High-altitude hanging basket operation safety monitoring method, system, equipment and medium

Abstract

The invention relates to a high-altitude hanging basket operation safety monitoring method, system and device and a medium, and belongs to the technical field of building construction safety, and the monitoring method comprises the steps: obtaining the operation state information of a plurality of to-be-monitored hanging baskets in high-altitude operation; performing data preprocessing on the operation state information and extracting state feature information; determining a corresponding risk threshold according to the parameter information of each to-be-monitored hanging basket; calculating a comprehensive risk grade score of each to-be-monitored hanging basket according to the state feature information on the basis of a risk threshold value; based on a preset risk monitoring mapping table, determining a preset monitoring frequency of each to-be-monitored hanging basket according to the comprehensive risk grade score; according to the preset monitoring frequency of each to-be-monitored hanging basket, abnormal behavior monitoring and inclination state monitoring are carried out on each to-be-monitored hanging basket; and generating an early warning signal in response to the abnormal monitoring result and sending the early warning signal to the management terminal. The safety monitoring requirements of a plurality of different hanging baskets can be met, and the operation safety is improved.

Description

Technical Field

[0001] The present application relates to the field of building construction safety technology, and in particular to a method, system, equipment and medium for monitoring the safety of aerial hanging basket operations. Background Art

[0002] With the rapid development of the construction industry, aerial work is one of the common forms of work in the construction, maintenance and other industries. Especially when cleaning, repairing or installing exterior walls, it is often necessary to use a hanging basket to carry workers and equipment. As an important aerial work tool, the safety of the hanging basket is directly related to the life safety of the workers. However, there are often many safety hazards in the process of hanging basket operation. These hazards may be caused by improper operation, aging equipment or other unforeseen factors. Therefore, the safety of the operation process has always been the focus of the industry.

[0003] At present, in the exterior wall decoration work of large-scale construction projects, it is usually necessary to use multiple hanging baskets for high-altitude work at the same time. Due to the different actual conditions of different hanging baskets (such as size, load capacity, working area, etc.), how to meet the safety monitoring needs of multiple hanging baskets at the same time according to the actual conditions of different hanging baskets to ensure the safety of all operations is an urgent problem to be solved. Summary of the invention

[0004] In order to meet the safety monitoring needs of multiple different hanging baskets and improve operation safety, the present application provides a method, system, equipment and medium for safety monitoring of high-altitude hanging basket operations.

[0005] In the first aspect, the present application provides a method for monitoring the safety of aerial hanging basket operations, which adopts the following technical solutions:

[0006] A method for monitoring the safety of aerial hanging basket operations, the monitoring method comprising:

[0007] Obtaining the operation status information of multiple suspended baskets to be monitored that are operating at high altitudes;

[0008] The operation status information is subjected to data preprocessing and status characteristic information is extracted; wherein the status characteristic information includes load information, operation height information and number of operators information;

[0009] Determine the corresponding risk threshold according to the parameter information of each suspended basket to be monitored; the risk threshold includes the upper limit of load, the upper limit of operating height and the upper limit of the number of operators;

[0010] Based on the risk threshold, a comprehensive risk level score of each suspended basket to be monitored is calculated according to the state characteristic information;

[0011] Based on a preset risk monitoring mapping table, a preset monitoring frequency of each suspended basket to be monitored is determined according to the comprehensive risk level score; wherein the preset risk monitoring mapping table includes a mapping relationship between multiple groups of comprehensive risk level score intervals and preset monitoring frequencies;

[0012] According to the preset monitoring frequency of each suspended basket to be monitored, each suspended basket to be monitored is monitored for abnormal behavior and tilt status;

[0013] In response to the abnormal monitoring results, an early warning signal is generated and sent to the management terminal.

[0014] By adopting the above technical scheme, when it is necessary to monitor multiple hanging baskets in high-altitude operations at the same time, the operation status information of each hanging basket is collected and processed in real time, and different risk thresholds are set according to the design parameters of different hanging baskets, including the upper limit of load, the upper limit of operation height and the upper limit of the number of operators, etc. These state characteristic information are compared with the preset risk threshold, and the comprehensive risk level score of each hanging basket is calculated by mathematical model, and then the abnormal monitoring frequency of each hanging basket is determined by referring to the preset risk monitoring mapping table, so as to ensure that high-risk hanging baskets are inspected more frequently to deal with abnormal risks in time, and low-risk hanging baskets are configured with a lower monitoring frequency to reduce the waste of system resources; and, by regularly performing abnormal behavior monitoring and tilt state monitoring, once the situation exceeding the threshold is found, the early warning mechanism is immediately triggered, and an early warning signal is sent to the management personnel so that corrective measures can be taken in time. This application combines risk assessment and dynamic monitoring strategies, dynamically adjusts the monitoring strategy according to the risk level of different hanging baskets, ensures that high-risk hanging baskets are inspected more frequently, and low-risk hanging baskets are inspected at a lower frequency to meet the safety monitoring needs of multiple hanging baskets, realizes effective management and control of risks in the hanging basket operation process, and improves the safety and efficiency of the entire system.

[0015] Optionally, based on the risk threshold, the step of calculating the comprehensive risk level score of each suspended basket to be monitored according to the state characteristic information includes:

[0016] Calculate the risk score of each state feature according to the ratio of the state feature information to the corresponding risk threshold;

[0017] Based on the preset weight of each status feature, the risk score of each feature is weighted averaged to calculate the comprehensive risk level score;

[0018] Among them, the calculation formula of the comprehensive risk level score RS is:

[0019]

[0020] In the above formula, W i is the preset weight of the i-th state feature, Ri is the risk score of the i-th state feature.

[0021] By adopting the above technical solution, the risk level of each feature is quantified, and the overall risk score is calculated using the weighted average method, so that the safety status of the hanging basket is effectively quantified and evaluated.

[0022] Optionally, the status characteristic information also includes maintenance record information, service life information and usage frequency information; the risk threshold also includes an upper limit on the maintenance interval duration, an upper limit on the service life and an upper limit on the usage frequency.

[0023] Optionally, the steps of monitoring abnormal behavior of each suspended basket to be monitored include:

[0024] Obtain image monitoring data of the suspended basket to be monitored and perform image preprocessing;

[0025] Performing target detection on the preprocessed image monitoring data, identifying the target person on the suspended basket to be monitored and marking the target frame, and locating the image area of ​​the person;

[0026] Extracting features according to the personnel image region, screening key feature vectors to form a feature subset; wherein the key feature vectors include personnel feature vectors and seat belt feature vectors;

[0027] Inputting the feature subset into a pre-trained abnormal behavior recognition model to obtain an abnormal behavior recognition result; wherein the abnormal behavior recognition result includes whether the wearing state of the seat belt of the personnel is abnormal and whether the operating behavior of the personnel is abnormal;

[0028] According to the abnormal behavior identification result, a normal or abnormal behavior monitoring result is output.

[0029] By adopting the above technical solution, advanced image processing technology and machine learning algorithms are used to realize automatic monitoring of the safety belt wearing status and operating behavior of people on the hanging basket. Through precise target detection and efficient feature extraction algorithms, combined with optimized abnormal behavior recognition models, abnormal behaviors can be identified quickly and accurately, and monitoring results can be output in a timely manner, effectively improving the safety management level of the construction site.

[0030] Optionally, the monitoring method further includes a training step of the abnormal behavior recognition model, and the training step includes:

[0031] Acquire a sample image data set and perform image preprocessing; wherein the sample image data set includes a plurality of historical sample image data of normal or abnormal behaviors and pre-labeled abnormal behavior labels;

[0032] Perform feature extraction on the sample image data set after data preprocessing, and screen key sample feature vectors to form a sample feature subset; wherein the key sample feature vectors include personnel sample feature vectors and seat belt sample feature vectors;

[0033] Dividing the sample feature subset into a training set, a validation set and a test set;

[0034] Inputting the training set into a pre-built neural network model for iterative training, optimizing the model parameters and calculating the cross entropy loss function of the model until the cross entropy loss function meets the preset conditions, thereby obtaining a trained abnormal behavior recognition model;

[0035] Validating the abnormal behavior recognition model based on the validation set, evaluating the model performance and adjusting the hyperparameters of the model according to the validation results;

[0036] The predictive ability of the adjusted abnormal behavior recognition model is tested based on the test set.

[0037] By adopting the above technical solution, a large amount of hanging basket sample image data is collected, and preprocessed, feature extracted and model trained. Finally, a model is obtained that can accurately identify whether the safety belt wearing status and operating behavior of people in the hanging basket are abnormal. It can effectively monitor abnormal behavior on the hanging basket in practical applications and improve the safety of the construction site.

[0038] Optionally, the step of monitoring the tilt status of each suspended basket to be monitored includes:

[0039] Obtain the inclination angle data of the suspended basket to be monitored;

[0040] Determine whether the tilt angle exceeds a preset threshold value according to the tilt angle data; if so, output an abnormal tilt state monitoring result;

[0041] If not, determining whether the tilt angle change value exceeds a preset change threshold value based on the tilt angle data;

[0042] If it exceeds, the abnormal tilt state monitoring result is output; if it does not exceed, the normal tilt state monitoring result is output.

[0043] By adopting the above technical solution, the inclination angle of the hanging basket is monitored in real time, and combined with the threshold judgment of the inclination angle and the inclination angle change value, safety accidents caused by the inclination of the hanging basket can be effectively prevented, which not only improves the safety of the hanging basket operation, but also reduces potential risks through a timely early warning mechanism, thereby ensuring the life safety and property safety of the staff.

[0044] In the second aspect, the present application provides a high-altitude hanging basket operation safety monitoring system, which adopts the following technical solution:

[0045] A high-altitude hanging basket operation safety monitoring system, the monitoring system comprising:

[0046] An acquisition module is used to acquire the operation status information of multiple suspended baskets to be monitored in high-altitude operations;

[0047] A data processing module, used for preprocessing the operation status information and extracting status characteristic information; wherein the status characteristic information includes load information, operation height information and number of operators information;

[0048] A risk threshold determination module is used to determine the corresponding risk threshold according to the parameter information of each suspended basket to be monitored; the risk threshold includes the upper limit of load, the upper limit of operating height and the upper limit of the number of operators;

[0049] A comprehensive risk level score calculation module, used to calculate the comprehensive risk level score of each suspended basket to be monitored based on the risk threshold and the state characteristic information;

[0050] A preset monitoring frequency determination module is used to determine the preset monitoring frequency of each suspended basket to be monitored according to the comprehensive risk level score based on a preset risk monitoring mapping table; wherein the preset risk monitoring mapping table includes a mapping relationship between multiple groups of comprehensive risk level score intervals and preset monitoring frequencies;

[0051] A monitoring module, used for performing abnormal behavior monitoring and tilt state monitoring on each suspended basket to be monitored according to a preset monitoring frequency of each suspended basket to be monitored;

[0052] The abnormal warning module is used to generate a warning signal in response to the abnormal monitoring result and send it to the management terminal.

[0053] Optionally, the monitoring module includes an abnormal behavior monitoring module, and the abnormal behavior monitoring module includes:

[0054] An image monitoring unit, used for acquiring image monitoring data of the hanging basket to be monitored;

[0055] A preprocessing unit, used for performing image preprocessing on the image monitoring data;

[0056] A target detection unit is used to perform target detection on the preprocessed image monitoring data, identify the target person on the suspended basket to be monitored, mark the target frame, and locate the image area of ​​the person;

[0057] A feature extraction unit, used to extract features according to the personnel image region, and screen key feature vectors to form a feature subset; wherein the key feature vectors include personnel feature vectors and seat belt feature vectors;

[0058] An abnormal behavior recognition unit is used to input the feature subset into a pre-trained abnormal behavior recognition model to obtain an abnormal behavior recognition result; wherein the abnormal behavior recognition result includes whether the wearing state of the personnel's seat belt is abnormal and whether the personnel's operating behavior is abnormal;

[0059] The behavior monitoring result output unit is used to output normal or abnormal behavior monitoring results according to the abnormal behavior identification results.

[0060] In a third aspect, the present application provides a computer device, which adopts the following technical solution:

[0061] A computer device comprises a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method as described in the first aspect.

[0062] In a fourth aspect, the present application provides a computer-readable storage medium, which adopts the following technical solution:

[0063] A computer-readable storage medium stores a computer program that can be loaded by a processor and execute any one of the methods in the first aspect.

[0064] To summarize, the present application includes at least one of the following beneficial technical effects: the present application combines risk assessment and dynamic monitoring strategies to dynamically adjust the monitoring strategies according to the risk levels of different hanging baskets, thereby ensuring more frequent inspections of high-risk hanging baskets and less frequent inspections of low-risk hanging baskets to meet the safety monitoring needs of multiple hanging baskets, thereby achieving effective management and control of risks during the hanging basket operation process and improving the safety and efficiency of the entire system. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a first flow chart of a method for safety monitoring of aerial hanging basket operations according to one of the embodiments of the present application.

[0066] Figure 2 It is a second flow chart of a method for safety monitoring of aerial hanging basket operations according to one of the embodiments of the present application.

[0067] Figure 3 It is a third flow chart of a method for safety monitoring of aerial hanging basket operations in one of the embodiments of the present application.

[0068] Figure 4 It is a fourth process diagram of a method for safety monitoring of aerial hanging basket operations according to one of the embodiments of the present application.

[0069] Figure 5 It is the fifth flow chart of a method for safety monitoring of aerial hanging basket operations according to one of the embodiments of the present application. DETAILED DESCRIPTION

[0070] In order to make the purpose, technical solutions and advantages of this application more clear, the following Figure 1-5 It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0071] With the acceleration of urbanization, the demand for construction and maintenance of high-rise buildings is increasing. As an important construction platform, aerial suspended baskets play an irreplaceable role in building exterior wall decoration, cleaning, and maintenance. However, aerial work itself has a high safety risk. Safety accidents often occur during aerial suspended basket operations due to overloading, improper operation of personnel, or equipment failure. Therefore, it is particularly important to develop a high-altitude suspended basket operation safety monitoring system that can effectively monitor the operating status of the suspended basket, promptly detect safety hazards, and take corresponding measures.

[0072] This application is aimed at this demand. Through real-time monitoring and data analysis of the operating status of the hanging basket, effective management and control of the risks of high-altitude operations are achieved.

[0073] The embodiment of the present application discloses a method for safety monitoring of aerial hanging basket operations.

[0074] Reference Figure 1 , a high-altitude hanging basket operation safety monitoring method, the monitoring method comprises:

[0075] Step S101, obtaining operation status information of multiple suspended baskets to be monitored in high-altitude operations;

[0076] In some embodiments, real-time operation status information can be collected by sensors installed on each hanging basket. These sensors include but are not limited to weight sensors (for detecting the load on the hanging basket), position sensors (for detecting the position and height of the hanging basket) and personnel counting sensors (for counting the number of workers in the hanging basket). Through wireless communication technology (such as Wi-Fi, Bluetooth, LoRa, etc.), these sensors transmit data to the central monitoring system to obtain accurate hanging basket operation status information, laying the foundation for subsequent data processing and risk assessment.

[0077] Step S102, preprocessing the job status information and extracting status feature information;

[0078] Among them, the state characteristic information includes load information, operation height information and number of operators information;

[0079] Specifically, the data preprocessing steps include data cleaning, denoising, standardization and other operations to remove invalid or abnormal data and improve data quality; at the same time, key status feature information is extracted, which can be used to evaluate the operational risks of different hanging baskets.

[0080] Step S103, determining a corresponding risk threshold according to parameter information of each suspended basket to be monitored;

[0081] Among them, the risk thresholds include the upper limit of load, the upper limit of operating height and the upper limit of the number of operators;

[0082] Specifically, the risk threshold is set according to the design parameters of different suspended baskets, such as the model, size, maximum load-bearing capacity, operating range, etc. of the suspended basket. Different risk thresholds are pre-set according to different suspended baskets, such as the maximum allowable load, maximum operating height and maximum number of operators, and stored in the database for subsequent use to ensure that the operation of the suspended basket meets the design requirements, avoid overload operation, and reduce the occurrence of safety accidents.

[0083] Step S104, based on the risk threshold, calculate the comprehensive risk level score of each suspended basket to be monitored according to the state characteristic information;

[0084] Specifically, by comparing the status characteristic information with the risk threshold, a comprehensive risk level score of the hanging basket can be calculated, for example, by using a weighted average or other mathematical model to quantify the risk level so that the safety status of the hanging basket can be intuitively evaluated and compared.

[0085] Step S105, based on the preset risk monitoring mapping table, determine the preset monitoring frequency of each suspended basket to be monitored according to the comprehensive risk level score;

[0086] Among them, the preset risk monitoring mapping table includes mapping relationships between multiple groups of comprehensive risk level scoring intervals and preset monitoring frequencies;

[0087] Specifically, the preset risk monitoring mapping table defines the monitoring frequencies corresponding to different risk levels; for example, a hanging basket with a low risk level may only require a lower monitoring frequency, while a hanging basket with a high risk level may require a higher monitoring frequency.

[0088] For example, four risk levels can be predefined: low risk, medium-low risk, medium-high risk and high risk. Each risk level corresponds to a different scoring interval and preset monitoring frequency. The preset monitoring frequency specifies the corresponding monitoring interval. For example, if the comprehensive risk level score of the hanging basket is 45, which belongs to the medium-high risk level, the preset monitoring frequency can be configured to be once every 30 minutes.

[0089] It is understandable that by implementing dynamic monitoring and rationally allocating resources, high-risk hanging baskets can be inspected more frequently and potential safety hazards can be reduced.

[0090] Step S106, performing abnormal behavior monitoring and tilt state monitoring on each suspended basket to be monitored according to the preset monitoring frequency of each suspended basket to be monitored;

[0091] Among them, abnormal behavior monitoring can include personnel safety belt wearing status monitoring and personnel operation behavior monitoring, and tilt status monitoring is to monitor the tilt status trend of the hanging basket;

[0092] For example, if the preset monitoring frequency is once every 30 minutes, abnormal behavior monitoring and tilt status monitoring can be performed every half an hour based on the video surveillance system to ensure that abnormal situations are discovered and handled in a timely manner.

[0093] Step S107: generating an early warning signal in response to the abnormal monitoring result and sending it to the management terminal.

[0094] Among them, the abnormal monitoring results include abnormal behavior monitoring results and abnormal tilt state monitoring results;

[0095] Specifically, through sensor data and preset rules, abnormal situations that may affect the stability of the hanging basket can be detected and handled in a timely manner. Once an abnormal situation is detected, the system will automatically trigger the early warning mechanism and send an alarm to the management personnel via SMS, email or dedicated applications, further improving safety.

[0096] In the above implementation mode, when it is necessary to monitor multiple hanging baskets in high-altitude operations at the same time, the operation status information of each hanging basket is collected and processed in real time, and different risk thresholds are set according to the design parameters of different hanging baskets, including load upper limit, operation height upper limit and operation number upper limit, etc. These status characteristic information are compared with the preset risk threshold, and the comprehensive risk level score of each hanging basket is calculated through a mathematical model, and then the abnormal monitoring frequency of each hanging basket is determined by referring to the preset risk monitoring mapping table, so as to ensure more frequent inspections of high-risk hanging baskets and respond to abnormal risks in time, and at the same time, a lower monitoring frequency is configured for low-risk hanging baskets to reduce the waste of system resources; and, by regularly performing abnormal behavior monitoring and tilt status monitoring, once a situation exceeding the threshold is found, the early warning mechanism is immediately triggered, and a warning signal is sent to the management personnel so that corrective measures can be taken in time.

[0097] Reference Figure 2 As an implementation method of step S104, based on the risk threshold, the step of calculating the comprehensive risk level score of each suspended basket to be monitored according to the state characteristic information includes:

[0098] Step S201, calculating the risk score of each state feature according to the ratio of the state feature information to the corresponding risk threshold;

[0099] Among them, according to the ratio of the state feature information to the corresponding risk threshold, the corresponding risk score can be obtained to indicate the risk level of the state feature;

[0100] Step S202, based on the preset weight of each state feature, the risk score of each feature is weighted averaged to calculate a comprehensive risk level score;

[0101] Specifically, the setting of each status feature should take into account the importance and impact of different features on the overall risk of the suspended basket. During actual configuration, the importance of different features can be evaluated and determined based on the experience and expertise of professionals or based on actual historical data to ensure the rationality of the preset weight setting.

[0102] Among them, the calculation formula of the comprehensive risk level score RS is:

[0103]

[0104] In the above formula, W i is the preset weight of the i-th state feature, R i is the risk score of the i-th state feature.

[0105] In the above implementation, the risk level of each feature is quantified, and the overall risk score is calculated using a weighted average method, so that the safety status of the hanging basket is effectively quantified and evaluated.

[0106] In some embodiments, the status characteristic information also includes maintenance record information, service life information, and usage frequency information, and the risk threshold also includes an upper limit on the maintenance interval, an upper limit on the service life, and an upper limit on the usage frequency.

[0107] Specifically, for example, the time interval of the most recent maintenance can be determined based on the maintenance record information, and the corresponding risk score can be determined by judging whether the time interval exceeds the upper limit of the maintenance interval. Similarly, the corresponding risk score can be determined by comparing the service life information and the frequency of use information with the corresponding upper limit of the service life and the upper limit of the frequency of use.

[0108] Reference Figure 3 As an implementation method of step S106, the step of monitoring abnormal behavior of each suspended basket to be monitored includes:

[0109] Step S301, obtaining image monitoring data of the hanging basket to be monitored and performing image preprocessing;

[0110] Among them, the image monitoring data of the hanging basket can be captured based on the video monitoring system, and the acquired images can be preprocessed, such as adjusting the image size, converting it into a grayscale image, performing histogram equalization, etc., for subsequent target detection processing.

[0111] Step S302, performing target detection on the preprocessed image monitoring data, identifying the target person on the suspended basket to be monitored and marking the target frame, and locating the image area of ​​the person;

[0112] For example, a target detection algorithm (such as YOLOv5 or Faster R-CNN) can be used to accurately identify all persons in the hanging basket, and the identified persons are marked with target frames to determine the specific location of the persons, reduce the false detection rate and missed detection rate, and improve the detection accuracy.

[0113] Step S303, extracting features based on the personnel image area, and screening key feature vectors to form a feature subset;

[0114] Among them, the key feature vectors include personnel feature vectors and seat belt feature vectors;

[0115] In some embodiments, a convolutional neural network (CNN) can be used to extract feature vectors related to personnel and their seat belt wearing in the personnel image area, and a feature selection algorithm (such as recursive feature elimination (RFE) or an entropy-based feature selection method) can be used to screen key features, and the features that best characterize the personnel's seat belt wearing status and personnel's operating behavior are selected to form a feature subset, so as to reduce computational complexity and reduce redundant features.

[0116] Step S304, inputting the feature subset into a pre-trained abnormal behavior recognition model to obtain an abnormal behavior recognition result;

[0117] Among them, the abnormal behavior recognition results include whether the personnel's seat belt wearing status is abnormal and whether the personnel's operating behavior is abnormal;

[0118] Exemplarily, the pre-trained abnormal behavior recognition model can adopt the architecture of convolutional neural network (CNN) combined with bidirectional long short-term memory network (BiLSTM). By combining the powerful feature extraction capability of CNN and the ability of BiLSTM to capture time series features, it can effectively monitor abnormal behavior on the hanging basket in practical applications and realize accurate identification of personnel's safety belt wearing status and operating behavior.

[0119] Step S305: outputting normal or abnormal behavior monitoring results according to the abnormal behavior recognition results.

[0120] Among them, the safety belt wearing status and behavior of the personnel on the monitored hanging basket are judged according to the abnormal behavior recognition results. If everything is normal, the normal behavior monitoring result is output; if there is an abnormality, the abnormal behavior monitoring result is output.

[0121] In the above implementation, advanced image processing technology and machine learning algorithms are used to realize automatic monitoring of the safety belt wearing status and operating behavior of people on the hanging basket. Through precise target detection and efficient feature extraction algorithms, combined with optimized abnormal behavior recognition models, abnormal behaviors can be identified quickly and accurately, and monitoring results can be output in a timely manner, effectively improving the safety management level of the construction site.

[0122] Reference Figure 4 As an implementation of the abnormal behavior recognition model in step S304, the training steps of the abnormal behavior recognition model include:

[0123] Step S401, obtaining a sample image data set and performing image preprocessing;

[0124] Exemplarily, the sample image dataset includes a plurality of historical sample image data of normal or abnormal behaviors and pre-labeled abnormal behavior labels;

[0125] Specifically, a large amount of hanging basket image data is collected. These images contain normal behaviors (wearing seat belts correctly, normal operation, etc.) and abnormal behaviors (not wearing seat belts, irregular operation, etc.). By annotating the videos, the seat belt wearing status of the personnel and whether the operating behavior is normal are marked, ensuring that the model can learn different types of behavior patterns from rich samples.

[0126] Among them, the preprocessing steps may include adjusting the image size, converting to grayscale, performing histogram equalization, etc., to ensure the consistency and high quality of the input data and reduce the impact of noise on model training.

[0127] Step S402, extracting features from the sample image data set after data preprocessing, and screening key sample feature vectors to form a sample feature subset; wherein the key sample feature vectors include personnel sample feature vectors and seat belt sample feature vectors;

[0128] Among them, the feature extraction step may include using a convolutional neural network (CNN) to extract sample feature vectors related to people and their seat belt wearing in the sample image data set, and using a feature selection algorithm to screen and select the features that best characterize the people's seat belt wearing status and people's operating behavior to form a sample feature subset, thereby improving the model training speed and accuracy.

[0129] Step S403, dividing the sample feature subset into a training set, a validation set and a test set;

[0130] Exemplarily, the division ratios may be 70%, 15%, and 15%, respectively, to ensure balanced data distribution, thereby providing sufficient training data for model training and providing an independent data set for model verification and testing.

[0131] Step S404, inputting the training set into a pre-built neural network model for iterative training, optimizing the model parameters and calculating the cross entropy loss function of the model until the cross entropy loss function meets the preset conditions, thereby obtaining a trained abnormal behavior recognition model;

[0132] In one of the embodiments of the present application, the pre-built neural network model may adopt a convolutional neural network (CNN) architecture. The hierarchical structure of the CNN model enables the model to learn multi-level feature representations from simple to complex, which helps to improve the generalization ability of the model, making it more efficient and accurate when processing abnormal behavior monitoring tasks on the hanging basket.

[0133] Among them, the network architecture of the abnormal behavior recognition model based on convolutional neural network (CNN) includes an input layer, a convolution layer, a maximum pooling layer, a fully connected layer, a Dropout layer and an output layer; among them, there are multiple convolution layers, and the number of convolution kernels in each layer gradually increases, which is used to extract local features in the image; the maximum pooling layer is used to reduce the spatial size of the feature map; the fully connected layer can use the ReLU activation function for further processing of the features; the Dropout layer is used to reduce overfitting, and the Dropout ratio can be set to 0.5; the output layer can use the Softmax activation function to output the probability distribution of normal behavior categories and abnormal behavior categories.

[0134] Specifically, during the model training process, the back-propagation algorithm can be used to continuously optimize the model parameters and improve the model performance. At the same time, the cross-entropy loss function can be used to measure the difference between the model prediction and the true label to guide the model optimization and continuously improve the generalization ability of the model.

[0135] Step S405, verifying the abnormal behavior recognition model based on the verification set, evaluating the model performance and adjusting the model's hyperparameters according to the verification results;

[0136] Exemplarily, the model is verified using validation set data to evaluate model performance indicators, such as accuracy, precision, recall, F1 score, etc. The model's hyperparameters (such as learning rate, batch size, regularization parameters, etc.) are continuously adjusted based on the validation results to ensure that the model performs well on unknown data.

[0137] Step S406: testing the prediction capability of the adjusted abnormal behavior recognition model based on the test set.

[0138] Among them, the adjusted model is finally tested on an independent test set to evaluate the model's predictive ability and generalization performance. The final test provides performance indicators of the model in practical applications and ensures the reliability and effectiveness of the model.

[0139] In the above implementation, a large amount of hanging basket sample image data is collected, and preprocessing, feature extraction and model training are performed on it, and finally a model is obtained that can accurately identify whether the safety belt wearing status and operating behavior of people in the hanging basket are abnormal. It can effectively monitor abnormal behavior on the hanging basket in practical applications and improve the safety of the construction site.

[0140] Reference Figure 5 As an implementation method of step S104, the step of monitoring the tilt state of each suspended basket to be monitored includes:

[0141] Step S501, obtaining the inclination angle data of the suspended basket to be monitored;

[0142] In some embodiments, a tilt sensor, such as a MEMS accelerometer or a gyroscope, can be installed at key locations of the monitored basket to monitor the tilt state of the basket. The sensor collects acceleration and angular velocity data that can reflect the direction of the basket relative to gravity.

[0143] Step S502, judging whether the tilt angle exceeds a preset threshold value according to the tilt angle data; if so, jumping to step S503; if not, jumping to step S504;

[0144] Step S503, outputting abnormal tilt state monitoring results;

[0145] Among them, a preset threshold of safe tilt angle is pre-set according to the design specifications and working environment of the hanging basket. By comparing the current tilt angle with the preset threshold, if it exceeds the preset threshold, it is judged as an abnormal tilt state. The system will automatically trigger the alarm mechanism and record key information such as the timestamp of the abnormal situation, the tilt angle value, etc., which is convenient for subsequent analysis, and send a warning signal to the management terminal at the same time.

[0146] Step S504, judging whether the tilt angle change value exceeds a preset change threshold value according to the tilt angle data; if it exceeds, jumping to step S505; if it does not exceed, jumping to step S506;

[0147] Specifically, based on the setting of the preset monitoring frequency, the difference between the current tilt angle and the tilt angle obtained by the previous monitoring, that is, the tilt angle change value, is calculated, and a change threshold is pre-set. The change threshold is used to determine whether the tilt angle change is significant. If the tilt angle change value exceeds the preset change threshold, it can also be determined as an abnormal tilt state.

[0148] It is understandable that by monitoring the changing trend of the tilt angle, possible problems such as unstable basket structure or uneven load distribution can be discovered early, and accidents caused by rapid changes in the tilt angle can be prevented.

[0149] Step S505, outputting abnormal tilt state monitoring results;

[0150] Step S506, outputting the normal tilt state monitoring result.

[0151] Specifically, if the tilt angle change value exceeds the preset change threshold, the abnormal tilt state monitoring result is output and the corresponding alarm mechanism is triggered; if the tilt angle change value does not exceed the preset change threshold, the normal tilt state monitoring result is output.

[0152] In the above implementation mode, the inclination angle of the hanging basket is monitored in real time, and combined with the threshold judgment of the inclination angle and the inclination angle change value, it is possible to effectively prevent safety accidents caused by the inclination of the hanging basket, which not only improves the safety of the hanging basket operation, but also reduces potential risks through a timely early warning mechanism, thereby ensuring the life safety and property safety of the staff.

[0153] In actual application, the technical solution of the present application combines risk assessment and dynamic monitoring strategies to dynamically adjust the monitoring strategy according to the risk levels of different hanging baskets, thereby ensuring more frequent inspections of high-risk hanging baskets and less frequent inspections of low-risk hanging baskets, so as to meet the safety monitoring needs of multiple hanging baskets, realize effective management and control of risks in the hanging basket operation process, and improve the safety and efficiency of the entire system.

[0154] The embodiment of the present application also discloses a high-altitude hanging basket operation safety monitoring system.

[0155] A high-altitude hanging basket operation safety monitoring system, the monitoring system comprises:

[0156] An acquisition module is used to acquire the operation status information of multiple suspended baskets to be monitored in high-altitude operations;

[0157] A data processing module is used to pre-process the operation status information and extract status characteristic information; wherein the status characteristic information includes load information, operation height information and number of operators information;

[0158] A risk threshold determination module is used to determine the corresponding risk threshold according to the parameter information of each suspended basket to be monitored; the risk threshold includes the upper limit of load, the upper limit of operating height and the upper limit of the number of operators;

[0159] A comprehensive risk level score calculation module is used to calculate the comprehensive risk level score of each suspended basket to be monitored based on the risk threshold and the state characteristic information;

[0160] A preset monitoring frequency determination module is used to determine the preset monitoring frequency of each suspended basket to be monitored according to the comprehensive risk level score based on the preset risk monitoring mapping table; wherein the preset risk monitoring mapping table includes a mapping relationship between multiple groups of comprehensive risk level score intervals and preset monitoring frequencies;

[0161] A monitoring module, used for performing abnormal behavior monitoring and tilt state monitoring on each suspended basket to be monitored according to a preset monitoring frequency of each suspended basket to be monitored;

[0162] The abnormal warning module is used to generate a warning signal in response to the abnormal monitoring result and send it to the management terminal.

[0163] In the above implementation mode, the operating status information of the hanging basket is collected in real time by the acquisition module, and the data processing module is used to perform data preprocessing and extract status feature information, thereby accurately reflecting the actual working condition of the hanging basket; by setting reasonable risk thresholds and calculating comprehensive risk level scores, the system can scientifically evaluate the safety risk level of each hanging basket, and dynamically adjust the monitoring frequency accordingly, thereby realizing effective allocation of resources; the coordinated use of the monitoring module and the abnormal warning module ensures timely monitoring and early warning of abnormal behavior and tilt status, greatly improving the safety of hanging basket operations and reducing the probability of accidents; in addition, the setting of the preset monitoring frequency also enables the system to avoid unnecessary frequent monitoring while ensuring safety, saving manpower and material resources and improving work efficiency.

[0164] To sum up, the technical solution of the present application can significantly improve the safety management level of high-altitude hanging basket operations in practical applications, reduce the occurrence rate of safety accidents, and provide strong protection for the life safety of construction workers.

[0165] As an implementation of the monitoring module, the monitoring module includes an abnormal behavior monitoring module, and the abnormal behavior monitoring module includes:

[0166] An image monitoring unit, used for acquiring image monitoring data of the hanging basket to be monitored;

[0167] A preprocessing unit, used for performing image preprocessing on the image monitoring data;

[0168] The target detection unit is used to perform target detection on the pre-processed image monitoring data, identify the target person on the suspended basket to be monitored, mark the target frame, and locate the image area of ​​the person;

[0169] A feature extraction unit is used to extract features according to the personnel image area and screen key feature vectors to form a feature subset; wherein the key feature vectors include personnel feature vectors and seat belt feature vectors;

[0170] An abnormal behavior recognition unit is used to input a feature subset into a pre-trained abnormal behavior recognition model to obtain an abnormal behavior recognition result; wherein the abnormal behavior recognition result includes whether the wearing state of the seat belt of the personnel is abnormal and whether the operating behavior of the personnel is abnormal;

[0171] The behavior monitoring result output unit is used to output normal or abnormal behavior monitoring results according to the abnormal behavior recognition results.

[0172] In the above implementation, advanced image processing technology and machine learning algorithms are used to automatically monitor the safety belt wearing status and operating behavior of people on the hanging basket, thereby improving the safety management level of the construction site.

[0173] A high-altitude hanging basket operation safety monitoring system in an embodiment of the present application can implement any of the above-mentioned high-altitude hanging basket operation safety monitoring methods, and the specific working process of each module in the high-altitude hanging basket operation safety monitoring system can refer to the corresponding process in the above-mentioned method embodiment.

[0174] In the several embodiments provided in this application, it should be understood that the provided methods and systems can be implemented in other ways. For example, the system embodiments described above are only illustrative; for example, the division of a module is only a logical function division, and there may be other division methods in actual implementation, such as multiple modules can be combined or integrated into another system, or some features can be ignored or not executed.

[0175] The embodiment of the present application also discloses a computer device.

[0176] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, a method for monitoring the safety of aerial hanging basket operations as described above is implemented.

[0177] The embodiment of the present application also discloses a computer-readable storage medium.

[0178] A computer-readable storage medium stores a computer program that can be loaded by a processor and execute any one of the above-mentioned methods for safety monitoring of aerial hanging basket operations.

[0179] Among them, computer-readable storage media can be any tangible medium that contains or stores a program, which can be used by or in conjunction with an instruction execution system, apparatus or device; the program code contained on the computer-readable medium can be transmitted using any appropriate medium, including but not limited to wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0180] It should be noted that in the above embodiments, the description of each embodiment has different emphases, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0181] The above are all preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. Any feature disclosed in this specification (including the abstract and drawings), unless otherwise stated, can be replaced by other equivalent or alternative features with similar purposes. That is, unless otherwise stated, each feature is only an example of a series of equivalent or similar features.

Claims

1. A method for monitoring the safety of aerial hanging basket operations, characterized in that: The monitoring method comprises: Obtaining the operation status information of multiple suspended baskets to be monitored that are operating at high altitudes; The operation status information is subjected to data preprocessing and status characteristic information is extracted; wherein the status characteristic information includes load information, operation height information and number of operators information; Determine the corresponding risk threshold according to the parameter information of each suspended basket to be monitored; the risk threshold includes the upper limit of load, the upper limit of operating height and the upper limit of the number of operators; Based on the risk threshold, a comprehensive risk level score of each suspended basket to be monitored is calculated according to the state characteristic information; Based on a preset risk monitoring mapping table, a preset monitoring frequency of each suspended basket to be monitored is determined according to the comprehensive risk level score; wherein the preset risk monitoring mapping table includes a mapping relationship between multiple groups of comprehensive risk level score intervals and preset monitoring frequencies; According to the preset monitoring frequency of each suspended basket to be monitored, each suspended basket to be monitored is monitored for abnormal behavior and tilt status; In response to the abnormal monitoring results, an early warning signal is generated and sent to the management terminal.

2. A method for monitoring safety of aerial hanging basket operations according to claim 1, characterized in that: Based on the risk threshold, the step of calculating the comprehensive risk level score of each suspended basket to be monitored according to the state characteristic information includes: Calculate the risk score of each state feature according to the ratio of the state feature information to the corresponding risk threshold; Based on the preset weight of each status feature, the risk score of each feature is weighted averaged to calculate the comprehensive risk level score; Among them, the calculation formula of the comprehensive risk level score RS is: In the above formula, W i is the preset weight of the i-th state feature, R i is the risk score of the i-th state feature.

3. A method for monitoring safety of aerial hanging basket operations according to claim 2, characterized in that: The status characteristic information also includes maintenance record information, service life information and usage frequency information; the risk threshold also includes the upper limit of the maintenance interval, the upper limit of the service life and the upper limit of the usage frequency.

4. A method for monitoring safety of aerial hanging basket operations according to claim 1, characterized in that: The steps for monitoring abnormal behavior of each suspended basket to be monitored include: Obtain image monitoring data of the suspended basket to be monitored and perform image preprocessing; Performing target detection on the preprocessed image monitoring data, identifying the target person on the suspended basket to be monitored and marking the target frame, and locating the image area of ​​the person; Extracting features according to the personnel image region, screening key feature vectors to form a feature subset; wherein the key feature vectors include personnel feature vectors and seat belt feature vectors; Inputting the feature subset into a pre-trained abnormal behavior recognition model to obtain an abnormal behavior recognition result; wherein the abnormal behavior recognition result includes whether the wearing state of the seat belt of the personnel is abnormal and whether the operating behavior of the personnel is abnormal; According to the abnormal behavior identification result, a normal or abnormal behavior monitoring result is output.

5. A method for monitoring safety of aerial hanging basket operations according to claim 4, characterized in that: The monitoring method further comprises a training step of the abnormal behavior recognition model, wherein the training step comprises: Acquire a sample image data set and perform image preprocessing; wherein the sample image data set includes a plurality of historical sample image data of normal or abnormal behaviors and pre-labeled abnormal behavior labels; Perform feature extraction on the sample image data set after data preprocessing, and screen key sample feature vectors to form a sample feature subset; wherein the key sample feature vectors include personnel sample feature vectors and seat belt sample feature vectors; Dividing the sample feature subset into a training set, a validation set and a test set; Inputting the training set into a pre-built neural network model for iterative training, optimizing the model parameters and calculating the cross entropy loss function of the model until the cross entropy loss function meets the preset conditions, thereby obtaining a trained abnormal behavior recognition model; Validating the abnormal behavior recognition model based on the validation set, evaluating the model performance and adjusting the hyperparameters of the model according to the validation results; The predictive ability of the adjusted abnormal behavior recognition model is tested based on the test set.

6. A method for monitoring safety of aerial hanging basket operations according to claim 1, characterized in that: The steps of monitoring the tilt status of each suspended basket to be monitored include: Obtain the inclination angle data of the suspended basket to be monitored; Determine whether the tilt angle exceeds a preset threshold value according to the tilt angle data; if so, output an abnormal tilt state monitoring result; If not, determining whether the tilt angle change value exceeds a preset change threshold value based on the tilt angle data; If it exceeds, the abnormal tilt state monitoring result is output; if it does not exceed, the normal tilt state monitoring result is output.

7. A high-altitude hanging basket operation safety monitoring system, characterized in that: The monitoring system comprises: An acquisition module is used to acquire the operation status information of multiple suspended baskets to be monitored in high-altitude operations; A data processing module, used for preprocessing the operation status information and extracting status characteristic information; wherein the status characteristic information includes load information, operation height information and number of operators information; A risk threshold determination module is used to determine the corresponding risk threshold according to the parameter information of each suspended basket to be monitored; the risk threshold includes the upper limit of load, the upper limit of operating height and the upper limit of the number of operators; A comprehensive risk level score calculation module, used to calculate the comprehensive risk level score of each suspended basket to be monitored based on the risk threshold and the state characteristic information; A preset monitoring frequency determination module is used to determine the preset monitoring frequency of each suspended basket to be monitored according to the comprehensive risk level score based on a preset risk monitoring mapping table; wherein the preset risk monitoring mapping table includes a mapping relationship between multiple groups of comprehensive risk level score intervals and preset monitoring frequencies; A monitoring module, used for performing abnormal behavior monitoring and tilt state monitoring on each suspended basket to be monitored according to a preset monitoring frequency of each suspended basket to be monitored; The abnormal warning module is used to generate a warning signal in response to the abnormal monitoring result and send it to the management terminal.

8. A high-altitude hanging basket operation safety monitoring system according to claim 7, characterized in that: The monitoring module includes an abnormal behavior monitoring module, and the abnormal behavior monitoring module includes: An image monitoring unit, used to obtain image monitoring data of the hanging basket to be monitored; A preprocessing unit, used for performing image preprocessing on the image monitoring data; A target detection unit is used to perform target detection on the preprocessed image monitoring data, identify the target person on the suspended basket to be monitored, mark the target frame, and locate the image area of ​​the person; A feature extraction unit, used to extract features according to the personnel image region, and screen key feature vectors to form a feature subset; wherein the key feature vectors include personnel feature vectors and seat belt feature vectors; An abnormal behavior recognition unit is used to input the feature subset into a pre-trained abnormal behavior recognition model to obtain an abnormal behavior recognition result; wherein the abnormal behavior recognition result includes whether the wearing state of the personnel's seat belt is abnormal and whether the personnel's operating behavior is abnormal; The behavior monitoring result output unit is used to output normal or abnormal behavior monitoring results according to the abnormal behavior identification results.

9. A computer device, characterized in that: The method comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method according to any one of claims 1 to 6 when executing the computer program.

10. A computer-readable storage medium, characterized in that: A computer program is stored which can be loaded by a processor and execute the method according to any one of claims 1 to 6.

Images