Near-electricity construction multifunctional integrated intelligent monitoring method and device

Abstract

The invention discloses a multifunctional integrated intelligent monitoring method and device for near-electricity construction. The method comprises the following steps: acquiring near-electricity construction historical data based on a laser radar, a camera and an electric field sensor; preprocessing the collected near-electricity construction historical data, and integrating the preprocessed data into multi-source heterogeneous data based on a preset data fusion method; constructing a data set based on the multi-source heterogeneous data; constructing a construction safety detection model based on a convolutional neural network CNN; training a construction safety detection model by adopting the data set; and collecting near-electricity construction data in real time, preprocessing the data, integrating the preprocessed data into multi-source heterogeneous data, inputting the multi-source heterogeneous data into the trained construction safety detection model, and detecting construction potential safety hazards and construction anomalies in real time. The method has the outstanding advantages of high safety, high real-time performance, high accuracy and the like.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of near-electricity operation intelligent monitoring, in particular to a near-electricity construction multifunctional integrated intelligent monitoring method and device. BACKGROUND

[0002] Near-electricity operation refers to all operations that may affect the safety of electrical equipment and personnel in the vicinity of power generation, power transformation, power transmission and distribution (within line protection area) and electrical equipment in operation with voltage level of 250V and above, which has a high safety risk. In order to protect the safety of near-electricity operation personnel, it is necessary to monitor the construction area in real time, discover potential safety threats in time, and avoid accidents. However, the current method or system for real-time monitoring of construction area has the disadvantages of high cost, low efficiency and low reliability.

[0003] CN117690239A discloses a near-electricity warning method and device. The method includes: picking up standard coordinate information of four forbidden zone points in the field; determining a straight line through two adjacent points, obtaining four straight lines through the four forbidden zone points, and setting the four straight lines as warning lines, and setting the area surrounded by the warning lines as a forbidden zone; obtaining four safety lines according to the distance formula of two parallel lines and the set warning area width, setting the area surrounded by the four safety lines as a warning area, and setting the area outside the warning area as a safety area; in response to a host positioning judgment request, the host enters a working state, real-time acquires host position information, judges whether the host enters the warning area / forbidden zone, and if the host enters the warning area / forbidden zone, the slave machine sends an alarm information. The patent monitors and warns potential electric danger areas by setting warning lines and safety lines. The method mentioned in the patent needs a lot of field work to pick up the standard coordinate information of the forbidden zone points, and the scheme has the disadvantages of low efficiency and easy to make mistakes. In addition, the patent mainly relies on accurate host positioning, and if the positioning system fails or has errors, it may cause false alarm or missed alarm.

[0004] CN114842616A provides a method for improving the safety supervision efficiency of power construction process. The method configures corresponding individual equipment for three roles (workers, guardians and work leaders) in the construction site and builds an information exchange center. The information exchange center is connected with individual equipment by 4G / 5G technology to realize information and voice interconnection and intercommunication among the three roles. The information exchange center collects real-time individual equipment information of workers, evaluates the risk degree of the information and generates a graded warning, and publishes the warning information to the three roles. The warning receiving confirmation mechanism among the three roles is established to avoid the situation that no one responds to the warning. The information exchange center is responsible for the whole process of supervision, warning release, receiving confirmation registration, process tracking and result summary. The patent realizes information and voice interconnection and intercommunication by connecting individual equipment with 4G / 5G technology, which depends on stable 4G / 5G network connection. In areas where the network is unstable or not covered, the system performance may be affected. In addition, all construction personnel need to be equipped with individual equipment, which will result in high initial cost and maintenance cost. If the system is in a complex electromagnetic environment, the stability and accuracy of signal transmission may be affected.

[0005] CN117422754A discloses a power substation near-electricity worker spatial distance calculation method based on instance segmentation, a readable storage medium and an electronic device. The method includes the following steps: performing segmentation operation on the target box obtained by target detection through instance segmentation algorithm to extract the real contour of the object. Based on the characteristics of the project itself, the contour obtained by segmentation is dilated, and the target function improvement algorithm based on MobileSAM is used to add binary cross-entropy loss according to the special application scenario to calculate the distance. The patent algorithm requires a large amount of computing resources, and the requirement for hardware devices is high, which may increase the cost and the accuracy of the algorithm may be affected by the result of the target detection stage. If the target detection is not accurate, the subsequent distance calculation will also be affected. Although the improved algorithm mentioned in the patent considers real-time performance, the real-time performance and accuracy in complex environments may be low. SUMMARY

[0006] To solve the problems of large workload, low efficiency, low accuracy, high cost, weak stability and poor real-time performance in the prior art, the present application provides a near-electricity construction multifunctional integrated intelligent monitoring method and device.

[0007] The present application adopts the following technical solutions.

[0008] The present application discloses a near-electricity construction multifunctional integrated intelligent monitoring method, which comprises: Step 1: Collecting near-electricity construction historical data based on laser radar, camera and electric field sensor; Step 2: preprocessing the collected near-electric construction history data, and integrating the preprocessed data into multi-source heterogeneous data based on a preset data fusion method; constructing a data set based on the multi-source heterogeneous data; Step 3: constructing a construction safety detection model based on a convolutional neural network (CNN); Step 4: training the construction safety detection model using the data set; Step 5: collecting near-electric construction data in real time, preprocessing the data according to Step 2 and integrating the preprocessed data into multi-source heterogeneous data, inputting the multi-source heterogeneous data into the construction safety detection model trained in Step 4, and real-time detecting construction safety hazards and construction abnormalities.

[0009] Further preferably, In Step 1, the near-electric construction history data includes laser point clouds collected by a laser radar, construction site images collected by a camera, and electric field intensities at different positions collected by an electric field sensor.

[0010] Further preferably, In Step 2, the preset data fusion method is: Grouping data collected by different data collectors / different times according to the same target; the target includes equipment and people in the construction site; Completing 3D reconstruction of the construction scene based on the grouping result, the construction site image and the laser point cloud, mapping the collected electric field intensity data at different positions to the 3D model, thereby completing data fusion.

[0011] Further preferably, In Step 2, the data set includes 5 types of labels, including safety accidents, illegal operations, equipment failures, approaching safety distance, and below safety distance.

[0012] Further preferably, In Step 4, the construction safety detection model uses an improved loss function as shown in the following formula: ; wherein, i represents the i-th sample; i is the total number of samples; N represents the c-th category of the label; c is the total number of label categories; c is the weight of category c when reading sample i; C is the true value for the c-th category; is the model's prediction for sample i; is the model's prediction for sample i; i is the true value for the c-th category; is the model's prediction for sample i; i is the model's prediction for sample i; cthe predicted probability of the class c for the sample i; log is a logarithm function; log is a logarithm function.

[0013] Further preferably, The calculation formula of the weight of the class c when reading the sample i is as follows: ; wherein, is the weight of the class c when reading the sample i-1.

[0014] Further preferably, In step 5, the construction safety hazards include illegal operation, the distance between the operating personnel and the live equipment is close to the safe distance, and the distance between the operating personnel and the live equipment is lower than the safe distance. The construction abnormalities include safety accidents and equipment failures.

[0015] Further preferably, In step 5, when the construction safety hazards or the construction abnormalities are detected, an alarm is issued. When illegal operation is detected, the operating personnel is informed to stop the illegal operation. When the distance between the operating personnel and the live equipment is close to the safe distance, the operating personnel is informed to move away from the live equipment. When the distance between the operating personnel and the live equipment is lower than the safe distance, power-off measures are taken, and the operating personnel is informed to move away from the live equipment. When a safety accident is detected, power-off measures are taken, and each operating personnel is informed to move away from the live equipment. When an equipment failure is detected, power-off measures are taken, the equipment is prompted to be checked, and each operating personnel is informed to move away from the equipment.

[0016] Another aspect of the present application discloses an intelligent monitoring device based on an intelligent monitoring method, comprising a perception module, a data fusion and processing module, an intelligent analysis module, and a warning system module: The perception module collects near-electricity construction historical data based on a laser radar, a camera, and an electric field sensor; The data fusion and processing module pre-processes the collected near-electricity construction historical data, integrates the pre-processed data into multi-source heterogeneous data based on a preset data fusion method, and constructs a data set based on the multi-source heterogeneous data; The intelligent analysis module collects near-electricity construction data in real time, pre-processes the data according to step 2, integrates the pre-processed data into multi-source heterogeneous data, inputs the multi-source heterogeneous data into the construction safety detection model trained in step 4, and detects construction safety hazards and construction abnormalities in real time; The warning system module issues an alarm corresponding to the detected construction safety hazards or construction abnormalities.

[0017] Another aspect of this application discloses an electronic device, including a processor and a storage medium; The storage medium is used to store instructions; The processor is used to operate according to the instructions to execute the aforementioned intelligent monitoring method.

[0018] This application also discloses a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the intelligent monitoring method.

[0019] The core technologies of this invention include data acquisition, data fusion and processing, intelligent analysis, and early warning systems. Each step will be described in detail below.

[0020] The beneficial effects of this invention are compared with those of the prior art: The multi-functional integrated intelligent monitoring device for near-electric construction of the present invention has the following advantages: (1) High safety: Through real-time monitoring and early warning system, the device of the present invention can detect and respond to potential safety risks in a timely manner, reducing the probability of accidents.

[0021] (2) Strong real-time performance: The device of the present invention can monitor the surrounding environment in real time and issue early warning signals in a timely manner to remind workers working near power to pay attention to safety.

[0022] (3) High accuracy: Various data are collected in real time through multiple sensing modules. By integrating and analyzing the information from each module, the accuracy and reliability of the overall monitoring results are improved.

[0023] (4) Strong data storage and analysis capabilities: Through data storage units and communication modules, long-term data storage, trend analysis and prediction can be carried out to provide a scientific basis for the safety management of near-electric work.

[0024] In summary, the multi-functional integrated intelligent monitoring method and device for near-power construction utilizes an information fusion technology based on multiple sensing modes. It aims to provide more comprehensive and accurate early warnings by integrating information from different modes, establishing a system that enables multi-dimensional perception, data fusion and processing, artificial intelligence self-learning and prediction of risks in the near-power construction work environment, as well as the design and application of an early warning system. This invention features high accuracy, real-time performance, and reliability, effectively ensuring the safety of personnel working near power lines. Attached Figure Description

[0025] Figure 1 A schematic diagram of a multi-functional integrated intelligent monitoring method for near-electric construction. Figure 2 This is a schematic diagram of a multi-functional integrated intelligent monitoring device for near-electricity construction. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, other embodiments obtained by those skilled in the art without creative effort are all within the protection scope of this invention.

[0027] This invention provides a multi-functional integrated intelligent monitoring method for near-electric construction, see appendix. Figure 1 This includes the following steps: Step 1: Collect historical data on near-electricity construction based on lidar, cameras, and electric field sensors; The historical data of near-electric construction includes laser point clouds collected by lidar, construction site images collected by cameras, and electric field strength at different locations collected by electric field sensors.

[0028] Step 2: Preprocess the collected historical data of near-electric construction, and integrate the preprocessed data into multi-source heterogeneous data based on the preset data fusion method; construct a dataset based on the multi-source heterogeneous data; The preset data fusion method is as follows: Data collected from different data acquisition devices / at different times is grouped according to the same target; the target includes equipment and personnel at the construction site; Based on the grouping results, construction site images, and laser point clouds, a 3D reconstruction of the construction scene was completed. The electric field intensity data collected at different locations were mapped onto the 3D model, thereby completing data fusion.

[0029] The dataset includes five categories of labels: safety incident, violation of regulations, equipment failure, approaching a safe distance, and falling below a safe distance.

[0030] Step 3: Construct a construction safety detection model based on a convolutional neural network (CNN); Step 4: Train the construction safety detection model using the dataset; The construction safety monitoring model adopts an improved loss function. As shown in the following formula: ; in, i Indicates the first i One sample; N The total number of samples; c Indicates the first of the tags c One category; C Total number of label categories; The weight of class c when reading sample i; For the sample i The true value of the cth category; For the model to the sample i For the category c The predicted probability; log is the logarithmic function; The base of the logarithmic function, preferably in the range (1, 10], and the skilled person can choose a suitable base, which will not be repeated here.

[0031] The calculation formula of the weight of category c when reading sample i is as follows: ; Wherein, The weight of category c when reading sample i-1.

[0032] Step 5: Real-time acquisition of near electric construction data, pre-processing of the data according to step 2 and integration of the pre-processed data into multi-source heterogeneous data, input of the multi-source heterogeneous data into the construction safety detection model trained in step 4, real-time detection of construction safety hazards and construction abnormalities.

[0033] The construction safety hazards include illegal operation, the distance between the operator and the live equipment close to the safe distance, and the distance between the operator and the live equipment below the safe distance; The construction abnormalities include safety accidents and equipment failures.

[0034] When the detected construction safety hazard or construction abnormality is detected, an alarm is issued; the alarm is an audible and visual alarm.

[0035] When illegal operation is detected, the operator is notified to stop illegal operation; When the distance between the operator and the live equipment is close to the safe distance, the operator is notified to stay away from the live equipment; When the distance between the operator and the live equipment is below the safe distance, the power is cut off, and the operator is notified to stay away from the live equipment; When a safety accident is detected, the power is cut off, and each operator is notified to stay away from the live equipment; When an equipment failure is detected, the power is cut off, the equipment is checked, and each operator is notified to stay away from the equipment.

[0036] Embodiment 1 Near electric construction multifunctional integrated intelligent monitoring method.

[0037] Step 1: Collect near electric construction historical data.

[0038] The near-electric construction historical data includes laser point cloud collected by the mountable three-dimensional laser radar, construction site images collected by the camera, and electric field intensity collected by the electric field sensor. These data can be used for subsequent analysis and decision support. The construction site images are presented in the form of videos, and the videos are composed of a sequence of construction site images. Each time's construction site image includes field area electrical equipment, surrounding environment, and operating personnel.

[0039] The skilled person should know that the electric field sensor can measure the electric field intensity at different positions around the target synchronously and reconstruct the three-dimensional electric field distribution. The electric field sensor can be selected from a MEMS electric field sensor array, an electric field probe for EMC testing, etc.

[0040] Step 2: Preprocess the near-electric construction historical data collected in step 1, and integrate the preprocessed near-electric construction historical data into multi-source heterogeneous data through data fusion technology. Construct a data set based on the multi-source heterogeneous data. Divide the data set into a training set, a validation set, and a test set.

[0041] In the near-electric construction multifunctional integrated intelligent monitoring method, the near-electric construction historical data collected in step 1 needs to be integrated into multi-source heterogeneous data through data fusion technology. First, preprocess the near-electric construction historical data, including normalizing the electric field intensity data, denoising the picture sequence collected by the camera, denoising the power line point cloud, and calibrating the power line point cloud based on the pictures collected by the camera at each time; The skilled person should know that calibrating the power line point cloud based on the pictures collected by the camera means matching and calibrating the power line point cloud data with the images actually collected by the camera. Specifically, calibrating the power line point cloud based on the pictures collected by the camera can be achieved by synchronously collecting point cloud and image data through a checkerboard calibration board, calculating the conversion matrix of camera internal and external parameters and the laser radar coordinate system, using strong features such as edges or corners in the environment to optimize the transformation matrix based on deep learning feature matching, etc. The skilled person should be able to select the required method according to the actual situation.

[0042] Then, group and associate the data collected by different sensors or at different times according to the same target to ensure the consistency and comparability of the data, and then complete the synthesis of multi-source heterogeneous data.

[0043] The grouping and association of data collected by different sensors or at different times according to the same target (device / person) to ensure consistency and comparability of the data, including grouping and association of data collected by different sensors or at different times according to the same target, including grouping and association of personnel image information collected at different times and corresponding to the construction personnel who have taken photos in advance, grouping and association of electric field intensity data collected at different times and corresponding to electrical equipment, grouping and association of transformer / equipment images collected by cameras, electric field intensity collected by electric field sensors, and laser point cloud of transformer / equipment images collected by laser radars and transformers / equipment, etc. To ensure that the distance between personnel and electrical equipment can be compared and judged over time. The device is the device on the construction site, and each construction site will be different, which can include cranes, elevators, excavators, pile drivers, concrete mixers, etc.

[0044] Those skilled in the art should know that the synthesis of the multi-source heterogeneous data is to fuse the laser point cloud, the construction site image and the electric field intensity at the same time; the available data fusion algorithms such as decision tree algorithm, Kalman filter, D-S evidence reasoning, multi-Bayesian estimation, etc. can be used to complete the synthesis of multi-source heterogeneous data, and those skilled in the art should know how to select the synthesis method of multi-source heterogeneous data according to the actual situation; the synthesis method of multi-source heterogeneous data proposed by the present application is only a preferred embodiment, and is not necessarily limited to the implementation of the present application. The synthesis method of multi-source heterogeneous data is as follows: Based on the grouping result, the construction site image and the laser point cloud, the 3D reconstruction of the construction scene is completed, and the electric field intensity data collected at different positions are mapped to the 3D model, thereby completing data fusion and providing a basis for subsequent judgment.

[0045] Specifically, first, the laser point cloud of each device is collected by a three-dimensional laser radar, and then the scene 3D reconstruction is completed based on the collected image information and the grouping result.

[0046] Those skilled in the art should know that the 3D reconstruction of the construction scene based on the grouping result, the construction site image and the laser point cloud is to restore the three-dimensional information through the correspondence between the construction site image and the laser point cloud (i.e. the grouping result), which can be realized by structure from motion (SfM) or multi-view stereo vision (MVS), etc. Those skilled in the art can construct according to the actual situation, which will not be repeated here.

[0047] Those skilled in the art should know that the mapping of the collected electric field intensity data of different positions to the 3D model is to present the electric field intensity of each position in the 3D model, specifically, the electric field intensity can be converted into a rainbow spectrum, a heat map, a 3D surface of a specific electric field intensity, etc. Those skilled in the art can construct according to the actual situation; preferably, the collected electric field intensity data of different positions can be mapped to the 3D model by using spatial interpolation combined with 3D volume rendering, including: First, through discrete electric field intensity sampling points, algorithms such as Kriging, inverse distance weighting (IDW) or radial basis function (RBF) are used to generate a continuous three-dimensional electric field intensity distribution field.

[0048] Then, volume rendering is performed, the interpolated 3D volume data (voxel grid) is rendered into a volume cloud chart in the 3D model through color mapping and transparency adjustment / isosurface generation, and the electric field intensity gradient is directly rendered. The color mapping is to directly display the electric field intensity in the point cloud by color coding (such as rainbow spectrum or heat map). The isosurface generation is to extract the isosurface from the scalar field by using the Marching Cubes algorithm to generate the 3D surface of a specific electric field intensity.

[0049] The multi-source heterogeneous data is labeled, the multi-source heterogeneous data and the label are taken as a data set, and the data set is divided into a training set, a validation set and a test set. The data set includes five types of labels, and the labels include safety accidents, illegal operations, equipment failures, proximity to a safe distance, and below a safe distance. Among them, the safety accidents include electric shock and mechanical injury; the illegal operations include not wearing a safety helmet and wearing an unqualified work uniform; the equipment failure includes equipment damage and equipment short circuit; the proximity to a safe distance means that the distance between the operating personnel and the live equipment is close to the safe distance, i.e. the operating personnel are located in the area between the safety threshold of the equipment and 1.25 times the safety threshold of the equipment; the below a safe distance means that the distance between the operating personnel and the live equipment is below the safe distance, which can also be said that the distance between the operating personnel and the equipment is below the safety threshold of the equipment.

[0050] The safe distance (safety threshold) specifies the closest distance between the operating personnel and the live body or electrical equipment; The closest distance between the operating personnel and the live body or electrical equipment needs to meet the safety distance of the live body or electrical equipment; the safety threshold is different in different operating environments and can be determined according to relevant live working specifications; for example, the safety threshold for 10kV / m and below live conventional operation is 0.7 meters to 1.0 meter, the safety threshold for indirect live working is 0.4 meters to 0.6 meters; the safety threshold for 110kV / m to 220kV / m live conventional operation is 3.0 meters to 4.0 meters, and the safety threshold for indirect live working is 1.8 meters to 2.5 meters.

[0051] Step 3: Construct a construction safety detection model based on a Convolutional Neural Network (CNN). The construction safety detection model includes multiple convolutional layers, activation layers, pooling layers, and fully connected layers. Finally, a softmax activation function can be used for multi-class classification output. Cross-entropy loss can be selected as the loss function, Adam can be selected as the optimizer, and the learning rate can be adjusted using a learning rate decay strategy.

[0052] Step 4: Use the dataset established in Step 2 to train the construction safety detection model constructed in Step 3.

[0053] The multi-source heterogeneous data and labels in the dataset are used as the input and output of the construction safety detection model for training. Based on the construction safety detection model, key features are extracted from the multi-source heterogeneous data to characterize the essential attributes of the observed data, providing support for subsequent data analysis and decision-making.

[0054] To better measure the difference between the model output and the true label, this invention improves the loss function. The improved loss function is calculated based on the weights of each category when reading each sample, the true value of each sample for each category, and the model's predicted probability of each sample belonging to each category, as shown in the following formula: ; in, This is the improved loss function of the present invention; i The integer represents the number of digits. i One sample; N The total number of samples; c It is an integer, representing the label's first digit. c One category; C Total number of label categories; The weight of category c when reading sample i is determined based on the probability of the c-th label category appearing in the first i samples; For the sample i For the true value of the c-th category (either 1 or 0 via one-hot encoding), when the sample i The tag is the first c When there are multiple categories, It is 1 if it is not 1, and 0 otherwise; For the model to sample i For category c The predicted probability; The base of the logarithmic function is preferably in the range of (1, 10]. Those skilled in the art can choose a suitable base, which will not be elaborated here.

[0055] Weights of class c when reading sample i The calculation formula is as follows: ; wherein, is the weight of class c for reading sample i-1; it is set that when i = 1, = 0.2.

[0056] Step 5: Real-time acquisition of near-electric construction data, pre-processing of the data according to step 2 and integration of the pre-processed data into multi-source heterogeneous data, input of the multi-source heterogeneous data into the construction safety detection model trained in step 4, real-time detection of construction safety hazards and construction abnormalities; when construction safety hazards and construction abnormalities are detected, an alarm is issued in time, and warning information is issued to alert near-electric operating personnel to take safety measures.

[0057] The near-electric construction data includes laser point clouds collected by a laser radar, construction site images collected by a camera, and electric field strengths at different positions collected by an electric field sensor.

[0058] The construction safety hazards include illegal operations, the distance between operating personnel and live equipment being close to a safe distance, and the distance between operating personnel and live equipment being below the safe distance. The construction abnormalities include safety accidents and equipment failures.

[0059] The present application analyzes through intelligent data processing technology to identify safety hazards and construction abnormalities in the construction process. Specifically, the construction safety detection model proposed in the present application identifies the edge contour of electrical equipment and operating personnel based on a convolutional neural network (CNN) combined with camera image information and laser point cloud data, judges whether a safety accident or illegal operation occurs, judges whether there is an equipment failure based on a convolutional neural network (CNN) combined with electric field strength data at different distances around the electrical equipment, and judges whether the distance between the operating personnel and the electrical equipment meets the relevant requirements.

[0060] Based on artificial intelligence and machine learning algorithms, the collected data are analyzed in depth to identify abnormal patterns or trends in the construction process. According to the analysis results, the system automatically evaluates the potential risk level to determine whether a warning needs to be triggered. When an abnormal situation or potential risk is detected based on the data analysis results, a warning signal is issued through a warning signal device to alert near-electric operating personnel to take safety measures in time. The warning signal can be transmitted in the form of sound, light or wireless communication, etc. to timely and effectively inform the operating personnel and take appropriate measures. The issuance of the warning signal is the last line of defense to protect the safety of near-electric operating personnel, and its stability and reliability are crucial.

[0061] In step 5, when a construction safety hazard or a construction abnormality is detected, an alarm is issued, and warning information is issued to alert near-electric operating personnel to take safety measures; the warning information includes: ​When a violation operation is detected, inform the operator to stop the violation operation; When the distance between the operator and the live equipment is detected to be close to the safe distance, the operator is informed to move away from the live equipment; When the distance between the operator and the live equipment is detected to be less than the safe distance, the power is cut off, and the operator is informed to move away from the live equipment; When a safety accident is detected, the power is cut off, and each operator is informed to move away from the live equipment; When a device failure is detected, the power is cut off, the device is prompted for inspection, and each operator is informed to move away from the device.

[0062] The application also discloses an intelligent monitoring device based on the intelligent monitoring method, comprising a perception module, a data fusion and processing module, an intelligent analysis module, and a warning system module: The perception module collects near-electricity construction historical data based on a laser radar, a camera and an electric field sensor; The data fusion and processing module pre-processes the collected near-electricity construction historical data, integrates the pre-processed data into multi-source heterogeneous data based on a preset data fusion method, and constructs a data set based on the multi-source heterogeneous data; The intelligent analysis module collects near-electricity construction data in real time, pre-processes the data according to step 2, integrates the pre-processed data into multi-source heterogeneous data, inputs the multi-source heterogeneous data into the construction safety detection model trained in step 4, and detects construction safety hazards and construction abnormalities in real time; The warning system module issues an alarm corresponding to the detected construction safety hazards or construction abnormalities.

[0063] The intelligent monitoring device disclosed by the application should be arranged at a position capable of collecting operators and facilities in the entire construction area. Those skilled in the art should know how to select the position of the intelligent monitoring device according to the detection range of the camera, the laser radar and the electric field intensity sensor.

[0064] Embodiment 2 A near-electricity construction multifunctional integrated intelligent monitoring device, as shown in Figure 2 The near-electricity construction multifunctional integrated intelligent monitoring device structure schematic diagram comprises a three-dimensional laser radar 1, a processor 2, a communication module 3, an alarm 4, an electric field sensor 5, a camera 6 and a battery 7. It covers the perception module, the data fusion and processing module, the intelligent analysis module and the warning system module.

[0065] The perception module is composed of a three-dimensional laser radar, a camera, and an electric field sensor. In near-electric construction, the laser radar can be used to monitor the three-dimensional changes of the construction area, detect the safety distance between construction machinery and power facilities, and perform 3D modeling of the construction site based on the point cloud data collected by the laser radar and the images collected by the camera. The camera is used to monitor the activities in the construction site in real time, detect unsafe behaviors, monitor the construction progress, and record important events (safety accidents, illegal operations, equipment failures, and emergencies) during the construction process. The electric field sensor can monitor the electromagnetic environment around the power facilities, warn of potential electrical faults, and ensure the safety of workers and equipment. Through the above monitoring methods, the safety of near-electric workers is ensured.

[0066] The data fusion and processing module plays a core role in the near-electric construction multifunctional integrated intelligent monitoring device, responsible for collecting laser point cloud, image sequence, and electric field strength changes from three-dimensional laser radar, camera, electric field sensor, and other devices in real time. The original data is preprocessed, such as cleaning, denoising, and standardization, to improve data quality. Data from different sensors is integrated to obtain a more comprehensive and consistent data representation.

[0067] The intelligent analysis module applies statistical analysis, machine learning, and artificial intelligence algorithms to conduct in-depth analysis of the data to identify abnormalities.

[0068] The warning system module should be able to send timely and accurate warning signals to alert near-electric workers to take safety measures. The warning signal device can choose appropriate sound, light, or vibration methods for warning according to actual needs. At the same time, the warning signal device should also have a communication interface with the data processing device to receive the warning information sent by the processing device, ensuring the accuracy and real-time performance of the signal.

[0069] In addition, the near-electric construction multifunctional integrated intelligent monitoring device proposed by the present application also has a data storage unit and a communication module, which can perform long-term data storage, trend analysis and prediction, and provide scientific basis for the safety management of near-electric operation.

[0070] The present disclosure can be a system, a method, and / or a computer program product. The computer program product can include a computer readable storage medium having computer readable program instructions loaded thereon for causing a processor to implement various aspects of the present disclosure.

[0071] Computer readable storage media can be tangible storage media which can retain and store instructions for use by an instruction execution device. Computer readable storage media can be, for example, but is not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer readable storage media include the following: a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanically encoded device such as punch-cards or raised structures in a groove having instructions recorded thereon, and any suitable combination of the foregoing. A computer readable storage medium, as used herein, is not to be construed as being transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission media (e.g., light pulses passing through a fiber-optic cable), or electrical signals transmitted through a wire.

[0072] Computer readable program instructions described herein can be downloaded to respective computing / processing devices from a computer readable storage medium or to an external computer or external storage device via a network, for example, the Internet, a local area network, a wide area network and / or a wireless network. The network can comprise copper transmission cables, optical transmission fibers, wireless transmission, routers, firewalls, switches, gateway computers and / or edge servers. A network adapter card or network interface in each computing / processing device receives computer readable program instructions from the network and forwards the computer readable program instructions for storage in a computer readable storage medium within the respective computing / processing device.

[0073] Computer readable program instructions for carrying out operations of the present disclosure can be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine dependent instructions, microcode, firmware instructions, state-setting data, or either source code or object code written in any combination of one or more programming languages, including an object oriented programming language such as Smalltalk, C++ or the like, and conventional procedural programming languages such as the "C" programming language or similar programming languages. The computer readable program instructions can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider). In some embodiments, electronic circuitry including, for example, programmable logic circuitry, field-programmable gate array (FPGA), or programmable logic array (PLA) can execute the computer readable program instructions by utilizing state information of the computer readable program instructions to personalize the electronic circuitry, in order to perform aspects of the present disclosure.

[0074] Finally, it should be noted that the above-mentioned embodiments are merely used to illustrate the technical solutions of the present application, but not to limit it. Although the present application has been described in detail with reference to the above-mentioned embodiments, those skilled in the art should understand that the specific embodiments of the present application can be modified or replaced, and any modification or replacement without departing from the spirit and scope of the present application should be covered in the protection scope of the claims of the present application.

Claims

1. A multi-functional integrated intelligent monitoring method for near-electric construction, characterized in that, include: Step 1: Collect historical data on near-electricity construction based on lidar, cameras, and electric field sensors; Step 2: Preprocess the collected historical data of near-electric construction, and integrate the preprocessed data into multi-source heterogeneous data based on the preset data fusion method; construct a dataset based on the multi-source heterogeneous data; Step 3: Construct a construction safety detection model based on a convolutional neural network (CNN); Step 4: Train the construction safety detection model using the dataset; Step 5: Collect near-electricity construction data in real time, preprocess the data according to Step 2, and integrate the preprocessed data into multi-source heterogeneous data. Input the multi-source heterogeneous data into the construction safety detection model trained in Step 4 to detect construction safety hazards and construction anomalies in real time.

2. The intelligent monitoring method according to claim 1, characterized in that: In step 1, the near-electric construction historical data includes laser point clouds collected by lidar, construction site images collected by cameras, and electric field strength at different locations collected by electric field sensors.

3. The intelligent monitoring method according to claim 1 or 2, characterized in that: In step 2, the preset data fusion method is as follows: Data collected from different data acquisition devices / at different times is grouped according to the same target; the target includes equipment and personnel at the construction site; Based on the grouping results, construction site images, and laser point clouds, a 3D reconstruction of the construction scene was completed. The electric field intensity data collected at different locations were mapped onto the 3D model, thereby completing data fusion.

4. The intelligent monitoring method according to claim 1, characterized in that: In step 2, the dataset includes 5 categories of labels, namely, safety incident, violation of operation, equipment failure, approaching the safe distance, and below the safe distance.

5. The intelligent monitoring method according to claim 1, characterized in that: In step 4, the construction safety monitoring model adopts an improved loss function. As shown in the following formula: ； in, i Indicates the first i One sample; N The total number of samples; c Indicates the first of the tags c One category; C Total number of label categories; The weight of class c when reading sample i; For the sample i For the true value of the c-th category; For the model to sample i For category c The predicted probability; log is the logarithmic function; is the base of the logarithmic function.

6. The intelligent monitoring method according to claim 5, characterized in that: The formula for calculating the weight of class c when reading sample i is as follows: ； in, The weight of class c when reading sample i-1.

7. The intelligent monitoring method according to claim 1, characterized in that: In step 5, the construction safety hazards include violations of regulations, workers being too close to the safe distance from live equipment, and workers being too far from the safe distance from live equipment. The construction anomalies include safety accidents and equipment malfunctions.

8. The intelligent monitoring method according to claim 1 or 7, characterized in that: In step 5, an alarm is issued when a construction safety hazard or construction abnormality is detected; When a violation is detected, the operator is notified to stop the violation. When it is detected that the distance between the operator and the live equipment is close to the safe distance, the operator is notified to move away from the live equipment; When it is detected that the distance between the operator and the live equipment is lower than the safe distance, power outage measures shall be taken, and the operator shall be notified to move away from the live equipment; When a safety incident is detected, power outage measures should be taken, and all personnel should be notified to stay away from energized equipment; When a equipment malfunction is detected, power should be cut off, the equipment should be inspected, and all personnel should be notified to stay away from the equipment.

9. An intelligent monitoring device utilizing the intelligent monitoring method according to any one of claims 1-8, characterized in that, It includes a perception module, a data fusion and processing module, an intelligent analysis module, and an early warning system module: The sensing module collects historical data on near-electric construction based on lidar, cameras, and electric field sensors; The data fusion and processing module preprocesses the collected historical data of near-electric construction and integrates the preprocessed data into multi-source heterogeneous data based on a preset data fusion method; and constructs a dataset based on the multi-source heterogeneous data. The intelligent analysis module collects near-electricity construction data in real time, preprocesses the data according to step 2, integrates the preprocessed data into multi-source heterogeneous data, and inputs the multi-source heterogeneous data into the construction safety detection model trained in step 4 to detect construction safety hazards and construction anomalies in real time. The early warning system module issues an alarm for any detected construction safety hazards or construction anomalies.

10. A terminal, comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the intelligent monitoring method according to any one of claims 1-8.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the intelligent monitoring method according to any one of claims 1-8.

Images