A crane safety distance detection and early warning device and its use method

Abstract

The present invention relates to the technical field of crane safety monitoring and early warning, and discloses a crane safety distance detection and early warning device and a method for using the device, including: a housing for accommodating and protecting various modules in the device; a sensor module, including a laser radar, an ultrasonic sensor and a video camera, for real-time measurement of the distance between the crane and the power conductor, and acquisition of environmental images; a data fusion module for fusing multi-source data from the sensor module, and optimizing distance estimation using Kalman filtering and particle filtering algorithms; and an early warning judgment module for automatically adjusting the safety distance threshold according to the real-time distance and dynamic data between the crane and the power conductor, and triggering an alarm signal if the distance is less than the set threshold. Data accuracy and stability are optimized through collaborative measurement of laser radar, ultrasonic sensor and video camera, combined with Kalman filtering and particle filtering algorithms.

Description

Technical Field

[0001] The present invention relates to the technical field of crane safety monitoring and early warning, and in particular to a crane safety distance detection and early warning device and a use method thereof. Background Art

[0002] As an important equipment in construction, power maintenance and port operations, cranes are widely used in aerial operations and heavy lifting. In complex environments, the operation safety of cranes is crucial, especially when operating near high-voltage power lines. The safe distance between cranes and power conductors must be strictly controlled to prevent arc discharge, equipment damage and even personal injury. With the development of automation and intelligent technology, more and more crane safety monitoring systems have been proposed to reduce the safety risks caused by human operating errors and improve the safety and efficiency of crane operations.

[0003] Existing crane safety warning systems mainly rely on a single sensor, such as a lidar or ultrasonic sensor, for distance measurement, and use a fixed threshold warning mechanism to determine the safe distance between the crane and the power conductor. When the crane approaches the set threshold, the system triggers an alarm to remind the operator to take action. However, this measurement method based on a single sensor is greatly affected by environmental factors. For example, the ranging accuracy of lidar decreases in rainy and foggy weather, and ultrasonic sensors are easily affected by wind speed, obstacle reflections, etc. In addition, the fixed threshold warning system cannot dynamically adapt to the operating status of the crane, which can easily lead to false alarms or missed alarms, affecting operating efficiency.

[0004] The main problem with the existing technology is the limitation of the single sensor distance measurement method, which leads to the measurement accuracy being greatly affected by the environment and cannot meet the safety detection requirements of the crane under different working conditions. In complex operating environments, such as high-voltage power construction or severe weather conditions, the safety distance monitoring of the crane requires a more accurate and stable detection scheme to ensure that the system can evaluate the relative position between the crane and the power conductor in real time and accurately, and issue early warning signals in time. To address this problem, the present invention proposes a safety distance detection and early warning device that integrates laser radar, ultrasonic sensors and video cameras, and combines deep reinforcement learning and augmented reality technology to improve the safety and intelligence level of crane operations. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides a crane safety distance detection and warning device and a method of using the same, which solves the problem that the existing crane safety distance detection system is susceptible to environmental interference in terms of the single sensor distance measurement accuracy and that the fixed threshold warning method is prone to false alarms or missed alarms.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a crane safety distance detection and early warning device, comprising:

[0007] A housing, used to accommodate and protect the modules in the device;

[0008] The sensor module includes a lidar, an ultrasonic sensor, and a video camera to measure the distance between the crane and the power conductors in real time and obtain an image of the environment;

[0009] The data fusion module fuses the multi-source data from the sensor module and uses Kalman filter and particle filter algorithms to optimize distance estimation;

[0010] The early warning judgment module automatically adjusts the safety distance threshold according to the real-time distance and dynamic data between the crane and the power conductor, and triggers an alarm signal if the distance is less than the set threshold;

[0011] The control module calculates and adjusts the optimal avoidance path through a deep reinforcement learning algorithm based on the signal from the early warning judgment module and the crane's motion status;

[0012] The alarm module and display module send out alarms through sound, light signals and vibration devices, and use augmented reality technology to display the relative position of the crane and power conductors in real time, providing three-dimensional visual feedback;

[0013] The power module integrates solar cells and external power interface, supports self-charging and ensures long-term stable operation.

[0014] Preferably, the sensor module comprises:

[0015] The laser radar sensor emits laser signals and receives the laser signals reflected from the crane and the power wires. It uses the speed of light formula to calculate the time difference of the reflected signals returning, and then calculates the distance between the crane and the power wires.

[0016] Ultrasonic sensor, which calculates the distance between the crane and the power conductor based on the propagation speed of the sound wave and the propagation time difference by emitting sound wave signals and receiving the sound wave signals reflected from between the crane and the power conductor;

[0017] The video camera collects the relative position images of the crane and the power conductors in real time, processes the image data through the image recognition algorithm, and assists in estimating the spatial relationship between the crane and the power conductors.

[0018] Preferably, the sensor module further comprises:

[0019] The data fusion module performs real-time fusion of the data acquired by the lidar sensor, ultrasonic sensor and video camera, optimizes the data using Kalman filter and particle filter algorithms, reduces the impact of noise on the measurement results, and outputs accurate distance data;

[0020] In the fusion processing step, the Kalman filter and particle filter algorithms dynamically correct the deviation of the sensor data according to the correlation between the sensor error model and the measured value, and optimize the final distance output data;

[0021] Output data: The precise distance data output by the data fusion module is used by the subsequent early warning judgment module and control module.

[0022] Preferably, the data fusion module includes:

[0023] A Kalman filter unit is used to filter the raw distance data obtained by the lidar sensor and the ultrasonic sensor, remove the deviation caused by environmental noise and sensor error, and output the filtered distance estimation data;

[0024] The particle filter unit is used to process the data from the Kalman filter unit, iteratively update the measurement data based on the crane's motion model and environmental parameters, correct the distance deviation caused by nonlinear errors, and output optimized data;

[0025] In the data fusion step, the Kalman filter unit and the particle filter unit process the data from different sensors respectively, and set weighting coefficients based on the signal-to-noise ratio and measurement accuracy of the sensors to merge the distance data of multiple sensors;

[0026] Output data: The data fusion module outputs the calculated fusion data and provides it to the early warning judgment module, control module and display module to support safety distance calculation, path planning and visual display.

[0027] Preferably, the data fusion module further comprises:

[0028] The data calibration unit is used to perform spatiotemporal calibration on the data from different sensors, ensuring that the data timestamps of the lidar, ultrasonic sensor, and video camera are consistent, and that the spatial position relationship between the sensors is accurately mapped;

[0029] The data weight adjustment step adjusts the weight of each sensor in the data fusion process according to its accuracy, sampling frequency and environmental adaptability, so as to increase the impact of high-precision sensor data on the final result and reduce the interference of low-precision sensors on the result;

[0030] The output results after fusion, the data results after optimization and calibration of the data fusion module will be used to generate the real-time safe distance between the crane and the power wire, and provide accurate data support for subsequent early warning and control.

[0031] Preferably, the early warning judgment module includes:

[0032] A data receiving unit, used for receiving distance data from the sensor module and the data fusion module, and comparing the distance data with a set safety distance threshold;

[0033] A safety distance calculation unit calculates the relative distance between the crane and the power conductor in real time based on the current position, speed, acceleration of the crane and the spatial distribution information of the power conductor, and generates safety distance data;

[0034] A threshold judgment unit compares the relative distance calculated in real time with the safety distance threshold, and triggers an early warning signal when the relative distance is less than or equal to the safety distance threshold;

[0035] An alarm level calculation unit calculates the warning level and generates an alarm signal based on the relative distance and the deviation from the safety threshold, which is used to control the alarm module to trigger different types of alarms;

[0036] The early warning triggering step, through the working process of the early warning judgment module, when the distance reaches or falls below the set threshold, the early warning is immediately triggered and the alarm signal is transmitted to the control module for subsequent processing.

[0037] Preferably, the early warning judgment module further includes:

[0038] A historical data learning unit is used to record the crane's historical operating data, environmental data, and warning triggering events, and calculate the typical distance range between the crane and the power conductor based on statistical analysis to adjust the safety distance threshold;

[0039] The real-time dynamic adjustment unit uses a dynamic weighted adjustment method to update the safety distance threshold according to the crane's acceleration, angular velocity, and operating environment parameters, and corrects the warning trigger conditions in combination with real-time measurement data;

[0040] The adaptive threshold update step calculates the safety distance based on the real-time operating status of the crane, according to the set safety factor and historical data, and adjusts the warning parameters according to environmental changes to ensure that the alarm signal is accurately triggered.

[0041] Preferably, the control module includes:

[0042] A data receiving unit, used for receiving the alarm signal from the early warning judgment module and the motion state data of the crane;

[0043] The path planning unit uses a deep reinforcement learning algorithm to generate an avoidance path based on the crane's current position, speed, acceleration, and target position;

[0044] The control decision unit calculates the control input of the crane, including speed and angular velocity, according to the path generated by the path planning unit;

[0045] Dynamic adjustment unit, which monitors the crane's motion status and environmental changes in real time and dynamically adjusts the control signal;

[0046] The output signal module transmits the control signal to the crane actuator to adjust the movement path of the crane;

[0047] Feedback mechanism, obtains real-time feedback on the crane's motion status and further optimizes the path.

[0048] The present invention also provides a method for using a crane safety distance detection and warning device, comprising the following steps:

[0049] Step 1: Start the crane safety distance detection and warning device, initialize the data receiving unit, and the sensor module starts to collect real-time distance data and environmental information between the crane and the power conductor;

[0050] Step 2: The lidar sensor, ultrasonic sensor and video camera respectively measure the distance between the crane and the power conductor, obtain the environmental image, and transmit the data to the data fusion module;

[0051] Step 3: The data fusion module performs spatiotemporal calibration on the sensor data, optimizes the data through Kalman filtering and particle filtering algorithms, reduces measurement errors, and generates optimized distance data;

[0052] Step 4: The early warning judgment module receives the optimized distance data, compares it with the set safety distance threshold, calculates the relative distance, and dynamically adjusts the safety distance threshold according to the crane movement state;

[0053] Step 5: The early warning judgment module determines whether the crane enters the safety threshold range based on the calculation results. If the relative distance is less than the safety distance threshold, an alarm signal is triggered and the alarm level is determined.

[0054] Preferably, the method further comprises the following steps:

[0055] Step 6: The alarm module receives the alarm signal and triggers sound, light or vibration signals according to the alarm level, and provides a real-time relative position image of the crane and the power conductor through the display module;

[0056] Step 7: The control module receives the alarm signal and the crane motion status data, generates an avoidance path through a deep reinforcement learning algorithm, calculates the control input parameters, and adjusts the crane's motion speed, steering angle, and acceleration;

[0057] Step 8: The control module sends a control signal to the crane actuator, and the crane adjusts the motion path according to the control command to avoid the dangerous area and updates the motion status in real time;

[0058] Step 9: The feedback mechanism obtains the crane motion status data and transmits it to the control module. The control module dynamically adjusts the avoidance path according to the feedback data to ensure the safe operation of the crane.

[0059] The present invention provides a crane safety distance detection and warning device and a method for using the same. It has the following beneficial effects:

[0060] 1. The present invention optimizes data accuracy and stability through collaborative measurement of laser radar, ultrasonic sensor and video camera, combined with Kalman filter and particle filter algorithm. Compared with the single sensor ranging method in the prior art, which is susceptible to environmental interference and large measurement error, the multi-source data fusion method of the present invention eliminates the limitation of single sensor, so that the crane can maintain accurate safety distance measurement in different operating environments.

[0061] 2. The present invention optimizes the avoidance path by combining the dynamic characteristics of the crane and using a deep reinforcement learning algorithm, so that the crane can dynamically adjust its driving trajectory according to the real-time environment. In the prior art, most path planning relies on rule settings or traditional algorithms, which have slow response speeds and inflexible path adjustments. The present invention makes the avoidance path more adaptable through continuous training of the machine learning model, and can autonomously optimize the crane's route in a complex operating environment, avoiding unreasonable adjustments caused by fixed path settings.

[0062] 3. The present invention dynamically adjusts the safety distance threshold by combining the crane's motion status, historical operation data and environmental factors through the early warning judgment module to ensure the accuracy of the alarm signal. Compared with the traditional fixed threshold early warning system, which is prone to false alarms or missed alarms, the present invention can adaptively adjust the early warning trigger conditions, so that the system can effectively identify potential risks under different working conditions and improve the safety protection level of the crane.

[0063] 4. The present invention uses an AR display module to present the spatial relationship between the crane and the power conductor in real time. The operator can directly see the three-dimensional visual warning information of the dangerous area and make a quick response decision. The traditional alarm system mainly relies on sound or light reminders, and the information transmission is single, which has the problem of operators not being able to understand intuitively. The present invention uses the visualization method of augmented reality to make safety information more intuitive, greatly improving the safety of crane operations and the human-computer interaction experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 A perspective view of the present invention;

[0065] Figure 2 It is a schematic diagram of the main framework flow chart of the present invention;

[0066] Figure 3 It is a schematic diagram of the data fusion module flow of the present invention;

[0067] Figure 4 It is a schematic diagram of the early warning judgment module flow of the present invention;

[0068] Figure 5 It is a schematic diagram of the alarm signal flow of the present invention;

[0069] Figure 6 A schematic diagram of a display module receiving a signal according to the present invention;

[0070] Figure 7 The figure is a flow chart of the method of the present invention.

[0071] Among them, 1. Shell; 2. Warning module; 3. Display module; 4. Battery module. DETAILED DESCRIPTION

[0072] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0073] Please see attached Figure 1 -Attached Figure 6 The embodiment of the present invention provides a crane safety distance detection and early warning device, comprising:

[0074] The housing 1 is used to accommodate and protect the modules in the device and has the functions of anti-electromagnetic interference, waterproof and dustproof;

[0075] The sensor module includes a lidar, an ultrasonic sensor, and a video camera to measure the distance between the crane and the power conductors in real time and obtain an image of the environment;

[0076] The data fusion module fuses the multi-source data from the sensor module and uses Kalman filter and particle filter algorithms to optimize distance estimation;

[0077] The early warning judgment module automatically adjusts the safety distance threshold according to the real-time distance and dynamic data between the crane and the power conductor, and triggers an alarm signal if the distance is less than the set threshold;

[0078] The control module calculates and adjusts the optimal avoidance path through a deep reinforcement learning algorithm based on the signal from the early warning judgment module and the crane's motion status;

[0079] The alarm module 2 and the display module 3 send out alarms through sound, light signals and vibration devices, and use augmented reality technology to display the relative position of the crane and the power conductor in real time, providing three-dimensional visual feedback;

[0080] The power module 4 integrates solar cells and an external power interface, supports self-charging and ensures long-term stable operation.

[0081] The sensor module includes:

[0082] The laser radar sensor emits laser signals and receives the laser signals reflected from the crane and the power wires. It uses the speed of light formula to calculate the time difference of the reflected signals returning, and then calculates the distance between the crane and the power wires.

[0083] Ultrasonic sensor, which calculates the distance between the crane and the power conductor based on the propagation speed of the sound wave and the propagation time difference by emitting sound wave signals and receiving the sound wave signals reflected from between the crane and the power conductor;

[0084] The video camera collects the relative position images of the crane and the power conductors in real time, processes the image data through the image recognition algorithm, and assists in estimating the spatial relationship between the crane and the power conductors.

[0085] The sensor module further comprises:

[0086] The data fusion module performs real-time fusion of data acquired by the lidar sensor, ultrasonic sensor and video camera, uses Kalman filter and particle filter algorithms to optimize the data, reduce the impact of noise on the measurement results, and output accurate distance data;

[0087] In the fusion processing step, the Kalman filter and particle filter algorithms dynamically correct the deviation of the sensor data according to the correlation between the sensor error model and the measured value, and optimize the final distance output data;

[0088] Output data: The precise distance data output by the data fusion module is used by the subsequent early warning judgment module and control module.

[0089] The data fusion module includes:

[0090] A Kalman filter unit is used to filter the raw distance data obtained by the lidar sensor and the ultrasonic sensor, remove the deviation caused by environmental noise and sensor error, and output the filtered distance estimation data;

[0091] The particle filter unit is used to process the data from the Kalman filter unit, iteratively update the measurement data based on the crane's motion model and environmental parameters, correct the distance deviation caused by nonlinear errors, and output optimized data;

[0092] In the data fusion step, the Kalman filter unit and the particle filter unit process the data from different sensors respectively, and set weighting coefficients based on the signal-to-noise ratio and measurement accuracy of the sensors to merge the distance data of multiple sensors;

[0093] Output data: The data fusion module outputs the calculated fusion data and provides it to the early warning judgment module, control module and display module to support safety distance calculation, path planning and visual display.

[0094] The data fusion module further includes:

[0095] The data calibration unit is used to perform spatiotemporal calibration on the data from different sensors, ensuring that the data timestamps of the lidar, ultrasonic sensor, and video camera are consistent, and that the spatial position relationship between the sensors is accurately mapped;

[0096] The data weight adjustment step adjusts the weight of each sensor in the data fusion process according to its accuracy, sampling frequency and environmental adaptability, so as to increase the impact of high-precision sensor data on the final result and reduce the interference of low-precision sensors on the result;

[0097] The output results after fusion, the data results after optimization and calibration of the data fusion module will be used to generate the real-time safe distance between the crane and the power wire, and provide accurate data support for subsequent early warning and control.

[0098] The early warning judgment module includes:

[0099] A data receiving unit, used for receiving distance data from the sensor module and the data fusion module, and comparing the distance data with a set safety distance threshold;

[0100] A safety distance calculation unit calculates the relative distance between the crane and the power conductor in real time based on the current position, speed, acceleration of the crane and the spatial distribution information of the power conductor, and generates safety distance data;

[0101] A threshold judgment unit compares the relative distance calculated in real time with the safety distance threshold, and triggers an early warning signal when the relative distance is less than or equal to the safety distance threshold;

[0102] An alarm level calculation unit calculates the warning level and generates an alarm signal based on the relative distance and the deviation from the safety threshold, which is used to control the alarm module to trigger different types of alarms;

[0103] The early warning triggering step, through the working process of the early warning judgment module, when the distance reaches or falls below the set threshold, the early warning is immediately triggered and the alarm signal is transmitted to the control module for subsequent processing.

[0104] The early warning judgment module further includes:

[0105] A historical data learning unit is used to record the crane's historical operating data, environmental data, and warning triggering events, and calculate the typical distance range between the crane and the power conductor based on statistical analysis to adjust the safety distance threshold;

[0106] The real-time dynamic adjustment unit uses a dynamic weighted adjustment method to update the safety distance threshold according to the crane's acceleration, angular velocity, and operating environment parameters, and corrects the warning trigger conditions in combination with real-time measurement data;

[0107] The adaptive threshold update step calculates the safety distance based on the real-time operating status of the crane, according to the set safety factor and historical data, and adjusts the warning parameters according to environmental changes to ensure that the alarm signal is accurately triggered.

[0108] The control module includes:

[0109] A data receiving unit, used for receiving the alarm signal from the early warning judgment module and the motion state data of the crane;

[0110] The path planning unit uses a deep reinforcement learning algorithm to generate an avoidance path based on the crane's current position, speed, acceleration, and target position;

[0111] The control decision unit calculates the control input of the crane, including speed and angular velocity, according to the path generated by the path planning unit;

[0112] Dynamic adjustment unit, which monitors the crane's motion status and environmental changes in real time and dynamically adjusts the control signal;

[0113] The output signal module transmits the control signal to the crane actuator to adjust the movement path of the crane;

[0114] Feedback mechanism, obtains real-time feedback on the crane's motion status and further optimizes the path.

[0115] Please see attached Figure 7 The present invention also provides a method for using a crane safety distance detection and warning device, comprising the following steps:

[0116] Step 1: Start the crane safety distance detection and warning device, initialize the data receiving unit, and the sensor module starts to collect real-time distance data and environmental information between the crane and the power conductor;

[0117] Step 2: The lidar sensor, ultrasonic sensor and video camera respectively measure the distance between the crane and the power conductor, obtain the environmental image, and transmit the data to the data fusion module;

[0118] Step 3: The data fusion module performs spatiotemporal calibration on the sensor data, optimizes the data through Kalman filtering and particle filtering algorithms, reduces measurement errors, and generates optimized distance data;

[0119] Step 4: The early warning judgment module receives the optimized distance data, compares it with the set safety distance threshold, calculates the relative distance, and dynamically adjusts the safety distance threshold according to the crane movement state;

[0120] Step 5: The early warning judgment module determines whether the crane enters the safety threshold range based on the calculation results. If the relative distance is less than the safety distance threshold, an alarm signal is triggered and the alarm level is determined.

[0121] The method further comprises the following steps:

[0122] Step 6: The alarm module receives the alarm signal and triggers sound, light or vibration signals according to the alarm level, and provides a real-time relative position image of the crane and the power conductor through the display module;

[0123] Step 7: The control module receives the alarm signal and the crane motion status data, generates an avoidance path through a deep reinforcement learning algorithm, calculates the control input parameters, and adjusts the crane's motion speed, steering angle, and acceleration;

[0124] Step 8: The control module sends a control signal to the crane actuator, and the crane adjusts the motion path according to the control command to avoid the dangerous area and updates the motion status in real time;

[0125] Step 9: The feedback mechanism obtains the crane motion status data and transmits it to the control module. The control module dynamically adjusts the avoidance path according to the feedback data to ensure the safe operation of the crane.

[0126] First, the LiDAR sensor is used to calculate the distance between the crane and the power conductor by emitting a laser beam and receiving the reflected laser signal. In general, the measurement principle of LiDAR is based on the time difference method, that is, the emitted laser signal propagates to the target and reflects back to the sensor, and the system calculates the distance by measuring the time difference from the laser emission to the reception. The propagation time of the laser signal can be calculated by the following formula:

[0127]

[0128] in:

[0129] d represents the distance between the crane and the power conductor;

[0130] c is the speed of light (about 3.0\t imes10 8 m / s);

[0131] Δt is the propagation time difference of the laser signal.

[0132] LiDAR uses this method to achieve high-precision measurement of distant targets, which is suitable for scenarios that require accurate measurement at long distances. Due to the high-frequency measurement and large range of LiDAR, it can provide real-time feedback on the distance changes of the target during crane operation.

[0133] As an option, an ultrasonic sensor measures the distance between the crane and the power conductor by emitting a sound wave signal and receiving the reflected sound wave signal. The propagation time of the ultrasonic wave is proportional to the propagation speed of the sound wave, which is calculated by the following formula:

[0134]

[0135] in:

[0136] v is the propagation speed of sound waves in air, which is about 343m / s (at 20°C);

[0137] Due to the working principle and environmental adaptability of ultrasonic sensors, they are suitable for close distances and environments not affected by light. Although the measurement accuracy is slightly inferior to that of lidar, its low cost and strong adaptability make it an effective auxiliary measurement method in crane operations.

[0138] The video camera is mainly used to collect images of the relative position of the crane and the power conductor. The image processing algorithm converts the video data into spatial data to assist in measuring distance and provide visual feedback. Specifically, the video camera analyzes the image data through a deep learning target detection algorithm to extract the relative position between the crane and the power conductor. By processing continuous frame images, the image recognition parameters can be dynamically adjusted to further improve the recognition accuracy in complex environments.

[0139] In one possible implementation, the image processing algorithm extracts the geometric features of the crane and power wires from each frame of video through an object detection model (such as YOLO or FasterR-CNN based on convolutional neural networks). These features are converted into three-dimensional coordinates through a spatial transformation algorithm so that they can be fused with the measurement data of the distance sensor to optimize the overall measurement results.

[0140] The sensor module transmits the raw data obtained by the laser radar, ultrasonic sensor and video camera to the data fusion module for processing. The data fusion module fuses these data in real time to ensure that the final output measurement results have high accuracy and stability. In order to eliminate the differences between different sensor data, the data fusion module optimizes the data through Kalman filtering and particle filtering algorithms.

[0141] Kalman filtering is used to deal with linear errors, mainly to smooth distance measurement data and reduce the impact of environmental noise, while particle filtering is suitable for dealing with nonlinear errors, especially in complex environments, and can more accurately estimate the relative position between the crane and the power conductor.

[0142] The data fusion module outputs the optimized and accurate distance data for subsequent warning judgment module and control module. These output data will directly affect the system's judgment of the safety distance threshold, the calculation of path planning, and the real-time feedback of the operator.

[0143] In this embodiment, the sensor module realizes high-precision real-time monitoring of the distance between the crane and the power conductor through multi-source data fusion and optimization processing technology. At the same time, the sensor module also has strong adaptability and can cope with dynamic changes in complex environments. This module not only improves the reliability of safety monitoring, but also provides accurate data support for subsequent intelligent control and real-time feedback.

[0144] First, the data fusion module receives raw measurement data from LiDAR, ultrasonic sensors, and video cameras. Each of these sensors has unique advantages: LiDAR is suitable for high-precision measurement at long distances, ultrasonic sensors are suitable for reliable measurement at close ranges, and video cameras provide the system with real-time image information. After the data is calibrated in time and space, it is processed by the data fusion module.

[0145] Kalman filtering is one of the core algorithms in the data fusion module. Kalman filtering is a recursive algorithm that optimizes the current state prediction based on the current measurement value and the previous state estimation. In the present invention, Kalman filtering is used to process the raw data from the lidar and ultrasonic sensors, especially when the sensors are noisy. The basic principle is to predict the system state based on the state equation and update it based on the measurement data, thereby reducing the deviation caused by the measurement error.

[0146] Specifically, the mathematical model of Kalman filtering can be expressed as the following two steps:

[0147]

[0148] in:

[0149] is the predicted state at the current time k;

[0150] A is the state transfer matrix, which describes the dynamic behavior of the system;

[0151] is the state estimate at the previous moment;

[0152] B is the control input matrix, uk For control input.

[0153] Update steps:

[0154]

[0155] in:

[0156] K k is the Kalman gain, and the calculation formula is:

[0157] K k =P k|k-1 ·H T ·(H·P k|k-1 ·H T +R) -1

[0158] P k|k-1 is the prediction error covariance matrix;

[0159] H is the observation matrix, which transforms the state space into the measurement space;

[0160] z k is the actual measured value.

[0161] Kalman filtering gradually reduces the estimation error of the system by fusing the predicted state with the actual measurement data, thereby improving the accuracy of the system.

[0162] Particle filtering is used to deal with nonlinear errors in data, especially in complex dynamic environments. Particle filtering is a recursive estimation method based on the Monte Carlo method. It represents the state of the system through a set of particles and assigns different weights to these particles. At each time step, the particles are updated according to the state transition model and the weights of the particles are updated through measurements. The advantage of particle filtering is that it can handle systems with nonlinearities and high dimensions. In the measurement of cranes and power conductors, particle filtering can effectively solve the nonlinear errors caused by irregular crane movement.

[0163] Specifically, the steps of particle filtering are as follows:

[0164] Sampling step: Generate the particle set at the current moment from the particle set at the previous moment according to the state transfer equation of the system.

[0165] Weight update step: Update the weight of each particle according to the sensor's measurement value. The weight update formula is:

[0166]

[0167] in:

[0168] is the weight of the i-th particle;

[0169] is the state of the ith particle;

[0170] is the likelihood function of the particle in the current state, indicating that in a given state, the measurement value z k probability.

[0171] Resampling step: Generate a new set of particles by weighted sampling to reduce the influence of low-weight particles and retain high-weight particles.

[0172] Through particle filtering, the system can cope with nonlinear dynamic environments and improve the accuracy and stability of estimation, which is especially suitable for scenarios with complex crane motion models and large nonlinear errors.

[0173] After the data fusion module completes the data optimization processing after Kalman filtering and particle filtering, the output distance data is transmitted to the subsequent early warning judgment module and control module. The early warning judgment module calculates the safe distance between the crane and the power conductor in real time and determines whether to trigger the early warning signal in combination with the safety threshold set by the system. If the distance is less than the safety threshold, the early warning judgment module will send an alarm signal to ensure the safety of the crane during operation.

[0174] At the same time, the control module receives the optimized data and calculates the optimal avoidance path. Through the deep reinforcement learning algorithm, the control module can adjust the crane's driving path in real time to ensure that the crane avoids dangerous areas and reduce the risk of misoperation.

[0175] The data receiving unit receives the processed distance data from the data fusion module. The data includes the real-time distance information between the crane and the power conductor, as well as the multi-source data fusion results provided by the lidar, ultrasonic sensor and video camera. The received distance data also includes the measurement error and dynamic changes of the sensor, and the system can perform real-time processing and monitoring based on this data.

[0176] Generally, the sensor's measurement data and the crane's motion state data need to be synchronized to ensure that the crane's current state can be accurately reflected when making judgments. To this end, the data receiving unit synchronizes the time and space of the data from different modules to ensure the consistency and accuracy of the data.

[0177] The safety distance calculation unit calculates the relative distance between the crane and the power conductor in real time based on the distance data received in real time, the crane's current position, speed, acceleration and other dynamic parameters, and the spatial position distribution of the power conductor. Specifically, the calculation formula is as follows:

[0178]

[0179] in:

[0180] d relative Indicates the relative distance between the crane and the power conductor;

[0181] (x 吊车 ,y 吊车 ,z 吊车 ) is the three-dimensional coordinate of the crane;

[0182] (x 导线 ,y 导线 ,z 导线 ) is the three-dimensional coordinate of the power conductor.

[0183] The dynamic parameters (position, speed, acceleration) of the crane and the spatial position of the power conductors can be updated in real time through sensor data and calculation models to ensure the accuracy of distance calculation in dynamic environments.

[0184] The threshold judgment unit is responsible for comparing the calculated relative distance with the set safety distance threshold to determine whether the current distance is less than or equal to the preset safety distance threshold. According to the threshold set by the system, when the relative distance between the crane and the power conductor reaches or falls below the safety distance, the early warning judgment module will consider that there is a safety risk and trigger the corresponding alarm signal.

[0185] Specifically, the set safety distance threshold can be adjusted according to the type of crane, the operating environment and other external factors. In a possible implementation, the threshold can be dynamically adjusted according to the speed and acceleration of the crane, so that the system can provide flexible early warning responses under different operating conditions.

[0186] The alarm level calculation unit calculates the current warning level according to the deviation between the relative distance and the safety distance threshold, and generates an alarm signal. The specific calculation formula is as follows:

[0187]

[0188] in:

[0189] Alarm level indicates the calculated alarm level, ranging from 000 to 111. A larger value indicates a higher risk.

[0190] d relative is the relative distance between the crane and the power conductor;

[0191] d threshold is the set safety distance threshold.

[0192] Depending on the alarm level, the alarm level calculation unit can generate different types of alarm signals. For example, when the alarm level is low, the system can issue a low-intensity alarm (such as a sound or light signal). In the case of severe danger, the intensity of the alarm signal will increase, and the system may also trigger a vibration alarm to remind the operator to take avoidance measures in time.

[0193] The early warning trigger unit determines whether to trigger an early warning based on the calculation result of the alarm level, and sends a signal to the subsequent alarm module and control module. In a possible implementation, the early warning trigger unit divides the alarm signal into multiple levels: mild warning, moderate warning and severe warning. The system can send corresponding alarm signals according to different situations and provide prompts through different feedback methods (such as sound, light signal, vibration).

[0194] In actual applications, the early warning judgment module can also dynamically adjust the safety threshold. Specifically, based on the crane's historical operation data, environmental conditions, the distribution of power facilities at the operation site and other information, the early warning judgment module can adjust the safety distance threshold in real time, making the early warning during the crane operation more accurate and intelligent.

[0195] The data receiving unit is used to receive the alarm signal from the early warning judgment module, as well as data related to the crane's motion status, including position, speed, angular velocity, acceleration, etc. In order to ensure the accuracy of the received data, the system uses a time synchronization mechanism to align the timestamps of the sensor measurements to avoid the influence of timing deviations between different data sources on the calculation results.

[0196] Generally, the data receiving unit also obtains the optimized distance data from the data fusion module and combines it with the crane's dynamic model to establish a more accurate motion state prediction. As an option, the unit can be connected to the crane control system through a wireless communication interface to ensure that the crane's motion state data can be transmitted to the control module in real time for processing.

[0197] The path planning unit uses the deep reinforcement learning (DRL) algorithm to calculate the optimal avoidance path based on the real-time state of the crane. Specifically, the path planning adopts a reinforcement learning framework based on the Markov decision process (MDP), in which the state, action and reward functions are defined as follows:

[0198] s t =(x t ,y t ,v t ,θ t )

[0199] in:

[0200] s t is the state of the system at time step t;

[0201] (x t ,y t ) is the current position of the crane;

[0202] v t is the crane speed;

[0203] θ t is the direction angle of the crane.

[0204] In this framework, the control module needs to select the appropriate action.

[0205] a t =(Δv t ,Δθ t )

[0206] in:

[0207] Δv t is the speed increment;

[0208] Δθ t is the direction angle adjustment amount.

[0209] The goal of path planning is to minimize the downtime that may be caused by the avoidance process, while ensuring that the crane stays away from the power wires and meets the safety distance requirements. Therefore, the reward function of the system can be defined as:

[0210] R t = -α·d t +β·v t -γ·c t

[0211] in:

[0212] R t is the reward value of the current action;

[0213] d t is the distance between the crane and the power conductor. The closer the distance, the greater the penalty value.

[0214] v t Represents the speed of the crane to encourage the crane to complete the avoidance as quickly as possible;

[0215] c t To avoid the control costs in the process, including energy consumption and extra time expenditure;

[0216] α, β, γ are weight parameters for adjusting the reward function.

[0217] In one possible implementation, the control module uses the DQN (DeepQ-Network) reinforcement learning algorithm to optimize the path. The DQN algorithm uses a neural network to approximate the Q-value function, and through continuous training and adjustment of strategies, the system can adaptively adjust the avoidance path in different environments.

[0218] The control decision unit generates specific control instructions based on the calculation results of the path planning unit to adjust the speed and direction of the crane to ensure that the crane moves according to the calculated optimal trajectory. The unit uses the PID (Proportional-Integral-Derivative) control algorithm to adjust the driving parameters of the crane so that it can avoid dangerous areas more smoothly.

[0219] The core equation of PID control is as follows:

[0220]

[0221] u(t) is the control signal;

[0222] K p ,K i ,K d is the PID control parameter;

[0223] e(t) is the error between the current position of the crane and the target path.

[0224] Generally, the control decision unit will optimize the PID parameters in combination with the crane's dynamic model to achieve a smoother control effect. In some embodiments, the unit will also constrain the control signal to prevent the crane's speed and direction from being adjusted too quickly, which could cause safety hazards.

[0225] The dynamic adjustment unit is responsible for monitoring the real-time motion status of the crane and adjusting the control strategy when necessary to adapt to changes in the operating environment. The unit continuously samples and adjusts feedback to ensure that the crane always maintains the best driving trajectory during the avoidance process.

[0226] In one possible implementation, the unit uses an adaptive optimization algorithm to dynamically update the path planning parameters based on the historical driving data of the crane. For example, based on the success rate of the historical avoidance path, the unit can automatically adjust the weight parameters in the reward function to improve the stability of the path planning.

[0227] The output signal module transmits the calculated control signal to the crane actuator, so that the crane can adjust its movement according to the calculated optimal trajectory. The module can be connected to the crane's control system through CAN bus, wireless communication or other data transmission methods to ensure that the control instructions can be accurately executed.

[0228] The feedback mechanism is used to monitor the crane's motion status in real time and transmit it back to the control module to further optimize path planning and control signals. The feedback data includes the crane's real-time position, speed, direction angle and other parameters, which can be used to train the deep reinforcement learning model, so that the system can adjust the path more efficiently in future avoidance processes.

[0229] The display module 3 provides the operator with real-time three-dimensional spatial feedback through visualization technology such as augmented reality (AR). After receiving the data from the early warning judgment module, the display module uses augmented reality technology to superimpose the relative position image between the crane and the power wire in the operator's field of view, as well as the real-time change information of the safety distance. This enables the operator to not only perceive the risk through sound and vibration alarms, but also confirm the specific location of the dangerous area through intuitive three-dimensional images, so as to make quick and accurate judgments.

[0230] Specifically, the display module collects the spatial relationship between the crane and the power wire in real time, and displays the current position of the crane and the relative position of the power wire on the screen or glasses.

[0231] Display content includes:

[0232] Real-time distance between crane and power conductors;

[0233] The current motion state of the crane (such as speed, angle, etc.);

[0234] The difference in distance safety thresholds, and the dynamically updated avoidance paths.

[0235] For example, the operator can see the relative position of the crane and the power line displayed in front of him through augmented reality glasses or displays, and the safe distance area is marked with different colors (such as green, yellow, and red). The green area indicates safety, the yellow area indicates a close distance, and the red area indicates extreme danger. The operator can use these intuitive feedback to determine whether to avoid it immediately.

[0236] In some embodiments, the display module also supports virtual button or gesture recognition functions, allowing the operator to confirm or adjust the alarm system directly through the virtual interface, further improving the convenience and response speed of operation.

[0237] Augmented reality technology enables operators to obtain immediate and accurate feedback in complex operating environments. During crane operation, the three-dimensional image obtained by the operator through AR technology seamlessly integrates the actual movement state of the crane with the surrounding environment, helping the operator to better understand the spatial relationship and take appropriate actions. Through AR display, the operator can clearly see the relative position of the crane and the power conductor, making it easy to judge whether the path needs to be adjusted or the machine needs to be shut down.

[0238] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A crane safety distance detection and early warning device, characterized in that: include: A housing (1) for accommodating and protecting the modules in the device; The sensor module includes a lidar, an ultrasonic sensor, and a video camera to measure the distance between the crane and the power conductors in real time and obtain an image of the environment; The data fusion module fuses the multi-source data from the sensor module and uses Kalman filter and particle filter algorithms to optimize distance estimation; The early warning judgment module automatically adjusts the safety distance threshold according to the real-time distance and dynamic data between the crane and the power conductor, and triggers an alarm signal if the distance is less than the set threshold; The control module calculates and adjusts the optimal avoidance path through a deep reinforcement learning algorithm based on the signal from the early warning judgment module and the crane's motion status; The alarm module (2) and the display module (3) send out an alarm through sound, light signal and vibration device, and use augmented reality technology to display the relative position of the crane and the power conductor in real time, providing three-dimensional visual feedback; The power module (4) integrates solar cells and an external power interface, supports self-charging and ensures long-term stable operation.

2. A crane safety distance detection and warning device according to claim 1, characterized in that: The sensor module comprises: The LiDAR sensor emits laser signals and receives the laser signals reflected from the crane and the power conductors. It uses the speed of light formula to calculate the time difference of the reflected signals returning, and then calculates the distance between the crane and the power conductors. Ultrasonic sensor, which calculates the distance between the crane and the power conductor based on the propagation speed of the sound wave and the propagation time difference by emitting sound wave signals and receiving the sound wave signals reflected from between the crane and the power conductor; The video camera collects the relative position images of the crane and the power conductors in real time, processes the image data through the image recognition algorithm, and assists in estimating the spatial relationship between the crane and the power conductors.

3. A crane safety distance detection and warning device according to claim 1, characterized in that: The sensor module further comprises: A data fusion module performs real-time fusion of the data acquired by the lidar sensor, ultrasonic sensor and video camera, optimizes the data using Kalman filtering and particle filtering algorithms, reduces the impact of noise on the measurement results, and outputs accurate distance data; In the fusion processing step, the Kalman filter and particle filter algorithms dynamically correct the deviation of the sensor data according to the correlation between the sensor error model and the measured value, and optimize the final distance output data; Output data: The precise distance data output by the data fusion module is used by the subsequent early warning judgment module and control module.

4. A crane safety distance detection and warning device according to claim 1, characterized in that: The data fusion module comprises: A Kalman filter unit is used to filter the raw distance data obtained by the lidar sensor and the ultrasonic sensor, remove the deviation caused by environmental noise and sensor error, and output the filtered distance estimation data; The particle filter unit is used to process the data from the Kalman filter unit, iteratively update the measurement data based on the crane's motion model and environmental parameters, correct the distance deviation caused by nonlinear errors, and output optimized data; In the data fusion step, the Kalman filter unit and the particle filter unit process the data from different sensors respectively, and set weighting coefficients based on the signal-to-noise ratio and measurement accuracy of the sensors to merge the distance data of multiple sensors; Output data: The data fusion module outputs the calculated fusion data and provides it to the early warning judgment module, control module and display module to support safety distance calculation, path planning and visual display.

5. The crane safety distance detection and warning device according to claim 1, characterized in that: The data fusion module further comprises: The data calibration unit is used to calibrate the data from different sensors in time and space, ensuring that the data timestamps of the lidar, ultrasonic sensor and video camera are consistent and the spatial position relationship between the sensors is accurately mapped; The data weight adjustment step adjusts the weight of each sensor in the data fusion process according to its accuracy, sampling frequency and environmental adaptability, so as to increase the impact of high-precision sensor data on the final result and reduce the interference of low-precision sensors on the result; The fused output results, the data results after optimization and calibration of the data fusion module will be used to generate the real-time safe distance between the crane and the power wire, and provide accurate data support for subsequent early warning and control.

6. A crane safety distance detection and warning device according to claim 1, characterized in that: The early warning judgment module includes: A data receiving unit, used for receiving distance data from the sensor module and the data fusion module, and comparing the distance data with a set safety distance threshold; A safety distance calculation unit calculates the relative distance between the crane and the power conductor in real time based on the current position, speed, acceleration of the crane and the spatial distribution information of the power conductor, and generates safety distance data; A threshold judgment unit compares the relative distance calculated in real time with the safety distance threshold, and triggers an early warning signal when the relative distance is less than or equal to the safety distance threshold; An alarm level calculation unit calculates the warning level and generates an alarm signal based on the relative distance and the deviation from the safety threshold, which is used to control the alarm module to trigger different types of alarms; The early warning triggering step, through the working process of the early warning judgment module, when the distance reaches or falls below the set threshold, the early warning is immediately triggered and the alarm signal is transmitted to the control module for subsequent processing.

7. The crane safety distance detection and warning device according to claim 1, characterized in that: The early warning judgment module further includes: A historical data learning unit is used to record the crane's historical operating data, environmental data, and warning triggering events, and calculate the typical distance range between the crane and the power conductor based on statistical analysis to adjust the safety distance threshold; The real-time dynamic adjustment unit uses a dynamic weighted adjustment method to update the safety distance threshold according to the crane's acceleration, angular velocity, and operating environment parameters, and corrects the warning trigger conditions in combination with real-time measurement data; The adaptive threshold update step calculates the safety distance based on the real-time operating status of the crane, according to the set safety factor and historical data, and adjusts the warning parameters according to environmental changes to ensure that the alarm signal is accurately triggered.

8. The crane safety distance detection and warning device according to claim 1 is characterized in that: The control module comprises: A data receiving unit, used for receiving the alarm signal from the early warning judgment module and the motion state data of the crane; The path planning unit uses a deep reinforcement learning algorithm to generate an avoidance path based on the crane's current position, speed, acceleration, and target position; The control decision unit calculates the control input of the crane, including speed and angular velocity, according to the path generated by the path planning unit; Dynamic adjustment unit, which monitors the crane's motion status and environmental changes in real time and dynamically adjusts the control signal; The output signal module transmits the control signal to the crane actuator to adjust the movement path of the crane; Feedback mechanism, obtains real-time feedback on the crane's motion status and further optimizes the path.

9. A method for using a crane safety distance detection and early warning device, according to any one of claims 1 to 8, characterized in that: The following steps are involved: Step 1: Start the crane safety distance detection and warning device, initialize the data receiving unit, and the sensor module starts to collect real-time distance data and environmental information between the crane and the power conductor; Step 2: The lidar sensor, ultrasonic sensor and video camera respectively measure the distance between the crane and the power conductor, obtain the environmental image, and transmit the data to the data fusion module; Step 3: The data fusion module performs spatiotemporal calibration on the sensor data, optimizes the data through Kalman filtering and particle filtering algorithms, reduces measurement errors, and generates optimized distance data; Step 4: The early warning judgment module receives the optimized distance data, compares it with the set safety distance threshold, calculates the relative distance, and dynamically adjusts the safety distance threshold according to the crane movement state; Step 5: The early warning judgment module determines whether the crane enters the safety threshold range based on the calculation results. If the relative distance is less than the safety distance threshold, an alarm signal is triggered and the alarm level is determined.

10. The method for using the crane safety distance detection and warning device according to claim 9, characterized in that: The method further comprises the following steps: Step 6: The alarm module receives the alarm signal and triggers sound, light or vibration signals according to the alarm level, and provides a real-time relative position image of the crane and the power conductor through the display module; Step 7: The control module receives the alarm signal and the crane motion status data, generates an avoidance path through a deep reinforcement learning algorithm, calculates the control input parameters, and adjusts the crane's motion speed, steering angle, and acceleration; Step 8: The control module sends a control signal to the crane actuator, and the crane adjusts the motion path according to the control command to avoid the dangerous area and updates the motion status in real time; Step 9: The feedback mechanism obtains the crane motion status data and transmits it to the control module. The control module dynamically adjusts the avoidance path according to the feedback data to ensure the safe operation of the crane.

Images