Safety early warning method, device and equipment for high-altitude operation of power distribution tower and storage medium

Abstract

The invention relates to the technical field of safety early warning, in particular to a power distribution tower aerial work safety early warning method, device and equipment and a storage medium, and the method comprises the steps: waking up an ultrasonic wave, infrared and double-frequency eddy current sensor and a camera device, and starting the infrared sensor if the wakeup is successful; if the real-time infrared information meets the ultrasonic triggering condition, the warning distance of the ultrasonic sensor is confirmed, and the ultrasonic sensor is started; if the feedback information of the ultrasonic sensor is matched with the preset multi-sensor triggering condition, the double-frequency eddy current sensor and the camera device start to work; pre-training a target recognition model to process a real-time image collected by a camera device to obtain detection target information; combining infrared, ultrasonic and metal type information with target information to confirm target comprehensive information, and generating an alarm instruction according to a risk level alarm mechanism; according to the method disclosed by the invention, real-time monitoring and timely early warning can be carried out on behaviors of personnel nearby high-altitude operation, and the safety of the high-altitude operation process is improved.

Description

Technical Field

[0001] The present invention relates to the field of safety warning technology, and in particular to a method, device, equipment and storage medium for safety warning of high-altitude operations on distribution towers. Background Art

[0002] Traditional safety precautions for high-altitude operations on distribution towers, such as setting up barrier nets, can isolate the construction area from the dangerous live electrical area to a certain extent, but they have obvious limitations. First, physical barrier devices such as barrier nets have a low level of intelligence and cannot monitor the behavior of construction workers in real time. Therefore, they cannot provide early warnings for potential dangerous actions. Second, physical barrier devices lack an effective alarm mechanism and cannot provide timely and accurate danger warnings to managers, making it difficult for managers to take timely intervention measures, thereby increasing the safety risks of high-altitude operations.

[0003] It can be seen that the existing technology still needs to be improved and enhanced. Summary of the Invention

[0004] In order to overcome the shortcomings of the existing technology, the purpose of the present invention is to provide a distribution tower high-altitude operation safety warning method, which can monitor the behavior of people near the high-altitude operation in real time and provide timely warnings, thereby improving the safety of the high-altitude operation process.

[0005] A first aspect of the present invention provides a safety warning method for high-altitude operations on a distribution tower, comprising: performing a wake-up check on a detection component, the detection component comprising an ultrasonic sensor, an infrared sensor, a dual-frequency eddy current sensor, and a camera device; if the detection component passes the wake-up check, controlling the infrared sensor to start working; when the real-time infrared information fed back by the infrared sensor meets the preset ultrasonic trigger condition, confirming the warning distance of the ultrasonic sensor based on the real-time environmental information, and controlling the ultrasonic sensor to start working; judging whether the preset multi-sensor trigger condition is met based on the real-time ultrasonic distance information fed back by the ultrasonic sensor and the confirmed warning distance, and if so, controlling the dual-frequency eddy current sensor and the camera device to start working; pre-training a target recognition model, inputting the real-time image captured by the camera device into the pre-trained target recognition model to obtain detection target information; confirming target comprehensive information based on the real-time infrared information fed back by the infrared sensor, the real-time ultrasonic distance information fed back by the ultrasonic sensor, the metal type information fed back by the dual-frequency eddy current sensor, and the detection target information; obtaining a preset risk level alarm mechanism, and generating an alarm instruction based on the target comprehensive information and the preset risk level alarm mechanism.

[0006] Optionally, in a first implementation method of the first aspect of the present invention, a wake-up check is performed on the detection component, and the detection component includes an ultrasonic sensor, an infrared sensor, a dual-frequency eddy current sensor and a camera device; if the detection component passes the wake-up check, the infrared sensor is controlled to start working, including: calling an interface to send a heartbeat packet to the ultrasonic sensor, the infrared sensor, the dual-frequency eddy current sensor and the camera device respectively, and confirming whether the detection component passes the wake-up check according to the received status code; if the detection component passes the wake-up check, the infrared sensor is controlled to start a scanning operation at a frequency of 2Hz; if the status code indicates that the ultrasonic sensor fails to wake up, an emergency alarm instruction is generated, and the detection component is set to a locked state; if the status code indicates that the infrared sensor, the dual-frequency eddy current sensor or the camera device fails to wake up, the wake-up failure information is obtained, and the corresponding preset backup plan is obtained based on the wake-up failure information.

[0007] Optionally, in a second implementation of the first aspect of the present invention, when the real-time infrared information fed back by the infrared sensor meets the preset ultrasonic trigger condition, the warning distance of the ultrasonic sensor is confirmed based on the real-time environmental information, and the ultrasonic sensor is controlled to start working, including: when the real-time infrared information fed back by the infrared sensor indicates that there are two or more consecutive pixel points whose temperatures continuously reach or exceed 35°C, real-time environmental information is obtained, and the real-time environmental information includes real-time humidity and real-time wind speed; environmental conditions are confirmed based on real-time humidity and real-time wind speed, and the environmental conditions include sunny days, rain and fog, and strong winds; when the environmental conditions are sunny, the warning distance of the ultrasonic sensor is adjusted to the first set warning value, and the ultrasonic sensor is controlled to start at a frequency of 10Hz. Updating real-time ultrasonic distance information; when the environmental condition is rain or fog, adjusting the warning distance of the ultrasonic sensor to a second set warning value, the second set warning value is less than the first set warning value, controlling the ultrasonic sensor to start updating the real-time ultrasonic distance information at a frequency of 20 Hz, and applying a 15% gain compensation to the real-time ultrasonic distance information; when the environmental condition is strong wind, adjusting the warning distance of the ultrasonic sensor to a third set warning value, the third set warning value is greater than the second set warning value and less than the first set warning value, and controlling the ultrasonic sensor to start updating the real-time ultrasonic distance information at a frequency of 15 Hz; using a discrete Kalman filter algorithm to process the real-time ultrasonic distance information to obtain the predicted moving speed of the target.

[0008] Optionally, in a third implementation method of the first aspect of the present invention, the real-time ultrasonic distance information fed back by the ultrasonic sensor and the confirmed warning distance are used to determine whether the preset multi-sensor trigger conditions are met. If so, the dual-frequency eddy current sensor and the camera device are controlled to start working, including: when the environmental condition is sunny and when the real-time ultrasonic distance information fed back by the ultrasonic sensor is ≤ the first set warning value, the dual-frequency eddy current sensor and the camera device are controlled to start working; when the environmental condition is rain and fog and when the real-time ultrasonic distance information fed back by the ultrasonic sensor is ≤ the second set warning value, the dual-frequency eddy current sensor and the camera device are controlled to start working, and the real-time image fed back by the camera device is pre-processed using the histogram equalization method; when the environmental condition is strong wind, the real-time ultrasonic distance information fed back by the ultrasonic sensor is ≤ the third set warning value and the predicted moving speed is > the preset speed threshold, the dual-frequency eddy current sensor and the camera device are controlled to start working.

[0009] Optionally, in a fourth implementation method of the first aspect of the present invention, the pre-trained target recognition model inputs the real-time image captured by the camera device into the pre-trained target recognition model to obtain detection target information, including: constructing a target recognition model based on the YOLOv5s algorithm, and setting the number of backbone network C3 modules of the target recognition model to 6; obtaining a historical high-altitude work scene image set, preprocessing and data enhancement processing on the historical high-altitude work scene image set to obtain preprocessed training data; using a labeling tool to label the preprocessed training data to label key features related to the staff, and obtain a labeled training data set; training the target recognition module based on the labeled training data set; inputting the real-time image captured by the camera device into the trained target recognition model to obtain detection target information.

[0010] Optionally, in the fifth implementation method of the first aspect of the present invention, the target comprehensive information is confirmed based on the real-time infrared information fed back by the infrared sensor, the real-time ultrasonic distance information fed back by the ultrasonic sensor, the metal type information fed back by the dual-frequency eddy current sensor, and the detection target information, including: obtaining a preset environmental adaptability coefficient based on the real-time environmental information, and calculating the ultrasonic weight and visual weight based on the preset environmental adaptability coefficient, the real-time infrared information fed back by the infrared sensor, the real-time ultrasonic distance information fed back by the ultrasonic sensor, the metal type information fed back by the dual-frequency eddy current sensor, and the detection target information; confirming the visual distance based on the real-time image fed back by the camera device, and calculating the final distance based on the calculated ultrasonic weight and visual weight, the real-time ultrasonic distance information fed back by the ultrasonic sensor, and the confirmed visual distance; generating target comprehensive information based on the metal type information fed back by the dual-frequency eddy current sensor, the detection target information, and the calculated final distance.

[0011] Optionally, in a sixth implementation method of the first aspect of the present invention, the obtaining of a preset risk level alarm mechanism and the generation of an alarm instruction based on the target comprehensive information and the preset risk level alarm mechanism include: obtaining a preset risk level alarm mechanism; when the target comprehensive information indicates that the target type is personnel and when the final distance is less than the preset personnel alarm distance, generating an emergency alarm instruction based on the risk level alarm mechanism; when the target comprehensive information indicates that the target type is a metal tool and when the final distance is less than the preset tool alarm distance, generating an intermediate alarm instruction based on the risk level alarm mechanism; when the target comprehensive information indicates that the target type is other types and when the final distance is less than the preset other object alarm distance, generating a primary alarm instruction based on the risk level alarm mechanism.

[0012] The second aspect of the present invention provides a safety warning device for high-altitude operation of a distribution tower, comprising: an initialization module for performing a wake-up check on a detection component, wherein the detection component includes an ultrasonic sensor, an infrared sensor, a dual-frequency eddy current sensor, and a camera device; if the detection component passes the wake-up check, the infrared sensor is controlled to start working; a first trigger module for confirming the warning distance of the ultrasonic sensor based on the real-time environmental information when the real-time infrared information fed back by the infrared sensor meets the preset ultrasonic trigger condition, and controlling the ultrasonic sensor to start working; a second trigger module for determining the warning distance of the ultrasonic sensor based on the real-time ultrasonic distance information fed back by the ultrasonic sensor and the confirmed warning distance Determine whether the preset multi-sensor trigger conditions are met. If so, control the dual-frequency eddy current sensor and the camera device to start working; the type recognition module is used to pre-train the target recognition model, and input the real-time image collected by the camera device into the pre-trained target recognition model to obtain the detection target information; the information confirmation module is used to confirm the target comprehensive information based on the real-time infrared information feedback from the infrared sensor, the real-time ultrasonic distance information feedback from the ultrasonic sensor, the metal type information feedback from the dual-frequency eddy current sensor, and the detection target information; the alarm module is used to obtain the preset risk level alarm mechanism, and generate an alarm instruction based on the target comprehensive information and the preset risk level alarm mechanism.

[0013] The third aspect of the present invention provides a distribution tower high-altitude operation safety warning device, the distribution tower high-altitude operation safety warning device includes: a memory and at least one processor, the memory storing instructions; at least one of the processors calls the instructions in the memory to enable the distribution tower high-altitude operation safety warning device to execute each step of the distribution tower high-altitude operation safety warning method described above.

[0014] A fourth aspect of the present invention provides a computer-readable storage medium having instructions stored thereon, which, when executed by a processor, implement the various steps of any of the above-mentioned methods for safety early warning of high-altitude operations on distribution towers.

[0015] In the technical solution of the present invention, firstly, infrared sensors, ultrasonic sensors, dual-frequency eddy current sensors and camera devices are used to work together, and in conjunction with the target recognition model, the behavior of people near the high-altitude work area can be monitored in real time, and timely warnings can be issued when potential dangers are detected, thereby improving the safety of the high-altitude work process; secondly, a hierarchical trigger mechanism is established to trigger the corresponding detection components to start working based on real-time data, reduce the working energy consumption of the detection components, and dynamically adjust the warning distance of the ultrasonic sensor according to the environment, thereby improving the adaptability of the safety warning method to the environment; in addition, the technical solution integrates the triple criteria of distance, target type, and environmental parameters to effectively avoid false alarms that may be generated by a single sensor, thereby improving the overall measurement accuracy and warning accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A logic flow chart of a safety early warning method for high-altitude operations on distribution towers provided by an embodiment of the present invention;

[0017] Figure 2 A schematic diagram of the structure of a safety warning device for high-altitude operations on a power distribution tower provided by an embodiment of the present invention;

[0018] Figure 3 A schematic structural diagram of a safety warning device for high-altitude operations on a distribution tower provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The present invention provides a method, device, equipment and storage medium for safety warning of high-altitude operations on distribution towers. In the present invention, the terms "first", "second", "third", "fourth" and the like (if any) in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or are inherent to these processes, methods, products or devices.

[0020] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 An embodiment of the power distribution tower high-altitude operation safety early warning method in the embodiment of the present invention includes:

[0021] 101. Perform a wake-up check on a detection component, the detection component including an ultrasonic sensor, an infrared sensor, a dual-frequency eddy current sensor, and a camera device; if the detection component passes the wake-up check, the infrared sensor is controlled to start working;

[0022] In this embodiment, the detection device is installed on the safety warning device, and the safety warning device is fixedly installed at an appropriate position of the distribution tower; by performing wake-up verification on the detection component, including verifying the ultrasonic sensor, infrared sensor, dual-frequency eddy current sensor and camera device one by one, it can effectively avoid the detection component from working blindly when it is not necessary, thereby significantly reducing energy consumption and extending the service life of the safety warning device.

[0023] 102. When the real-time infrared information fed back by the infrared sensor meets the preset ultrasonic triggering condition, the warning distance of the ultrasonic sensor is confirmed based on the real-time environmental information, and the ultrasonic sensor is controlled to start working;

[0024] 103. Determine whether a preset multi-sensor trigger condition is met based on the real-time ultrasonic distance information fed back by the ultrasonic sensor and the confirmed warning distance. If so, control the dual-frequency eddy current sensor and the camera device to start working;

[0025] In this embodiment, through the coordinated work of infrared sensors, ultrasonic sensors, dual-frequency eddy current sensors and camera devices, all-round, multi-level comprehensive monitoring of the detection target is achieved; the infrared sensor first captures real-time infrared information, and when the real-time infrared information meets the preset ultrasonic trigger conditions, the ultrasonic sensor's warning distance is accurately confirmed based on the real-time environmental information, and it is controlled to start working; through multi-sensor fusion, not only the accuracy and reliability of detection are improved, but also the probability of false alarms and missed alarms is greatly reduced, so that the safety warning device can still maintain an efficient working state in various complex environments.

[0026] 104. Pre-training a target recognition model, inputting the real-time image captured by the camera device into the pre-trained target recognition model to obtain detection target information;

[0027] 105. Confirm comprehensive target information based on real-time infrared information fed back by the infrared sensor, real-time ultrasonic distance information fed back by the ultrasonic sensor, metal type information fed back by the dual-frequency eddy current sensor, and detected target information;

[0028] 106. Obtain a preset risk level alarm mechanism and generate an alarm instruction based on the target comprehensive information and the preset risk level alarm mechanism;

[0029] In this embodiment, based on the real-time infrared information fed back by the infrared sensor, the real-time ultrasonic distance information fed back by the ultrasonic sensor, the metal type information fed back by the dual-frequency eddy current sensor, and the detection target information, the target comprehensive information is comprehensively confirmed and combined with the preset risk level alarm mechanism to generate an alarm instruction; through multi-level and multi-dimensional information fusion and processing methods, the safety warning device can perform flexible risk assessment and alarm processing according to actual conditions, greatly improving the safety of the high-altitude operation process.

[0030] The present application discloses a safety early warning method for high-altitude operations on distribution towers. First, an infrared sensor, an ultrasonic sensor, a dual-frequency eddy current sensor and a camera device are used to work together, and in conjunction with a target recognition model, the behavior of personnel near the high-altitude operation area can be monitored in real time. When potential danger is detected, a timely early warning is issued to improve the safety of the high-altitude operation process; secondly, a hierarchical trigger mechanism is established to trigger the corresponding detection components to start working based on real-time data, thereby reducing the working energy consumption of the detection components, and dynamically adjusting the warning distance of the ultrasonic sensor according to the environment, thereby improving the adaptability of the safety early warning method to the environment; in addition, the technical solution integrates the triple criteria of distance, target type and environmental parameters to effectively avoid false alarms that may be generated by a single sensor, thereby improving the overall measurement accuracy and early warning accuracy.

[0031] Furthermore, in an embodiment of the present invention, the detection component is waked up and verified, and the detection component includes an ultrasonic sensor, an infrared sensor, a dual-frequency eddy current sensor, and a camera device; if the detection component passes the wake-up verification, the infrared sensor is controlled to start working, including:

[0032] 201. Calling an interface to send a heartbeat packet to the ultrasonic sensor, infrared sensor, dual-frequency eddy current sensor, and camera device respectively, and confirming whether the detection component passes the wake-up check based on the received status code;

[0033] In this embodiment, when the status code is 0x00, it indicates that the corresponding sensor or camera device is normal; when the status code is 0x01, it indicates that the corresponding sensor or camera device is faulty; by calling the interface to send heartbeat packets to each sensor and camera device, and performing wake-up verification based on the received status code, the working status of the detection component can be judged efficiently and accurately, ensuring the stability and reliability of the safety warning device during operation.

[0034] 202. If the detection component passes the wake-up check, the infrared sensor is controlled to start scanning at a frequency of 2 Hz;

[0035] In this embodiment, in the initial stage, only the infrared sensor is controlled to start working, and the ultrasonic sensor, dual-frequency eddy current sensor and camera device are in sleep state, which can effectively reduce the working energy consumption of the detection component and balance the real-time detection and the endurance of the safety warning device.

[0036] 203. If the status code indicates that the ultrasonic sensor fails to wake up, an emergency alarm instruction is generated and the detection component is set to a locked state;

[0037] In this embodiment, if the ultrasonic sensor fails to wake up, an emergency alarm instruction is immediately generated and the detection component is locked, which can respond to the fault in a timely manner, prevent misjudgment or missed judgment due to sensor failure, and improve the safety and emergency handling capabilities of the safety warning device during operation.

[0038] 204. If the status code indicates that the infrared sensor, the dual-frequency eddy current sensor, or the camera device fails to wake up, then obtaining wake-up failure information, and obtaining a corresponding preset backup plan based on the wake-up failure information;

[0039] In this embodiment, when the infrared sensor fails to wake up, visual temperature estimation is performed based on the real-time image fed back by the camera device, and the temperature data is inferred from the color data of the real-time image using the RGB-IR algorithm. Although the detection distance is shortened, the infrared sensor can temporarily take over its function when it fails, thereby ensuring the continuity and reliability of the operation of the safety warning device. When the camera device fails to wake up, the working mode of the camera device is switched to the black and white image mode, and the device is re-awakened. By adjusting the working mode of the camera device, the robustness of the safety warning device can be improved and the risk of device shutdown due to failure of a single mode can be reduced. When the dual-frequency eddy current sensor fails to wake up, the texture features of the target metal are obtained based on the real-time image fed back by the camera device, and compared with the known metal type feature library to confirm the metal type. Although the detection accuracy is reduced, the camera device can also temporarily take over its function when the dual-frequency eddy current sensor fails, thereby ensuring the continuity and reliability of the operation of the safety warning device.

[0040] Furthermore, in an embodiment of the present invention, when the real-time infrared information fed back by the infrared sensor meets a preset ultrasonic trigger condition, determining the warning distance of the ultrasonic sensor based on the real-time environmental information and controlling the ultrasonic sensor to start working includes:

[0041] 301. When the real-time infrared information fed back by the infrared sensor indicates that the temperature of two or more consecutive pixels continuously reaches or exceeds 35° C., obtain real-time environmental information, the real-time environmental information including real-time humidity and real-time wind speed;

[0042] In this embodiment, the infrared information fed back in real time by the infrared sensor can promptly detect and respond to areas with abnormal temperatures, which helps to provide early warning of potential safety hazards.

[0043] 302. Determine environmental conditions based on real-time humidity and real-time wind speed, where the environmental conditions include sunny days, rainy fog, and strong winds;

[0044] In this embodiment, when the real-time humidity is greater than 85%, it indicates that the environmental condition is rain and fog; when the real-time wind speed is greater than 10m / s, it indicates that the environmental condition is strong wind; combining the real-time humidity and real-time wind speed information, the current environmental conditions can be accurately judged to adjust the warning distance and update frequency of the ultrasonic sensor, which not only improves the flexibility and adaptability of the safety warning device during operation, but also maintains high monitoring accuracy and stability under different environmental conditions.

[0045] 303. When the environmental condition is sunny, adjust the warning distance of the ultrasonic sensor to a first set warning value, and control the ultrasonic sensor to start updating the real-time ultrasonic distance information at a frequency of 10 Hz;

[0046] In this embodiment, the first set warning value is 3 meters; under clear weather conditions, updating the real-time ultrasonic distance information at a frequency of 10 Hz helps to quickly capture the movement of the target and improve monitoring efficiency.

[0047] 304. When the environmental condition is rain or fog, adjust the warning distance of the ultrasonic sensor to a second set warning value, where the second set warning value is less than the first set warning value, control the ultrasonic sensor to start updating the real-time ultrasonic distance information at a frequency of 20 Hz, and apply a 15% gain compensation to the real-time ultrasonic distance information;

[0048] In this embodiment, the second set warning value is 2.2 meters. In rainy and foggy conditions, since ultrasonic signals are easily affected by water vapor scattering, resulting in a decrease in ranging accuracy, the warning distance is adjusted to a smaller second set warning value to reduce false alarms. The update frequency is adjusted to 20 Hz and gain compensation is applied to compensate for ultrasonic attenuation in rainy and foggy environments and maintain the effective detection distance, effectively improving the detection capability and accuracy of the safety warning device in adverse weather conditions.

[0049] 305. When the environmental condition is strong wind, adjust the warning distance of the ultrasonic sensor to a third set warning value, wherein the third set warning value is greater than the second set warning value and less than the first set warning value, and control the ultrasonic sensor to begin updating real-time ultrasonic distance information at a frequency of 15 Hz; use a discrete Kalman filter algorithm to process the real-time ultrasonic distance information to obtain a predicted moving speed of the target;

[0050] In this embodiment, the third set warning value is 2.5 meters; under strong wind conditions, the real-time ultrasonic distance information is updated at a moderate frequency, which not only takes into account the impact of strong wind on the target movement speed, but also avoids the waste of resources caused by excessively high frequency; further, the real-time ultrasonic distance information is processed by a discrete Kalman filter algorithm, which can eliminate measurement noise in a strong wind environment, improve the accuracy and reliability of the target predicted movement speed, and provide strong data support for subsequent decision-making and control; specifically, the discrete Kalman filter algorithm analyzes the changing trend of the real-time ultrasonic distance information, combines the preset state transfer equation and observation equation, predicts and updates the position information of the target object, and then further predicts the target movement speed through continuous analysis of the position information.

[0051] Furthermore, in an embodiment of the present invention, the process of determining whether a preset multi-sensor trigger condition is met based on the real-time ultrasonic distance information fed back by the ultrasonic sensor and the confirmed warning distance, and if so, controlling the dual-frequency eddy current sensor and the camera device to start working, includes:

[0052] 401. When the environmental condition is sunny and the real-time ultrasonic distance information fed back by the ultrasonic wave is less than or equal to the first set warning value, control the dual-frequency eddy current sensor and the camera device to start working;

[0053] 402. When the environmental condition is rain and fog and the real-time ultrasonic distance information fed back by the ultrasonic sensor is less than or equal to the second set warning value, the dual-frequency eddy current sensor and the camera device are controlled to start operating, and the real-time image fed back by the camera device is pre-processed using a histogram equalization method;

[0054] In this embodiment, the histogram equalization method is used to pre-process the real-time image fed back by the camera device, which can effectively improve the image quality, enhance the contrast of the image in the rainy and foggy environment, make the detail information in the image more prominent, facilitate the subsequent analysis and processing of the image, and improve the success rate of target recognition.

[0055] 403. When the environmental conditions are strong wind, the real-time ultrasonic distance information fed back by the ultrasonic wave is less than or equal to the third set warning value, and the predicted moving speed is greater than the preset speed threshold, the dual-frequency eddy current sensor and the camera device are controlled to start working;

[0056] In this embodiment, the preset speed threshold is 0.5 m / s.

[0057] In this embodiment, the dual-frequency eddy current sensor is used to detect the type of metal. The dual-frequency eddy current sensor includes two working modes, namely high-frequency mode and low-frequency mode; in high-frequency mode, the dual-frequency eddy current sensor detects small-sized metal within 0.8m, such as wrenches, and distinguishes iron and aluminum through the impedance phase angle, where the phase angle of iron is >45° and the phase angle of aluminum is ≤45°; in low-frequency mode, the dual-frequency eddy current sensor detects large-sized metal within 1.2m, such as scaffolding. If a metal signal is detected three times in a row, it is marked as "high-risk object approaching".

[0058] Furthermore, in an embodiment of the present invention, the pre-trained target recognition model inputs the real-time image captured by the camera device into the pre-trained target recognition model to obtain the detection target information, including:

[0059] 501. Build a target recognition model based on the YOLOv5s algorithm and set the number of C3 modules in the backbone network of the target recognition model to 6;

[0060] In this embodiment, a target recognition model based on the YOLOv5s algorithm is constructed, and the number of C3 modules in the backbone network is optimized to 6, which significantly improves the speed and accuracy of target recognition. The YOLOv5s algorithm is lightweight and efficient, and adjusting the number of C3 modules further adapts to the real-time and accuracy requirements in high-altitude operation scenarios, ensuring timely response to warnings.

[0061] 502. Obtain a historical high-altitude work scene image set, perform preprocessing and data enhancement processing on the historical high-altitude work scene image set, and obtain preprocessed training data;

[0062] In this embodiment, the historical high-altitude work scene image set covers environmental conditions such as sunny days, rain, fog, and strong winds, and includes construction workers wearing and not wearing safety helmets, as well as targets such as metal tools and birds; by preprocessing and data enhancement processing of the historical high-altitude work scene image set, the data enhancement processing includes geometric transformation processing and lighting enhancement processing; it effectively enriches the diversity and robustness of the training data, helps the target recognition model to better learn the complex background and changes in the high-altitude work scenes, and improves the generalization ability of the model in practical applications.

[0063] 503. Using a labeling tool to label the pre-processed training data to label key features related to the staff, thereby obtaining a labeled training data set;

[0064] In this embodiment, the LabelMe annotation tool is used to annotate the preprocessed training data to mark key features related to the workers, such as safety helmets, safety belts, wrenches, etc., so that the target recognition model can more accurately identify the workers and their related features, reduce the misjudgment rate of construction workers, and improve the accuracy of early warning.

[0065] 504. Training the target recognition module based on the labeled training data set;

[0066] 505. Input the real-time image captured by the camera device into the trained target recognition model to obtain detection target information, where the detection target information includes the target type and its corresponding confidence level.

[0067] Furthermore, in an embodiment of the present invention, the target comprehensive information confirmed based on the real-time infrared information fed back by the infrared sensor, the real-time ultrasonic distance information fed back by the ultrasonic sensor, the metal type information fed back by the dual-frequency eddy current sensor, and the detected target information includes:

[0068] 601. Obtain a preset environmental adaptability coefficient based on real-time environmental information, and calculate an ultrasonic weight and a visual weight based on the preset environmental adaptability coefficient, real-time infrared information fed back by the infrared sensor, real-time ultrasonic distance information fed back by the ultrasonic sensor, metal type information fed back by the dual-frequency eddy current sensor, and detection target information;

[0069] In this embodiment, the calculation formula of the ultrasound weight is:

[0070]

[0071] Among them, S in the molecule i is the sensitivity of the ultrasonic sensor, C in the molecule i is the environmental adaptability coefficient of the ultrasonic sensor; and S in the denominator i Refers to the sensitivity of the three types of sensors and the sensitivity of the camera device. The C in the denominator i Refers to the environmental adaptability coefficient corresponding to the sensor and the environmental adaptability coefficient corresponding to the camera device; where S i The quantitative indicators are as follows:

[0072] 1. The sensitivity of the ultrasonic sensor = 1 / (ranging error). If the ranging error is ±3cm, then S 超声 is 33.3;

[0073] 2. The sensitivity of the camera device is the confidence level of the target recognition model output;

[0074] 3. Sensitivity of infrared sensor = effective number of thermal imaging pixels / total number of pixels. If 5 effective pixels are detected, S 红外=5 / 192≈0.026;

[0075] 4. The sensitivity of the dual-frequency eddy current sensor is the metal detection type reliability, such as 1 for iron, 0.8 for aluminum, and 0 for others;

[0076] C i The environmental adaptability coefficient ranges from 0.5 to 1.5. For example, in rainy and foggy weather, the environmental adaptability coefficient of the ultrasonic sensor is 0.8, and the environmental adaptability coefficient of the camera is 1.2. In night scenes, the environmental adaptability coefficient of the infrared sensor is 1.5, and the environmental adaptability coefficient of the camera is 0.7.

[0077] For example, suppose S 超声 The value of the human confidence level is 0.9, and the corresponding environmental adaptability coefficient is 1. The infrared sensor detects 8 valid pixels, and the corresponding environmental adaptability coefficient is 1. The ultrasonic weight is:

[0078]

[0079] The visual weight = 1-ultrasound weight.

[0080] In this embodiment, the environmental adaptability coefficient is obtained based on real-time environmental information, and the detection parameters can be automatically adjusted according to different environmental conditions, thereby improving the detection accuracy and stability in different environments.

[0081] 602. Confirm the visual distance based on the real-time image fed back by the camera device, and calculate the final distance based on the calculated ultrasonic weight and visual weight, the real-time ultrasonic distance information fed back by the ultrasonic sensor, and the confirmed visual distance;

[0082] In this embodiment, the visual distance of the target can be confirmed based on the real-time image feedback from the camera device. Specifically, the real-time image containing the object and its background information is captured by the camera device, and the object and its outline are identified using image processing technology. Then, a depth perception algorithm or stereo vision technology is used, combined with the parameters of the camera device, such as focal length, field of view, etc., as well as the scale in the image or the size of the known object, to analyze the position of the object in three-dimensional space, thereby inferring the actual distance between the target and the camera device.

[0083] In this embodiment, the final distance = ultrasonic weight * real-time ultrasonic distance information + visual weight * visual distance; by calculating the ultrasonic weight and visual weight, the advantages of multiple detection methods are complementary, which can effectively improve the distance measurement accuracy and provide a reliable foundation for subsequent target recognition and positioning.

[0084] 603. Generate target comprehensive information based on the metal type information fed back by the dual-frequency eddy current sensor, the detection target information, and the calculated final distance, where the target comprehensive information includes the target type and the final distance.

[0085] Furthermore, in an embodiment of the present invention, the step of obtaining a preset risk level warning mechanism and generating an alarm instruction based on the target comprehensive information and the preset risk level warning mechanism includes:

[0086] 701. Obtain the preset risk level warning mechanism;

[0087] 702. When the comprehensive target information indicates that the target type is a person and the final distance is less than the preset person warning distance, an emergency alarm instruction is generated based on the risk level alarm mechanism;

[0088] In this embodiment, the preset personnel warning distance is 1 meter, and the emergency alarm instruction includes: controlling the red light to flash, using the voice broadcast device to cyclically broadcast the voice prompt "High voltage danger! Evacuate immediately", and controlling the buzzer to operate at a frequency of 1000Hz.

[0089] 703. When the target comprehensive information indicates that the target type is a metal tool and when the final distance is less than the preset tool warning distance, a medium-level alarm instruction is generated based on the risk level alarm mechanism;

[0090] In this embodiment, the preset tool warning distance is 0.8 meters; the intermediate alarm instructions include: controlling the yellow light to be always on, using the voice broadcast device to cyclically broadcast the voice prompt "Metal object approaching, please pay attention", and controlling the buzzer to operate at a frequency of 500Hz.

[0091] 704. When the comprehensive target information indicates that the target type is other and the final distance is less than the preset warning distance for other objects, a primary alarm instruction is generated based on the risk level alarm mechanism;

[0092] In this embodiment, the preset other object warning distance is 2 meters; the primary alarm instruction includes: controlling the blue light to flash intermittently, using the voice broadcast device to cyclically broadcast the voice prompt "An object is approaching, please confirm", and controlling the buzzer to operate at a frequency of 300Hz.

[0093] In this embodiment, intelligent risk identification and assessment are performed for different types of targets to achieve precise responses, which helps to improve the accuracy and efficiency of security monitoring and reduce false positives and missed alerts.

[0094] The above describes the power distribution tower high altitude operation safety warning method according to the embodiment of the present invention. The following describes the power distribution tower high altitude operation safety warning device according to the embodiment of the present invention. Figure 2In one embodiment of the present invention, a safety warning device for high-altitude operation of a power distribution tower includes:

[0095] Initialization module 801 is used to perform a wake-up check on the detection component, which includes an ultrasonic sensor, an infrared sensor, a dual-frequency eddy current sensor, and a camera device; if the detection component passes the wake-up check, the infrared sensor is controlled to start working;

[0096] The first trigger module 802 is configured to determine the warning distance of the ultrasonic sensor based on the real-time environmental information and control the ultrasonic sensor to start working when the real-time infrared information fed back by the infrared sensor meets the preset ultrasonic trigger condition;

[0097] The second trigger module 803 is used to determine whether the preset multi-sensor trigger condition is met based on the real-time ultrasonic distance information fed back by the ultrasonic sensor and the confirmed warning distance. If so, it controls the dual-frequency eddy current sensor and the camera device to start working;

[0098] The type recognition module 804 is used to pre-train the target recognition model, input the real-time image captured by the camera device into the pre-trained target recognition model, and obtain the detection target information;

[0099] An information confirmation module 805 is configured to confirm comprehensive target information based on real-time infrared information fed back by the infrared sensor, real-time ultrasonic distance information fed back by the ultrasonic sensor, metal type information fed back by the dual-frequency eddy current sensor, and detected target information;

[0100] The alarm module 806 is used to obtain a preset risk level alarm mechanism and generate an alarm instruction based on the target comprehensive information and the preset risk level alarm mechanism.

[0101] Based on the same idea as the method in the above embodiment, the device provided in this application can implement the method in the above embodiment.

[0102] above Figure 2 The distribution tower high-altitude operation safety warning device in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The distribution tower high-altitude operation safety warning device in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0103] Figure 3FIG. 1 is a schematic diagram of the structure of a safety warning device for high-altitude operations on distribution towers, provided in an embodiment of the present invention. The safety warning device 900 for high-altitude operations on distribution towers may vary significantly depending on configuration or performance. The device may include one or more processors (central processing units, CPUs) 910 (e.g., one or more processors), a memory 920, and one or more storage media 930 (e.g., one or more mass storage devices) storing application programs 933 or data 932. The memory 920 and storage medium 930 may be either transient or persistent storage. The program stored in the storage medium 930 may include one or more modules (not shown), each of which may include a series of instructions and operations within the safety warning device 900 for high-altitude operations on distribution towers. Furthermore, the processor 910 may be configured to communicate with the storage medium 930, executing the series of instructions and operations stored in the storage medium 930 on the safety warning device 900 for high-altitude operations on distribution towers, thereby implementing the steps of the safety warning method for high-altitude operations on distribution towers provided in the aforementioned method embodiments.

[0104] The power distribution tower aerial work safety warning device 900 may also include one or more power supplies 940, one or more wired or wireless network interfaces 950, one or more input and output interfaces 960, and / or one or more operating systems 931, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. It will be understood by those skilled in the art that Figure 3 The structure of the distribution tower high-altitude operation safety warning device shown does not constitute a limitation of the distribution tower high-altitude operation safety warning device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0105] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the steps of a safety warning method for high-altitude operations on distribution towers.

[0106] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0107] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, etc., various media that can store program code.

[0108] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A safety early warning method for high-altitude operation of distribution towers, characterized in that: include: Performing a wake-up check on the detection component, wherein the detection component includes an ultrasonic sensor, an infrared sensor, a dual-frequency eddy current sensor, and a camera device; If the detection component passes the wake-up check, the infrared sensor is controlled to start working; When the real-time infrared information fed back by the infrared sensor meets the preset ultrasonic trigger conditions, the warning distance of the ultrasonic sensor is confirmed based on the real-time environmental information, and the ultrasonic sensor is controlled to start working; Based on the real-time ultrasonic distance information fed back by the ultrasonic sensor and the confirmed warning distance, it is determined whether the preset multi-sensor trigger conditions are met. If so, the dual-frequency eddy current sensor and the camera device are controlled to start working; Pre-training the target recognition model, inputting the real-time image captured by the camera device into the pre-trained target recognition model to obtain the detection target information; Confirm comprehensive target information based on real-time infrared information fed back by infrared sensors, real-time ultrasonic distance information fed back by ultrasonic sensors, metal type information fed back by dual-frequency eddy current sensors, and detected target information; Obtain the preset risk level alarm mechanism and generate an alarm instruction based on the target comprehensive information and the preset risk level alarm mechanism.

2. The power distribution tower high-altitude operation safety early warning method according to claim 1 is characterized in that: The detection component is waked up and checked, and the detection component includes an ultrasonic sensor, an infrared sensor, a dual-frequency eddy current sensor, and a camera device; if the detection component passes the wake-up check, the infrared sensor is controlled to start working, including: The call interface sends heartbeat packets to the ultrasonic sensor, infrared sensor, dual-frequency eddy current sensor and camera device respectively, and confirms whether the detection component passes the wake-up check based on the received status code; If the detection component passes the wake-up check, the infrared sensor is controlled to start scanning at a frequency of 2Hz; If the status code indicates that the ultrasonic sensor fails to wake up, an emergency alarm command is generated and the detection component is set to a locked state; If the status code indicates that the infrared sensor, the dual-frequency eddy current sensor or the camera device fails to wake up, then wake-up failure information is obtained, and a corresponding preset backup plan is obtained based on the wake-up failure information.

3. The power distribution tower high-altitude operation safety early warning method according to claim 1 is characterized in that: When the real-time infrared information fed back by the infrared sensor meets the preset ultrasonic trigger condition, the warning distance of the ultrasonic sensor is confirmed based on the real-time environmental information, and the ultrasonic sensor is controlled to start working, including: When the real-time infrared information fed back by the infrared sensor indicates that the temperature of two or more consecutive pixels continuously reaches or exceeds 35° C., real-time environmental information is obtained, wherein the real-time environmental information includes real-time humidity and real-time wind speed; confirming environmental conditions based on real-time humidity and real-time wind speed, wherein the environmental conditions include sunny days, rainy fog, and strong winds; When the environmental condition is sunny, the warning distance of the ultrasonic sensor is adjusted to a first set warning value, and the ultrasonic sensor is controlled to start updating the real-time ultrasonic distance information at a frequency of 10 Hz; When the environmental condition is rain or fog, the warning distance of the ultrasonic sensor is adjusted to a second set warning value, which is less than the first set warning value, and the ultrasonic sensor is controlled to start updating the real-time ultrasonic distance information at a frequency of 20 Hz, and a 15% gain compensation is applied to the real-time ultrasonic distance information; When the environmental condition is strong wind, the warning distance of the ultrasonic sensor is adjusted to a third set warning value, which is greater than the second set warning value and less than the first set warning value, and the ultrasonic sensor is controlled to start updating real-time ultrasonic distance information at a frequency of 15 Hz; the real-time ultrasonic distance information is processed using a discrete Kalman filter algorithm to obtain the predicted movement speed of the target.

4. The power distribution tower high-altitude operation safety early warning method according to claim 3 is characterized in that: The method of determining whether a preset multi-sensor trigger condition is met based on the real-time ultrasonic distance information fed back by the ultrasonic sensor and the confirmed warning distance, and if so, controlling the dual-frequency eddy current sensor and the camera device to start working, includes: When the environmental condition is sunny and the real-time ultrasonic distance information fed back by the ultrasonic wave is less than or equal to the first set warning value, the dual-frequency eddy current sensor and the camera device are controlled to start working; When the environmental condition is rain and fog and the real-time ultrasonic distance information fed back by the ultrasonic wave is less than or equal to the second set warning value, the dual-frequency eddy current sensor and the camera device are controlled to start working, and the real-time image fed back by the camera device is pre-processed using a histogram equalization method; When the environmental conditions are strong wind, the real-time ultrasonic distance information fed back by the ultrasonic wave is less than or equal to the third set warning value, and the predicted moving speed is greater than the preset speed threshold, the dual-frequency eddy current sensor and the camera device are controlled to start working.

5. The power distribution tower high-altitude operation safety early warning method according to claim 1 is characterized in that: The pre-trained target recognition model inputs the real-time image captured by the camera device into the pre-trained target recognition model to obtain detection target information, including: Build a target recognition model based on the YOLOv5s algorithm and set the number of C3 modules in the backbone network of the target recognition model to 6; Obtain a historical high-altitude work scene image set, perform preprocessing and data enhancement on the historical high-altitude work scene image set, and obtain preprocessed training data; Use annotation tools to annotate the pre-processed training data to annotate key features related to the staff and obtain an annotated training data set; Training the target recognition module based on the labeled training data set; The real-time image captured by the camera device is input into the trained target recognition model to obtain the detection target information.

6. The power distribution tower high-altitude operation safety early warning method according to claim 1 is characterized in that: The target comprehensive information confirmed based on the real-time infrared information fed back by the infrared sensor, the real-time ultrasonic distance information fed back by the ultrasonic sensor, the metal type information fed back by the dual-frequency eddy current sensor, and the detected target information includes: Obtaining a preset environmental adaptability coefficient based on real-time environmental information, and calculating an ultrasonic weight and a visual weight based on the preset environmental adaptability coefficient, real-time infrared information fed back by the infrared sensor, real-time ultrasonic distance information fed back by the ultrasonic sensor, metal type information fed back by the dual-frequency eddy current sensor, and detection target information; confirming the visual distance based on the real-time image fed back by the camera device, and calculating the final distance based on the calculated ultrasonic weight and visual weight, the real-time ultrasonic distance information fed back by the ultrasonic sensor, and the confirmed visual distance; Target comprehensive information is generated based on the metal type information fed back by the dual-frequency eddy current sensor, the detection target information and the calculated final distance.

7. The power distribution tower high-altitude operation safety early warning method according to claim 1 is characterized in that: The step of obtaining a preset risk level warning mechanism and generating an alarm instruction based on the target comprehensive information and the preset risk level warning mechanism includes: Obtain the preset risk level warning mechanism; When the comprehensive target information indicates that the target type is a person and when the final distance is less than the preset person warning distance, an emergency alarm instruction is generated based on the risk level alarm mechanism; When the comprehensive target information indicates that the target type is a metal tool and the final distance is less than the preset tool warning distance, an intermediate alarm instruction is generated based on the risk level alarm mechanism; When the comprehensive target information indicates that the target type is other types and when the final distance is less than the preset warning distance for other objects, a primary alarm instruction is generated based on the risk level alarm mechanism.

8. A safety warning device for high-altitude operation of distribution tower, characterized in that: include: An initialization module is used to perform a wake-up check on the detection component, which includes an ultrasonic sensor, an infrared sensor, a dual-frequency eddy current sensor, and a camera device; If the detection component passes the wake-up check, the infrared sensor is controlled to start working; The first trigger module is used to determine the warning distance of the ultrasonic sensor based on the real-time environmental information and control the ultrasonic sensor to start working when the real-time infrared information fed back by the infrared sensor meets the preset ultrasonic trigger condition; The second trigger module is used to determine whether the preset multi-sensor trigger conditions are met based on the real-time ultrasonic distance information fed back by the ultrasonic sensor and the confirmed warning distance. If so, it controls the dual-frequency eddy current sensor and the camera device to start working; The type recognition module is used to pre-train the target recognition model, input the real-time image captured by the camera device into the pre-trained target recognition model, and obtain the detection target information; An information confirmation module is used to confirm comprehensive target information based on real-time infrared information fed back by the infrared sensor, real-time ultrasonic distance information fed back by the ultrasonic sensor, metal type information fed back by the dual-frequency eddy current sensor, and detected target information; The alarm module is used to obtain a preset risk level alarm mechanism and generate an alarm instruction based on the target comprehensive information and the preset risk level alarm mechanism.

9. A safety warning device for high-altitude operation of distribution tower, characterized in that: The power distribution tower high-altitude operation safety warning device includes: a memory and at least one processor, wherein the memory stores instructions; At least one of the processors calls the instructions in the memory to enable the distribution tower aerial work safety warning device to execute each step of the distribution tower aerial work safety warning method as described in any one of claims 1-7.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the various steps of the distribution tower high-altitude operation safety early warning method as described in any one of claims 1-7 are implemented.

Images