measured crack length or damage level. During monitoring, the continuous-time evolution curve of the crack damage degree can reflect the dynamic changes in the initiation, stable propagation, and rapid failure stages of cracks in the post insulator in real time.

[0024] S3. Construct a damage history prediction model based on the degree of crack damage to calculate the damage evolution factor, predict the crack damage evolution trend of the post insulator with cracks, and set health threshold and failure threshold for dual-layer early warning.

[0025] In this embodiment, for the post insulator whose crack damage degree has been quantified through recursive graph analysis, the process proceeds to step S3 to predict and warn of crack development trends. First, the first stress wave signals acquired by the piezoelectric sensor array deployed during the insulator's operating cycle are organized according to monitoring points and time sequence to form a first signal library. Simultaneously, the stress wave signals of post insulators at the same time point, the same monitoring point, and confirmed to have cracks are organized into a third signal library for comparative analysis.

[0026] Subsequently, stress variation curves were plotted for the first and third signal libraries, and the first and third stress values ​​were extracted. Based on the crack damage degree, the first stress value, and the third stress value, a damage history prediction model was constructed. The damage evolution factor was calculated using the formula for different time series. The damage evolution factor comprehensively reflects the amplitude and velocity of stress wave propagation characteristics in the crack region, and is used to characterize the dynamic process of crack initiation, stable propagation, and accelerated propagation stages.

[0027] After obtaining the damage evolution factor, it is compared with the health threshold and failure threshold calculated in advance using historical samples to implement a two-layer early warning strategy: when the damage evolution factor is greater than the health threshold, the insulator is determined to be in a healthy state and can operate normally; when the damage evolution factor is less than or equal to the health threshold but greater than or equal to the failure threshold, the insulator is determined to be in a sub-healthy state, requiring increased monitoring frequency and close tracking of its damage evolution; when the damage evolution factor is less than the failure threshold, the insulator is determined to have reached a failure state, requiring immediate power outage for repair or replacement. For example, in the actual operation of a substation, the damage evolution factor of a post insulator was initially 0.65, higher than the health threshold of 0.6, and was determined to be in a healthy state; after two months, it dropped to 0.58, between 0.6 and 0.4, triggering a level one early warning, and the maintenance personnel changed the monitoring cycle from once a month to once a week; after another month, the factor dropped to 0.35, lower than the failure threshold of 0.4, triggering a level two early warning and recommending immediate replacement. Subsequent disassembly and inspection revealed that the insulator crack was over 20 mm long, with significant stress concentration at the crack tip, verifying the effectiveness of the prediction model and the dual-layer early warning mechanism.

[0028] Preferably, based on the deployment of a voltage sensor array on the post insulator, specifically, the first stress wave signal is acquired by uniformly arranging piezoelectric sensors from top to bottom on the post insulator as monitoring points to construct a first signal library, the first stress wave signals of all post insulators at the same monitoring point at the same time are statistically analyzed to construct a second signal library, and the first signal library and the second signal library are preprocessed respectively.

[0029] This embodiment is based on uniformly distributing a voltage-type sensor array on the surface of the post insulator structure to achieve real-time monitoring and analysis of the internal stress wave characteristics of the post insulator. Specifically, several piezoelectric sensors are uniformly installed along the length of each post insulator from top to bottom, serving as multi-point monitoring and acquisition units. Through these sensors, the first stress wave signal at each monitoring point is collected in real time, forming a first signal library for a single insulator. Furthermore, to improve the robustness and accuracy of the monitoring data, the first stress wave signals collected from all monitored post insulators at the same time point and corresponding to the same monitoring point location are statistically summarized to construct a second signal library. The second signal library contains stress wave characteristic information of multiple post insulators under the same monitoring conditions, facilitating lateral comparative analysis and anomaly detection.

[0030] In the data processing stage, the first and second signal libraries are preprocessed respectively. Preprocessing includes noise filtering, signal normalization, time synchronization correction, and feature extraction to eliminate the influence of environmental interference and equipment errors on the signals, thereby improving the accuracy and reliability of subsequent feature analysis and crack prediction models.

[0031] Specifically, the mechanical characteristics of the post insulator under load are analyzed based on stress wave signals. These characteristics include average stress intensity and local stress concentration intensity. The average stress intensity includes a first average stress intensity and a second average stress intensity, and the local stress concentration intensity includes a first local stress concentration intensity and a second local stress concentration intensity. The specific calculation process is as follows: The first stress variation curve is plotted using the first signal library, and the second stress variation curve is plotted using the second signal library. The stress value is then extracted based on the stress variation curve. The first average stress intensity is calculated based on the second signal library of the post insulator at the same time. The formula for calculating the first average stress intensity is as follows: In the formula: The first average stress intensity at time t, Let be the number of piezoelectric sensor monitoring points on the i-th post insulator. This refers to the number of post insulators. Let t be the stress value at the j-th piezoelectric sensor monitoring point on the i-th post insulator.

[0032] The first average stress intensity is used to calculate the overall average intensity of the stress values ​​at all monitored post insulators and all piezoelectric sensor monitoring points at time t. By summing and averaging the stress values ​​at all monitoring points of all posts, the "global" average stress intensity of the entire monitoring system at that time is obtained, reflecting the overall load level.

[0033] The second average stress intensity is calculated based on the first signal library of the post insulator. The formula for calculating the second average stress intensity is as follows: In the formula: Let be the second average stress intensity of the i-th post insulator at time t; The second average stress intensity reflects the overall load condition of a single post insulator, which helps to compare the load differences between different posts.

[0034] The first local stress concentration intensity of each support insulator is calculated based on the stress value and the first average stress intensity. The calculation formula is as follows: In the formula: Let be the first local stress concentration strength of the i-th post insulator. For the i-th post insulator and The maximum ratio; The first local stress concentration intensity is used to calculate the first local stress concentration intensity of the i-th post insulator at time t. Specifically, it is the maximum value among the ratios of the stress values ​​at all sensor monitoring points within the post to the global average stress intensity. It reflects the degree of peak concentration of stress at a certain point in the post relative to the overall average stress. The larger the value, the more obvious the local stress concentration, which may be a risk area for potential damage or cracking.

[0035] The second local stress concentration intensity of each support insulator is calculated based on the stress value and the second average stress intensity. The calculation formula is as follows: In the formula: This is the second local stress concentration intensity. This refers to the maximum ratio of the stress value to the corresponding second average stress intensity among all post insulators and all their corresponding piezoelectric sensor monitoring points.

[0036] The second local stress concentration strength calculation is the maximum value of the ratio of stress value to the corresponding single-post second average stress strength among all post insulators and their monitoring points. This reflects the most significant local stress concentration point. At this time, the ratio is relative to the average stress level of each post, focusing more on the local concentration within a single post. This helps to find the location of extreme local stress concentration among the posts and provides a basis for accurately locating crack risks.

[0037] Preferably, a dynamic crack survival model is constructed based on the mechanical characteristics of the load to predict the existence of cracks. The specific process includes: A stress wave propagation model for post insulators was established, and the dynamic stress concentration factor at the crack tip was obtained based on finite element simulation. ; The sensor response matrix is ​​pre-calculated based on the weighting function under several reference crack lengths and several far-field stress conditions. , The first crack tip stress is determined based on the reference crack length. With sensor response matrix Establish weight function mapping as simulation samples In the formula: Let be the sensor hotspot vector at time t. The stress at the first crack tip at time t; The second crack tip stress is obtained by correcting the output of the weight function mapping model using linear regression. The corrected model formula is as follows: In the formula: The stress at the second crack tip at time t is... The first crack tip stress at time t is the weighting function output. These are the feature weight coefficients. This represents the k-th mechanical characteristic under load at time t. For data correction items; It should be noted that, Let t be the k-th load mechanical characteristic at time t, including the first average stress intensity, the second average stress intensity, the first local stress concentration intensity, and the second local stress concentration intensity.

[0038] A dynamic crack survival model is constructed based on the dynamic stress concentration factor at the crack tip and the stress at the second crack tip. The formula for the dynamic crack survival model is as follows: In the formula: For crack dynamic survival factors, The dynamic stress concentration factor at the crack tip at time t-1. The stress at the second crack tip at time t-1, To prevent adjustment factors with a denominator of zero, This refers to the number of data collection sessions. It should be noted that the dynamic stress concentration factor at the crack tip and the stress at the second crack tip used in the construction of the crack dynamic survival model have been dimensionless.

[0039] Extract the crack dynamic survival factor and compare it with a preset threshold. If the crack dynamic growth factor is greater than the preset threshold, the prediction output indicates that a crack exists; if the crack dynamic growth factor is less than the preset threshold, the prediction output indicates that no crack exists.

[0040] In this embodiment, considering the structural characteristics of the post insulator, a stress wave propagation model was established using the finite element method, focusing on simulating the stress field distribution in the crack tip region. Through simulation, the dynamic stress concentration factor at the crack tip was obtained. This factor reflects the stress amplification effect at the crack tip as time and load change, providing key mechanical parameters for crack evolution analysis.

[0041] Subsequently, under different reference crack lengths and various far-field stress conditions, the response matrix of the sensor was calculated in advance through finite element simulation. These pre-calculated response data provided rich training samples for the establishment of the weight function mapping model, enabling accurate mapping from hot spot vectors collected by real-time sensors to crack tip stress.

[0042] To further improve the accuracy of crack tip stress estimation, a linear regression method is used to correct the first crack tip stress output by the weighted function model. The corrected model combines various load-bearing mechanical characteristics, such as the first and second crack tip stresses. Next, based on the dimensionless crack tip dynamic stress concentration factor and the second crack tip stress, a crack dynamic survival model is constructed, and the crack dynamic survival factor is calculated. This factor quantifies the dynamic evolution trend and survival status of the crack by accumulating the ratio of crack tip stress changes over time, providing a quantitative basis for crack risk assessment.

[0043] Finally, the calculated crack dynamic survival factor is compared with a preset threshold. If the factor value exceeds the threshold, it is predicted that there is a risk of cracking in the post insulator, indicating that corresponding maintenance or monitoring measures need to be taken. If the factor value is lower than the threshold, it is determined that there is no obvious risk of cracking, thereby realizing dynamic monitoring and early warning of the health status of the post insulator.

[0044] Preferably, for post insulators predicted to have cracks, the optimal phase space reconstruction parameters are determined based on the first stress wave signal, and a recursion diagram is plotted. Multi-scale damage characteristics are determined based on the changing trend of the recursion diagram, and the degree of crack damage is quantified based on these multi-scale damage characteristics. The specific process includes: Based on the recursion graph, the recursion rate, entropy, and average diagonal length are calculated as multi-scale damage features. A crack damage severity assessment model is then established based on these multi-scale damage features to quantify the crack damage characteristics. The formula for the crack damage severity assessment model is as follows: In the formula: Let t represent the degree of crack damage at time t. The number of multi-scale damage features, Let d be the weight of the d-th feature. For the d-th preprocessed and normalized multiscale damage feature in The value at any given moment.

[0045] In this embodiment, for a post insulator predicted to have cracks, the optimal phase space reconstruction parameters are first determined based on the acquired first stress wave signal. The phase space reconstruction method maps the time-series signal to a multi-dimensional space, revealing the signal's inherent dynamic characteristics. Subsequently, a recursion graph is drawn using the reconstruction result. This graph reflects the dynamic changes of the system by showing the repeated visits of the system state in the phase space.

[0046] Based on the recursion graph, indices such as recursion rate, entropy, and average diagonal length are calculated. These indices, as multi-scale damage characteristics, can characterize the dynamic anomalies caused by cracks at different levels and scales. The recursion rate measures the repeatability of the system state, entropy reflects the complexity of the system state, and the average diagonal length reveals the duration and stability of the system state. This damage assessment model can quantitatively reflect the degree of damage caused by cracks at different time points, providing a scientific basis and data support for health monitoring and maintenance decisions of post insulators.

[0047] Preferably, a damage history prediction model is constructed based on the degree of crack damage to calculate the damage evolution factor and predict the crack damage evolution trend of the post insulator with cracks. The specific process includes: Obtain the first signal library of post insulators with cracks, and classify the first stress wave signals of post insulators with cracks at the same monitoring point and monitored at the same time point into the third signal library; A first stress variation curve is plotted using a first signal library, and a first stress value is extracted based on the first stress variation curve; a third stress variation curve is plotted using a third signal library, and a third stress value is extracted based on the third stress variation curve. A damage history prediction model is established based on the crack damage degree, the first signal library, and the third signal library to calculate the damage evolution factor and predict the crack damage evolution trend of the post insulator with cracks. The calculation formula of the damage history prediction model is as follows: In the formula: As a damage evolution factor, Let be the first stress value at the (j+1)th piezoelectric sensor monitoring point on the i-th post insulator at time t. Let be the first stress value at the j-th piezoelectric sensor monitoring point on the i-th post insulator at time t. The maximum first stress value in the first stress variation curve. This represents the minimum first stress value in the first stress variation curve. Let be the third stress value at the j-th piezoelectric sensor monitoring point on the i-th post insulator at time t+1. Let be the third stress value at the j-th piezoelectric sensor monitoring point on the i-th post insulator at time t. This represents the maximum third stress value in the third stress variation curve. This represents the minimum third stress value in the third stress variation curve.

[0048] For cracked post insulators, firstly, data from the first signal library is acquired, and then the first stress wave signals corresponding to all cracked post insulators at the same monitoring point and time point are aggregated into a third signal library. Subsequently, corresponding stress variation curves are plotted based on the first and third signal libraries, and the first and third stress values ​​are extracted. These stress values ​​reflect the load variations caused by the cracks at different posts and monitoring points.

[0049] Based on the extracted crack damage degree and stress characteristics from two signal libraries, a damage history prediction model is established to calculate the damage evolution factor, thereby quantifying the evolution trend of crack damage. The model comprehensively reflects the damage evolution process caused by stress changes by combining the crack damage degree with the normalized stress difference. Specifically, the damage evolution factor is obtained by weighted summing the differences between the normalized first stress change amplitude between adjacent sensor monitoring points in the first signal library and the normalized third stress change amplitude at the same monitoring point before and after a certain time in the third signal library within the cumulative time series. Normalization uses the maximum and minimum values ​​of each stress change curve to avoid dimensional influence, and a small adjustment factor is introduced to prevent division by zero errors. The damage history prediction model can effectively capture the dynamic evolution characteristics of crack damage and predict crack propagation trends, thus providing a scientific basis and decision support for the health monitoring and maintenance of post insulators.

[0050] Preferably, a dual-layer early warning system is implemented by setting a health threshold and a failure threshold. The specific process includes: obtaining historical damage evolution factors of the post insulator cracks as threshold classification samples; determining the health threshold and failure threshold based on the threshold classification samples; and comparing the damage evolution factors with the health threshold and failure threshold respectively to determine the damage state of the post insulator. Specifically: If the damage evolution factor is greater than the health threshold, the post insulator can still be used normally; If the damage evolution factor is less than or equal to the health threshold and greater than or equal to the failure threshold, the post insulator can still be used, but the detection frequency needs to be increased. At this time, a level one warning is triggered. If the damage evolution factor is less than the failure threshold, the post insulator cannot continue to be used, and a level two warning is triggered.

[0051] It should be noted that the health threshold is determined by taking the mean value of the historical damage evolution factors of post insulators that have no abnormalities during use but have cracks. The failure threshold is determined by taking the mean value of the historical damage evolution factors of post insulators that have abnormalities and cracks during use.

[0052] Among the criteria for determining whether there are abnormalities in the use of post insulators with cracks, the following are included: sudden changes in piezoelectric sensor signals that continuously exceed limits; significant abnormalities in recursion graph entropy and recursion rate; continuous increase in the dynamic stress concentration factor at the crack tip; and exceeding the standard for crack length detected on-site.

[0053] In this embodiment of the invention, a dual-layer early warning mechanism is employed to dynamically assess the damage status of cracked post insulators. Specifically, historical damage evolution factor data of post insulator cracks is first collected, and threshold samples are then defined based on this data. The health threshold is determined by the average historical damage evolution factor of post insulators with cracks but no abnormal behavior during use; the failure threshold is derived from the average historical damage evolution factor of post insulators with abnormal behavior and cracks during use. The criteria for determining these abnormalities include abrupt and continuous exceedance of piezoelectric sensor signals, significant anomalies in the entropy and recursion rate of the recursion graph, a continuous increase in the dynamic stress concentration factor at the crack tip, and exceeding the on-site crack length detection limit, among other indicators.

[0054] During actual monitoring, the real-time calculated damage evolution factor is compared with preset health and failure thresholds to determine the damage status of the post insulator. Specifically, three scenarios are defined: when the damage evolution factor is greater than the health threshold, the post insulator is considered to be in good condition and can continue to be used normally; when the damage evolution factor is between the health and failure thresholds, the post insulator is still usable, but it is recommended to increase the detection frequency, at which point the system triggers a level one warning; and when the damage evolution factor is lower than the failure threshold, it indicates that the post insulator is in a severely damaged state and cannot continue to be used, triggering a level two warning to remind timely repair or replacement measures. This dual-layer warning mechanism enables early detection and graded management of crack damage in post insulators, effectively ensuring the safe operation of equipment.

[0055] A specific example of a sudden change and continuous exceedance of the piezoelectric sensor signal is as follows: The stress wave signal amplitude of the No. 5 voltage sensor on a certain support insulator jumped from an average of 100 units to 300 units in a short period of time, and the high amplitude was maintained in multiple consecutive samples, exceeding the historical normal fluctuation range. This indicates that a local crack is rapidly expanding or a crack is generated at this time, resulting in a sharp increase in local stress.

[0056] A specific example of significant anomalies in recursion graph entropy and recursion rate is as follows: Analysis of the recursion graph constructed based on stress wave signals shows that the entropy value of a certain support insulator has recently jumped from an average of 0.5 to over 0.9, while the recursion rate has decreased by more than 20%. This indicates that the dynamic characteristics of the stress wave have become more complex and chaotic, representing an aggravation of crack damage.

[0057] A specific example of the continuous increase in the dynamic stress concentration factor at the crack tip is as follows: the stress concentration factor at the crack tip, estimated by finite element and sensor data, increased from 0.8 to 1.5 and continued to fluctuate at a high level in subsequent sampling; the increase in crack tip stress indicates that the crack tip is subjected to a greater load, at which point the risk of crack propagation is obvious.

[0058] A specific example of a crack length exceeding the standard during on-site detection is as follows: During regular manual inspections, it was found that the crack length of a certain post insulator exceeded the preset safety limit, for example, more than 20mm, and the crack depth also increased significantly. Even if the sensor signal did not show extreme changes, the visible crack expansion indicated that the structure was already in a dangerous state.

[0059] A device for detecting and warning of crack damage in a post insulator, applied to the aforementioned method for detecting and warning of crack damage in a post insulator, the device comprising: a voltage sensor, a crack prediction unit, a crack damage quantification unit, and a damage evolution prediction unit; Voltage sensors are used to acquire stress wave signals; The crack prediction unit is used to analyze the mechanical characteristics of the post insulator under load based on stress wave signals, and to construct a dynamic crack survival model based on the mechanical characteristics under load to predict the existence of cracks. The crack damage quantification unit is used to determine the optimal phase space reconstruction parameters based on the first stress wave signal for the post insulator predicted to have cracks, and to draw a recursion graph. Based on the changing trend of the recursion graph, the multi-scale damage characteristics are determined, and the degree of crack damage is quantified based on the multi-scale damage characteristics. The damage evolution prediction unit is used to construct a damage history prediction model based on the degree of crack damage, calculate the damage evolution factor, predict the damage evolution trend of cracked post insulators, and set health threshold and failure threshold for dual-layer early warning.

[0060] like Figure 2 The diagram shown is a structural schematic of a post insulator crack damage detection and early warning device according to an embodiment of the present invention, which can be used to perform... Figure 1 The steps in the method embodiments shown are implemented in a similar manner and have similar technical effects, and will not be repeated here.

[0061] A crack damage detection and early warning system for post insulators includes: Memory, used to store programs; A processor is used to execute the program stored in the memory. When the program is executed, the processor is used to execute the above-described method for detecting and warning of crack damage in post insulators.

[0062] The term "processor" can also refer to other general-purpose processors, digital signal processors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor serves as the control center of a post insulator crack damage detection and early warning system, connecting various parts of the entire overhead line fault traveling wave diagnostic system of the distribution network through various interfaces and lines.

[0064] The memory can be used to store computer-readable instructions. The processor implements various functions of a post insulator crack damage detection and early warning system by running or executing the computer-readable instructions or modules stored in the memory, and by calling data stored in the memory. The memory can mainly include a program storage area and a data storage area. The program storage area can store the operating system, at least one application program required for a given function (such as sound playback, image playback, etc.); the data storage area can store data created for use by the post insulator crack damage detection and early warning system. Furthermore, the memory can include a hard disk, RAM, plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, at least one disk storage device, flash memory device, Read-Only Memory (ROM), Random Access Memory (RAM), or other non-volatile / volatile storage devices.

[0065] If a module integrating a post insulator crack damage detection and early warning system is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by instructing related hardware through computer-readable instructions. These computer-readable instructions can be stored in a computer-readable storage medium, and when executed by a processor, they can implement the steps of the various method embodiments described above.

[0066] Through the above embodiments, this invention analyzes the load-bearing mechanical characteristics of post insulators based on stress wave signals, constructs a dynamic crack survival model based on these characteristics to predict the existence of cracks, determines the optimal phase space reconstruction parameters and draws a recursion graph for post insulators predicted to have cracks, determines multi-scale damage characteristics based on the changing trends of the recursion graph, and quantifies the degree of crack damage based on these multi-scale damage characteristics, constructs a damage history prediction model based on the degree of crack damage to calculate damage evolution factors, predicts the crack damage evolution trend of post insulators with cracks, and sets health thresholds and failure thresholds for dual-layer early warning. This enables early and accurate diagnosis and trend prediction of crack damage in post insulators, improving monitoring efficiency and the reliability of early warning.

[0067] This invention utilizes a voltage-type sensor array to acquire stress wave signals in real time, enabling precise perception of the load state of post insulators and improving the real-time and comprehensiveness of monitoring. Based on the load-bearing mechanical characteristics extracted from the stress wave signals, a dynamic crack survival model is constructed, which can scientifically and accurately predict the existence of cracks, avoiding false alarms and missed detections caused by relying solely on surface detection. For post insulators predicted to have cracks, phase space reconstruction and recursive graph analysis methods are used to mine the multi-scale dynamic characteristics of the stress wave signals, achieving a quantitative assessment of the crack damage degree and enhancing the ability to understand the internal evolution process of cracks, effectively improving the accuracy and stability of crack detection. A damage history prediction model based on the crack damage degree dynamically reflects the crack evolution trend by calculating damage evolution factors. Combined with set health and failure thresholds, a dual-layer early warning mechanism is implemented, ensuring the safe operation of equipment and optimizing maintenance resource allocation, avoiding blind replacement or delayed repairs.

[0068] The above is only a specific implementation method of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed in this application, which should be covered by the scope of protection of the present application.

[0069] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for detecting and warning of crack damage in post insulators, characterized in that, Includes the following steps: By deploying a voltage sensor array on the post insulator to acquire stress wave signals, analyzing the load-bearing mechanical characteristics of the post insulator based on the stress wave signals, and constructing a dynamic crack survival model based on the load-bearing mechanical characteristics to predict the existence of cracks; For post insulators predicted to have cracks, the optimal phase space reconstruction parameters are determined based on stress wave signals, and a recursion diagram is drawn. Multi-scale damage characteristics are determined based on the changing trend of the recursion diagram, and the degree of crack damage is quantified based on the multi-scale damage characteristics. A damage history prediction model is constructed by constructing a damage history prediction model based on the degree of crack damage to calculate the damage evolution factor, predict the crack damage evolution trend of the post insulator with cracks, and set a health threshold and a failure threshold for dual-layer early warning. The process of predicting the crack damage evolution trend of cracked post insulators includes: The first signal library of post insulators with cracks is obtained, and the first stress wave signal of post insulators with cracks at the same monitoring point and at the same time point is classified into the third signal library. A first stress variation curve is plotted using a first signal library, and a first stress value is extracted based on the first stress variation curve; a third stress variation curve is plotted using a third signal library, and a third stress value is extracted based on the third stress variation curve. A damage history prediction model is established based on the degree of crack damage, the first signal library, and the third signal library to calculate the damage evolution factor and predict the crack damage evolution trend of the post insulator with cracks.

2. The method for detecting and warning of crack damage in post insulators according to claim 1, characterized in that, Based on the deployment of voltage sensor arrays on post insulators, specifically by uniformly arranging piezoelectric sensors from top to bottom on the post insulators as monitoring points to acquire first stress wave signals and construct a first signal library, and statistically analyzing the first stress wave signals of all post insulators at the same monitoring point at the same time to construct a second signal library, and preprocessing the first signal library and the second signal library respectively.

3. The method for detecting and warning of crack damage in post insulators according to claim 2, characterized in that, The mechanical characteristics of the post insulator under load are analyzed based on stress wave signals. These characteristics include average stress intensity and local stress concentration intensity. The average stress intensity includes a first average stress intensity and a second average stress intensity, while the local stress concentration intensity includes a first local stress concentration intensity and a second local stress concentration intensity. The specific process is as follows: The first stress variation curve is plotted using the first signal library, and the second stress variation curve is plotted using the second signal library. The stress value is then extracted based on the stress variation curve. The first average stress intensity is obtained based on the second signal library of the post insulator at the same time. The second average stress intensity is determined based on the first signal library of the post insulator; The first local stress concentration strength of each post insulator is calculated based on the stress value and the first average stress intensity. The second local stress concentration strength of each post insulator is calculated based on the stress value and the second average stress intensity.

4. The method for detecting and warning of crack damage in post insulators according to claim 3, characterized in that, A dynamic crack survival model is constructed based on the mechanical characteristics of the load to predict the existence of cracks. The specific process includes: Based on the pre-calculation of the weight function at several reference crack lengths Sensor response matrix under several far-field stress conditions , The first crack tip stress is determined based on the reference crack length. With sensor response matrix Establish weight function mapping as simulation samples In the formula: Let be the sensor hotspot vector at time t. The stress at the first crack tip at time t; The second crack tip stress is obtained by correcting the output of the weight function mapping model through linear regression. A dynamic crack survival model is built based on the dynamic stress concentration factor at the crack tip and the stress at the second crack tip, and the dynamic crack survival factor is calculated. Extract the crack dynamic survival factor and compare it with a preset threshold. If the crack dynamic growth factor is greater than the preset threshold, the prediction output indicates that a crack exists; if the crack dynamic growth factor is less than the preset threshold, the prediction output indicates that no crack exists.

5. The method for detecting and warning of crack damage in post insulators according to claim 1, characterized in that, The multi-scale damage characteristics are determined based on the changing trend of the recursion graph, and the degree of crack damage is quantified based on the multi-scale damage characteristics. The specific process includes: calculating the recursion rate, entropy and average diagonal length based on the recursion graph as multi-scale damage characteristics, and establishing a crack damage degree assessment model based on the multi-scale damage characteristics to calculate the degree of crack damage.

6. The method for detecting and warning of crack damage in post insulators according to claim 1, characterized in that, A two-tiered early warning system is implemented by setting health and failure thresholds. The specific process includes: obtaining historical damage evolution factors of post insulator cracks as threshold classification samples; determining health and failure thresholds based on these samples; and comparing the damage evolution factors with both the health and failure thresholds to determine the damage state of the post insulator. If the damage evolution factor is greater than the health threshold, the post insulator can still be used normally; If the damage evolution factor is less than or equal to the health threshold and greater than or equal to the failure threshold, the post insulator can still be used, but the detection frequency needs to be increased. At this time, a level one warning is triggered. If the damage evolution factor is less than the failure threshold, the post insulator cannot continue to be used, and a level two warning is triggered.

7. The method for detecting and warning of crack damage in post insulators according to claim 1, characterized in that, The health threshold is determined by taking the mean value of the historical damage evolution factors of post insulators that have no abnormalities during use but have cracks. The failure threshold is determined by taking the mean value of the historical damage evolution factors of post insulators that have abnormalities and cracks during use.

8. A device for detecting and warning of crack damage in post insulators, applied to a method for detecting and warning of crack damage in post insulators as described in any one of claims 1-7, characterized in that, The device includes: a voltage sensor, a crack prediction unit, a crack damage quantification unit, and a damage evolution prediction unit. Voltage sensors are used to acquire stress wave signals; The crack prediction unit is used to analyze the mechanical characteristics of the post insulator under load based on stress wave signals, and to construct a dynamic crack survival model based on the mechanical characteristics under load to predict the existence of cracks. The crack damage quantification unit is used to determine the optimal phase space reconstruction parameters based on the first stress wave signal for the post insulator predicted to have cracks, and to draw a recursion graph. Based on the changing trend of the recursion graph, the multi-scale damage characteristics are determined, and the degree of crack damage is quantified based on the multi-scale damage characteristics. The damage evolution prediction unit is used to construct a damage history prediction model based on the degree of crack damage, calculate the damage evolution factor, predict the damage evolution trend of cracked post insulators, and set health threshold and failure threshold for dual-layer early warning.

9. A system for detecting and warning of crack damage in post insulators, characterized in that, include: Memory, used to store programs; A processor for executing the program stored in the memory, wherein when the program is executed, the processor is configured to perform the method as described in any one of claims 1-7.

Images