Safety monitoring and early warning method and platform for electric tower base

Abstract

The invention discloses a safety monitoring and early warning method and platform for an electric tower base, and relates to the technical field of safety monitoring, and the method comprises the steps: carrying out the monitoring feature mining according to multivariate safety risk factors, and building a base monitoring feature distribution network; the base is monitored in real time through the Internet of Things sensing array, and a monitoring data stream is obtained; performing error compensation by using the sensing correction multi-level model, and extracting base structure, environment and external force feature information; activating a security risk detection channel, and detecting a risk coefficient in combination with the feature information; if the risk coefficient reaches a threshold value, the early warning device is activated, and a graded early warning signal is generated, so that the technical problems of low base safety risk assessment accuracy and low monitoring response speed caused by lack of comprehensive analysis capability on information in multiple aspects such as base structure state, environmental influence and external force action in existing electric tower base safety monitoring are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of safety monitoring, and in particular to a safety monitoring and early warning method and platform for a power tower base. Background Art

[0002] The base of the power tower is the core bearing structure of the power transmission line, and its safety is directly related to the stable operation of the power transmission system. However, the existing technology still has many problems in the safety monitoring and early warning of the power tower base. First, because the power tower base is often in a complex and changeable environment, its structural state is easily damaged by the long-term influence of multiple factors such as wind, precipitation, and geological conditions. The existing monitoring technology is difficult to fully perceive the multi-dimensional state characteristics of the base, especially the inability to effectively extract multiple risk factors such as the base structure, environment, and external forces, which leads to low accuracy of safety monitoring. Secondly, the existing monitoring system mostly relies on single sensor data, and the accumulation of sensor errors will affect the reliability of the monitoring results, and the abnormal state of the sensor equipment cannot be corrected in real time. In addition, the data fusion analysis for large-scale IoT sensor arrays is still insufficient, and it is impossible to achieve real-time tracking and accurate capture of dynamic features. Thirdly, the algorithm model of the existing safety risk assessment and early warning system is generally relatively simple, and fails to comprehensively consider the correlation between different risk factors, resulting in delayed or distorted risk assessment results. At the same time, the response speed of the early warning system is slow, which makes it difficult to meet the needs of quickly locating risk sources and taking timely intervention. Summary of the invention

[0003] The present application provides a safety monitoring and early warning method and platform for power tower bases, which is used to solve the technical problems that the existing power tower base safety monitoring lacks comprehensive analysis capabilities for multiple information such as the base structure status, environmental impacts, and external forces, resulting in low accuracy in base safety risk assessment and slow monitoring response speed.

[0004] In view of the above problems, the present application provides a safety monitoring and early warning method and platform for power tower bases.

[0005] In a first aspect of the present application, a safety monitoring and early warning method for a power tower base is provided, the method comprising: mining monitoring features of the power tower base according to multiple safety risk factors, and establishing a base monitoring feature distribution network; based on the base monitoring feature distribution network, real-time monitoring of the power tower base is performed according to a pre-deployed Internet of Things sensor array to obtain a base monitoring data stream; based on a sensor correction multi-level model, sensor error compensation is performed on the base monitoring data stream to obtain base structure state characteristic information, base environment state characteristic information and base external force action characteristic information of the power tower base; activating a base safety risk detection multi-channel, combining the base structure state characteristic information, the base environment state characteristic information and the base external force action characteristic information to perform safety risk detection on the power tower base to obtain a base safety risk coefficient; if the base safety risk coefficient is greater than or equal to a base safety risk threshold, activating a base safety early warning device; inputting the base safety risk coefficient into the base safety early warning device to generate a base safety graded early warning signal.

[0006] The second aspect of the present application provides a safety monitoring and early warning platform for power tower bases, the platform comprising: a monitoring feature mining module: mining monitoring features of power tower bases according to multiple safety risk factors, and establishing a base monitoring feature distribution network; a real-time monitoring module: based on the base monitoring feature distribution network, the power tower base is monitored in real time according to a pre-deployed IoT sensor array to obtain a base monitoring data stream; a sensor error compensation module: based on a sensor correction multi-level model, the sensor error compensation of the base monitoring data stream is performed to obtain the base structure status feature information of the power tower base , base environmental status characteristic information and base external force characteristic information; safety risk detection module: activate the base safety risk detection multi-channel, combine the base structure status characteristic information, the base environmental status characteristic information and the base external force characteristic information to perform safety risk detection on the power tower base, and obtain the base safety risk coefficient; early warning device activation module: if the base safety risk coefficient is greater than or equal to the base safety risk threshold, activate the base safety early warning device; early warning signal generation module: input the base safety risk coefficient into the base safety early warning device to generate a base safety graded early warning signal.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: According to the multivariate safety risk factors, the monitoring characteristics of the power tower base are mined to establish a base monitoring characteristic distribution network; based on the base monitoring characteristic distribution network, the power tower base is monitored in real time according to the pre-deployed Internet of Things sensor array to obtain the base monitoring data stream; based on the sensor correction multi-level model, the sensor error compensation is performed on the base monitoring data stream to obtain the base structure state characteristic information, base environment state characteristic information and base external force action characteristic information of the power tower base; the base safety risk detection multi-channel is activated, and the power tower base is tested for safety risk in combination with the base structure state characteristic information, the base environment state characteristic information and the base external force action characteristic information to obtain the base safety risk coefficient; if the base safety risk coefficient is greater than or equal to the base safety risk threshold, the base safety early warning device is activated; the base safety risk coefficient is input into the base safety early warning device to generate a base safety graded early warning signal. The technical effect of accurately identifying the base safety risk level and improving the monitoring reliability and early warning efficiency by combining the sensor error compensation model with the multi-channel safety risk detection is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0009] Figure 1 A schematic flow chart of a safety monitoring and early warning method for a power tower base provided in an embodiment of the present application.

[0010] Figure 2 A schematic diagram of the structure of a safety monitoring and early warning platform for a power tower base provided in an embodiment of the present application.

[0011] Explanation of the accompanying drawings: monitoring feature mining module 1, real-time monitoring module 2, sensor error compensation module 3, safety risk detection module 4, early warning device activation module 5, early warning signal generation module 6. DETAILED DESCRIPTION

[0012] This application provides a safety monitoring and early warning method and platform for power tower bases, which is used to solve the technical problems that the existing power tower base safety monitoring lacks comprehensive analysis capabilities for multiple information such as the base structure status, environmental impacts, and external forces, resulting in low accuracy of base safety risk assessment and slow monitoring response speed.

[0013] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0014] Embodiment 1, as Figure 1 As shown, the present application provides a safety monitoring and early warning method for a power tower base, the method comprising: Based on multiple safety risk factors, the monitoring characteristics of the power tower base are mined and a base monitoring feature distribution network is established.

[0015] In one embodiment, for various safety risks that the power tower base may be subject to, key risk factors are determined. These factors mainly include factors related to the base structure state, environmental state factors, and external force factors. These factors work together to determine the overall safety state of the base. For each risk factor, by collecting historical monitoring data and risk events in typical cases, monitoring event features related to the risk factor are extracted. For example, for structural state risks, monitoring parameters related to crack expansion can be extracted. For environmental state risks, the impact of temperature changes on base stress can be extracted. For external force risks, the intensity of wind load on the base structure can be extracted. ; Subsequently, the monitoring events in each type of risk factor are deeply mined, and reliable monitoring indicators are screened out by setting confidence thresholds. These indicators represent the key characteristic data of different risk factors and can effectively reflect their respective risk characteristics; After that, these monitoring indicators are organized in the form of a network, and characteristic distribution sub-networks of structural status, environmental status and external force are constructed respectively. These sub-networks are then integrated through parallel nodes to form an overall base monitoring characteristic distribution network. This network can comprehensively describe the multi-dimensional safety risk characteristics of the power tower base, serve as the basis for subsequent real-time monitoring and risk assessment, and provide data support for accurate safety warnings.

[0016] In one possible implementation, monitoring features of power tower bases are mined based on multiple safety risk factors, and a base monitoring feature distribution network is established, including: The multivariate safety risk factors include structural state risk, environmental state risk and external force risk; monitoring events of the power tower base are retrieved according to the multivariate safety risk factors to obtain a set of structural state risk monitoring events, a set of environmental state risk monitoring events and a set of external force risk monitoring events; monitoring indicators of the structural state risk monitoring event set are mined based on a predetermined confidence level to construct a structural state monitoring feature distribution subnetwork; monitoring indicators of the environmental state risk monitoring event set are mined based on the predetermined confidence level to construct an environmental state monitoring feature distribution subnetwork; monitoring indicators of the external force risk monitoring event set are mined based on the predetermined confidence level to construct an external force monitoring feature distribution subnetwork; the structural state monitoring feature distribution subnetwork, the environmental state monitoring feature distribution subnetwork and the external force monitoring feature distribution subnetwork are connected as parallel nodes to generate the base monitoring feature distribution network.

[0017] Optionally, according to the operating environment and structural characteristics of the power tower base, the multiple safety risk factors are divided into three categories, namely structural state risk, environmental state risk and external force risk. Among them, the structural state risk includes aging of base materials, crack extension, change of inclination angle, etc., which directly reflect the problem of the integrity of the base structure. The environmental state risk includes changes in temperature, humidity, rainfall and geological conditions, etc. These environmental factors may have an indirect impact on the physical state of the base. The external force risk includes wind load, seismic force, vibration impact, etc. These directly applied external forces may cause abnormal stress on the base. For the above three types of risk factors, retrieve the historical monitoring data of the power tower base and typical events that occur in actual operation, forming a structural state risk monitoring event set, an environmental state risk monitoring event set and an external force risk monitoring event set, wherein the structural state risk monitoring event set includes historical monitoring records of foundation cracks, deformation measurement results, etc., the environmental state risk monitoring event set includes historical meteorological data, environmental stress test data, etc., and the external force risk monitoring event set includes wind load, vibration intensity monitoring data, etc.; subsequently, data analysis and mining are performed on each monitoring event set, and effective monitoring indicators are screened out based on a predetermined confidence level (for example, the reliability of a certain monitoring indicator on the risk factor reaches a certain threshold); by analyzing the structural state risk By mining monitoring indicators, we can extract high-confidence indicators such as crack length, material stress distribution, and inclination angle, and then associate these indicators in a network form to construct a structural state monitoring feature distribution subnetwork; by mining environmental state risk monitoring indicators, we can extract high-confidence indicators such as the impact of humidity on base stress and the impact of rainfall on foundation changes, and associate these indicators in a network form to construct an environmental state monitoring feature distribution subnetwork; by mining external force risk monitoring indicators, we can extract high-confidence indicators such as the effect of wind load on base deformation and the relationship between vibration frequency and structural response, and associate these indicators in a network form to construct an external force monitoring feature distribution subnetwork. Distributed subnetworks; each subnetwork node of these subnetworks represents a key monitoring indicator, and the lines between the nodes represent the correlation or synergy between the indicators; then, the structural state monitoring feature distribution subnetwork, the environmental state monitoring feature distribution subnetwork and the external force monitoring feature distribution subnetwork are connected through parallel nodes to form a comprehensive base monitoring feature distribution network. In this network, different subnetworks exist in a parallel structure, and each node and line provides a complete description of multi-dimensional feature information. The base monitoring feature distribution network can serve as the core foundation for real-time monitoring and risk assessment, and can be used for subsequent data analysis and risk warning to support accurate safety monitoring decisions.

[0018] In a possible implementation, monitoring index mining is performed on the structure state risk monitoring event set based on a predetermined confidence level to construct a structure state monitoring feature distribution subnetwork, including: Monitoring indicators are identified based on the structural state risk monitoring event set to obtain multiple sample structural state risk monitoring indicators; confidence evaluation is performed on the multiple sample structural state risk monitoring indicators based on the structural state risk monitoring event set to obtain multiple sample indicator confidences; based on the multiple sample indicator confidences, confidence features are selected for the multiple sample structural state risk monitoring indicators according to the predetermined confidences to generate a structural state monitoring confidence indicator distribution; based on the structural state monitoring confidence indicator distribution, the structural state monitoring feature distribution subnetwork is constructed.

[0019] Optionally, multiple sample monitoring indicators related to the structural state are extracted from the structural state risk monitoring event set as sample structural state risk monitoring indicators. These indicators reflect the characteristics of different structural state risks. For example, the crack length can reflect the degree of damage to the base structure, the inclination angle can show the stability problem of the base, the stress distribution value can indicate the distribution of the bearing capacity of the base, and the vibration amplitude can show the response of the base to dynamic external forces. Subsequently, based on the structural state risk monitoring event set, the confidence of the sample monitoring indicators extracted above is evaluated. In this process, the sample structural state risk monitoring indicators are normalized by the maximum and minimum methods, and these indicator data are converted into Convert it to a comparable unified scale, and then use the Pearson correlation coefficient to calculate the linear correlation between the indicator and the frequency of risk events to obtain the linear correlation coefficient. If the absolute value of this linear correlation coefficient is lower than the preset correlation threshold, there may be a nonlinear correlation. At this time, use mutual information to calculate the nonlinear relationship between the indicator and the risk event to obtain the nonlinear correlation coefficient; in addition, the standard deviation of the normalized sample structure status risk monitoring indicator is used to calculate the ratio of the mean, and then use 1 to subtract this ratio and compare it with 0, and the larger one is used as the indicator stability coefficient; then, the calculated linear correlation coefficient or nonlinear correlation coefficient is weighted with the indicator stability coefficient The sample indicator confidence of each indicator is calculated, and then the sample indicator confidence is selected with the predetermined confidence to select the indicators that are greater than or equal to the predetermined confidence. For example, the predetermined confidence is 0.9, the crack length (0.92) and the stress distribution value (0.94) meet the requirements and are set as the structural state monitoring confidence indicators. The inclination angle (0.87) and the vibration amplitude (0.89) do not meet the requirements and are not included in the structural state monitoring confidence indicators. The qualified sample structural state risk monitoring indicators are defined as structural state monitoring confidence indicators, and a set of structural state monitoring confidence indicators is generated. According to this set of structural state monitoring confidence indicators, the indicators are distributed in a multidimensional In the distribution diagram, each indicator is a node, and the weight of the node can be expressed by the confidence of the indicator. For example, the crack length and stress distribution value occupy a key position in the distribution diagram, and the node weights are 0.92 and 0.94 respectively, indicating their importance in monitoring. Then, based on the distribution of confidence indicators for structural state monitoring, a feature distribution subnetwork is constructed. The nodes represent confidence indicators (such as crack length and stress distribution value), and the lines between nodes represent the correlation or synergy between different indicators (such as crack length and stress distribution value may have a positive correlation). This feature distribution subnetwork will be used as a feature distribution subnetwork for structural state monitoring, and will be used for feature support and analysis in subsequent risk detection. Through the above process, high-confidence monitoring features can be effectively mined from the structural state risk monitoring event set, and a scientific and reasonable feature distribution subnetwork can be built to provide reliable support for accurate structural state risk assessment and early warning.

[0020] Based on the base monitoring feature distribution network, the power tower base is monitored in real time according to the pre-deployed Internet of Things sensor array to obtain a base monitoring data stream.

[0021] In one embodiment, multiple types of sensors are pre-arranged at key locations of the power tower base. For example, structural status monitoring sensors are installed at the load-bearing locations of the base to monitor information such as cracks, tilt angles, and stress distribution. Environmental status monitoring sensors are arranged around the base to collect environmental data such as temperature, humidity, and rainfall. External force monitoring sensors are installed on or near the base surface to record external force data such as wind speed, wind direction, and impact force. These sensors are connected to the Internet of Things gateway via wireless communications (such as Bluetooth, WiFi, LoRa, etc.) to form an Internet of Things sensor array to achieve real-time data transmission. Monitoring data is continuously collected through the Internet of Things sensor array and sent to the system terminal at a predetermined frequency (such as every second or every minute). These data include sensor locations, monitoring index values, and timestamps to form a complete set of monitoring data packets. After receiving these sensor data, the system terminal distributes the data to the network according to the sensor type and monitoring characteristics. The data are classified and integrated to obtain data on structural status (such as time series data on crack extension and stress distribution), data on environmental status (such as environmental condition sequences such as temperature and humidity changes and rainfall intensity), and data on external forces (such as dynamic external force information such as wind load impact force and vibration frequency). These data streams together constitute the base monitoring data stream and are stored in a structured form, marking the feature type corresponding to the node in the feature distribution network. That is, each piece of data in the monitoring data stream is mapped to the corresponding monitoring feature in real time according to the node definition of the base monitoring feature distribution network. For example, the sensor data of crack extension is associated with the crack length node in the structural status monitoring subnetwork, and the temperature and humidity data are mapped to the corresponding node in the environmental status monitoring subnetwork; according to the real-time collected monitoring data stream, the status of the base monitoring feature distribution network can be dynamically updated, ensuring the timely perception of the safety status of the power tower base, and providing a basis for subsequent risk assessment and early warning. Through this process, the IoT sensor array is closely integrated with the base monitoring feature distribution network, realizing real-time monitoring and data stream generation of the power tower base, and effectively improving the accuracy and timeliness of safety monitoring.

[0022] Based on the sensor correction multi-level model, sensor error compensation is performed on the base monitoring data stream to obtain base structure state characteristic information, base environment state characteristic information and base external force action characteristic information of the power tower base.

[0023] In one embodiment, after obtaining the base monitoring data stream, the data in the base monitoring data stream and the state information of the corresponding sensor equipment are input into the sensor correction multi-level model. The sensor correction multi-level model will detect abnormal values ​​or inconsistencies in the monitoring data stream, analyze their potential impact on the overall monitoring results, and determine whether compensation is required; if compensation is required, the monitoring data will be error corrected through the sensor correction layer inside the model, and the corrected monitoring data will be correspondingly integrated to obtain the base structure state characteristic information, base environment state characteristic information, and base external force action characteristic information, providing highly reliable data support for subsequent risk assessment and early warning. Through this process, the sensor correction multi-level model can effectively identify and compensate for sensor errors, thereby generating accurate structural state, environmental state, and external force action characteristic information, significantly improving the reliability and data quality of monitoring.

[0024] In a possible implementation, the base monitoring data stream is subjected to sensor error compensation based on a sensor correction multi-level model to obtain base structure state characteristic information, base environment state characteristic information, and base external force action characteristic information of the power tower base, including: The base monitoring data stream includes base structure monitoring data, base environment monitoring data and base external force monitoring data; the base structure monitoring data is traversed to extract the first parameter of base structure monitoring; the first parameter sensor device status information corresponding to the first parameter of base structure monitoring is loaded; based on the first parameter sensor device status information, the first parameter of base structure monitoring is corrected according to the sensor correction multi-level model to obtain the first correction parameter of base structure monitoring, and the first correction parameter of base structure monitoring is added to the base structure state characteristic information; based on the sensor correction multi-level model, the base environment monitoring data and the base external force monitoring data are corrected respectively to generate the base environment state characteristic information and the base external force action characteristic information.

[0025] Optionally, the base monitoring data stream includes base structure monitoring data, base environment monitoring data and base external force monitoring data, wherein the base structure monitoring data includes crack length, tilt angle, etc., the base environment monitoring data includes temperature, humidity, rainfall intensity, etc., and the base external force monitoring data includes wind load impact force, vibration frequency, etc.; for the base structure monitoring data, the first parameter traversed is used as the first parameter of the base structure monitoring, and then the status information of the corresponding sensor device is loaded as the first parameter sensor device status information. Each monitoring parameter corresponds to a collected sensor device, and its status information includes device identification (such as device ID, model, etc.), working status (whether there is a fault at present), historical error data, operating environment information (such as whether the equipment is subject to high temperature, high humidity and other environmental impacts); then, the first parameter sensing device status information and the first parameter of the base structure monitoring are input into the sensing correction multi-level model for correction. The abnormal identification layer of the sensing correction multi-level model will detect whether the parameters are abnormal, the error impact evaluation layer will calculate the potential impact of the abnormality on the monitoring results, the error judgment layer will confirm whether correction is required, and the sensing correction layer will correct the abnormal data; after the sensor correction layer is corrected, the corrected parameters are output as the first correction parameters of the base structure monitoring, and added to the base structure status characteristic information; based on the same method, the sensing correction multi-level model is also used to analyze and correct the base environment monitoring data and the base external force monitoring data one by one, and the base environment status characteristic information and the base external force action characteristic information are obtained. Through the above steps, the sensing error compensation of the structure, environment and external force monitoring data in the base monitoring data stream is fully realized, providing high-quality and reliable data support for subsequent risk assessment and safety warning.

[0026] In a possible implementation, based on the first parameter sensing device state information, the first parameter of the base structure monitoring is corrected according to the sensing correction multi-level model to obtain the first correction parameter of the base structure monitoring, including: The sensor correction multi-level model includes a sensor anomaly recognition layer, a sensor anomaly impact evaluation layer, an abnormal impact judgment layer and a sensor correction layer; the first parameter sensor device state information is input into the sensor anomaly recognition layer to obtain a first sensor device state abnormality recognition result; the first sensor device state abnormality recognition result is input into the sensor anomaly impact evaluation layer to obtain a first sensor anomaly impact evaluation coefficient; the first sensor anomaly impact evaluation coefficient is input into the abnormal impact judgment layer to determine whether the first sensor anomaly impact evaluation coefficient is greater than or equal to the sensor anomaly impact evaluation threshold in the abnormal impact judgment layer; if the first sensor anomaly impact evaluation coefficient is greater than or equal to the sensor anomaly impact evaluation threshold, the sensor correction layer is activated; the first sensor device state abnormality recognition result and the first parameter of the base structure monitoring are input into the sensor correction layer to generate the first correction parameter of the base structure monitoring.

[0027] Optionally, the sensor correction multi-level model includes a sensor anomaly identification layer, a sensor anomaly impact evaluation layer, an abnormal impact judgment layer and a sensor correction layer. When the sensor correction multi-level model receives the first parameter sensor device status information and the first parameter of the base structure monitoring, the first parameter sensor device status information will be transmitted to the internal sensor anomaly identification layer. The sensor anomaly identification layer will identify the abnormality of the device status according to the first parameter sensor device status information, such as voltage, frequency signal, temperature, humidity, etc., that is, determine whether these information parameters are within the normal working range. If there is an abnormality, a first sensor device status abnormality identification result is generated. This result includes information parameters that are not within the normal working range, and the type of abnormality (such as exceeding the upper limit, exceeding the lower limit, etc.); then, the first sensor device status abnormality identification result is transmitted to the sensor anomaly impact evaluation layer. The sensor anomaly impact evaluation layer will extract the information parameters in the first sensor device status abnormality identification result, and obtain the normal working range corresponding to the information parameters, by calculating the ratio of the absolute value of the information parameter minus the upper limit or lower limit of the normal working range to the difference between the upper limit and the lower limit of the normal working range. The influence coefficient of the information parameter is obtained, wherein the upper limit or lower limit of the normal working range is subtracted depending on the type of abnormality, and then the influence coefficients of all information parameters are averaged to obtain the first sensor abnormality impact evaluation coefficient, which reflects the degree of deviation of the abnormality on the monitoring data; then, the evaluation coefficient is input into the abnormality impact judgment layer to compare the sensor abnormality impact evaluation threshold, which is the maximum acceptable degree of influence of the abnormality on the monitoring data; if the evaluation coefficient is less than the threshold, it is considered that the influence of the abnormality on the data is within an acceptable range and no correction is required; if the evaluation coefficient is greater than or equal to the threshold, it means that the abnormal influence exceeds the allowable range and needs to be corrected. At this time, the sensor correction layer is activated, and the first sensor device state abnormality identification result and the first base structure monitoring parameter are input into the sensor correction layer together. The sensor correction layer multiplies the first sensor abnormality impact evaluation coefficient with the deviation value of the information parameter in the first sensor device state abnormality identification result, and subtracts the obtained product from the first base structure monitoring parameter to obtain the first correction parameter for base structure monitoring, and adds this corrected parameter to the base structure state characteristic information for subsequent risk assessment. Through the above steps, the sensor correction multi-level model can accurately identify sensor anomalies, quantify the impact of anomalies, determine whether correction is needed, and ultimately generate highly reliable correction parameters, effectively improving the accuracy and reliability of monitoring data.

[0028] Activate the multi-channel of base safety risk detection, perform safety risk detection on the power tower base in combination with the base structure state characteristic information, the base environment state characteristic information and the base external force action characteristic information, and obtain the base safety risk coefficient.

[0029] In one embodiment, after obtaining the characteristic information of the base structure state, the characteristic information of the base environment state and the characteristic information of the base external force, the multi-channel for base safety risk detection is activated, which will comprehensively analyze the base structure state, the environmental state around the base, and the external force on the base and other information, and integrate the output results of the internal structure state risk detection channel, the environmental state risk detection channel and the external force risk detection channel by weighted calculation to obtain a base safety risk coefficient, which is used to indicate the safety state of the base under the current conditions. If the risk coefficient is high, it means that the base has a high risk and needs to take early warning measures or perform maintenance.

[0030] In a possible implementation, activating the base safety risk detection multi-channel, combining the base structure state characteristic information, the base environment state characteristic information and the base external force action characteristic information to perform safety risk detection on the power tower base, and obtaining the base safety risk coefficient, including: The base safety risk detection multi-channel includes a structural state risk detection channel, an environmental state risk detection channel, an external force risk detection channel and a base safety risk integration channel; the base structural state characteristic information is input into the structural state risk detection channel to obtain a structural state risk detection coefficient; the base environmental state characteristic information is input into the environmental state risk detection channel to obtain an environmental state risk detection coefficient; the base external force characteristic information is input into the external force risk detection channel to obtain an external force risk detection coefficient; the structural state risk detection coefficient, the environmental state risk detection coefficient and the external force risk detection coefficient are input into the base safety risk integration channel to output the base safety risk coefficient, wherein the base safety risk integration channel includes a predetermined weight for structural state risk, a predetermined weight for environmental state risk and a predetermined weight for external force risk.

[0031] Optionally, the base safety risk detection multi-channel includes a structural state risk detection channel, an environmental state risk detection channel, an external force risk detection channel and a base safety risk integration channel, wherein the structural state risk detection channel is used to evaluate the structural stability of the power tower base, the environmental state risk detection channel is used to analyze the environmental conditions around the base, the external force risk detection channel is used to evaluate the external force impact on the base, and the base safety risk integration channel is used to integrate the risk detection results of multiple channels to output an overall safety risk coefficient; the structural state characteristic information of the base is input into the structural state risk detection channel, and the structural state risk detection channel analyzes the input information through multiple internal structural state risk detection models to evaluate the risk level of the base at the structural level and obtain multiple structural state risk coefficients, and then these structural state risk coefficients are weighted to obtain a structural state risk detection coefficient, which is used to quantify the stability of the base structure in the current state; similarly The characteristic information of the foundation environment state is input into the environmental state risk detection channel, and the potential impact of environmental conditions on the foundation safety is analyzed in the same way to generate an environmental state risk detection coefficient to reflect the impact of the environment on the foundation stability. The characteristic information of the foundation external force is input into the external force risk detection channel to evaluate the direct impact of external force on the foundation safety and generate an external force risk detection coefficient to quantify the risk impact of external forces on the foundation. Subsequently, the obtained structural state risk detection coefficient, environmental state risk detection coefficient, and external force risk detection coefficient are transmitted to the foundation safety risk integration channel. The influence of different risk detection coefficients is combined through the predetermined weights of the structural state risk, the predetermined weights of the environmental state risk, and the predetermined weights of the external force risk in the foundation safety risk integration channel to obtain the final foundation safety risk coefficient. Among them, the predetermined weights of each channel are usually set according to the actual scenario, such as the structure may have a greater weight, the environment is second, and the external force may be slightly lighter. This foundation safety risk coefficient is used to characterize the overall safety of the foundation under the current structural state, environmental conditions, and external forces. This process of multi-channel detection and weighted integration ensures that all important factors affecting foundation safety are considered, providing a more comprehensive safety assessment.

[0032] In a possible implementation, inputting the base structure state characteristic information into the structure state risk detection channel to obtain a structure state risk detection coefficient includes: The structural state risk detection channel includes K structural state risk detection models, wherein K is a positive integer greater than 1; the base structure state characteristic information is input into the K structural state risk detection models to obtain K structural state risk coefficients; a proportion calculation is performed according to the K structural state risk detection accuracies corresponding to the K structural state risk detection models to obtain K structural state risk incentive coefficients; a weighted calculation is performed on the K structural state risk coefficients according to the K structural state risk incentive coefficients to output the structural state risk detection coefficients.

[0033] Optionally, the structural state risk detection channel includes K different structural state risk detection models, which are used to evaluate the health of the base structure from different angles, where K is a positive integer greater than 1. These structural state risk detection models can be constructed based on decision trees, support vector machines, neural networks, random forests, etc., and the combination method can be the common use of multiple algorithms or only based on one algorithm. Taking the neural network as an example, random sampling with replacement is performed from the sample safety risk data set to construct K sample training sets, K sample verification sets, and K sample test sets, each of which contains sample base structure state feature information and sample structure state risk coefficient. For any structural state risk detection model, the model structure is designed based on the neural network, including the input layer, hidden layer, output layer, etc. The sample base structure state feature information in the sample training set is calculated from the input layer to the output layer in sequence by forward propagation. The predicted risk coefficient is obtained by weighted summation and activation function processing of each layer. The loss of the current batch of samples is calculated by the mean square error (MSE) to measure the model prediction value and the corresponding sample structure state. The difference in risk coefficient is calculated by the chain rule, and the gradient of the loss relative to the weights and biases of each layer is calculated. The optimizer (such as Adam) is used to update the weights and biases of each layer to reduce the value of the loss function. The model will perform multiple iterations on the entire sample training set until the value of the loss function converges or reaches the predetermined number of iterations. After each training batch is completed, the sample validation set is used to evaluate the model performance to see if the loss is decreasing to determine whether the model is overfitting. If the validation set loss no longer decreases in several consecutive batches, the training is stopped to prevent the model from overfitting. After the model training is completed, the accuracy on the test set is calculated to verify the generalization ability of the model. If the accuracy does not meet the preset expectations, the learning rate is adjusted so that the model can find the optimal solution more stably, or the number of layers or nodes of the hidden layer is adjusted to find the best model structure. If the accuracy meets the preset expectations, the current structural state risk detection model is output and the current accuracy is stored as the structural state risk detection accuracy of the model. The above process is repeated until K structural state risk detection models are constructed.

[0034] Optionally, the characteristic information of the foundation structure state is input into the structural state risk detection channel, which distributes it to each structural state risk detection model. Each structural state risk detection model analyzes and evaluates the input data to generate a risk coefficient, thereby obtaining K structural state risk coefficients; each structural state risk detection model has its corresponding detection accuracy, which is used to represent the reliability of the model. According to the detection accuracy of each model, the corresponding structural state risk incentive coefficient is calculated to represent the degree of influence of each model on the final result. The calculation method is to calculate the ratio of the structural state risk incentive coefficient to the sum of all structural state risk incentive coefficients; then, these structural state risk incentive coefficients are used to perform weighted calculation on the K structural state risk coefficients to output the final structural state risk detection coefficient, which is used to quantify the structural health of the foundation. This multi-model integration method can improve the comprehensiveness and accuracy of structural state assessment and ensure accurate assessment of the health of the foundation from multiple angles.

[0035] If the base safety risk coefficient is greater than or equal to the base safety risk threshold, the base safety early warning device is activated; the base safety risk coefficient is input into the base safety early warning device to generate a base safety graded early warning signal.

[0036] In one embodiment, if the base safety risk coefficient is greater than or equal to the preset base safety risk threshold, the base safety warning device is activated. First, the safety status of the current base is determined by comparing the base safety risk coefficient with the preset threshold. If the base safety risk coefficient is greater than or equal to the base safety risk threshold, it means that there is a greater risk of hidden danger in the safety status of the base, and further countermeasures need to be taken. At this time, the base safety warning device is activated. The activation process includes starting the alarm device, such as an alarm sound, an alarm light, or sending a communication signal to the remote monitoring system to remind maintenance personnel to respond in time; when the warning device is activated, the base safety risk coefficient is input into the base safety warning device to generate a corresponding graded warning signal; the base safety warning device classifies the risk level according to the size of the risk coefficient and according to pre-set rules. For example, the risk coefficient can be divided into three levels, corresponding to low-level warning, intermediate-level warning and high-level warning. The low-level warning indicates that the base has a certain risk and needs continuous attention. The intermediate warning indicates that the base has a greater risk. If the risk is high, it is recommended to check and take appropriate maintenance measures as soon as possible. The advanced warning indicates that there are serious safety hazards in the base and reinforcement must be carried out immediately. The early warning device performs conditional judgment on the value of the risk coefficient. If the base safety risk coefficient is within the low-level warning range, a low-level warning signal is generated. If the base safety risk coefficient is within the medium-level warning range, an medium-level warning signal is generated. If the base safety risk coefficient is within the high-level warning range, a high-level warning signal is generated. These graded warning signals will be notified to maintenance personnel locally through display screens, warning lights or other means, and will also be transmitted to the remote monitoring center through the communication network to centrally manage the safety status of the base. In addition, these warning signals will also be stored in the database, including warning levels, timestamps and other information, to facilitate subsequent tracking and analysis.

[0037] Embodiment 2, based on the same inventive concept as the safety monitoring and early warning method for power tower base in the above embodiment, Figure 2As shown, the present application provides a safety monitoring and early warning platform for power tower bases. The platform and method embodiments in the embodiments of the present application are based on the same inventive concept. Among them, the platform includes: monitoring feature mining module 1: mining monitoring features of the power tower base according to multiple safety risk factors, and establishing a base monitoring feature distribution network; real-time monitoring module 2: based on the base monitoring feature distribution network, the power tower base is monitored in real time according to the pre-deployed Internet of Things sensor array to obtain a base monitoring data stream; sensor error compensation module 3: based on the sensor correction multi-level model, the sensor error compensation of the base monitoring data stream is performed to obtain the base structure state feature information, base environment state feature information and base of the power tower base. External force action characteristic information; Safety risk detection module 4: Activate the base safety risk detection multi-channel, combine the base structure status characteristic information, the base environment status characteristic information and the base external force action characteristic information to perform safety risk detection on the power tower base, and obtain the base safety risk coefficient; Early warning device activation module 5: If the base safety risk coefficient is greater than or equal to the base safety risk threshold, activate the base safety early warning device; Early warning signal generation module 6: Input the base safety risk coefficient into the base safety early warning device to generate a base safety graded early warning signal.

[0038] Furthermore, the monitoring feature mining module 1 also includes: The multivariate safety risk factors include structural state risk, environmental state risk and external force risk; monitoring events of the power tower base are retrieved according to the multivariate safety risk factors to obtain a set of structural state risk monitoring events, a set of environmental state risk monitoring events and a set of external force risk monitoring events; monitoring indicators of the structural state risk monitoring event set are mined based on a predetermined confidence level to construct a structural state monitoring feature distribution subnetwork; monitoring indicators of the environmental state risk monitoring event set are mined based on the predetermined confidence level to construct an environmental state monitoring feature distribution subnetwork; monitoring indicators of the external force risk monitoring event set are mined based on the predetermined confidence level to construct an external force monitoring feature distribution subnetwork; the structural state monitoring feature distribution subnetwork, the environmental state monitoring feature distribution subnetwork and the external force monitoring feature distribution subnetwork are connected as parallel nodes to generate the base monitoring feature distribution network.

[0039] Furthermore, the monitoring feature mining module 1 also includes: Monitoring indicators are identified based on the structural state risk monitoring event set to obtain multiple sample structural state risk monitoring indicators; confidence evaluation is performed on the multiple sample structural state risk monitoring indicators based on the structural state risk monitoring event set to obtain multiple sample indicator confidences; based on the multiple sample indicator confidences, confidence features are selected for the multiple sample structural state risk monitoring indicators according to the predetermined confidences to generate a structural state monitoring confidence indicator distribution; based on the structural state monitoring confidence indicator distribution, the structural state monitoring feature distribution subnetwork is constructed.

[0040] Furthermore, the sensor error compensation module 3 also includes: The base monitoring data stream includes base structure monitoring data, base environment monitoring data and base external force monitoring data; the base structure monitoring data is traversed to extract the first parameter of base structure monitoring; the first parameter sensor device status information corresponding to the first parameter of base structure monitoring is loaded; based on the first parameter sensor device status information, the first parameter of base structure monitoring is corrected according to the sensor correction multi-level model to obtain the first correction parameter of base structure monitoring, and the first correction parameter of base structure monitoring is added to the base structure state characteristic information; based on the sensor correction multi-level model, the base environment monitoring data and the base external force monitoring data are corrected respectively to generate the base environment state characteristic information and the base external force action characteristic information.

[0041] Furthermore, the sensor error compensation module 3 also includes: The sensor correction multi-level model includes a sensor anomaly recognition layer, a sensor anomaly impact evaluation layer, an abnormal impact judgment layer and a sensor correction layer; the first parameter sensor device state information is input into the sensor anomaly recognition layer to obtain a first sensor device state abnormality recognition result; the first sensor device state abnormality recognition result is input into the sensor anomaly impact evaluation layer to obtain a first sensor anomaly impact evaluation coefficient; the first sensor anomaly impact evaluation coefficient is input into the abnormal impact judgment layer to determine whether the first sensor anomaly impact evaluation coefficient is greater than or equal to the sensor anomaly impact evaluation threshold in the abnormal impact judgment layer; if the first sensor anomaly impact evaluation coefficient is greater than or equal to the sensor anomaly impact evaluation threshold, the sensor correction layer is activated; the first sensor device state abnormality recognition result and the first parameter of the base structure monitoring are input into the sensor correction layer to generate the first correction parameter of the base structure monitoring.

[0042] Furthermore, the security risk detection module 4 also includes: The base safety risk detection multi-channel includes a structural state risk detection channel, an environmental state risk detection channel, an external force risk detection channel and a base safety risk integration channel; the base structural state characteristic information is input into the structural state risk detection channel to obtain a structural state risk detection coefficient; the base environmental state characteristic information is input into the environmental state risk detection channel to obtain an environmental state risk detection coefficient; the base external force characteristic information is input into the external force risk detection channel to obtain an external force risk detection coefficient; the structural state risk detection coefficient, the environmental state risk detection coefficient and the external force risk detection coefficient are input into the base safety risk integration channel to output the base safety risk coefficient, wherein the base safety risk integration channel includes a predetermined weight for structural state risk, a predetermined weight for environmental state risk and a predetermined weight for external force risk.

[0043] Furthermore, the security risk detection module 4 also includes: The structural state risk detection channel includes K structural state risk detection models, wherein K is a positive integer greater than 1; the base structure state characteristic information is input into the K structural state risk detection models to obtain K structural state risk coefficients; a proportion calculation is performed according to the K structural state risk detection accuracies corresponding to the K structural state risk detection models to obtain K structural state risk incentive coefficients; a weighted calculation is performed on the K structural state risk coefficients according to the K structural state risk incentive coefficients to output the structural state risk detection coefficients.

[0044] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0045] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0046] This specification and the drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A safety monitoring and early warning method for a power tower base, characterized in that: The method comprises: Mining the monitoring characteristics of power tower bases based on multiple safety risk factors and establishing a base monitoring characteristic distribution network; Based on the base monitoring characteristic distribution network, the power tower base is monitored in real time according to a pre-deployed IoT sensor array to obtain a base monitoring data stream; Based on the sensor correction multi-level model, the sensor error compensation is performed on the base monitoring data stream to obtain the base structure state characteristic information, base environment state characteristic information and base external force action characteristic information of the power tower base; Activate the base safety risk detection multi-channel, combine the base structure state characteristic information, the base environment state characteristic information and the base external force action characteristic information to perform safety risk detection on the power tower base, and obtain the base safety risk coefficient; If the base safety risk coefficient is greater than or equal to the base safety risk threshold, activating the base safety early warning device; The base safety risk coefficient is input into the base safety early warning device to generate a base safety graded early warning signal.

2. The safety monitoring and early warning method for a power tower base according to claim 1, characterized in that: Based on multiple safety risk factors, the monitoring characteristics of the power tower base are mined and a base monitoring characteristic distribution network is established, including: The multiple safety risk factors include structural state risk, environmental state risk and external force risk; Perform monitoring event retrieval on the power tower base according to the multivariate safety risk factors to obtain a structure state risk monitoring event set, an environment state risk monitoring event set, and an external force risk monitoring event set; Based on a predetermined confidence level, monitoring index mining is performed on the structural state risk monitoring event set to construct a structural state monitoring feature distribution subnetwork; Based on the predetermined confidence level, monitoring indicator mining is performed on the environmental state risk monitoring event set to construct an environmental state monitoring feature distribution subnetwork; Based on the predetermined confidence level, monitoring index mining is performed on the external force risk monitoring event set to construct an external force monitoring feature distribution subnetwork; The structural state monitoring characteristic distribution subnetwork, the environmental state monitoring characteristic distribution subnetwork and the external force action monitoring characteristic distribution subnetwork are connected as parallel nodes to generate the base monitoring characteristic distribution network.

3. The safety monitoring and early warning method for power tower base according to claim 2, characterized in that: Based on a predetermined confidence level, monitoring index mining is performed on the structural state risk monitoring event set to construct a structural state monitoring feature distribution subnetwork, including: Identify monitoring indicators based on the structural state risk monitoring event set to obtain multiple sample structural state risk monitoring indicators; Performing confidence evaluation on the plurality of sample structural state risk monitoring indicators based on the structural state risk monitoring event set to obtain a plurality of sample indicator confidences; Based on the confidence levels of the multiple sample indicators, selecting confidence features of the multiple sample structural state risk monitoring indicators according to the predetermined confidence levels to generate a structural state monitoring confidence indicator distribution; According to the structure state monitoring confidence indicator distribution, the structure state monitoring feature distribution subnetwork is established.

4. The safety monitoring and early warning method for a power tower base according to claim 1, characterized in that: Based on the sensor correction multi-level model, the sensor error compensation is performed on the base monitoring data stream to obtain the base structure state characteristic information, base environment state characteristic information and base external force action characteristic information of the power tower base, including: The base monitoring data stream includes base structure monitoring data, base environment monitoring data and base external force monitoring data; Traversing the base structure monitoring data, and extracting a first parameter of the base structure monitoring; Loading first parameter sensing device status information corresponding to the first parameter monitored by the base structure; Based on the first parameter sensing device state information, the first base structure monitoring parameter is corrected according to the sensing correction multi-level model to obtain a first correction parameter for base structure monitoring, and the first correction parameter for base structure monitoring is added to the base structure state characteristic information; Based on the sensor correction multi-level model, the base environment monitoring data and the base external force monitoring data are corrected respectively to generate the base environment state characteristic information and the base external force action characteristic information.

5. The safety monitoring and early warning method for power tower base according to claim 4, characterized in that: Based on the first parameter sensing device state information, the first parameter of the base structure monitoring is corrected according to the sensing correction multi-level model to obtain the first correction parameter of the base structure monitoring, including: The sensor correction multi-level model includes a sensor anomaly identification layer, a sensor anomaly impact evaluation layer, an abnormal impact judgment layer and a sensor correction layer; Inputting the first parameter sensing device state information into the sensing abnormality identification layer to obtain a first sensing device state abnormality identification result; Inputting the first sensor device state abnormality identification result into the sensor abnormality impact evaluation layer to obtain a first sensor abnormality impact evaluation coefficient; Inputting the first sensor abnormality impact evaluation coefficient into the abnormality impact judgment layer, and judging whether the first sensor abnormality impact evaluation coefficient is greater than or equal to the sensor abnormality impact evaluation threshold in the abnormality impact judgment layer; If the first sensor abnormality impact evaluation coefficient is greater than or equal to the sensor abnormality impact evaluation threshold, activating the sensor correction layer; The abnormal state identification result of the first sensing device and the first parameter of the base structure monitoring are input into the sensing correction layer to generate the first correction parameter of the base structure monitoring.

6. The safety monitoring and early warning method for a power tower base according to claim 1, characterized in that: Activate the multi-channel of base safety risk detection, perform safety risk detection on the power tower base in combination with the base structure state characteristic information, the base environment state characteristic information and the base external force action characteristic information, and obtain the base safety risk coefficient, including: The base safety risk detection multi-channel includes a structural state risk detection channel, an environmental state risk detection channel, an external force risk detection channel and a base safety risk integration channel; Inputting the base structure state characteristic information into the structure state risk detection channel to obtain a structure state risk detection coefficient; Inputting the base environment state characteristic information into the environment state risk detection channel to obtain an environment state risk detection coefficient; Inputting the external force action characteristic information of the base into the external force action risk detection channel to obtain an external force action risk detection coefficient; The structural state risk detection coefficient, the environmental state risk detection coefficient and the external force risk detection coefficient are input into the base safety risk integration channel, and the base safety risk coefficient is output, wherein the base safety risk integration channel includes a predetermined weight for the structural state risk, a predetermined weight for the environmental state risk and a predetermined weight for the external force risk.

7. The safety monitoring and early warning method for power tower base according to claim 6, characterized in that: Inputting the foundation structure state characteristic information into the structure state risk detection channel to obtain the structure state risk detection coefficient includes: The structural state risk detection channel includes K structural state risk detection models, wherein K is a positive integer greater than 1; Inputting the foundation structure state characteristic information into the K structure state risk detection models to obtain K structure state risk coefficients; According to the K structural state risk detection accuracies corresponding to the K structural state risk detection models, a proportion calculation is performed to obtain K structural state risk incentive coefficients; The K structural state risk coefficients are weightedly calculated according to the K structural state risk incentive coefficients, and the structural state risk detection coefficient is output.

8. A safety monitoring and early warning platform for power tower bases, characterized in that: The platform is used to execute the safety monitoring and early warning method for a power tower base according to any one of claims 1 to 7, comprising: Monitoring feature mining module: mining monitoring features of power tower bases based on multiple safety risk factors, and establishing a base monitoring feature distribution network; Real-time monitoring module: Based on the base monitoring feature distribution network, the power tower base is monitored in real time according to the pre-deployed IoT sensor array to obtain the base monitoring data stream; Sensor error compensation module: performs sensor error compensation on the base monitoring data stream based on a sensor correction multi-level model to obtain base structure state characteristic information, base environment state characteristic information and base external force action characteristic information of the power tower base; Safety risk detection module: activating the base safety risk detection multi-channel, combining the base structure state characteristic information, the base environment state characteristic information and the base external force action characteristic information to perform safety risk detection on the power tower base, and obtaining the base safety risk coefficient; Early warning device activation module: if the base safety risk coefficient is greater than or equal to the base safety risk threshold, activate the base safety early warning device; Warning signal generation module: inputs the base safety risk coefficient into the base safety warning device to generate a base safety graded warning signal.

Images