Power transmission tower attitude online monitoring system based on big data analysis

Abstract

The invention relates to the field of power transmission tower monitoring, in particular to a power transmission tower attitude on-line monitoring system based on big data analysis, which comprehensively constructs quantitative evaluation indexes in a power transmission tower monitoring process through fusion of multi-source heterogeneous data, and then performs real-time early warning according to multi-dimensional static real-time early warning, dynamic trend early warning and early development early warning. According to the method, the risk classification and classification management can be realized, slow development and potential trend faults can be captured, the risk of missing report and lagging alarm can be reduced, a specific early warning coping decision based on historical verification can be automatically provided according to a knowledge graph by constructing a historical database and the knowledge graph, high-efficiency decision support can be provided for operation and maintenance personnel, and the operation and maintenance efficiency can be improved. And the autonomous learning module can collect and feed back the final implementation effect of the decision, so that the knowledge graph is optimized.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of power transmission tower monitoring, in particular to a power transmission tower posture online monitoring system based on big data analysis. BACKGROUND

[0002] As the core supporting structure of power transmission line, the structural health and posture stability of power transmission tower are directly related to the safe and reliable operation of the entire power system. The power transmission tower is usually distributed in complex geographical environments such as open country, mountainous area, river, etc. and is subjected to natural external forces such as wind load, icing, earthquake, etc. for a long time, as well as the influence of material aging, foundation settlement, etc. which may cause the tower to tilt, twist, settle and other posture changes, and further cause structural damage or even collapse accidents, resulting in huge economic losses and social impact. Therefore, it is of great significance to monitor the posture of the power transmission tower in real time and online, and to realize early warning of faults to ensure the safety of the power grid.

[0003] At present, the power transmission tower monitoring refers to a system for monitoring the structural state of the power transmission tower in real time or periodically by installing various sensors, data acquisition and communication equipment on the tower body, combined with a back-end data processing platform. The traditional power transmission tower posture monitoring system is basically a static alarm mode based on fixed threshold, which is not sensitive to slow developing hidden dangers and often alarms only when the hidden danger develops into a significant fault, missing the best opportunity for early warning. At the same time, due to the lack of in-depth analysis of data time sequence characteristics and cross-validation of multi-source information, the system is easily affected by instantaneous interference and generates false alarms, reducing the credibility of the monitoring results and causing waste of operation and maintenance resources. In addition, the system also lacks the ability to predict the trend of posture changes and cannot provide precise early warning and response decision suggestions based on historical experience and multi-dimensional data analysis for operation and maintenance personnel.

[0004] In view of the above technical problems, the present application provides a solution. SUMMARY

[0005] The present application is to construct quantitative evaluation indexes in the process of power transmission tower monitoring by fusing multi-source heterogeneous data, and to realize the grading and classification management of risks according to multi-dimensional static immediate early warning, dynamic trend early warning and early development early warning, so as to capture slow developing and potential trend faults, reduce the risk of missed and delayed alarms, and provide specific early warning response decisions based on historical verification by constructing a historical database and a knowledge graph, and provide efficient decision support for operation and maintenance personnel. The autonomous learning module can collect and feedback the implementation effect of the final decision, optimize the knowledge graph, so as to solve the technical defects proposed in the background art.

[0006] The purpose of the application can be realized by the technical scheme: a power transmission tower posture online monitoring system based on big data analysis, comprising a multi-source information acquisition module, a posture detection module, a dynamic evaluation module, a historical data comparison module and an autonomous learning judgment module:

[0007] The multi-source information acquisition module collects the posture information of the power transmission tower through multiple sensors deployed on the power transmission tower, and obtains a multi-source heterogeneous data set reflecting the posture of the power transmission tower;

[0008] The posture monitoring module obtains the multi-source heterogeneous data set through the multi-element information acquisition module, and performs fusion calculation through the multi-element heterogeneous data to obtain the posture quantitative index of the power transmission tower;

[0009] The dynamic evaluation module obtains the posture quantitative index of the power transmission tower through the posture monitoring module, and performs threshold judgment on the posture quantitative index, and makes a static abnormality early warning decision according to the judgment result, and the dynamic evaluation module accumulates and counts the posture quantitative index when no static abnormality early warning is generated, and makes a secondary threshold judgment according to the accumulation statistical result, and makes a dynamic abnormality early warning decision;

[0010] The historical data comparison module obtains the posture quantitative index, the static abnormality early warning and the dynamic abnormality early warning, mobilizes the knowledge graph composed of historical posture data and decision means in the database, imports the posture quantitative index, the static abnormality early warning and the dynamic abnormality early warning into the knowledge graph, and makes early warning response decision through the knowledge graph;

[0011] The autonomous learning judgment module outputs the early warning response decision, obtains processing scheme feedback and processing result feedback, obtains early warning response decision effectiveness evaluation through processing scheme feedback and processing result feedback, and performs incremental update of the knowledge graph through the early warning response decision.

[0012] As a preferred embodiment of the application, the multi-source heterogeneous data set created by the multi-source information acquisition module includes inclination data, vibration data, three-dimensional displacement data and environmental load data;

[0013] Among them, the inclination data is the angle between the power transmission tower and the vertical line, the vibration data is the vibration frequency and amplitude of the power transmission tower body, the three-dimensional displacement data is determined by mixed calculation of sensors and ground reference station, and the environmental load data is the wind load data borne by the power transmission tower, and when the wind load data is counted, the wind load data is filtered by low value.

[0014] As a preferred embodiment of the application, before the posture monitoring module fuses the multi-source heterogeneous data, the multi-source heterogeneous data is time series filtered by Kalman filtering to obtain pure heterogeneous data;

[0015] The inclination data in the pure isomer data, the height information of the sensor monitoring the inclination data, and the three-dimensional displacement data are processed through coordinate transformation and data fitting to generate the shape index including the tower body inclination, the torsion angle, and the settlement amount;

[0016] The posture monitoring module fuses the vibration data and the wind load data to obtain the tower body load index.

[0017] The posture monitoring module records the tower body load index and the shape index as the posture quantization index.

[0018] As a preferred embodiment of the present application, when the dynamic evaluation module makes a static judgment on the posture quantization index, the shape index and the tower body load index are compared with corresponding threshold values respectively, and a decision of normal shape or abnormal shape is made according to the comparison result of the shape index, and a decision of normal load or abnormal load is made according to the tower body load index.

[0019] When the dynamic evaluation module generates an abnormal shape or an abnormal load, a static abnormal early warning decision is output, and when a normal shape and a normal load are generated simultaneously, no static abnormal early warning decision is output.

[0020] As a preferred embodiment of the present application, when the dynamic evaluation module makes accumulation statistics, the posture quantization index obtained initially is recorded, and when the posture quantization index is obtained subsequently, it is compared with the posture quantization index obtained previously to obtain an interval growth ratio.

[0021] The dynamic evaluation module makes an interval threshold judgment on the interval growth ratio, and makes an interval abnormal decision according to the judgment result.

[0022] When the dynamic evaluation module obtains an interval abnormal decision, a dynamic abnormal early warning decision is made.

[0023] As a preferred embodiment of the present application, when the dynamic evaluation module does not make an interval abnormal decision, the posture quantization index is compared with a low-level threshold value, if the posture quantization index does not exceed the low-level threshold value, no reaction is made, and if the posture quantization index exceeds the low-level threshold value, a cumulative record is made once.

[0024] The dynamic evaluation module makes a threshold judgment on the number of continuous cumulative records, and if the number exceeds a set limit, a dynamic abnormal early warning decision is made.

[0025] As a preferred embodiment of the present application, the historical data comparison module imports the posture quantization index into a knowledge graph, takes the posture quantization index as a node in the knowledge graph, takes the static abnormal early warning or the dynamic abnormal early warning as a constraint condition, obtains a coping scheme for the same situation in the knowledge graph to output, and completes the generation of the early warning coping decision.

[0026] As a preferred embodiment of the present application, the specific process of the autonomous learning judgment module for knowledge graph incremental updating is as follows: taking the posture data, early warning response decision, processing scheme feedback and final processing result in each early warning event as a new learning sample, and optimizing the weight and association relationship of nodes and edges in the knowledge graph through association rule mining and pattern recognition algorithm.

[0027] Compared with the prior art, the present application has the following beneficial effects:

[0028] 1. In the present application, through the integration of multi-source information, the organic integration of multi-source heterogeneous data is realized, so that the quantitative evaluation index in the transmission tower monitoring process is comprehensively constructed, and according to the multi-dimensional static real-time early warning, dynamic trend early warning and early development early warning, the classification and management of risks are realized, not only the sudden serious abnormality can be responded, but also the slowly developing and potential trend failure can be acutely captured through the analysis of data accumulation and change trend, so that early warning in the transmission tower monitoring process is realized, and the risk of missed and delayed alarm is greatly reduced.

[0029] 2. In the present application, by constructing a historical database and creating a related knowledge graph, the real-time monitored data and early warning conditions are analyzed by the knowledge graph, so that the specific early warning response decision based on historical verification is automatically provided according to the knowledge graph, efficient decision support is provided for the operation and maintenance personnel, the experience requirement for the operation and maintenance personnel is effectively reduced, the judgment difference between different operation and maintenance personnel is reduced, and the operation and maintenance efficiency and scientificity are improved.

[0030] 3. In the present application, the response decision provided by the knowledge graph, the response decision finally taken by the operation and maintenance personnel and the implementation effect of the decision are collected and fed back by the autonomous learning module, the incremental updating of the knowledge graph is realized, so that the system can continuously self-optimize learning when generating the response decision, can adaptively match according to the characteristics and aging law of different transmission towers, and can provide more intelligent and more targeted monitoring and decision means for different types of transmission towers in different regions. BRIEF DESCRIPTION OF DRAWINGS

[0031] In order to facilitate the understanding of those skilled in the art, the present application will be further described below with reference to the accompanying drawings.

[0032] Figure 1 is a system block diagram of the present application;

[0033] Figure 2 is a system flowchart of the present application. DETAILED DESCRIPTION

[0034] The technical solutions of the present application will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of the present application.

[0035] Embodiment one: please refer to Figure 1 Figure 2 As shown in the figure, a power transmission tower posture online monitoring system based on big data analysis includes a multi-source information acquisition module, a posture detection module, a dynamic evaluation module, a historical data comparison module, and an autonomous learning and judgment module.

[0036] The multi-source information acquisition module collects the posture information of the power transmission tower through various sensors deployed on the power transmission tower, obtains a multi-source heterogeneous data set reflecting the posture of the power transmission tower, and creates a multi-source heterogeneous data set including inclination data, vibration data, three-dimensional displacement data, and environmental load data.

[0037] The inclination data is the angle between the power transmission tower and the vertical line, measured by a high-precision dual-axis inclination sensor, used to determine the angle between the main material of the power transmission tower and the vertical line in two orthogonal directions. The vibration data is the vibration frequency and amplitude of the tower body, captured by a three-axis acceleration sensor, used to analyze the vibration frequency and amplitude of the tower body under external forces such as wind load. The three-dimensional displacement data is determined by a hybrid solution of sensors and ground reference stations, i.e., a high-precision global navigation satellite system receiver is used for monitoring, the carrier phase difference data between it and the ground reference station is solved, and the three-dimensional coordinate change of the tower top with millimeter-level precision is determined. The environmental load data is the wind load data borne by the power transmission tower, collected by an ultrasonic wind speed and direction instrument installed on the tower top. To improve data effectiveness, the system will filter out low-value data before data processing, thereby avoiding the recording of low wind load and thereby interfering with the accuracy of subsequent power transmission tower posture quantitative index generation.

[0038] After data collection, the posture monitoring module performs deep fusion and solution. First, this module uses Kalman filtering algorithm to perform time series filtering on the above multi-source heterogeneous data. This process can effectively suppress sensor noise and transient environmental interference, smooth the data curve, and thus obtain pure heterogeneous data that better reflects the true state of the tower body.

[0039] Subsequently, the posture monitoring module performs fusion and solution on the multi-element heterogeneous data to obtain the posture quantitative index of the power transmission tower

[0040] ​The inclination data in the pure isomer data, the height information of the sensor monitoring the inclination data, and the three-dimensional displacement data are processed through coordinate transformation and data fitting, and then a shape index including the inclination of the tower body, the torsion angle, and the settlement amount is generated through spatial geometric calculation;

[0041] The posture monitoring module fuses the vibration data and the wind load data, analyzes the dynamic response characteristics of the tower body under a specific wind load, and obtains a tower body load index;

[0042] The posture monitoring module records the tower body load index and the shape index as a posture quantization index, and analyzes the dynamic response characteristics of the tower body under a specific wind load;

[0043] The dynamic evaluation module adopts a double-layer early warning mechanism to evaluate the posture quantization index;

[0044] Firstly, static judgment, when the dynamic evaluation module performs static judgment on the posture quantization index, the shape index and the tower body load index are compared with the corresponding threshold values, a decision of normal shape or abnormal shape is made according to the comparison result of the shape index, and a decision of abnormal load or normal load is made according to the tower body load index;

[0045] When the dynamic evaluation module generates an abnormal shape or an abnormal load, a static abnormal early warning decision is output, and when the dynamic evaluation module generates a normal shape and a normal load, no static abnormal early warning decision is output. Static judgment is generally used to deal with sudden and serious structural risks;

[0046] When the static early warning is not triggered, the module will start a more detailed accumulation statistical analysis. When the dynamic evaluation module performs accumulation statistics, the posture quantization index obtained initially is recorded, and when the posture quantization index is obtained subsequently, it is compared with the posture quantization index obtained previously to obtain an interval growth ratio;

[0047] The dynamic evaluation module performs interval threshold judgment on the interval growth ratio. If the ratio exceeds the set interval threshold value, it indicates that the tower body state is accelerating deterioration, even if the absolute value does not exceed the limit, an interval abnormal decision will be triggered and a dynamic abnormal early warning will be generated, and an interval abnormal decision is made according to the judgment result;

[0048] When the dynamic evaluation module obtains the interval abnormal decision, a dynamic abnormal early warning decision is made;

[0049] In addition, when the interval abnormal decision is not made, the dynamic evaluation module also performs static comparison of the posture quantization index with the low-level threshold value. If the posture quantization index does not exceed the low-level threshold value, no reaction is made. If the posture quantization index exceeds the low-level threshold value but does not trigger the static abnormal early warning decision, a cumulative record is made;

[0050] The dynamic evaluation module performs threshold judgment on the continuous cumulative number of records, and if the number exceeds the set limit, it reflects a high-frequency mild abnormal state, and a dynamic abnormal early warning decision is made, thereby effectively realizing early capture of slow accumulation risk and potential trend failure.

[0051] Embodiment two: please refer to Figure 1 - Figure 2 As shown, the historical data comparison module obtains the posture quantization index, static abnormal early warning and dynamic abnormal early warning, and mobilizes the knowledge graph composed of historical posture data and decision means in the database. The knowledge graph associates posture data, early warning records, artificial disposal schemes and disposal effects in history, forming a structured experience database.

[0052] The posture quantization index, static abnormal early warning and dynamic abnormal early warning are imported into the knowledge graph. The posture quantization index is used as a node in the knowledge graph, and the static abnormal early warning or dynamic abnormal early warning is used as a constraint condition. The graph quickly retrieves the successful coping scheme under similar circumstances in history through node matching and path traversal, obtains the coping scheme under the same circumstances in the knowledge graph for output, and completes the generation of early warning and coping decisions.

[0053] Finally, the autonomous learning and judgment module outputs the early warning and coping decision to the operation and maintenance personnel, and collects subsequent processing scheme feedback and processing result feedback, such as the processing scheme adopted and the processing result after processing. Each complete early warning and disposal event will be used as a new learning sample. The autonomous learning and judgment module analyzes these samples using association rule mining and pattern recognition algorithms, such as strengthening effective disposal paths and correcting ineffective association relationships, thereby realizing incremental updating of the knowledge graph, so that the decision-making ability of the system can evolve and become more accurate and intelligent with the increase of running time.

[0054] The threshold or preset value, preset range, etc. is set for result comparison and analysis to determine good and bad. The size of the value is determined based on large model analysis of sample data and artificial experience, and is recorded and stored. It can also be adjusted appropriately according to seasonal or rational influence conditions.

[0055] The weight proportion coefficient and influence factor are set according to the influence of each parameter on the result, and the specific numerical value is allocated to finally reflect the influence of the result. The value is recorded and stored by combining large model analysis of sample data and artificial experience. It can also be adjusted appropriately according to seasonal or rational influence conditions.

[0056] The preferred embodiments of the application disclosed above are only to facilitate the elucidation of the application. The preferred embodiments do not describe all the details of the application and limit the application to the specific embodiments. Obviously, many modifications and variations can be made in light of the teachings above. The description is chosen and described in order to provide the best illustration of the application and its practical application to those skilled in the art and to enable those skilled in the art to best utilize the application. The application is limited only by the claims and their full scope and equivalents.

Claims

1. A transmission tower attitude online monitoring system based on big data analysis, characterized in that, It includes a multi-source information acquisition module, a posture detection module, a dynamic evaluation module, a historical data comparison module, and a self-learning judgment module. The multi-source information acquisition module collects the attitude information of the transmission tower through various sensors deployed on the transmission tower, and obtains a multi-source heterogeneous data set reflecting the attitude of the transmission tower. The attitude monitoring module acquires a multi-source heterogeneous data set through the multi-source information acquisition module, and performs fusion calculation on the multi-source heterogeneous data to obtain the attitude quantification index of the transmission tower. After obtaining the attitude quantification index of the transmission tower through the attitude monitoring module, the dynamic evaluation module performs threshold judgment on the attitude quantification index and makes a static anomaly warning decision based on the judgment result. When no static anomaly warning is generated, the dynamic evaluation module accumulates and statistically analyzes the attitude quantification index and performs a secondary threshold judgment based on the accumulated statistical result to make a dynamic anomaly warning decision. The historical data comparison module acquires attitude quantification indicators, static anomaly warnings, and dynamic anomaly warnings, and mobilizes the knowledge graph composed of historical attitude data and decision-making methods in the database to import attitude quantification indicators, static anomaly warnings, and dynamic anomaly warnings into the knowledge graph for warning response decisions. The autonomous learning judgment module outputs the early warning response decision and obtains the processing plan feedback and processing result feedback. It obtains the effectiveness evaluation of the early warning response decision through the processing plan feedback and processing result feedback, and performs incremental updates to the knowledge graph based on the early warning response decision.

2. The online monitoring system for transmission tower attitude based on big data analysis according to claim 1, characterized in that, The multi-source heterogeneous data set created by the multi-source information acquisition module includes tilt angle data, vibration data, three-dimensional displacement data, and environmental load data; Among them, the tilt angle data is the angle between the transmission tower and the vertical line, the vibration data is the vibration frequency and amplitude of the transmission tower body, the three-dimensional displacement data is determined by the hybrid calculation of the sensor and the ground reference station, and the environmental load data is the wind load data borne by the transmission tower. When statistically analyzing the wind load data, the low value of the wind load data is filtered out.

3. The online monitoring system for transmission tower attitude based on big data analysis according to claim 1, characterized in that, Before fusing multi-source heterogeneous data, the attitude monitoring module performs time-series filtering on the multi-source heterogeneous data using Kalman filtering to obtain clean heterogeneous data. After processing the tilt angle data, the height information of the sensor monitoring the tilt angle data, and the three-dimensional displacement data in the pure heterogeneous data through coordinate transformation and data fitting, a morphological index containing tower tilt angle, torsion angle, and settlement is generated. The attitude monitoring module fuses vibration data and wind load data to obtain tower load indicators; The attitude monitoring module records the tower load index and shape index together as attitude quantification index.

4. The online monitoring system for transmission tower attitude based on big data analysis according to claim 1, characterized in that, When performing static judgment on attitude quantification indicators, the dynamic evaluation module compares the morphological indicators and tower load indicators with the corresponding thresholds. Based on the comparison results of the morphological indicators, it makes a decision on whether the morphology is normal or abnormal, and based on the tower load indicators, it makes a decision on whether the load is abnormal or normal. When generating abnormal shape or load, the dynamic evaluation module outputs a static abnormality warning decision; when generating both normal shape and normal load, it does not output a static abnormality warning decision.

5. The online monitoring system for transmission tower attitude based on big data analysis according to claim 1, characterized in that, When the dynamic evaluation module performs cumulative statistics, it records the attitude quantification index obtained for the first time, and compares it with the attitude quantification index obtained the previous time when it obtains attitude quantification index in subsequent times to obtain the interval growth ratio. The dynamic evaluation module will determine the interval growth rate based on the interval threshold and make an interval anomaly decision based on the determination result. When the dynamic evaluation module obtains an interval anomaly decision, it makes a dynamic anomaly early warning decision.

6. The online monitoring system for transmission tower attitude based on big data analysis according to claim 5, characterized in that, When no abnormal interval decision is made, the dynamic evaluation module statically compares the attitude quantification index with the low-level threshold. If the attitude quantification index does not exceed the low-level threshold, no reaction is made. If the attitude quantification index exceeds the low-level threshold, an cumulative record is made. The dynamic evaluation module will perform threshold judgment on the number of consecutive cumulative records. If the number exceeds the set limit, a dynamic anomaly warning decision will be made.

7. The online monitoring system for transmission tower attitude based on big data analysis according to claim 1, characterized in that, After importing the attitude quantification index into the knowledge graph, the historical data comparison module uses the attitude quantification index as a node in the knowledge graph and static or dynamic anomaly warnings as constraints to obtain and output corresponding solutions for the same situation in the knowledge graph, thus generating the warning response decision.

8. The online monitoring system for transmission tower attitude based on big data analysis according to claim 1, characterized in that, The specific process of the autonomous learning judgment module for incremental updating of the knowledge graph is as follows: the attitude data, early warning response decision, processing scheme feedback and final processing result in each early warning event are taken as a new learning sample, and the weights and association relationships of nodes and edges in the knowledge graph are optimized through association rule mining and pattern recognition algorithms.

Images