Construction elevator safety control method and device based on multi-sensing

Abstract

This application provides a method and device for safety management of construction elevators based on multi-sensor perception. It achieves accurate verification of personnel identity and protective equipment status through multi-modal identity recognition using binocular liveness detection and voiceprint acquisition, combined with a deep neural network. An innovative equipment status assessment model based on multi-sensor fusion is designed to achieve coordinated monitoring of key parameters such as load, position, and speed. The system employs intelligent control strategies for adaptive adjustment of operating parameters, establishes a real-time fault protection mechanism based on threshold judgment, and achieves remote monitoring through IoT technology. This method breaks through the limitations of traditional elevator management models and provides an intelligent overall solution for construction safety management.

Description

Technical Field

[0001] This application relates to the field of data processing, specifically to a method and device for safety management of construction elevators based on multi-sensor perception. Background Technology

[0002] Traditional safety management methods for construction hoists mainly rely on simple mechanical safety devices and basic manual supervision, which are insufficient to meet the intelligent safety management needs of modern construction. Existing technologies have significant shortcomings in personnel identity verification and safety protection detection, and lack the ability to monitor the qualifications of construction workers and the wearing status of protective equipment in real time.

[0003] Meanwhile, existing systems also have limitations in equipment status monitoring and fault early warning. Traditional methods often use a single sensor for simple parameter monitoring, failing to achieve collaborative analysis of multi-dimensional data and lacking a comprehensive assessment mechanism for equipment operating status. The system is relatively lagging in fault identification and safety early warning, making it difficult to achieve timely fault detection and rapid response.

[0004] Furthermore, existing technologies need improvement in intelligent control and safety protection. They lack precise operational control strategies and adaptive adjustment mechanisms, failing to achieve dynamic optimization of equipment operating parameters. Safety protection systems often rely on fixed threshold judgments, lacking flexible fault protection strategies and remote monitoring capabilities.

[0005] In terms of human-computer interaction and remote monitoring, existing systems generally suffer from problems such as unintuitive information display and untimely alarm response. The lack of a robust IoT communication architecture prevents real-time transmission and remote monitoring of construction elevator safety information. Solving these problems is crucial for improving the safety and management efficiency of construction elevators. Summary of the Invention

[0006] To address the problems in the existing technology, this application provides a construction hoist safety management method and device based on multi-sensor perception, which can break through the limitations of the traditional hoist management mode and provide an intelligent overall solution for construction safety management.

[0007] To solve at least one of the above problems, this application provides the following technical solution:

[0008] Firstly, this application provides a safety management method for construction elevators based on multi-sensor perception, including:

[0009] Facial images of construction workers are captured by a binocular liveness detection camera, and voiceprint information is collected by a voiceprint collector. The facial images are input into a personnel verification model for identity feature extraction, and the voiceprint information is input into a voiceprint verification model for voiceprint feature extraction. Based on a deep neural network, the identity features and voiceprint features are verified against preset information of the construction workers. At the same time, images of the construction workers wearing protective equipment are captured, and these images are input into a posture recognition model. Based on the identity verification result, the voiceprint verification result, and the posture recognition result, the construction workers' permission information is generated and input into the elevator control authorization unit.

[0010] Weight sensors, position sensors, speed sensors, vibration sensors, limit sensors, door opening and closing sensors, and safety clamp sensors are installed at the bottom of the car, the side walls of the guide rails, and the edges of the door frame of the construction hoist to collect load data, position data, speed data, vibration data, travel data, door status data, and braking data of the construction hoist. The collected multi-dimensional data is input into a data preprocessing unit for noise reduction, filtering, and standardization. The preprocessed data is then input into an equipment status assessment model according to preset weights. Based on the status parameters output by the equipment status assessment model, the operating status information of the construction hoist is generated.

[0011] The operating status information is displayed on the monitoring touch screen, and the operating status information and the construction personnel's permission information are input into the intelligent control unit. The intelligent control unit generates operating control commands based on the load data, position data and speed data of the construction elevator, and adjusts the braking distance and leveling accuracy of the construction elevator according to the operating control commands. When the operating status information deviates from the preset threshold, a fault protection mechanism is triggered. The fault protection mechanism controls the safety clamp sensor to perform braking and sends an alarm signal to the safety supervision system through the Internet of Things communication module.

[0012] Furthermore, the process involves acquiring facial images of construction workers using a binocular liveness detection camera and voiceprint information using a voiceprint collector. The facial images are input into a personnel verification model for identity feature extraction, and the voiceprint information is input into a voiceprint verification model for voiceprint feature extraction. A deep neural network is then used to perform dual verification of the identity features and voiceprint features against preset information about the construction workers. This includes:

[0013] The facial images captured by the binocular liveness detection camera are preprocessed with image enhancement. The preprocessed facial images are then input into a convolutional neural network for facial key point detection. The coordinates of facial feature points, including the contours of the eyes, nose, and mouth, are extracted. A facial geometric feature vector is constructed based on the facial feature point coordinates. The facial geometric feature vector is then input into a personnel verification model to generate an identity feature code. The identity feature code is then encrypted and stored.

[0014] Continuous voice segments of construction workers are collected using a voiceprint collector. The voice segments are then processed by frame segmentation and windowing. Mel-frequency cepstral coefficient features are extracted from the voice segments. The extracted cepstral coefficient features are input into a voiceprint verification model to generate a voiceprint feature vector. The identity feature code and the voiceprint feature vector are then imported into a deep neural network. The deep neural network calculates the feature similarity score with a preset identity information database. The similarity score is then compared with a preset matching threshold to generate a verification result.

[0015] Further, the process of collecting images of construction workers wearing protective equipment, inputting these images into a posture recognition model, generating worker access information based on the identity verification result, the voiceprint verification result, and the posture recognition result, and then inputting this access information into the elevator control authorization unit includes:

[0016] Collect front and side image sequences of construction workers, perform target detection and image segmentation on the image sequences, extract the outline features of protective equipment including safety helmets, reflective vests and safety belts, input the outline features into a posture recognition model based on human skeletal key points, calculate the spatial positional relationship between the protective equipment and human skeletal key points through the posture recognition model, and generate a protective equipment wearing status feature map.

[0017] The protective equipment wearing status feature image, the identity verification result, and the voiceprint verification result are input into the permission evaluation model. The permission evaluation model is used to comprehensively analyze the construction personnel's identity level and compliance with protection specifications. Based on the analysis results, permission information containing operation permission level and operation time limit is generated. The permission information is written into the data register of the elevator control authorization unit to establish a mapping relationship between the construction personnel's identity and operation permission.

[0018] Furthermore, weight sensors, position sensors, speed sensors, vibration sensors, limit sensors, door opening / closing sensors, and safety gear sensors are installed at the bottom of the construction elevator car, the side walls of the guide rails, and the edges of the door frame to collect load data, position data, speed data, vibration data, travel data, door status data, and braking data of the construction elevator. The collected multi-dimensional data is then input into a data preprocessing unit for noise reduction filtering and standardization processing, including:

[0019] A weight sensor is installed at the bottom of the construction elevator car, and position sensors, speed sensors, and vibration sensors are installed on the side walls of the guide rails. Limit sensors, door opening and closing sensors, and safety clamp sensors are installed on the edge of the door frame. The weight sensor collects load data, the position sensor collects position data, the speed sensor collects speed data, the vibration sensor collects vibration data, the limit sensor collects travel data, the door opening and closing sensor collects door status data, and the safety clamp sensor collects braking data.

[0020] The collected multidimensional data is input into the data preprocessing unit, where wavelet transform is used to denoise the multidimensional data, a Butterworth low-pass filter is used to filter out high-frequency interference signals, zero-mean standardization is performed on the filtered data, and the standardized data is time-aligned according to the sampling timestamp to generate a preprocessed data matrix.

[0021] Further, the step of inputting the preprocessed data into the equipment status assessment model according to preset weights, and generating construction hoist operation status information based on the status parameters output by the equipment status assessment model, includes:

[0022] A preset weight matrix for constructing an equipment condition assessment model is used. Weight coefficients are assigned according to the degree of influence of each sensor data on the operating status of the construction elevator. The preprocessed data matrix and the preset weight matrix are multiplied by matrix to generate a weighted feature vector. The weighted feature vector is then extracted using a recurrent neural network. The extracted time-series features are then input into the equipment condition assessment model.

[0023] Based on the load parameters, operating speed parameters, position parameters, vibration parameters, door status parameters, and braking parameters output by the equipment condition assessment model, the parameters are combined to construct a condition feature space. The condition feature space is then nonlinearly mapped by a deep neural network to generate operating condition information that includes equipment operating status, component wear status, and safety threshold status.

[0024] Furthermore, the step of displaying the operating status information on the monitoring touch screen, and simultaneously inputting the operating status information and the construction personnel permission information into the intelligent control unit, wherein the intelligent control unit generates operating control commands based on the load data, position data, and speed data of the construction elevator, and adjusts the braking distance and leveling accuracy of the construction elevator according to the operating control commands, including:

[0025] The operation status display interface of the construction elevator is built on the monitoring touch screen. The equipment operation status, component wear status and safety threshold status in the operation status information are converted into graphical display data through the data visualization module. At the same time, the operation status information and the operation permission level and operation time limit in the construction personnel permission information are imported into the data cache area of ​​the intelligent control unit.

[0026] The load data, position data, and speed data are input into the data analysis module of the intelligent control unit. The data is analyzed in real time through a fuzzy neural network to establish a speed control model based on load and displacement. The calculation results of the speed control model are compared with the rated parameters of the construction elevator to generate operation control commands that include braking torque and leveling compensation values.

[0027] Furthermore, when the operating status information deviates from a preset threshold, a fault protection mechanism is triggered. This fault protection mechanism controls the safety clamp sensor to perform braking and sends an alarm signal to the safety monitoring system via the IoT communication module, including:

[0028] The system monitors the equipment operation status, component wear status, and safety threshold status in the operation status information in real time. It compares the monitoring data with the preset fault judgment threshold. When any status parameter exceeds the preset threshold, the fault protection mechanism is activated, and a braking control command is sent to the safety clamp sensor. The brake is then driven by the safety clamp sensor to perform emergency braking.

[0029] The deviation value between the status parameter that triggers the fault protection mechanism and the preset threshold is input into the IoT communication module. The deviation data is encrypted and encapsulated by the IoT communication module, a communication link with the safety supervision system is established, and the encrypted deviation data and the device number information are combined to form an alarm data packet. The alarm data packet is sent to the safety supervision system through the communication link.

[0030] Secondly, this application provides a safety control device for construction elevators based on multi-sensor perception, comprising:

[0031] The biometric detection module is used to collect facial images of construction workers through a binocular liveness detection camera and voiceprint information through a voiceprint collector. The facial images are input into a personnel verification model for identity feature extraction, and the voiceprint information is input into a voiceprint verification model for voiceprint feature extraction. Based on a deep neural network, the identity features and voiceprint features are verified against preset information of the construction workers. At the same time, images of the construction workers wearing protective equipment are collected and input into a posture recognition model. Based on the identity verification result, the voiceprint verification result, and the posture recognition result, the construction workers' permission information is generated and input into the elevator control authorization unit.

[0032] The status monitoring module is used to install weight sensors, position sensors, speed sensors, vibration sensors, limit sensors, door opening and closing sensors, and safety clamp sensors on the bottom of the car, the side walls of the guide rails, and the edges of the door frame of the construction hoist. It collects load data, position data, speed data, vibration data, travel data, door status data, and braking data of the construction hoist. The collected multi-dimensional data is input into the data preprocessing unit for noise reduction filtering and standardization processing. The preprocessed data is input into the equipment status assessment model according to the preset weights. Based on the status parameters output by the equipment status assessment model, the operating status information of the construction hoist is generated.

[0033] The safety management module displays the operating status information on the monitoring touch screen and inputs the operating status information and the construction personnel's permission information into the intelligent control unit. The intelligent control unit generates operating control commands based on the load data, position data, and speed data of the construction elevator. According to the operating control commands, it adjusts the braking distance and leveling accuracy of the construction elevator. When the operating status information deviates from a preset threshold, a fault protection mechanism is triggered. The fault protection mechanism controls the safety clamp sensor to perform braking and sends an alarm signal to the safety supervision system through the Internet of Things communication module.

[0034] Thirdly, this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the construction elevator safety management method based on multi-sensor perception.

[0035] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the construction elevator safety management method based on multi-sensor perception.

[0036] Fifthly, this application provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of the aforementioned multi-sensor-based construction elevator safety management method.

[0037] As described above, this application provides a method and device for safety management of construction elevators based on multi-sensor perception. It achieves accurate verification of personnel identity and protective equipment status through multi-modal identity recognition using binocular liveness detection and voiceprint acquisition, combined with deep neural networks. An innovative equipment status assessment model based on multi-sensor fusion is designed to achieve coordinated monitoring of key parameters such as load, position, and speed. The system employs intelligent control strategies for adaptive adjustment of operating parameters, establishes a real-time fault protection mechanism based on threshold judgment, and achieves remote monitoring through IoT technology. This method breaks through the limitations of traditional elevator management models and provides an intelligent overall solution for construction safety management. Attached Figure Description

[0038] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 This is one of the flowcharts illustrating the multi-sensor-based safety management method for construction elevators in this application embodiment;

[0040] Figure 2 This is the second flowchart illustrating the multi-sensor-based safety management method for construction elevators in this application.

[0041] Figure 3 This is the third flowchart illustrating the multi-sensor-based safety management method for construction elevators in this application.

[0042] Figure 4 This is the fourth flowchart illustrating the multi-sensor-based safety management method for construction elevators in this application.

[0043] Figure 5 This is the fifth flowchart illustrating the multi-sensor-based safety management method for construction elevators in this application.

[0044] Figure 6 This is the sixth flowchart illustrating the multi-sensor-based safety management method for construction elevators in this application.

[0045] Figure 7 This is the seventh flowchart illustrating the multi-sensor-based safety management method for construction elevators in this application.

[0046] Figure 8 This is a structural diagram of the construction elevator safety control device based on multi-sensor perception in the embodiments of this application;

[0047] Figure 9 This is a schematic diagram of the structure of the electronic device in the embodiments of this application.

[0048] Figure label:

[0049] Electronic device 9600, central processing unit 9100, memory 9140, communication module 9110, input unit 9120, audio processor 9130, display 9160, power supply 9170, buffer memory 9141, application / function storage unit 9142, data storage unit 9143, driver storage unit 9144, antenna 9111, speaker 9131, microphone 9132. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0051] The acquisition, storage, use, and processing of data in this application all comply with the relevant provisions of national laws and regulations.

[0052] To address the problems existing in current technologies, this application provides a method and device for safety management of construction elevators based on multi-sensor perception. It achieves accurate verification of personnel identity and protective equipment status through multi-modal identity recognition using binocular liveness detection and voiceprint acquisition, combined with deep neural networks. An innovative equipment status assessment model based on multi-sensor fusion is designed to achieve coordinated monitoring of key parameters such as load, position, and speed. The system employs intelligent control strategies for adaptive adjustment of operating parameters, establishes a real-time fault protection mechanism based on threshold judgment, and enables remote monitoring through IoT technology. This method overcomes the limitations of traditional elevator management models and provides an intelligent overall solution for construction safety management.

[0053] To overcome the limitations of traditional elevator management models and provide an intelligent overall solution for construction safety management, this application provides an embodiment of a multi-sensor-based construction elevator safety control method. See [link to embodiment]. Figure 1 The construction elevator safety management method based on multi-sensor perception specifically includes the following:

[0054] Step S101: Facial images of construction workers are captured by a binocular liveness detection camera, and voiceprint information is collected by a voiceprint collector. The facial images are input into a personnel verification model for identity feature extraction, and the voiceprint information is input into a voiceprint verification model for voiceprint feature extraction. The identity features and voiceprint features are then double-verified against the preset information of the construction workers based on a deep neural network. Simultaneously, images of the construction workers wearing protective equipment are captured, and these images are input into a posture recognition model. Based on the identity verification result, the voiceprint verification result, and the posture recognition result, construction worker permission information is generated, and the permission information is input into the elevator control authorization unit.

[0055] Optionally, in this embodiment, a binocular liveness detection camera is installed at the entrance of the construction elevator, employing binocular depth imaging technology to achieve 3D facial reconstruction. Each camera has a resolution of 2048×1536 pixels and a field of view of 85 degrees. The baseline distance between the two cameras is 12 centimeters, and binocular calibration ensures that the optical axes are parallel. The system acquires visible light images and 850nm infrared images, combining them with depth information to achieve liveness detection. For example, when a discontinuity in facial depth or an abnormal depth value is detected, it is determined to be planar deception such as from a photograph.

[0056] In this embodiment, facial image preprocessing first involves adaptive histogram equalization to enhance contrast. For images with uneven lighting, they are divided into multiple 8×8 local regions for individual enhancement. Then, a Gaussian low-pass filter is used to eliminate noise, with a standard deviation of 1.5 to ensure that facial details are preserved while denoising. Next, coarse localization of the face region is performed based on the YCbCr skin color model, and precise segmentation is achieved through edge detection using the Sobel operator.

[0057] This embodiment employs an improved ResNet-50 network structure in the personnel verification model. The network comprises five stages, each using a different number of residual blocks. Residual connections effectively mitigate the vanishing gradient problem. The attention mechanism is implemented through a combination of spatial attention and channel attention modules, focusing on key facial regions such as the eyes, nose, and mouth. The model is trained using the ArcFace loss function, and inter-class distances are increased by adding angular boundaries.

[0058] In this embodiment, an 8-channel circular microphone array with a microphone spacing of 4.5 cm is used for voiceprint acquisition. A delay-summation beamforming algorithm is employed to enhance the sound source signal in the target direction while suppressing interference from other directions. Construction personnel are required to read preset commands such as "Construction safety first, safe production is important." A dual-threshold endpoint detection algorithm is used, with the energy threshold set to twice the background noise energy and the zero-crossing rate threshold set to 25 times per frame, to accurately segment effective speech segments.

[0059] In this embodiment, the speech signal is pre-emphasized during the voiceprint feature extraction process, with the pre-emphasis coefficient set to 0.97. A 25ms frame length and a 10ms frame shift are used for framing, and a Hamming window is employed to reduce spectral leakage. Thirteen-dimensional Mel-frequency cepstral coefficients are extracted using a 32-triangular Mel filter bank, and first-order and second-order difference coefficients are calculated to obtain a 39-dimensional feature vector. To improve robustness, cepstral mean subtraction is used for channel compensation.

[0060] In this embodiment, a ResNet-based temporal convolutional network is used in the voiceprint verification model. The network contains multiple residual blocks, each containing two one-dimensional convolutional layers with a kernel size of 3. Dilated convolutions expand the receptive field, effectively modeling long-term temporal dependencies. The model is trained using the GE2E loss function, simultaneously optimizing speaker recognition and speech verification tasks. The verification threshold is determined using an ROC curve to balance the false alarm rate and the false negative rate.

[0061] This embodiment employs an improved YOLOv3 algorithm for protective equipment detection. The backbone network uses DarkNet-53, achieving multi-scale detection through a feature pyramid structure. The detection head outputs feature maps at three different scales to detect large, medium, and small targets respectively. For the protective equipment detection scenario, suitable anchor box sizes were designed, and nine optimal anchor boxes were obtained through K-means clustering. The model can simultaneously detect protective equipment such as helmets, reflective vests, and seat belts, outputting bounding box coordinates and confidence scores.

[0062] This embodiment employs an improved HRNet network in the pose recognition model. Multi-scale features are extracted through parallel multi-resolution sub-networks to maintain high-resolution representation. Seventeen human keypoints are designed, including the head, torso, and limbs. Spatial relationships between keypoints are modeled using a graph convolutional network, with the adjacency matrix of the graph predefined based on the human skeletal structure. The model outputs a heatmap representing the probability distribution of keypoint locations, and Gaussian kernel smoothing is used to improve localization accuracy.

[0063] In this embodiment, an attention-based multimodal feature fusion network is used in the permission information generation stage. Facial features, voiceprint features, and pose features are first mapped to the same dimension through independent fully connected layers, and then the correlation weights between features are calculated through a self-attention mechanism. According to the safety management requirements of the construction site, the permission levels are divided into three levels: administrator, operator, and ordinary user, with each level corresponding to different operation scopes and time limits.

[0064] This embodiment implements fine-grained access control based on RBAC in the elevator control authorization unit. Access information includes fields such as user ID, access level, operation scope, and validity period, and is stored using RSA encryption. The authorization unit maintains an access policy table, defining the set of operations that can be executed at different access levels, supporting dynamic adjustment and real-time verification of permissions. This scheme achieves reliable verification of construction personnel identity and protective equipment, ensuring the safe use of the elevator.

[0065] Step S102: Install weight sensors, position sensors, speed sensors, vibration sensors, limit sensors, door opening / closing sensors, and safety clamp sensors at the bottom of the car, the side walls of the guide rails, and the edge of the door frame of the construction hoist. Collect load data, position data, speed data, vibration data, travel data, door status data, and braking data of the construction hoist. Input the collected multidimensional data into the data preprocessing unit for noise reduction filtering and standardization. Input the preprocessed data into the equipment status assessment model according to the preset weights. Generate the construction hoist operating status information based on the status parameters output by the equipment status assessment model.

[0066] Optionally, in this embodiment, four high-precision tension-type weight sensors are installed at the bottom of the construction elevator car. The sensors employ a full-bridge strain gauge design, with a measurement range of 0-5000 kg and a resolution of 0.1 kg. The sensors are positioned at the four corners of the car bottom, eliminating the influence of off-center loading through multi-point measurement. The output signal of each sensor is sampled by a 24-bit AD converter at a sampling frequency of 100 Hz, enabling real-time acquisition of load data.

[0067] In this embodiment, an optical encoder is installed on the side wall of the guide rail as a position sensor, employing a combination of incremental and absolute encoders. The incremental encoder has a resolution of 4096 pulses / revolution for high-precision displacement measurement; the absolute encoder uses 17-bit Gray code for power-on positioning. The signals from both encoders are acquired through a dedicated interface circuit, achieving redundant backup of the position data.

[0068] In this embodiment, a Hall effect speed sensor is installed on the side wall of the guide rail. The sensor uses a differential output method to improve anti-interference capability. The speed signal is processed by a bandpass filter with a cutoff frequency of 0.5-50Hz to filter out power frequency interference and high-frequency noise. The sensor sampling frequency is 200Hz, and speed and acceleration information are obtained through digital integration and differentiation operations.

[0069] In this embodiment, a triaxial MEMS vibration sensor is installed at the bottom of the car and the connection with the guide rail. The measurement range is ±16g, and the bandwidth is 5kHz. The sensor adopts a distributed arrangement scheme, with multiple measuring points deployed at key structural locations. The spatial distribution characteristics of the vibration signal are collected through a sensor array. The signal acquisition adopts a synchronous trigger mode to ensure the temporal consistency of multi-channel data.

[0070] In this embodiment, travel limit sensors are installed at the top and bottom of the elevator guide rails, employing a normally closed safety contact design. The limit sensors work in conjunction with a mechanical buffer device; when the car approaches its limit position, the sensors trigger emergency braking. The sensor output signal undergoes opto-isolation processing to improve electrical safety.

[0071] In this embodiment, Hall effect door switch sensors are installed at the car door and landing door, employing a dual-redundancy design to ensure reliability. The sensors detect the engagement state of the door lock mechanism, and the output signal is processed by a logic judgment circuit to achieve real-time monitoring of the door status. Simultaneously, current detection is incorporated into the door operator drive circuit to monitor the door operator's operating status.

[0072] In this embodiment, a displacement sensor and a force sensor are installed on the safety clamp mechanism to monitor the operating status and braking force of the safety clamp in real time. The displacement sensor uses a linear potentiometer to detect the position of the safety clamp wedge; the force sensor uses a piezoelectric design to monitor the magnitude of the braking force. The sensor signals are processed by a signal conditioning circuit to achieve precise monitoring of the braking process.

[0073] In this embodiment, wavelet packet decomposition is used for signal noise reduction during data preprocessing. The vibration signal undergoes a 5-level wavelet decomposition, with the db4 wavelet basis selected, and high-frequency coefficients are processed using a soft thresholding method. The position and velocity signals are processed using a Kalman filtering algorithm, with filter parameters dynamically adjusted based on real-time noise estimation. The preprocessed data is then normalized in amplitude and time-aligned.

[0074] This embodiment employs a deep convolutional neural network structure in the equipment status assessment model. The model input includes time-series data from seven sensor channels, and temporal features are extracted through one-dimensional convolutional layers. The network incorporates residual connections and attention mechanisms to enhance its ability to identify key patterns. Model training utilizes historical equipment operating data, with labels including normal, warning, and fault states.

[0075] In this embodiment, the importance weight of each sensor data point is determined based on expert experience and data analysis. Load and position data have higher weights, directly related to operational safety; vibration and velocity data have lower weights, used for status monitoring; door status and limit switch data serve as auxiliary judgment criteria. The weight values ​​are optimized and adjusted through model validation.

[0076] This embodiment achieves accurate assessment of the elevator's operating status through multi-sensor collaborative sensing and deep learning analysis. This solution can promptly detect equipment anomalies, predict potential faults, and provide a basis for preventative maintenance decisions. In practical applications, it significantly improves the reliability and safety of equipment operation.

[0077] Step S103: Display the operating status information on the monitoring touch screen, and simultaneously input the operating status information and the construction personnel permission information into the intelligent control unit. The intelligent control unit generates operating control commands based on the load data, position data, and speed data of the construction elevator, and adjusts the braking distance and leveling accuracy of the construction elevator according to the operating control commands. When the operating status information deviates from the preset threshold, a fault protection mechanism is triggered. The fault protection mechanism controls the safety clamp sensor to perform braking and sends an alarm signal to the safety supervision system through the Internet of Things communication module.

[0078] Optionally, in this embodiment, a 15.6-inch capacitive touchscreen display with a resolution of 1920×1080 pixels is installed in the control room of the construction elevator, and it adopts an industrial-grade dustproof and waterproof design. The display interface adopts a partitioned layout. The main area displays real-time operating parameters, including current load, operating speed, location information, and door status; the status area displays equipment health and fault warning information; and the operation area provides emergency braking and remote alarm buttons.

[0079] This embodiment employs a dual-CPU redundancy design in the intelligent control unit. The main CPU is responsible for operation and control, while the backup CPU performs status monitoring. The controller uses a real-time operating system with a task scheduling cycle of 10ms to ensure timely response to control commands. The control unit communicates with each sensor module via a CAN bus, and differential signal transmission is used to improve anti-interference capabilities.

[0080] In this embodiment, an adaptive control algorithm based on a fuzzy neural network is implemented in the operation control strategy. The controller input includes load data, position data, and speed data, and the optimal control quantity is calculated through fuzzy inference rules. For example, when the load is large, the acceleration and deceleration are automatically reduced; when approaching the target layer, the braking torque is dynamically adjusted according to the distance to achieve a smooth stop.

[0081] In this embodiment, a feedforward-feedback combined control strategy is adopted during braking control. Feedforward control estimates the braking force requirement based on load characteristics and position information, while feedback control adjusts the braking force based on real-time speed deviation. The controller outputs a PWM signal to drive the brake; the PWM frequency is 20kHz, and the braking force is continuously adjusted by changing the duty cycle.

[0082] In this embodiment, a two-stage deceleration strategy is designed for leveling control. When the distance to the target floor is 2 meters, a coarse adjustment phase begins, reducing the speed to 30% of the rated speed. When the distance is 0.3 meters, a fine adjustment phase begins, using a crawling speed to approach the target position. Position sensor feedback signals are used to correct stopping errors in real time, ensuring leveling accuracy.

[0083] This embodiment establishes a multi-level threshold judgment standard in the fault protection mechanism. Warning thresholds and alarm thresholds are set for operating parameters. When a parameter exceeds the warning threshold, the controller initiates preventative protection measures; when it exceeds the alarm threshold, emergency braking is immediately triggered. For example, when the speed or load exceeds the limit, the safety brake is automatically activated.

[0084] In this embodiment, the safety clamp control circuit employs a watchdog circuit and redundant power supply design. Control signals are protected by both opto-isolation and relays to ensure reliable execution of braking commands. After the safety clamp actuates, the braking effect is monitored by displacement and force sensors, forming a closed-loop control system.

[0085] In this embodiment, the IoT communication module adopts a 4G / 5G dual-mode communication scheme, supporting automatic switching of communication networks. Alarm data uses an encrypted transmission protocol, and the data packet includes the device ID, fault type, fault parameters, and timestamp. The communication module is equipped with a heartbeat mechanism to monitor the connection status with the monitoring platform in real time.

[0086] This embodiment implements a tiered early warning mechanism in the early warning information processing flow. Based on the severity of the fault, it is divided into three levels: alert, warning, and alarm, with each level corresponding to a different handling strategy. Alert-level faults are logged and operation continues; warning-level faults limit the operating speed; and alarm-level faults trigger emergency braking and are reported to the monitoring platform.

[0087] This embodiment implements permission-based operation control in its human-computer interaction design. Operators with different permission levels have different permitted operation scopes; for example, administrators can modify control parameters, while ordinary operators can only perform basic operational tasks. Operation logs are recorded in real time and uploaded to the monitoring platform.

[0088] This embodiment effectively ensures the safe operation of the construction hoist through a multi-layered safety protection mechanism and intelligent control strategy. The solution can promptly detect and handle various abnormal situations, and achieve full monitoring of equipment operation through remote supervision, significantly improving the reliability and safety of the construction hoist and providing reliable protection for the transportation of personnel and materials at the construction site.

[0089] As described above, the construction elevator safety management method based on multi-sensor perception provided in this application can accurately verify personnel identity and protective equipment status through multi-modal identity recognition using binocular liveness detection and voiceprint acquisition, combined with deep neural networks. It innovatively designs an equipment status assessment model based on multi-sensor fusion to achieve collaborative monitoring of key parameters such as load, position, and speed. The system adopts an intelligent control strategy for adaptive adjustment of operating parameters, establishes a real-time fault protection mechanism based on threshold judgment, and achieves remote monitoring through IoT technology. This method breaks through the limitations of traditional elevator management models and provides an intelligent overall solution for construction safety management.

[0090] In one embodiment of the construction hoist safety management method based on multi-sensoring of this application, see [link to relevant documentation]. Figure 2 It can also specifically include the following:

[0091] Step S201: Perform image enhancement preprocessing on the facial images captured by the binocular liveness detection camera, input the preprocessed facial images into a convolutional neural network for facial key point detection, extract the coordinates of facial feature points including the contours of the eyes, nose and mouth, construct a facial geometric feature vector based on the facial feature point coordinates, input the facial geometric feature vector into a personnel verification model to generate an identity feature code, and encrypt and store the identity feature code.

[0092] Step S202: Collect continuous speech segments of construction personnel using a voiceprint collector, perform frame-by-frame windowing processing on the speech segments, extract Mel-frequency cepstral coefficient features from the speech segments, input the extracted cepstral coefficient features into the voiceprint verification model to generate a voiceprint feature vector, import the identity feature code and the voiceprint feature vector into a deep neural network, calculate the feature similarity score with the preset identity information database through the deep neural network, and compare the similarity score with a preset matching threshold to generate a verification result.

[0093] Optionally, this embodiment first uses a binocular liveness detection camera to simultaneously acquire frontal and side images of construction workers. The camera employs an infrared supplementary lighting design to ensure image quality in low-light environments. The binocular camera obtains intrinsic and extrinsic parameter matrices through calibration to reconstruct depth information, effectively preventing deception methods such as using flat photographs. The camera resolution is 2048×1536 pixels, with a frame rate of 30fps, and supports autofocus.

[0094] In this embodiment, during the image enhancement preprocessing stage, illumination compensation is first performed using an adaptive enhancement algorithm based on the brightness histogram. For backlit or shadowed areas, gamma correction is used to enhance local details. Then, bilateral filtering is used for noise reduction while preserving edge information. The spatial domain standard deviation of the filter is set to 3, and the value domain standard deviation is set to 25, achieving a balance between noise suppression and edge preservation.

[0095] In this embodiment, an improved MTCNN cascaded convolutional network is used for the face detection and alignment stages. The network consists of three sub-networks: P-Net, R-Net, and O-Net, which respectively perform candidate box generation, candidate box optimization, and key point localization. Each sub-network contains convolutional layers, pooling layers, and fully connected layers, and detects faces of different sizes through a multi-scale feature pyramid.

[0096] In this embodiment, 68 facial feature points are located during keypoint detection, including 12 points for the eye contour, 10 points for the eyebrows, 9 points for the nose, 20 points for the mouth, and 17 points for the face contour. Heatmap regression is used to predict the feature point locations, and a deep residual network is used to extract multi-level features. The network uses the Wing loss function to increase the training weights for samples with small biases.

[0097] In this embodiment, geometric features are calculated based on detected key points during the feature vector construction stage. These features include the inter-eye distance ratio, the nose-to-mouth distance ratio, and facial contour symmetry. Simultaneously, local texture features are extracted, and the LBP operator is used to describe facial details. The geometric and texture features are then concatenated to form a 512-dimensional feature vector.

[0098] In this embodiment, a high-fidelity digital microphone with a sampling rate of 16kHz and quantization precision of 24 bits is used for voiceprint acquisition. A dual-channel noise reduction algorithm is employed to suppress ambient noise, and adaptive beamforming technology is used to enhance the target sound source. Construction personnel are required to read a preset command, and a 3-5 second voice clip is collected.

[0099] In this embodiment, framing is performed using a 25ms frame length and a 10ms frame shift during speech preprocessing. A Hamming window is used to reduce spectral leakage, and the pre-emphasis coefficient is set to 0.97. Valid speech segments are detected using a dual threshold of short-time energy and zero-crossing rate, and silent segments are removed. Endpoint detection and silence removal are performed on the speech signal.

[0100] In this embodiment, during feature extraction, the power spectrum is calculated using a 32-point FFT, and Mel-spectral features are extracted using a 40-triangular filter bank. A 13-dimensional Mel-frequency cepstral coefficient is obtained through discrete cosine transform. First-order and second-order difference coefficients are calculated to form a 39-dimensional feature vector. Channel compensation is performed using cepstral mean subtraction.

[0101] In this embodiment, a ResNet-based temporal convolutional network is used in the voiceprint verification model. The network contains multiple residual blocks, each containing two one-dimensional convolutional layers and a batch normalization layer. Dilated convolutions are used to expand the receptive field and model long-term temporal dependencies. The model training uses a triplet loss function to increase inter-class distance and decrease intra-class distance.

[0102] In this embodiment, facial features and voiceprint features are input into a deep neural network during the feature fusion stage. The network uses an attention mechanism to dynamically adjust the weights of the two features and adaptively fuses them based on feature reliability. The cosine similarity with a preset feature library is calculated to generate a similarity score. The verification threshold is determined through ROC curve optimization to balance the false alarm rate and the missed alarm rate.

[0103] This embodiment uses the AES-256 algorithm to encrypt and store the verified identity signature code, and the key is managed through a hardware encryption module. The stored record includes the signature code, timestamp, and verification result, supporting both offline and online verification mechanisms. This scheme achieves reliable verification of construction personnel's identities, effectively preventing impersonation and substitution of work personnel.

[0104] In one embodiment of the construction hoist safety management method based on multi-sensoring of this application, see [link to relevant documentation]. Figure 3 It can also specifically include the following:

[0105] Step S301: Collect front and side image sequences of construction workers, perform target detection and image segmentation processing on the image sequences, extract the outline features of protective equipment including safety helmets, reflective vests and safety belts, input the outline features into a posture recognition model based on human skeletal key points, calculate the spatial positional relationship between the protective equipment and the human skeletal key points through the posture recognition model, and generate a protective equipment wearing status feature map.

[0106] Step S302: Input the protective equipment wearing status feature image, the identity verification result, and the voiceprint verification result into the permission evaluation model. Through the permission evaluation model, comprehensively analyze the construction personnel's identity level and compliance with protection specifications. Based on the analysis results, generate permission information containing operation permission level and operation time limit. Write the permission information into the data register of the elevator control authorization unit to establish a mapping relationship between the construction personnel's identity and operation permission.

[0107] Optionally, in this embodiment, two industrial-grade high-speed cameras are deployed at the entrance of the construction elevator to capture frontal and side images of construction workers, respectively. The cameras employ a global shutter design and are equipped with image-stabilized gyroscopes to achieve clear imaging even in high-speed motion scenarios. The illumination compensation system includes an adaptive LED array that dynamically adjusts the supplementary lighting parameters according to the ambient light intensity, ensuring consistent image quality at different times.

[0108] This embodiment first preprocesses the acquired image sequence, employing adaptive histogram equalization to enhance contrast and locally enhancing shadow and overexposed areas. Bilateral filtering is used for noise reduction, with the spatial domain standard deviation varying with the image gradient, effectively suppressing noise while preserving edge details. Gamma correction optimizes the image's dynamic range and enhances shadow details.

[0109] This embodiment employs an improved YOLOv5 network in the target detection phase, optimizing the network structure for the detection scenario of protective equipment at construction sites. The backbone network uses CSPDarknet53, and the computational load is reduced through cross-stage local networks. Three detection heads are designed: a large-scale detection head handles large targets such as reflective vests, a medium-scale detection head handles safety helmets, and a small-scale detection head handles delicate components such as safety belt buckles.

[0110] In this embodiment, a dataset of protective equipment with varying lighting, angles, and occlusions was constructed during the training of the detection network. A multi-scale training strategy was employed, with the input image size randomly varying between 416 and 832 pixels. The loss function comprehensively considered bounding box regression, confidence prediction, and category classification, with increased weighting for the detection of key parts. Transfer learning was used to pre-train the detection network, improving the model's generalization ability in small-sample scenarios.

[0111] In this embodiment, an improved Mask R-CNN model is used to achieve accurate segmentation of protective equipment during image segmentation. The network uses ResNet101-FPN as the feature extraction backbone and fuses multi-scale features through a feature pyramid network. The ROIAlign layer maintains the spatial accuracy of the feature map and improves the segmentation effect of small targets. The segmentation head adopts a four-layer convolutional structure and outputs a high-precision 28×28 mask.

[0112] In this embodiment, morphological processing and contour extraction are performed on the segmentation mask during the contour feature extraction stage. Noise is removed through opening operations, holes are filled through closing operations, and the reconstruction operation preserves the target shape. A curvature-based contour simplification algorithm is employed to retain key contour points. A shape descriptor, including geometric features such as perimeter ratio, compactness, and eccentricity, is calculated to evaluate the standardization of protective equipment.

[0113] In this embodiment, an improved HRNet architecture is used in the pose recognition model to detect key points on the human skeleton. The network contains four parallel branches, maintaining feature maps at different resolutions. Multi-scale contextual information is fused onto the high-resolution feature maps through repeated multi-scale fusion modules. A heatmap of 17 key points is output, and Gaussian kernel smoothing is used to improve localization accuracy.

[0114] This embodiment designs a wearing status evaluation method based on spatial geometric constraints. A spatial correlation model between protective equipment and key points on the human body is established, such as the relative position and tilt angle of the helmet and key points on the head, the overlap degree of the reflective vest with the torso contour, and the fixation relationship between the seat belt and the hip position. The degree of proper wearing is evaluated by calculating the deviation values ​​of these geometric features.

[0115] This embodiment employs a multimodal feature fusion architecture based on graph neural networks in its access control assessment model. Protective equipment status features, authentication results, and voiceprint features are constructed as nodes, and the correlation strength between nodes is calculated using an attention mechanism. The model iteratively updates node features, ultimately outputting an access control level score and operation time limit suggestions. Differentiated access control standards are set according to construction site safety management regulations.

[0116] This embodiment implements dynamic permission management based on spatiotemporal constraints in the elevator control authorization process. Permission information includes fields such as user identity, level, operation scope, and timeliness, and a distributed storage structure is used to improve access efficiency. The authorization strategy supports multi-dimensional control based on time windows, work areas, and equipment status, ensuring the safe and standardized use of construction elevators. This solution intelligently associates the wearing status of protective equipment with operating permissions, establishing a complete access management mechanism for construction personnel.

[0117] In one embodiment of the construction hoist safety management method based on multi-sensoring of this application, see [link to relevant documentation]. Figure 4 It can also specifically include the following:

[0118] Step S401: Install a weight sensor at the bottom of the construction elevator car, and install position sensors, speed sensors, and vibration sensors on the side walls of the guide rails. Install limit sensors, door opening / closing sensors, and safety clamp sensors on the edge of the door frame. Collect load data through the weight sensor, position data through the position sensor, speed data through the speed sensor, vibration data through the vibration sensor, travel data through the limit sensor, door status data through the door opening / closing sensor, and braking data through the safety clamp sensor.

[0119] Step S402: Input the collected multidimensional data into the data preprocessing unit, use wavelet transform to perform noise reduction on the multidimensional data, filter out high-frequency interference signals through Butterworth low-pass filter, perform zero-mean standardization on the filtered data, and align the standardized data according to the sampling timestamp to generate a preprocessed data matrix.

[0120] Optionally, in this embodiment, high-precision tension-type weight sensors are installed at the four corners of the bottom of the construction elevator car. The sensors adopt a full-bridge strain gauge design and are equipped with a temperature compensation circuit to eliminate the influence of temperature drift. The signal from each sensor is sampled by a 24-bit ADC at a sampling frequency of 100Hz, enabling real-time acquisition of load data from multiple points. The sensor signals are processed by an amplification and conditioning circuit, and differential transmission is used to improve anti-interference capability.

[0121] This embodiment uses a combined position sensor mounted on the sidewall of the guide rail, including an incremental photoelectric encoder and an absolute magnetic scale. The incremental encoder is used for high-precision displacement measurement with a resolution of 0.1 mm; the absolute magnetic scale is used for position calibration and power-on positioning. Signals from both sensors are acquired through a dedicated interface circuit, enabling redundant backup of the position data.

[0122] In this embodiment, a Hall effect speed sensor is installed on the side wall of the guide rail, and a differential output method is used to improve signal quality. The sensor output is processed by a bandpass filter with a cutoff frequency of 0.5-50Hz, effectively filtering out power frequency interference and high-frequency noise. The sampling frequency is set to 200Hz, and speed and acceleration information are obtained simultaneously through digital integration and differentiation operations.

[0123] In this embodiment, a triaxial MEMS vibration sensor array is deployed at the bottom of the car and the connection with the guide rail. The sensors adopt a distributed arrangement to acquire the spatial distribution characteristics of the vibration signal. The vibration signal sampling frequency is 1kHz, and synchronous triggering ensures the timing consistency of multi-channel data. The signal conditioning circuit uses a programmable gain amplifier to dynamically adjust the amplification factor according to the vibration amplitude.

[0124] In this embodiment, travel limit sensors are installed at the top and bottom of the elevator guide rails, employing a normally closed safety contact design. The limit sensors work in conjunction with a mechanical buffer device, triggering a braking signal when the car approaches its limit position. The sensor outputs are photoelectrically isolated to ensure electrical safety, and redundant contacts are provided to improve reliability.

[0125] In this embodiment, Hall effect door switch sensors are installed at the car door and landing door, employing a dual-redundancy design to ensure safety. The sensors detect the engagement state of the door lock mechanism, and the output signal is processed by a logic judgment circuit to achieve real-time monitoring of the door status. Simultaneously, current detection is incorporated into the door operator drive circuit to monitor the door operator's operating status.

[0126] In this embodiment, a displacement sensor and a force sensor are installed on the safety clamp mechanism to achieve precise monitoring of the braking state. The displacement sensor uses a linear potentiometer to detect the position of the safety clamp wedge; the force sensor uses a piezoelectric design to monitor the magnitude of the braking force. The sensor signals are processed by a signal conditioning circuit to achieve closed-loop control of the braking process.

[0127] In this embodiment, wavelet transform is first used for noise reduction during data preprocessing. The vibration signal is decomposed into five levels of wavelets, and the db4 wavelet basis function is selected. High-frequency coefficients are processed using a soft thresholding method. An improved thresholding function is used when reconstructing the signal to avoid over-smoothing that could lead to the loss of useful information.

[0128] This embodiment designs a fourth-order Butterworth low-pass filter, with the cutoff frequency dynamically set according to the characteristics of different sensor signals. The filter employs a bidirectional filtering strategy to eliminate phase delay and uses a buffer queue for real-time processing. For abrupt signals, an adaptive filtering algorithm is used to maintain the signal's fast response characteristics.

[0129] In this embodiment, a sliding window is used to calculate the local mean and standard deviation during data standardization, with a window length of 1 second. Zero-mean standardization eliminates dimensional differences between different sensors, improving data comparability. Outliers are handled using the median replacement method to maintain data continuity.

[0130] This embodiment implements a timestamp-based multidimensional data alignment mechanism. A unified clock source is used to provide the sampling trigger signal, ensuring the synchronization of multi-channel data. An interpolation algorithm is used to process data with inconsistent sampling frequencies, generating an equally spaced time-series data matrix. This scheme achieves high-quality acquisition and preprocessing of multi-source sensor data from construction elevators, providing a reliable data foundation for subsequent condition assessment.

[0131] In one embodiment of the construction hoist safety management method based on multi-sensoring of this application, see [link to relevant documentation]. Figure 5 It can also specifically include the following:

[0132] Step S501: Construct a preset weight matrix for the equipment status assessment model, assign weight coefficients according to the influence of each sensor data on the operating status of the construction elevator, perform matrix multiplication operation between the preprocessed data matrix and the preset weight matrix to generate a weighted feature vector, extract time-series features from the weighted feature vector through a recurrent neural network, and input the extracted time-series features into the equipment status assessment model.

[0133] Step S502: Based on the load parameters, operating speed parameters, position parameters, vibration parameters, door status parameters and braking parameters output by the equipment status assessment model, combine the parameters to construct a status feature space, and perform nonlinear mapping on the status feature space through a deep neural network to generate operating status information including equipment operating status, component wear status and safety threshold status.

[0134] Optionally, this embodiment first constructs a weight allocation mechanism based on expert knowledge, designing an 8×8 preset weight matrix. The weight coefficients are determined according to the degree of influence of each sensor data on the safety of the elevator operation, and a judgment matrix is ​​established using the analytic hierarchy process (AHP). For example, load data and speed data have a greater impact on safety and are assigned higher weights; vibration data is more critical for assessing equipment wear status and is given higher weights in the corresponding dimensions.

[0135] This embodiment considers the correlation and reliability of sensor data in the weight matrix design. Correlation coefficients between sensors are calculated through historical data analysis, and the weights of highly correlated data are appropriately reduced to avoid information redundancy. Simultaneously, a reliability compensation coefficient is set based on the sensor failure rate and measurement error, dynamically adjusting the weight values.

[0136] This embodiment employs an improved LSTM recurrent neural network for temporal feature extraction. The network consists of three LSTM layers with 128, 64, and 32 nodes per layer, respectively, and selectively memorizes long-term dependencies through a gating mechanism. The input layer receives weighted feature vectors, the hidden layers use dropout to prevent overfitting, and the output layer extracts temporal feature representations.

[0137] In this embodiment, a sequence-to-sequence training mode is used in the LSTM network training. The input sequence length is 60 seconds, the stride is 1 second, and training samples are generated through a sliding window. The loss function combines mean squared error and Huber loss to balance the impact of outliers. The Adam optimizer is used to dynamically adjust the learning rate to improve the model's convergence speed.

[0138] This embodiment employs an attention-enhanced deep neural network in the state evaluation model. The attention mechanism dynamically adjusts the importance of different temporal features based on the current state, improving the model's sensitivity to abnormal states. The network contains multiple residual blocks, each consisting of two fully connected layers and a batch normalization layer. Skip connections mitigate the vanishing gradient problem.

[0139] In this embodiment, different parameters are mapped to a unified metric space during feature space construction. Min-max normalization is used to scale parameter values ​​to the [0,1] interval, and a kernel function is used to map low-dimensional features to a high-dimensional space, enhancing the expressive power of the features. At the same time, the physical constraints between parameters are maintained to ensure the physical meaning of the feature space.

[0140] This embodiment designs a multi-task learning framework that simultaneously predicts equipment operating status, component wear status, and safety threshold status. Each task branch employs an independent fully connected layer while sharing the underlying feature extraction network. An adaptive task weight adjustment strategy balances the training difficulty of different tasks, improving the model's generalization ability.

[0141] In this embodiment, a multi-level status classification standard is established for operational status assessment. Based on the combined characteristics of parameters such as load, speed, and position, the operational status is divided into three levels: normal, warning, and fault. A soft voting mechanism is used to fuse the prediction results of multiple classifiers, improving the reliability of status judgment.

[0142] In this embodiment, a wear degree prediction model is established based on vibration signal characteristics for component wear condition assessment. Characteristic frequency components are extracted through spectral analysis and combined with time-domain statistical features to predict the wear degree of key components such as guide rails, gears, and bearings. The model outputs a wear condition score and an estimate of remaining service life.

[0143] This embodiment employs a dynamic threshold adjustment strategy in safety threshold status assessment. Based on historical equipment operating data, a parameter change trend model is established to predict future parameter change trends. When a parameter approaches the safety threshold, the system issues an early warning signal, allowing sufficient response time.

[0144] This embodiment achieves intelligent mapping from multi-dimensional sensor data to equipment operating status, establishing a complete status assessment system. By using a deep learning model to mine the potential correlations between data features, it achieves accurate assessment of the elevator's operating status. This solution can promptly detect equipment anomalies and predict component failures, providing a reliable guarantee for the safe operation of construction elevators.

[0145] In one embodiment of the construction hoist safety management method based on multi-sensoring of this application, see [link to relevant documentation]. Figure 6 It can also specifically include the following:

[0146] Step S601: Construct a construction elevator operation status display interface on the monitoring touch screen, and convert the equipment operation status, component wear status and safety threshold status in the operation status information into graphical display data through the data visualization module. At the same time, import the operation status information and the operation permission level and operation time limit in the construction personnel permission information into the data cache area of ​​the intelligent control unit.

[0147] Step S602: Input the load data, the position data and the speed data into the data analysis module of the intelligent control unit, perform real-time analysis of the data through a fuzzy neural network, establish a speed control model based on load and displacement, compare the calculation results of the speed control model with the rated parameters of the construction elevator, and generate an operation control command including braking torque and leveling compensation value.

[0148] Optionally, this embodiment features a hierarchical status display interface on the monitoring touchscreen. The main interface uses a fan-shaped dashboard to display core operating parameters, including current load, operating speed, and location information. The dashboard uses color-coded stripes to indicate safety threshold ranges: green for normal operation, yellow for warning, and red for danger. The interface layout is responsive and supports display devices with different resolutions.

[0149] This embodiment employs a WebGL graphics engine in the data visualization module to achieve a 3D dynamic display of the elevator's operating status. Skeletal animation simulates the elevator's trajectory, updating the car's position and posture in real time. Key components such as guide rails, wire ropes, and brakes are rendered semi-transparently to visually display their wear status, with different colors representing different degrees of wear.

[0150] This embodiment designs a multi-level data caching mechanism, including a high-speed cache and persistent storage. The high-speed cache adopts a circular buffer structure to store the runtime status data of the most recent 5 minutes, supporting fast read and write access. The persistent storage uses a time-series database to compress and encode the status data, achieving efficient storage and retrieval of long-term data.

[0151] This embodiment designs two layers of fuzzy rules within a fuzzy neural network. The first layer classifies operating conditions based on load weight, dividing it into four levels: unloaded, lightly loaded, medium-loaded, and heavily loaded. The second layer classifies operating stages based on displacement, including three stages: acceleration, constant speed operation, and deceleration / stopping. Control strategies for different operating conditions and stages are determined through fuzzy inference.

[0152] This embodiment employs a hybrid learning algorithm in network training. First, an initial fuzzy rule base is constructed based on expert experience. Then, the membership function parameters are optimized using a backpropagation (BP) algorithm. Finally, a genetic algorithm is used to optimize the rule weights. The training data includes running trajectories under different load conditions to ensure the model's adaptability to various working conditions.

[0153] This embodiment implements an adaptive control strategy in the speed control model. The acceleration limit is dynamically adjusted based on the current load weight, reducing acceleration to avoid impact under heavy loads. Speed ​​planning is achieved through position feedback, initiating deceleration in advance as the model approaches the target floor. The model outputs the desired speed curve, which serves as the setpoint for the PID controller.

[0154] This embodiment designs a model-predictive braking control algorithm. Precise control is achieved through a dual-loop structure of speed and position; the outer loop uses a fuzzy controller, and the inner loop uses a PID controller. Braking torque commands are calculated based on speed and position errors to achieve smooth braking. As the vehicle approaches the target floor, compensation control eliminates the influence of mechanical backlash, improving leveling accuracy.

[0155] This embodiment realizes a real-time comparison mechanism for operating parameters. A database containing parameters such as rated load, rated speed, and acceleration limit is established. Before generating a control instruction, safety verification is performed to ensure that each parameter does not exceed the limit range. When approaching the limit, an early warning mechanism is activated, and safety protection actions are executed if necessary.

[0156] This embodiment adopts a distributed architecture during the execution of control instructions. The main controller generates reference control instructions, and the field controller performs real-time compensation adjustment. Control instructions and status feedback are transmitted through a real-time bus network to ensure the real-time performance of the control loop. This solution realizes the intuitive display and intelligent control of the operating state of the lift, improving the safety and reliability of equipment operation.

[0157] This embodiment establishes a complete human-machine interaction system. The operator can monitor the equipment status in real time through the touch interface, and the system dynamically adjusts the control authority according to the operation permission level. When the equipment status is abnormal or approaches the safety threshold, the system reminds the operator in a dual way of vision and sound and automatically records abnormal events. This solution improves the intelligent level of the construction lift and provides reliable vertical transportation guarantee for the construction site.

[0158] In an embodiment of the safety control method for a construction lift based on multi-sensor perception of the present application, see Figure 7 , it may also specifically include the following content:

[0159] Step S701: Real-time monitor the equipment operation state, component wear state and safety threshold state in the operation state information, compare the monitoring data with a preset fault determination threshold, and activate the fault protection mechanism when any state parameter exceeds the preset threshold, send a braking control instruction to the safety clamp sensor, and drive the brake to implement emergency braking through the safety clamp sensor;

[0160] Step S702: Input the deviation value between the state parameter that triggers the fault protection mechanism and the preset threshold into the Internet of Things communication module, encrypt and package the deviation data through the Internet of Things communication module, establish a communication link with the safety supervision system, form an alarm data packet with the encrypted deviation data and the equipment number information, and send the alarm data packet to the safety supervision system through the communication link.

[0161] Optionally, this embodiment designs a multiple fault monitoring mechanism to ensure the operation safety of the construction lift through hierarchical early warning and linkage protection of state parameters. The monitoring indicators include key parameters such as car load, running speed, braking torque, vibration amplitude, and door lock status. Three-level early warning thresholds are set for each parameter, corresponding to attention, warning, and danger states respectively, and the threshold settings are based on equipment specification requirements and historical operation data statistics.

[0162] This embodiment employs a fuzzy comprehensive evaluation method for fault determination. An evaluation index system is established, encompassing equipment operating status, component wear status, and safety threshold status. The weights of each index are determined using the analytic hierarchy process (AHP), considering the correlation and importance between indices. The evaluation results are categorized into three levels: normal, minor anomaly, and severe fault, with different levels triggering different protection responses.

[0163] This embodiment employs a redundant design in the safety clamp control system. The controller uses a dual-CPU architecture: the main control CPU handles normal control logic, while the monitoring CPU handles safety protection functions. The two CPUs exchange data via independent communication buses, monitoring each other's operating status. When a fault is detected, both CPUs must simultaneously confirm before triggering emergency braking to prevent malfunctions.

[0164] This embodiment implements adaptive braking force adjustment in braking control. The required braking torque is calculated based on the current operating speed and load, and the braking pressure is precisely controlled via a proportional solenoid valve. The braking process is divided into two stages: rapid clamping and pressure holding, ensuring a smooth and controllable braking process. Simultaneously, the brake temperature is monitored to prevent over-braking and brake failure.

[0165] This embodiment employs a multi-layered security protection mechanism in the IoT communication module. Data encryption uses the AES-256 algorithm, and the key is securely exchanged using asymmetric encryption. The communication protocol uses an improved MQTT protocol, supporting message hierarchy and priority transmission. VPN technology is used at the network layer to establish a secure channel, preventing data from being illegally intercepted or tampered with.

[0166] This embodiment designs an adaptive data compression algorithm. Appropriate compression methods are selected based on the characteristics of different types of parameters; for example, differential coding is used for continuously changing parameters, and run-length encoding is used for discrete states. The compression algorithm significantly reduces the amount of data transmitted and improves communication efficiency while ensuring data accuracy.

[0167] This embodiment implements automatic reconnection and link backup mechanisms during the communication link establishment process. The primary communication link uses a 4G network, while WiFi is provided as a backup link. When a degradation in the quality of the primary link is detected, the system automatically switches to the backup link. Local data caching is enabled during link switching to ensure that alarm data is not lost.

[0168] This embodiment employs a modular design in constructing the alarm data packet. The data packet contains three parts: basic device information, fault status information, and system configuration information. Device information includes fixed information such as device number, model, and installation location; fault information includes parameters that trigger the fault, thresholds, and timestamps; and system information includes the current software version and configuration parameters.

[0169] This embodiment implements an alarm priority management mechanism. Alarm data packets are assigned different priorities based on the severity of the fault; high-priority packets can interrupt the current transmission and are sent to the monitoring system first. Upon receiving an alarm, the monitoring system automatically notifies relevant personnel according to preset rules and initiates the emergency response process.

[0170] This embodiment establishes a complete traceability mechanism for fault data processing. Detailed status data is recorded for each fault trigger, including operating parameters and operation logs prior to the fault. This data is stored in a local database and simultaneously uploaded to a cloud server for subsequent fault analysis and preventative maintenance. This solution enables timely detection and rapid response to construction elevator faults, improving the safety and reliability of equipment operation.

[0171] To overcome the limitations of traditional elevator management models and provide an intelligent overall solution for construction safety management, this application provides an embodiment of a multi-sensor-based construction elevator safety control device for implementing all or part of the aforementioned multi-sensor-based construction elevator safety control method. See [link to embodiment]. Figure 8 The construction elevator safety control device based on multi-sensor perception specifically includes the following components:

[0172] The biometric detection module 10 is used to collect facial images of construction workers through a binocular liveness detection camera and voiceprint information through a voiceprint collector. The facial images are input into a personnel verification model for identity feature extraction, and the voiceprint information is input into a voiceprint verification model for voiceprint feature extraction. Based on a deep neural network, the identity features and voiceprint features are verified against preset information of the construction workers. At the same time, images of the construction workers wearing protective equipment are collected and input into a posture recognition model. Based on the identity verification result, the voiceprint verification result, and the posture recognition result, the construction workers' permission information is generated and input into the elevator control authorization unit.

[0173] The status monitoring module 20 is used to install weight sensors, position sensors, speed sensors, vibration sensors, limit sensors, door opening and closing sensors, and safety clamp sensors on the bottom of the car, the side walls of the guide rails, and the edge of the door frame of the construction hoist. It collects load data, position data, speed data, vibration data, travel data, door status data, and braking data of the construction hoist. The collected multi-dimensional data is input into the data preprocessing unit for noise reduction filtering and standardization processing. The preprocessed data is input into the equipment status assessment model according to the preset weights. The construction hoist operation status information is generated based on the status parameters output by the equipment status assessment model.

[0174] The safety management module 30 is used to display the operating status information on the monitoring touch screen, and simultaneously input the operating status information and the construction personnel permission information into the intelligent control unit. The intelligent control unit generates operating control commands based on the load data, position data and speed data of the construction elevator, and adjusts the braking distance and leveling accuracy of the construction elevator according to the operating control commands. When the operating status information deviates from the preset threshold, a fault protection mechanism is triggered. The fault protection mechanism controls the safety clamp sensor to perform braking and sends an alarm signal to the safety supervision system through the Internet of Things communication module.

[0175] As described above, the construction elevator safety management device based on multi-sensor perception provided in this application can accurately verify personnel identity and protective equipment status through multi-modal identity recognition using binocular liveness detection and voiceprint acquisition, combined with deep neural networks. It innovatively designs an equipment status assessment model based on multi-sensor fusion to achieve coordinated monitoring of key parameters such as load, position, and speed. The system employs intelligent control strategies for adaptive adjustment of operating parameters, establishes a real-time fault protection mechanism based on threshold judgment, and achieves remote monitoring through IoT technology. This method breaks through the limitations of traditional elevator management models and provides an intelligent overall solution for construction safety management.

[0176] From a hardware perspective, in order to overcome the limitations of traditional elevator management models and provide an intelligent overall solution for construction safety management, this application provides an embodiment of an electronic device for implementing all or part of the aforementioned multi-sensor-based construction elevator safety control method. The electronic device specifically includes the following components:

[0177] The system comprises a processor, memory, a communication interface, and a bus; wherein the processor, memory, and communication interface communicate with each other via the bus; the communication interface is used to realize information transmission between the multi-sensor-based construction elevator safety management device and core business systems, user terminals, and related databases and other related equipment; the logic controller can be a desktop computer, tablet computer, or mobile terminal, etc., and this embodiment is not limited to these. In this embodiment, the logic controller can be implemented with reference to the embodiments of the multi-sensor-based construction elevator safety management method and the multi-sensor-based construction elevator safety management device in the embodiments, the content of which is incorporated herein, and repeated details will not be described again.

[0178] It is understood that the user terminal may include smartphones, tablet computers, network set-top boxes, portable computers, desktop computers, personal digital assistants (PDAs), in-vehicle devices, smart wearable devices, etc. Among these, the smart wearable devices may include smart glasses, smartwatches, smart bracelets, etc.

[0179] In practical applications, some parts of the multi-sensor-based construction hoist safety management method can be executed on the electronic device side as described above, or all operations can be completed in the client device. The choice can be made based on the processing power of the client device and the limitations of the user's usage scenario. This application does not impose any limitations on this. If all operations are completed in the client device, the client device may further include a processor.

[0180] The aforementioned client device may have a communication module (i.e., a communication unit) that can communicate with a remote server to achieve data transmission with the server. The server may include a server on the task scheduling center side; in other implementation scenarios, it may also include a server on an intermediate platform, such as a server on a third-party server platform that has a communication link with the task scheduling center server. The server may include a single computer device, a server cluster consisting of multiple servers, or a distributed server structure.

[0181] Figure 9 This is a schematic block diagram illustrating the system configuration of the electronic device 9600 according to an embodiment of this application. Figure 9 As shown, the electronic device 9600 may include a central processing unit 9100 and a memory 9140; the memory 9140 is coupled to the central processing unit 9100. It is worth noting that... Figure 9 This is an example; other types of structures can also be used to supplement or replace this structure to achieve telecommunications functions or other functions.

[0182] In one embodiment, the safety management method for construction hoists based on multi-sensor perception can be integrated into the central processing unit 9100. The central processing unit 9100 can be configured to perform the following control:

[0183] Step S101: Facial images of construction workers are captured by a binocular liveness detection camera, and voiceprint information is collected by a voiceprint collector. The facial images are input into a personnel verification model for identity feature extraction, and the voiceprint information is input into a voiceprint verification model for voiceprint feature extraction. The identity features and voiceprint features are then double-verified against the preset information of the construction workers based on a deep neural network. Simultaneously, images of the construction workers wearing protective equipment are captured, and these images are input into a posture recognition model. Based on the identity verification result, the voiceprint verification result, and the posture recognition result, construction worker permission information is generated, and the permission information is input into the elevator control authorization unit.

[0184] Step S102: Install weight sensors, position sensors, speed sensors, vibration sensors, limit sensors, door opening / closing sensors, and safety clamp sensors at the bottom of the car, the side walls of the guide rails, and the edge of the door frame of the construction hoist. Collect load data, position data, speed data, vibration data, travel data, door status data, and braking data of the construction hoist. Input the collected multidimensional data into the data preprocessing unit for noise reduction filtering and standardization. Input the preprocessed data into the equipment status assessment model according to the preset weights. Generate the construction hoist operating status information based on the status parameters output by the equipment status assessment model.

[0185] Step S103: Display the operating status information on the monitoring touch screen, and simultaneously input the operating status information and the construction personnel permission information into the intelligent control unit. The intelligent control unit generates operating control commands based on the load data, position data, and speed data of the construction elevator, and adjusts the braking distance and leveling accuracy of the construction elevator according to the operating control commands. When the operating status information deviates from the preset threshold, a fault protection mechanism is triggered. The fault protection mechanism controls the safety clamp sensor to perform braking and sends an alarm signal to the safety supervision system through the Internet of Things communication module.

[0186] As described above, the electronic device provided in this application embodiment achieves accurate verification of personnel identity and protective equipment status through multimodal identity recognition using binocular liveness detection and voiceprint acquisition, combined with a deep neural network. It innovatively designs an equipment status assessment model based on multi-sensor fusion, enabling collaborative monitoring of key parameters such as load, position, and speed. The system employs intelligent control strategies for adaptive adjustment of operating parameters, establishes a real-time fault protection mechanism based on threshold judgment, and achieves remote monitoring through IoT technology. This method breaks through the limitations of traditional elevator management models, providing an intelligent overall solution for construction safety management.

[0187] In another embodiment, the multi-sensor-based construction elevator safety management and control device can be configured separately from the central processing unit 9100. For example, the multi-sensor-based construction elevator safety management and control device can be configured as a chip connected to the central processing unit 9100, and the multi-sensor-based construction elevator safety management and control method function can be realized through the control of the central processing unit.

[0188] like Figure 9 As shown, the electronic device 9600 may further include: a communication module 9110, an input unit 9120, an audio processor 9130, a display 9160, and a power supply 9170. It is worth noting that the electronic device 9600 does not necessarily need to include these components. Figure 9 All components shown; in addition, the electronic device 9600 may also include Figure 9 For components not shown, please refer to existing technologies.

[0189] like Figure 9 As shown, the central processing unit 9100, sometimes also referred to as a controller or operating control, may include a microprocessor or other processor device and / or logic device, which receives inputs and controls the operation of various components of the electronic device 9600.

[0190] The memory 9140 may be, for example, one or more of a cache, flash memory, hard drive, removable media, volatile memory, non-volatile memory, or other suitable devices. It may store the aforementioned failure-related information, and also store a program for executing that information. The central processing unit 9100 may execute the program stored in the memory 9140 to perform information storage or processing, etc.

[0191] Input unit 9120 provides input to central processing unit 9100. Input unit 9120 may be, for example, a keypad or touch input device. Power supply 9170 provides power to electronic device 9600. Display 9160 displays images and text. Display may be, for example, an LCD display, but is not limited thereto.

[0192] The memory 9140 can be a solid-state memory, such as a read-only memory (ROM), random access memory (RAM), a SIM card, etc. It can also be a memory that retains information even when power is off, can be selectively erased, and contains more data; examples of this type of memory are sometimes referred to as EPROMs. The memory 9140 can also be some other type of device. The memory 9140 includes a buffer memory 9141 (sometimes referred to as a buffer). The memory 9140 may include an application / function storage unit 9142 for storing application programs and function programs or processes for executing the operation of the electronic device 9600 via the central processing unit 9100.

[0193] The memory 9140 may also include a data storage unit 9143 for storing data, such as contacts, digital data, pictures, sounds, and / or any other data used by the electronic device. The driver storage unit 9144 of the memory 9140 may include various drivers for the electronic device for communication functions and / or for performing other functions of the electronic device (such as messaging applications, address book applications, etc.).

[0194] The communication module 9110 is a transmitter / receiver that sends and receives signals via the antenna 9111. The communication module 9110 (transmitter / receiver) is coupled to the central processing unit 9100 to provide input signals and receive output signals, which is the same as in a conventional mobile communication terminal.

[0195] Based on different communication technologies, multiple communication modules 9110 can be configured in the same electronic device, such as cellular network modules, Bluetooth modules, and / or wireless LAN modules. The communication module 9110 (transmitter / receiver) is also coupled to a speaker 9131 and a microphone 9132 via an audio processor 9130 to provide audio output via the speaker 9131 and receive audio input from the microphone 9132, thereby realizing typical telecommunications functions. The audio processor 9130 may include any suitable buffer, decoder, amplifier, etc. Additionally, the audio processor 9130 is coupled to a central processing unit 9100, enabling on-device recording via the microphone 9132 and on-device playback of stored audio via the speaker 9131.

[0196] Embodiments of this application also provide a computer-readable storage medium capable of implementing all steps of the multi-sensor-based construction elevator safety management method with a server or client as the execution subject in the above embodiments. The computer-readable storage medium stores a computer program that, when executed by a processor, implements all steps of the multi-sensor-based construction elevator safety management method with a server or client as the execution subject in the above embodiments. For example, when the processor executes the computer program, it implements the following steps:

[0197] Step S101: Facial images of construction workers are captured by a binocular liveness detection camera, and voiceprint information is collected by a voiceprint collector. The facial images are input into a personnel verification model for identity feature extraction, and the voiceprint information is input into a voiceprint verification model for voiceprint feature extraction. The identity features and voiceprint features are then double-verified against the preset information of the construction workers based on a deep neural network. Simultaneously, images of the construction workers wearing protective equipment are captured, and these images are input into a posture recognition model. Based on the identity verification result, the voiceprint verification result, and the posture recognition result, construction worker permission information is generated, and the permission information is input into the elevator control authorization unit.

[0198] Step S102: Install weight sensors, position sensors, speed sensors, vibration sensors, limit sensors, door opening / closing sensors, and safety clamp sensors at the bottom of the car, the side walls of the guide rails, and the edge of the door frame of the construction hoist. Collect load data, position data, speed data, vibration data, travel data, door status data, and braking data of the construction hoist. Input the collected multidimensional data into the data preprocessing unit for noise reduction filtering and standardization. Input the preprocessed data into the equipment status assessment model according to the preset weights. Generate the construction hoist operating status information based on the status parameters output by the equipment status assessment model.

[0199] Step S103: Display the operating status information on the monitoring touch screen, and simultaneously input the operating status information and the construction personnel permission information into the intelligent control unit. The intelligent control unit generates operating control commands based on the load data, position data, and speed data of the construction elevator, and adjusts the braking distance and leveling accuracy of the construction elevator according to the operating control commands. When the operating status information deviates from the preset threshold, a fault protection mechanism is triggered. The fault protection mechanism controls the safety clamp sensor to perform braking and sends an alarm signal to the safety supervision system through the Internet of Things communication module.

[0200] As described above, the computer-readable storage medium provided in this application embodiment achieves accurate verification of personnel identity and protective equipment status through multimodal identity recognition using binocular liveness detection and voiceprint acquisition, combined with deep neural networks. It innovatively designs an equipment status assessment model based on multi-sensor fusion, enabling collaborative monitoring of key parameters such as load, position, and speed. The system employs intelligent control strategies for adaptive adjustment of operating parameters, establishes a real-time fault protection mechanism based on threshold judgment, and achieves remote monitoring through IoT technology. This method breaks through the limitations of traditional elevator management models, providing an intelligent overall solution for construction safety management.

[0201] Embodiments of this application also provide a computer program product capable of implementing all steps in the multi-sensor-based construction elevator safety management method described above, where the execution subject is a server or client. When executed by a processor, this computer program / instruction implements the steps of the multi-sensor-based construction elevator safety management method. For example, the computer program / instruction implements the following steps:

[0202] Step S101: Facial images of construction workers are captured by a binocular liveness detection camera, and voiceprint information is collected by a voiceprint collector. The facial images are input into a personnel verification model for identity feature extraction, and the voiceprint information is input into a voiceprint verification model for voiceprint feature extraction. The identity features and voiceprint features are then double-verified against the preset information of the construction workers based on a deep neural network. Simultaneously, images of the construction workers wearing protective equipment are captured, and these images are input into a posture recognition model. Based on the identity verification result, the voiceprint verification result, and the posture recognition result, construction worker permission information is generated, and the permission information is input into the elevator control authorization unit.

[0203] Step S102: Install weight sensors, position sensors, speed sensors, vibration sensors, limit sensors, door opening / closing sensors, and safety clamp sensors at the bottom of the car, the side walls of the guide rails, and the edge of the door frame of the construction hoist. Collect load data, position data, speed data, vibration data, travel data, door status data, and braking data of the construction hoist. Input the collected multidimensional data into the data preprocessing unit for noise reduction filtering and standardization. Input the preprocessed data into the equipment status assessment model according to the preset weights. Generate the construction hoist operating status information based on the status parameters output by the equipment status assessment model.

[0204] Step S103: Display the operating status information on the monitoring touch screen, and simultaneously input the operating status information and the construction personnel permission information into the intelligent control unit. The intelligent control unit generates operating control commands based on the load data, position data, and speed data of the construction elevator, and adjusts the braking distance and leveling accuracy of the construction elevator according to the operating control commands. When the operating status information deviates from the preset threshold, a fault protection mechanism is triggered. The fault protection mechanism controls the safety clamp sensor to perform braking and sends an alarm signal to the safety supervision system through the Internet of Things communication module.

[0205] As described above, the computer program product provided in this application embodiment achieves accurate verification of personnel identity and protective equipment status through multimodal identity recognition using binocular liveness detection and voiceprint acquisition, combined with deep neural networks. It innovatively designs an equipment status assessment model based on multi-sensor fusion, enabling collaborative monitoring of key parameters such as load, position, and speed. The system employs intelligent control strategies for adaptive adjustment of operating parameters, establishes a real-time fault protection mechanism based on threshold judgment, and achieves remote monitoring through IoT technology. This method breaks through the limitations of traditional elevator management models, providing an intelligent overall solution for construction safety management.

[0206] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0207] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (devices), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0208] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0209] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0210] Specific embodiments are used in the present invention to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core ideas. At the same time, for those skilled in the art, according to the ideas of the present invention, there may be changes in the specific implementation methods and application scopes. In summary, the contents of this specification should not be understood as limiting the present invention.

Claims

1. A safety management method for construction hoists based on multi-sensor perception, characterized in that, The method includes: Facial images of construction workers are captured by a binocular liveness detection camera, and voiceprint information is collected by a voiceprint collector. The facial images are input into a personnel verification model for identity feature extraction, and the voiceprint information is input into a voiceprint verification model for voiceprint feature extraction. Based on a deep neural network, the identity features and voiceprint features are verified against preset information of the construction workers. Simultaneously, images of the construction workers wearing protective equipment are captured, and these images are input into a posture recognition model to obtain protective equipment status features. Based on the identity verification result, the voiceprint verification result, and the protective equipment status features, the correlation strength between nodes is determined through an attention mechanism to generate construction worker permission information. This permission information is input into the elevator control authorization unit. The permission information includes user identity, level, operating range, and timeliness. The permission information is obtained based on multi-dimensional control of time window, work area, and equipment status to ensure the safe and standardized use of the construction elevator. Weight sensors, position sensors, speed sensors, vibration sensors, limit sensors, door opening / closing sensors, and safety clamp sensors are installed at the bottom of the car, the side walls of the guide rails, and the edges of the door frame of the construction hoist. These sensors collect data on the hoist's load, position, speed, vibration, travel, door status, and braking. The collected multidimensional data is then input into a data preprocessing unit for noise reduction, filtering, and standardization. A preset weight matrix for the equipment condition assessment model is constructed. Weight coefficients are assigned based on the degree of influence of each sensor's data on the hoist's operating status. The preprocessed data matrix and the preset weight matrix are then multiplied to generate a weighted feature vector. A recurrent neural network is used to extract time-series features from the weighted feature vector, and the extracted time-series features are input into the equipment condition assessment model. Based on the output load parameters, operating speed parameters, position parameters, vibration parameters, door status parameters and braking parameters of the equipment status assessment model, the parameters are combined to construct a status feature space. The status feature space is then nonlinearly mapped by a deep neural network to generate operating status information that includes equipment operating status, component wear status and safety threshold status. The operating status information is displayed on the monitoring touch screen, and the operating status information and the construction personnel's permission information are input into the intelligent control unit. The intelligent control unit generates operating control commands based on the load data, position data and speed data of the construction elevator, and adjusts the braking distance and leveling accuracy of the construction elevator according to the operating control commands. When the operating status information deviates from the preset threshold, a fault protection mechanism is triggered. The fault protection mechanism controls the safety clamp sensor to perform braking and sends an alarm signal to the safety supervision system through the Internet of Things communication module.

2. The construction hoist safety management method based on multi-sensoring according to claim 1, characterized in that, The process involves acquiring facial images of construction workers using a binocular liveness detection camera and voiceprint information using a voiceprint collector. The facial images are input into a personnel verification model for identity feature extraction, and the voiceprint information is input into a voiceprint verification model for voiceprint feature extraction. A deep neural network is then used to perform dual verification of the identity features and voiceprint features against preset information about the construction workers, including: The facial images captured by the binocular liveness detection camera are preprocessed with image enhancement. The preprocessed facial images are then input into a convolutional neural network for facial key point detection. The coordinates of facial feature points, including the contours of the eyes, nose, and mouth, are extracted. A facial geometric feature vector is constructed based on the facial feature point coordinates. The facial geometric feature vector is then input into a personnel verification model to generate an identity feature code. The identity feature code is then encrypted and stored. Continuous voice segments of construction workers are collected using a voiceprint collector. The voice segments are then processed by frame segmentation and windowing. Mel-frequency cepstral coefficient features are extracted from the voice segments. The extracted cepstral coefficient features are input into a voiceprint verification model to generate a voiceprint feature vector. The identity feature code and the voiceprint feature vector are then imported into a deep neural network. The deep neural network calculates the feature similarity score with a preset identity information database. The similarity score is then compared with a preset matching threshold to generate a verification result.

3. The construction hoist safety management method based on multi-sensoring according to claim 1, characterized in that, The process includes collecting images of construction workers wearing protective equipment, inputting these images into a posture recognition model, generating worker access information based on the identity verification result, voiceprint verification result, and posture recognition result, and then inputting this access information into the elevator control authorization unit. Collect front and side image sequences of construction workers, perform target detection and image segmentation on the image sequences, extract the outline features of protective equipment including safety helmets, reflective vests and safety belts, input the outline features into a posture recognition model based on human skeletal key points, calculate the spatial positional relationship between the protective equipment and human skeletal key points through the posture recognition model, and generate a protective equipment wearing status feature map. The protective equipment wearing status feature image, the identity verification result, and the voiceprint verification result are input into the permission evaluation model. The permission evaluation model is used to comprehensively analyze the construction personnel's identity level and compliance with protection specifications. Based on the analysis results, permission information containing operation permission level and operation time limit is generated. The permission information is written into the data register of the elevator control authorization unit to establish a mapping relationship between the construction personnel's identity and operation permission.

4. The construction hoist safety management method based on multi-sensoring according to claim 1, characterized in that, Weight sensors, position sensors, speed sensors, vibration sensors, limit sensors, door opening / closing sensors, and safety gear sensors are installed at the bottom of the construction elevator car, the side walls of the guide rails, and the edges of the door frame. These sensors collect data on the construction elevator's load, position, speed, vibration, travel, door status, and braking. The collected multi-dimensional data is then input into a data preprocessing unit for noise reduction, filtering, and standardization, including: A weight sensor is installed at the bottom of the construction elevator car, and position sensors, speed sensors, and vibration sensors are installed on the side walls of the guide rails. Limit sensors, door opening and closing sensors, and safety clamp sensors are installed on the edge of the door frame. The weight sensor collects load data, the position sensor collects position data, the speed sensor collects speed data, the vibration sensor collects vibration data, the limit sensor collects travel data, the door opening and closing sensor collects door status data, and the safety clamp sensor collects braking data. The collected multidimensional data is input into the data preprocessing unit, where wavelet transform is used to denoise the multidimensional data, a Butterworth low-pass filter is used to filter out high-frequency interference signals, zero-mean standardization is performed on the filtered data, and the standardized data is time-aligned according to the sampling timestamp to generate a preprocessed data matrix.

5. The construction hoist safety management method based on multi-sensoring according to claim 1, characterized in that, The process involves displaying the operating status information on a monitoring touchscreen, simultaneously inputting the operating status information and the construction personnel's permission information into an intelligent control unit. The intelligent control unit generates operating control commands based on the construction elevator's load data, position data, and speed data, and adjusts the construction elevator's braking distance and leveling accuracy according to these commands. This includes: The operation status display interface of the construction elevator is built on the monitoring touch screen. The equipment operation status, component wear status and safety threshold status in the operation status information are converted into graphical display data through the data visualization module. At the same time, the operation status information and the operation permission level and operation time limit in the construction personnel permission information are imported into the data cache area of ​​the intelligent control unit. The load data, position data, and speed data are input into the data analysis module of the intelligent control unit. The data is analyzed in real time through a fuzzy neural network to establish a speed control model based on load and displacement. The calculation results of the speed control model are compared with the rated parameters of the construction elevator to generate operation control commands that include braking torque and leveling compensation values.

6. The construction hoist safety management method based on multi-sensoring according to claim 1, characterized in that, When the operating status information deviates from a preset threshold, a fault protection mechanism is triggered. This mechanism controls the safety clamp sensor to perform braking and sends an alarm signal to the safety monitoring system via the IoT communication module. The mechanism includes: The system monitors the equipment operation status, component wear status, and safety threshold status in the operation status information in real time. It compares the monitoring data with the preset fault judgment threshold. When any status parameter exceeds the preset threshold, the fault protection mechanism is activated, and a braking control command is sent to the safety clamp sensor. The brake is then driven by the safety clamp sensor to perform emergency braking. The deviation value between the status parameter that triggers the fault protection mechanism and the preset threshold is input into the IoT communication module. The deviation data is encrypted and encapsulated by the IoT communication module, a communication link with the safety supervision system is established, and the encrypted deviation data and the device number information are combined to form an alarm data packet. The alarm data packet is sent to the safety supervision system through the communication link.

7. A safety control device for construction elevators based on multi-sensor perception, characterized in that, The device includes: The biometric detection module is used to collect facial images of construction workers through a binocular liveness detection camera and voiceprint information through a voiceprint collector. The facial images are input into a personnel verification model for identity feature extraction, and the voiceprint information is input into a voiceprint verification model for voiceprint feature extraction. Based on a deep neural network, the identity features and voiceprint features are verified against preset information of the construction workers. Simultaneously, images of the construction workers wearing protective equipment are collected and input into a posture recognition model to obtain protective equipment status features. Based on the identity verification result, the voiceprint verification result, and the protective equipment status features, the module determines the correlation strength between nodes through an attention mechanism to generate construction worker permission information. The permission information is input into the elevator control authorization unit. The permission information includes user identity, level, operating range, and timeliness. The permission information is obtained based on multi-dimensional control of time window, work area, and equipment status to ensure the safe and standardized use of the construction elevator. The status monitoring module is used to install weight sensors, position sensors, speed sensors, vibration sensors, limit sensors, door opening and closing sensors, and safety clamp sensors on the bottom of the car, the side walls of the guide rails, and the edges of the door frame of the construction hoist. It collects load data, position data, speed data, vibration data, travel data, door status data, and braking data of the construction hoist. The collected multidimensional data is input into the data preprocessing unit for noise reduction filtering and standardization processing to construct a preset weight matrix of the equipment status assessment model. The weight coefficients are assigned according to the degree of influence of each sensor data on the operating status of the construction hoist. The preprocessed data matrix and the preset weight matrix are multiplied to generate a weighted feature vector. The time-series features of the weighted feature vector are extracted through a recurrent neural network and input into the equipment status assessment model. Based on the output load parameters, operating speed parameters, position parameters, vibration parameters, door status parameters and braking parameters of the equipment status assessment model, the parameters are combined to construct a status feature space. The status feature space is then nonlinearly mapped by a deep neural network to generate operating status information that includes equipment operating status, component wear status and safety threshold status. The safety management module displays the operating status information on the monitoring touch screen and inputs the operating status information and the construction personnel's permission information into the intelligent control unit. The intelligent control unit generates operating control commands based on the load data, position data, and speed data of the construction elevator. According to the operating control commands, it adjusts the braking distance and leveling accuracy of the construction elevator. When the operating status information deviates from a preset threshold, a fault protection mechanism is triggered. The fault protection mechanism controls the safety clamp sensor to perform braking and sends an alarm signal to the safety supervision system through the Internet of Things communication module.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the construction elevator safety management method based on multiple perception as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the construction elevator safety management method based on multiple perception as described in any one of claims 1 to 6.

Images