Factory crane steel member hoisting safety monitoring and analysis method based on deep learning algorithm

Abstract

The invention discloses a factory crane steel member hoisting safety monitoring and analysis method based on a deep learning algorithm, and the method obtains and analyzes field data through a visual perception technology, and specifically comprises the following steps: (1) data collection; (2) feature extraction; (3) safety risk assessment; (4) real-time monitoring and early warning; and (5) data analysis and optimization. The system has high-automation and intelligent safety monitoring capability, the operation safety is remarkably improved, the industrial injury risk is reduced, and the production efficiency and the working quality are improved through continuous learning and optimization. The system is easy to implement, high in cost effectiveness and suitable for factory environments of various scales and complexity, and has important application value for promoting safety production of the modern manufacturing industry.

Description

Technical Field

[0001] The present invention relates to the field of industrial automation, and in particular to a method for monitoring and analyzing the safety of hoisting steel components of a factory crane based on a deep learning algorithm. Background Art

[0002] The traditional factory crane system is mainly composed of a control system, a drive system and a steel frame. The control system controls the speed and tension of the crane; the drive system is powered by a motor or a hydraulic pump to move the steel frame; and the steel frame is the main part of the load-bearing components and equipment. During the factory crane hoisting process, safety hazards may occur due to various factors (such as equipment failure or human error), such as crane tipping over and component falling. If these hazards are not discovered and handled in time, they may cause injuries to personnel or damage to components.

[0003] In order to ensure safety and improve work efficiency, the safety of the factory driving system needs to be analyzed and evaluated. Current evaluation methods include qualitative judgment based on empirical rules and quantitative calculations through experimental simulation. However, these methods have their limitations: the former relies on subjective judgment and lacks accuracy; the latter requires a lot of time and cost investment. In addition, large errors may occur when predicting specific environmental conditions, and the response speed to complex situations is slow, which can easily miss the best time for intervention and lead to greater losses. Summary of the invention

[0004] In order to solve the technical problems existing in the above-mentioned prior art, the present invention provides a deep learning-based factory crane steel structure hoisting safety monitoring and analysis method. The method introduces visual technology into the safety control and protection process of steel structure hoisting, and uses images as a carrier to judge and evaluate the workers' movements and behaviors.

[0005] To achieve the above purpose, the present invention adopts the following technical solution: a method for safety monitoring and analysis of steel component hoisting in factory cranes based on deep learning, which introduces visual technology into the safety control and protection process of steel component hoisting, and uses images as carriers to judge and evaluate the actions and behaviors of workers. The structure and composition of the present invention include the following:

[0006] 1. Data acquisition module: It consists of high-definition cameras and sensors, which are used to collect video data and environmental parameters (such as humidity, temperature, brightness, noise, etc.) of the hoisting area. The cameras are installed at key locations in the hoisting area to ensure that the entire hoisting process can be fully covered. The sensors are used to monitor environmental parameters in real time and provide data support in more dimensions.

[0007] 2. Feature extraction module: Convolutional neural network (CNN) is used to pre-process the collected video data. Feature information that can characterize dangerous behaviors, such as the posture of workers and the movement trajectory of objects, is extracted. Features are classified and labeled through deep learning models to identify potential dangerous behaviors.

[0008] 3. Safety risk assessment module: Combine deep reinforcement learning to simulate the worker's action process. Use the deep learning model to analyze the extracted feature information, identify possible dangerous situations and their causes, generate a safety assessment report, and provide preventive measures.

[0009] 4. Real-time monitoring and early warning module: Send warning signals to operators through smart terminals (such as mobile phones, tablets, etc.). When abnormal conditions occur, timely automatic intervention measures are taken, such as stopping the lifting operation, adjusting the lifting path, etc. Provide a real-time video monitoring interface to facilitate operators to check the situation in the lifting area at any time.

[0010] 5. Data analysis and optimization module: Use machine learning algorithms to mine historical data and find regular change patterns. Based on the analysis and prediction of the current scenario, continuously adjust and improve the system to improve the accuracy and response speed of the model. Provide data analysis reports to guide decision makers to develop reasonable and efficient operation plans.

[0011] The core principle of the present invention is to use deep learning algorithms to analyze and process video data and environmental parameters during the lifting process, thereby achieving real-time monitoring and risk assessment of workers' behavior.

[0012] Compared with the prior art, the present invention has the following beneficial effects:

[0013] The present invention uses deep learning algorithms to analyze video data and environmental parameters, which can accurately identify potential dangerous behaviors, improve the accuracy of safety monitoring, and reduce errors in human judgment. The real-time monitoring and early warning module can send warning signals to operators in a timely manner and take automatic intervention measures to ensure rapid response when dangerous situations occur and avoid accidents. The data analysis and optimization module can continuously adjust and improve the system based on the analysis and prediction of historical data and current scenes, improve the adaptability and response speed of the system, and ensure that it can work effectively in different environments. The system has a simple structure, is easy to install and maintain, has low cost, and is suitable for factory workshops of various types, sizes and complexities. The system uses advanced computer vision technology and big data analysis methods to improve monitoring efficiency and safety levels, provide more accurate information and more diversified services, and ensure production safety and employee health.

[0014] In summary, the present invention provides a method for monitoring and analyzing the safety of steel structure hoisting in factory cranes based on a deep learning algorithm. The method uses deep learning technology to accurately analyze and judge the behavior of workers, prevent and reduce hoisting accidents, and protect the life safety and health of employees. The method is highly reliable, easy to promote and implement, and has important significance and role in improving the safety level of production in modern manufacturing. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a flow chart of the method of the present invention, showing a complete closed-loop process from data collection, feature extraction, risk assessment to real-time monitoring and data analysis;

[0016] Figure 2 This is the overall architecture diagram of the system of the present invention, which describes the interrelationships and data flows of the five core modules: data collection, feature extraction, security risk assessment, real-time monitoring and early warning, and data analysis and optimization;

[0017] Figure 3 This is a multi-source data collection and fusion flow chart of the present invention, which explains how to collect data from multiple sources (such as high-definition cameras, environmental sensors, equipment status sensors, etc.) and perform fusion processing;

[0018] Figure 4 This is a diagram of the deep learning feature extraction network structure of the present invention, showing the combined architecture of a convolutional neural network (CNN) and a recurrent neural network (RNN) for extracting key features from multi-source data;

[0019] Figure 5 The safety risk assessment decision tree of the present invention describes how the system assesses the risk level of the current operation based on factors such as equipment status, environmental conditions, operator status, and safety of the work area;

[0020] Figure 6 This is a real-time monitoring and automatic intervention flow chart of the present invention, which shows how the system can continuously monitor, issue warnings or perform automatic intervention measures based on real-time risk assessment results. DETAILED DESCRIPTION

[0021] Reference Figures 1 to 6 The specific implementation method of the factory crane steel structure hoisting safety monitoring and analysis method based on deep learning of the present invention is further explained.

[0022] A deep learning-based method for monitoring and analyzing the safety of steel component hoisting in a factory. This method uses advanced visual perception technology to achieve real-time monitoring and risk assessment of the hoisting process. The method specifically includes the following steps:

[0023] 1. Data collection: Start the data collection module, and the high-definition camera and sensors start working to collect video data and environmental parameters of the hoisting area in real time. The high-definition camera should have a resolution of at least 1080p and support a frame rate of at least 30fps to ensure smooth and clear video capture. Environmental sensors should include but are not limited to temperature, humidity, light intensity and noise level sensors to fully monitor the working conditions of the hoisting area. All data is transmitted to the central processing unit via a wireless network to ensure the real-time and stability of data transmission.

[0024] 2. Feature extraction: The central processing unit uses a convolutional neural network to pre-process the video data and extract feature information that can characterize dangerous behaviors. The feature extraction process includes steps such as image denoising, background segmentation, target recognition and tracking. The CNN model will be trained to identify the workers' safety equipment, operating posture, and the motion state of the hoisted objects. In addition, feature extraction will also include real-time analysis of environmental parameters to evaluate the impact of environmental factors on the safety of hoisting operations.

[0025] 3. Safety risk assessment: The deep reinforcement learning model simulates the worker's movement process and analyzes it in combination with feature information. The model will evaluate the relative position between the worker and the hoisted object, the coordination of the movement, and the potential risk of collision. The safety risk assessment will use machine learning algorithms to quantify the potential hazards based on the extracted features and environmental parameters, and generate a detailed safety assessment report, including the hazard level, possible consequences, and recommended preventive measures.

[0026] 4. Real-time monitoring and early warning: The real-time monitoring and early warning module sends warning signals to operators through smart terminals. The smart terminal will display the real-time video stream and alert the operator through visual and auditory signals when anomalies are detected. The system will also automatically trigger preset intervention measures, such as suspending the lifting operation, adjusting the lifting path, or initiating emergency plans. Operators can quickly access safety assessment reports and historical data through smart terminals to make timely decisions.

[0027] 5. Data analysis and optimization: The data analysis and optimization module mines historical data to find regular change patterns. The module will apply data mining techniques such as cluster analysis and association rule mining to identify common risk factors and potential security vulnerabilities in lifting operations. Based on the mining results, the system will automatically adjust the parameters of the deep learning model, optimize the feature extraction and risk assessment algorithms, and improve the accuracy and response speed of the model. In addition, the module will also generate regular data analysis reports to provide decision support for management.

[0028] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A method for monitoring and analyzing the safety of steel structure hoisting in factory cranes based on deep learning algorithms, characterized in that: Introducing visual technology into the safety control and protection process of steel component hoisting, using images as a carrier to judge and evaluate the workers' actions and behaviors, specifically including the following steps: a. Data acquisition, collecting visual, environmental, equipment status and personnel data in industrial environments through multi-source sensors; b. Feature extraction: use deep learning algorithms to extract features from the collected multimodal data; c. Security risk assessment: Based on the extracted features, a machine learning model is used to perform real-time security risk assessment; d. Real-time monitoring and early warning: Based on the risk assessment results, real-time monitoring and early warning are carried out, and automatic intervention is carried out when necessary; e. Data analysis and optimization: continuously analyze historical data to optimize system performance and predictive models.

2. The method for safety monitoring and analysis of steel structure hoisting in factory cranes based on deep learning algorithm according to claim 1 is characterized in that: The data acquisition module will integrate more diverse sensors, such as stress, vibration, temperature and humidity sensors, to obtain more comprehensive lifting environment and equipment status data. At the same time, high-precision and high-frequency data acquisition technology will be adopted to ensure the real-time and accuracy of the data.

3. The method for safety monitoring and analysis of steel structure hoisting in factory cranes based on deep learning algorithm according to claim 1 is characterized in that: The extraction module will adopt multiple deep learning architectures, including CNN, RNN and GNN, to process the temporal and relational features in the video data, which will improve the recognition accuracy of workers' postures, object motion trajectories and environmental changes.

4. The method for safety monitoring and analysis of steel structure hoisting in factory cranes based on deep learning algorithm according to claim 1 is characterized in that: The safety risk assessment module will use multi-model fusion technology, combining different types of deep learning models, to improve the accuracy and robustness of risk assessment. In addition, deep reinforcement learning will be introduced to simulate the worker's action process to more accurately identify potential hazards.

5. The method for safety monitoring and analysis of factory crane steel component hoisting based on deep learning algorithm according to claim 1 is characterized in that: The real-time monitoring and early warning module will add an adaptive early warning system to automatically adjust the degree of urgency according to different risk levels. It will also integrate voice recognition functions, allowing operators to interact with the system through voice commands, improving response speed and operational convenience.

6. The method for safety monitoring and analysis of steel structure hoisting in factory cranes based on deep learning algorithm according to claim 1 is characterized in that: The data analysis and optimization module will introduce online learning and incremental learning mechanisms to enable the system to continuously optimize based on new data. At the same time, reinforcement learning algorithms will be applied to enable the system to learn the optimal monitoring strategy through interaction with the environment.

Images