Icing jump monitoring and alarming method and system

Abstract

The invention discloses an icing jump monitoring and alarming method and system, and relates to the technical field of power online monitoring, and the method comprises the steps: marking historical icing monitoring pictures, generating a first data set and a second data set, and building a lead detection model and a shockproof hammer target detection model through a deep learning model. And detecting a picture shot in real time through the wire detection model, and if an iced wire is detected, starting a pre-recording mode. And detecting the video frame by using the vibration damper target detection model, analyzing the posture of the vibration damper, judging the icing jump condition of the lead, and controlling the start and stop of video recording. The energy consumption and the weight of the monitoring device are reduced. By utilizing the characteristic that the stockbridge damper is parallel to the lead, a lead jumping track is restored through an image recognition technology without an additional sensor; a deicing jump monitoring function is rapidly added on an existing intelligent camera through remote upgrading, and power consumption in a non-icing state is effectively reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of power online monitoring, and in particular to an ice jump monitoring and alarm method and system. Background Art

[0002] In the field of power transmission, the ice shedding jump phenomenon of transmission lines is an important research topic. Ice shedding jump refers to the sudden fall of ice and snow covering on transmission lines under certain conditions, causing violent vibration and displacement of the conductors, which may have a serious impact on the safety and stability of the line. Therefore, accurate monitoring and evaluation of the ice shedding jump behavior of transmission lines is crucial to ensure the reliable operation of the power system.

[0003] Existing monitoring methods and systems have made some progress in monitoring ice shedding and ice jumping on transmission lines, but there are still some problems and challenges. For example, some existing monitoring devices are complex in structure and require the installation of multiple sensors on the transmission lines, which may increase the cost and difficulty of deployment and maintenance in large-scale applications. In addition, some methods rely on modifications to the suspension structure of existing transmission line towers, which may lead to additional engineering work and potential risks. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is: how to provide a method and system for monitoring the ice jumping of transmission line conductors, to solve the problem that the existing monitoring methods are complex to install and are not suitable for large-scale promotion, and to provide a low-energy and easy-to-install implementation method and system for solving the online monitoring of ice jumping of transmission line conductors.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present invention provides an ice jumping monitoring and alarm method, comprising:

[0008] Annotate historical ice monitoring photos to obtain a first data set and a second data set, and use a deep learning model to establish a first model and a second model;

[0009] Using the first data set to train the first model to obtain a wire detection model;

[0010] The second model is trained using the second data set to obtain an anti-vibration hammer target detection model;

[0011] Use the wire detection model to detect the photos taken in real time, and if an ice-covered wire target is detected, perform the first operation;

[0012] The anti-shock hammer target detection model is used to detect the video frames extracted from the real-time video, and then the second operation is performed.

[0013] As a preferred solution for ice jump monitoring and alarm method, wherein:

[0014] The first operation includes:

[0015] When the monitoring point enters the ice-covered state, the video pre-recording mode is turned on, and the pre-recorded content is saved for a preset time length;

[0016] Set the initial pre-recording time length, adjustment length, alarm trigger threshold and safety threshold. When the ice thickness exceeds the alarm trigger threshold, the pre-recording time will be increased by an adjustment length; when the ice thickness is lower than the safety threshold, the pre-recording time will be reduced by an adjustment length.

[0017] As a preferred solution for ice jump monitoring and alarm method, wherein:

[0018] The use of the anti-vibration hammer target detection model to detect the video frames extracted from the real-time video includes:

[0019] The extracted video frame content is subjected to target monitoring, the identified anti-vibration hammer position coordinates and angle are recorded, and the data of a preset time length is cyclically stored.

[0020] As a preferred solution for ice jump monitoring and alarm method, wherein:

[0021] The first data set includes: selecting ice-covered and non-ice-covered photos from historical ice-covered monitoring photos for annotation to obtain the first data set; the second data set includes: selecting shock-absorbing hammer photos from historical ice-covered monitoring photos for annotation to obtain the second data set.

[0022] As a preferred solution for ice jump monitoring and alarm method, wherein:

[0023] The second operation includes:

[0024] Calculate the trajectory of the wire according to the posture of the anti-vibration hammer;

[0025] Determine whether ice jump occurs on the conductor: When periodic changes are detected in both the position sequence data and the angle sequence data, it is determined that ice jump occurs on the transmission line.

[0026] As a preferred solution for ice jump monitoring and alarm method, wherein:

[0027] The second operation further includes:

[0028] When ice jumping is detected, the alarm mechanism is triggered;

[0029] When ice jumping is detected and the video recording function is not in progress, the pre-recorded content is saved and the video recording function is turned on;

[0030] When no periodic change is detected in the cyclically stored data, it is determined that the ice jump is completed.

[0031] As a preferred solution for ice jump monitoring and alarm method, wherein:

[0032] The second operation further includes:

[0033] After judging that the ice-covered jump is finished, the ongoing video recording is terminated and the video file is uploaded to the upper-level main station;

[0034] At the same time, the alarm records of ice-covered jumping will be uploaded.

[0035] In a second aspect, an embodiment of the present invention provides an ice jumping monitoring and alarm system, comprising:

[0036] A data annotation and model building module is used to annotate historical ice monitoring photos to obtain a first data set and a second data set, and to build a first model and a second model using a deep learning model;

[0037] A wire detection model training module, used to train the first model using the first data set to obtain a wire detection model;

[0038] A shock-absorbing hammer target detection model training module, used to train the second model using the second data set to obtain a shock-absorbing hammer target detection model;

[0039] A wire ice monitoring and response module is used to detect the real-time photographs using a wire detection model, and if an ice-covered wire target is detected, perform a first operation;

[0040] The ice-covered jump monitoring and video recording control module is used to use the anti-shock hammer target detection model to detect the video frames extracted from the real-time video and perform the second operation.

[0041] In a third aspect, an embodiment of the present invention provides a computing device, including:

[0042] Memory and processor;

[0043] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the one or more programs are executed by the one or more processors, the one or more processors implement the ice jump monitoring and alarm method as described in any embodiment of the present invention.

[0044] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the ice jump monitoring and alarm method.

[0045] Beneficial effects of the present invention: The present invention introduces image intelligent analysis technology into the transmission line, and dynamically starts and controls the camera to perform pre-shooting based on the principle that deicing jump may occur only after the conductor is covered with ice, thereby reducing the energy consumption required for long-term operation of the camera and effectively reducing the overall weight of the monitoring device; the anti-shock hammer is parallel to the conductor, and the image recognition technology is used to detect the running posture of the anti-shock hammer, thereby determining the running posture of the conductor, analyzing the end conditions of the ice jump, and further controlling the device to stop shooting; the present invention can quickly add the deicing jump monitoring function to the existing smart camera through remote upgrade; in response to the high power consumption problem caused by long-term recording, the present invention adopts a method of intelligently identifying the ice scene and dynamically starting the recording function, so that the operating power consumption of the system in non-icing conditions is greatly reduced. On the other hand, the posture of the anti-shock hammer is detected by the rotating target detection method, and the wire jump trajectory is restored by using the characteristic that the anti-shock hammer is parallel to the wire, achieving the goal of restoring the deicing jump of the wire without additional sensors and markers. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0047] Figure 1 It is an overall flow chart of the ice jump monitoring and alarm method described in the present invention. DETAILED DESCRIPTION

[0048] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0049] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0050] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0051] Example 1

[0052] Reference Figure 1 , which is the first embodiment of the present invention, provides an ice jumping monitoring and alarm method, comprising:

[0053] S1: annotate historical ice monitoring photos to obtain a first data set and a second data set, and use a deep learning model to establish a first model and a second model;

[0054] S2: training the first model using the first data set to obtain a wire detection model;

[0055] S3: training the second model using the second data set to obtain a shock-absorbing hammer target detection model;

[0056] S4: Using the wire detection model to detect the real-time photograph, if an ice-covered wire target is detected, performing a first operation;

[0057] S5: Use the anti-shock hammer target detection model to detect the video frames extracted from the real-time video and perform the second operation.

[0058] It should be noted that through steps S1-S5, intelligent monitoring and response to the icing condition of the transmission line can be achieved, which not only improves the accuracy and efficiency of monitoring, but also can timely discover and deal with potential risks such as icing jumps, thereby ensuring the safe and stable operation of the transmission line.

[0059] Example 2

[0060] Reference Figure 1 , which is an embodiment of the present invention, provides an ice jumping monitoring and alarm method based on the previous embodiment, including:

[0061] In the embodiment of the present application, the first data set in the above step S1 includes:

[0062] Ice-covered and non-ice-covered photos were selected from historical ice-covered monitoring photos for annotation and training.

[0063] Specifically, a large number of transmission line conductor images are collected, including ice-covered and non-ice-covered states; these images are annotated to mark which images are ice-covered and which are non-ice-covered.

[0064] In the embodiment of the present application, the second data set in the above step S1 includes:

[0065] Anti-vibration hammer photos are selected from historical ice monitoring photos for annotation and training, and the annotation method adopts the rotating target annotation method.

[0066] Specifically, an image of a transmission line conductor installed with a shock-absorbing hammer is collected.

[0067] Accurately mark the shock absorber in the image. Since the shock absorber may have different angles and postures, a rotating target marking method is required to record its position and angle information.

[0068] In the embodiment of the present application, the use of the deep learning model to establish the first model and the second model in the above step S1 includes:

[0069] Select a suitable machine learning or deep learning model, such as a classification model in a convolutional neural network (CNN). Subsequently, use the labeled ice-covered and non-ice-covered image dataset to train the first model and optimize the first model parameters so that it can accurately distinguish between ice-covered and non-ice-covered states.

[0070] The second model is trained using a dataset of labeled anti-shock hammer images, enabling it to accurately detect and locate the anti-shock hammer in the image and estimate its angle.

[0071] In another possible implementation, a deep learning model suitable for target detection is selected, such as Faster R-CNN, YOLO (You Only Look Once), etc.

[0072] Due to the rotational characteristics of the shock-absorbing hammer, it may be necessary to use a detection model that can handle rotating targets, or to adjust and optimize the model appropriately.

[0073] In another possible implementation, in order to improve the generalization ability of the model, data enhancement may be performed on the labeled data, such as rotation, scaling, cropping, adding noise, etc.

[0074] For the annotation of the shock-absorbing hammer, in addition to the position and angle, its scale information can also be annotated so that the model can better understand the size changes of the target.

[0075] In the embodiment of the present application, the first model is trained using the first data set in the above step S2 to obtain the wire detection model, including:

[0076] Fine-tune uses a model pre-trained on a large-scale dataset as a starting point for the task of ice detection; combines multiple models with different architectures to improve detection accuracy through voting or weighted averaging; focuses on samples that are difficult for the model to distinguish, conducts targeted training, and improves the robustness of the model.

[0077] In the embodiment of the present application, the second model is trained using the second data set in the above step S3 to obtain the shock-absorbing hammer target detection model, which includes:

[0078] Use a detection model that can handle rotated targets, such as the Rotated Bounding Box detection model. In addition to the position of the target, the model also needs to learn to predict the angle of the target. A regression head can be used to accurately estimate the angle. If the proportion of the shock absorber in the image is small, sampling techniques can be used to balance positive and negative samples to prevent the model from being biased towards the majority class.

[0079] In the embodiment of the present application, the wire detection model is used in the above step S4 to detect the real-time photograph. If an ice-covered wire target is detected, the first operation includes:

[0080] It should be noted that this embodiment conventionally collects images and videos of transmission line conductors: a camera is installed on a pole tower, and the conductor is photographed from the side to ensure that the anti-vibration hammer installed on the conductor can be photographed, and a photo of the conductor is taken every hour. The photographed photo is detected using a conductor detection model, and if an ice-covered conductor target is detected in the photo, the monitoring point is marked as entering an ice-covered state.

[0081] The first operation includes:

[0082] When the monitoring point enters the ice-covered state, start the video pre-recording mode: turn on the camera, start pre-recording, and save the pre-recording content for a preset length of time;

[0083] Set the initial pre-recording time length, adjustment length, alarm trigger threshold and safety threshold. When the ice thickness exceeds the alarm trigger threshold, the pre-recording time will be increased by an adjustment length; when the ice thickness is lower than the safety threshold, the pre-recording time will be reduced by an adjustment length.

[0084] In the embodiment of the present application, the storage time is preferably 5S.

[0085] Exemplary, initial parameter settings:

[0086] Initial pre-recording time length: 10 minutes.

[0087] Adjusted length: 2 minutes.

[0088] Alarm trigger threshold: 5mm.

[0089] Safety threshold: 2mm.

[0090] Dynamic adjustment rules:

[0091] When the ice thickness exceeds the alarm trigger threshold:

[0092] If the ice thickness is >5mm, increase the pre-recording time by an adjustment length (2 minutes).

[0093] The initial pre-recording time is 10 minutes, and the ice thickness is 6 mm, so the adjusted pre-recording time is 12 minutes.

[0094] When the ice thickness is below the safety threshold:

[0095] If the ice thickness is less than 2 mm, the pre-recording time is reduced by an adjustment length (2 minutes).

[0096] The initial pre-recording time is 10 minutes, and the ice thickness is 1.5 mm, so the adjusted pre-recording time is 8 minutes.

[0097] In another possible implementation, the first operation may further include:

[0098] After detecting ice-covered wires, the system can record the start time of icing, the severity of icing (such as ice thickness or density), and monitoring point information.

[0099] Ice cover alarm information is sent to relevant operation and maintenance personnel or monitoring centers through SMS, email, mobile phone APP push or sound and light alarm, reminding them to pay attention to the ice cover of transmission lines.

[0100] In the embodiment of the present application, the above step S5 uses the anti-vibration hammer target detection model to detect the video frames extracted from the real-time video, and the second operation includes:

[0101] The extracted video frame content is subjected to target monitoring, the identified anti-vibration hammer position coordinates Xn, Yn and angle An are recorded, and the data of a preset time length is cyclically stored.

[0102] The second operation includes:

[0103] Calculate the wire motion trajectory according to the posture of the anti-vibration hammer: perform periodic data analysis on the position sequence and angle sequence saved in the loop, and record the position and change amplitude of the periodic data;

[0104] Determine whether the conductor has ice jump: When periodic changes are detected in the position sequence data and the angle sequence data at the same time, it is determined that ice jump has occurred in the transmission line;

[0105] When ice jump is detected, the alarm mechanism is triggered: ice jump alarm information is sent to the operation and maintenance personnel or the monitoring center through SMS, email, sound and light alarm, etc. The alarm information includes detailed data of the ice jump, such as the time when the jump occurred, the duration, the intensity, etc.

[0106] When ice jumping is detected and the video recording function is not in progress, the pre-recorded content is saved and the video recording function is turned on;

[0107] When no periodic changes are detected in the cyclically stored data, it is determined that the ice jump is finished;

[0108] After determining that the ice-covered jump is completed, the ongoing video recording is terminated and the video file is uploaded to the upper-level main station.

[0109] At the same time, the alarm records and related data of ice jumping are uploaded for subsequent analysis and processing.

[0110] In another possible implementation, the second operation may further include:

[0111] By detecting the shape changes of the shock-absorbing hammer, it is possible to identify whether it is damaged or deformed, so that the damaged shock-absorbing hammer can be discovered and replaced in time.

[0112] Based on the position and posture changes of the shock-absorbing hammer, it is evaluated whether its shock-absorbing performance meets expectations, providing a basis for preventive maintenance.

[0113] When ice-covered jumps are detected, the resolution and frame rate of the video are automatically increased to ensure that more detailed dynamic changes can be captured.

[0114] Dynamically adjust the length of the video recording based on the duration and intensity of the ice jump to ensure that the entire event is fully recorded.

[0115] Example 3

[0116] The above is a schematic scheme of the ice jump monitoring and alarm method of this embodiment. It should be noted that the technical scheme of the ice jump monitoring and alarm system and the technical scheme of the above ice jump monitoring and alarm method belong to the same concept, and the details not described in detail in the technical scheme of the ice jump monitoring and alarm system in this embodiment can be referred to the description of the technical scheme of the above ice jump monitoring and alarm method.

[0117] This embodiment also provides a system based on ice jumping monitoring and alarm method, including:

[0118] A data annotation and model building module is used to annotate historical ice monitoring photos to obtain a first data set and a second data set, and to build a first model and a second model using a deep learning model;

[0119] A wire detection model training module, used to train the first model using the first data set to obtain a wire detection model;

[0120] A shock-absorbing hammer target detection model training module, used to train the second model using the second data set to obtain a shock-absorbing hammer target detection model;

[0121] A wire ice monitoring and response module is used to detect the real-time photographs using a wire detection model, and if an ice-covered wire target is detected, perform a first operation;

[0122] The ice-covered jump monitoring and video recording control module is used to use the anti-shock hammer target detection model to detect the video frames extracted from the real-time video and perform the second operation.

[0123] This embodiment also provides a computing device, which is applicable to the case of ice-covered jump monitoring and alarm method, including:

[0124] Memory and processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the ice jump monitoring and alarm method proposed in the above embodiment.

[0125] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the ice jump monitoring and alarm method proposed in the above embodiment is implemented.

[0126] The storage medium proposed in this embodiment and the ice jump monitoring and alarm method proposed in the above embodiment belong to the same inventive concept. The technical details not fully described in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0127] Example 4

[0128] Reference Figure 1 , which is an embodiment of the present invention, provides an ice jump monitoring and alarm method. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through simulation experiments.

[0129] 1. Simulation data generation:

[0130] Ice cover data:

[0131] Based on the ice-covered images of actual transmission lines, simulated images of different ice thickness and density are generated. The gradual accumulation of ice is simulated by computer graphics technology, and a dynamic image sequence of ice-covered conductors is generated.

[0132] Shockproof hammer data:

[0133] The motion of the shock-absorbing hammer at different angles and postures is simulated to generate an image sequence containing the rotation, tilt and shaking of the shock-absorbing hammer. Combined with the dynamic scene of ice-covered jumping, the posture changes of the shock-absorbing hammer during the jumping process are simulated.

[0134] Background:

[0135] Simulate the transmission line background under different weather conditions (such as sunny, rainy, and snowy days) to ensure that the model can work normally in complex environments.

[0136] 2. Simulation platform:

[0137] Use simulation software (such as MATLAB Simulink) to build a physical model of ice jumping on transmission lines.

[0138] The dynamic process of conductor icing and jumping is simulated to generate dynamic data of icing and jumping.

[0139] Combined with the deep learning model, a simulation system for ice detection and shock hammer monitoring is built.

[0140] Simulation process and steps:

[0141] Use the simulated ice data to annotate (ice-covered and non-ice-covered states) and establish the first dataset. Train the wire detection model based on a deep learning model (such as FasterRCNN or YOLO) and evaluate the model's accuracy, recall, and F1 score.

[0142] Shock-absorbing hammer detection model training:

[0143] The simulated anti-vibration hammer data is annotated (position, angle and posture) to establish the second data set. A rotation target detection model (such as the RotatedBoundingBox detection model) is selected for training, and the mAP and angle estimation error of the model are evaluated.

[0144] Ice detection simulation:

[0145] The simulated ice-covered images are input into the wire detection model to verify whether the model can accurately identify ice-covered wires. The time when ice-covered wires start and the dynamic process of ice-covered thickness changes are set in the simulation environment.

[0146] Ice jump detection simulation:

[0147] Trigger an ice-covered jump event in a simulated environment to generate data on the posture change of the shock absorber. Use the shock absorber target detection model to detect the video frames during the jump and record the position and angle changes of the shock absorber.

[0148] Alarm mechanism simulation:

[0149] When ice cover or ice jump is detected, the alarm mechanism is triggered to simulate the sending process of SMS, email or sound and light alarm. The response time of the alarm is recorded to evaluate the real-time performance of the alarm mechanism.

[0150] Simulation results and analysis:

[0151] 1. Ice detection results:

[0152] In simulation tests, the wire detection model achieved an accuracy rate of 98.5% for ice-covered images, a recall rate of 97.2%, and a false alarm rate of 0.8%.

[0153] The estimated error of ice thickness is within ±2mm, which meets the actual monitoring needs.

[0154] 2. Ice-covered jump test results:

[0155] The average detection accuracy (mAP) of the shock-absorbing hammer target detection model during jumping is 96.7%, and the angle estimation error is within ±1.5°.

[0156] The deviation between the detected ice-covered jump occurrence time and the jump time set in the simulation is less than 0.5 seconds, which has high real-time performance.

[0157] The periodic analysis results of the shock-absorbing hammer's posture change are highly consistent with the jumping mode set in the simulation.

[0158] 3. Alarm mechanism results:

[0159] The delay time for sending alarm information is less than 2 seconds, which meets the needs of real-time monitoring.

[0160] The accuracy of the alarm information (i.e., the alarm is triggered only when real ice-covered jumps are detected) is 99.5%, and the false alarm rate is less than 0.5%.

[0161] 4. System performance optimization:

[0162] Through simulation tests, it was found that data enhancement (such as rotation, scaling, etc.) significantly improved the generalization ability of the model, allowing the system to maintain a high detection accuracy under complex weather conditions.

[0163] Multi-model fusion (such as the combination of FasterRCNN and YOLO) further improves the robustness of detection and reduces false detections and missed detections.

[0164] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An ice jumping monitoring and alarm method, characterized in that: include: Annotate historical ice monitoring photos to obtain a first data set and a second data set, and use a deep learning model to establish a first model and a second model; Using the first data set to train the first model to obtain a wire detection model; The second model is trained using the second data set to obtain an anti-vibration hammer target detection model; Use the wire detection model to detect the photos taken in real time, and if an ice-covered wire target is detected, perform the first operation; The anti-shock hammer target detection model is used to detect the video frames extracted from the real-time video, and then the second operation is performed.

2. The ice jumping monitoring and alarm method according to claim 1, characterized in that: The first operation includes: When the monitoring point enters the ice-covered state, the video pre-recording mode is turned on, and the pre-recorded content is saved for a preset time length; Set the initial pre-recording time length, adjustment length, alarm trigger threshold and safety threshold. When the ice thickness exceeds the alarm trigger threshold, the pre-recording time will be increased by an adjustment length; when the ice thickness is lower than the safety threshold, the pre-recording time will be reduced by an adjustment length.

3. The ice jumping monitoring and alarm method according to claim 2, characterized in that: The use of the anti-vibration hammer target detection model to detect the video frames extracted from the real-time video includes: The extracted video frame content is subjected to target monitoring, the identified anti-vibration hammer position coordinates and angle are recorded, and the data of a preset time length is cyclically stored.

4. The ice jumping monitoring and alarm method according to claim 3, characterized in that: The first data set includes: iced and non-iced photos are selected from historical ice monitoring photos for annotation to obtain a first data set; the second data set includes: anti-vibration hammer photos are selected from historical ice monitoring photos for annotation to obtain a second data set.

5. The ice jump monitoring and alarm method according to claim 4, characterized in that: The second operation includes: Calculate the trajectory of the wire according to the posture of the anti-vibration hammer; Determine whether ice jump occurs on the conductor: When periodic changes are detected in both the position sequence data and the angle sequence data, it is determined that ice jump occurs on the transmission line.

6. The ice jumping monitoring and alarm method according to claim 5, characterized in that: The second operation further includes: When ice jumping is detected, the alarm mechanism is triggered; When ice jumping is detected and the video recording function is not in progress, the pre-recorded content is saved and the video recording function is turned on; When no periodic change is detected in the cyclically stored data, it is determined that the ice jump is completed.

7. The ice jumping monitoring and alarm method according to claim 6, characterized in that: The second operation further includes: After determining that the ice-covered jump is completed, the ongoing video recording is terminated and the video file is uploaded to the upper-level main station; At the same time, the alarm records of ice-covered jumping will be uploaded.

8. A system using the ice jump monitoring and alarm method according to any one of claims 1 to 7, characterized in that: include: A data annotation and model building module is used to annotate historical ice monitoring photos to obtain a first data set and a second data set, and to build a first model and a second model using a deep learning model; A wire detection model training module, used to train the first model using the first data set to obtain a wire detection model; A shock-absorbing hammer target detection model training module, used to train the second model using the second data set to obtain a shock-absorbing hammer target detection model; A wire ice monitoring and response module is used to detect the real-time photographs using a wire detection model, and if an ice-covered wire target is detected, perform a first operation; The ice-covered jump monitoring and video recording control module is used to use the anti-shock hammer target detection model to detect the video frames extracted from the real-time video and perform the second operation.

9. A computing device comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the ice jump monitoring and alarm method described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, can implement the steps of the ice jump monitoring and alarm method according to any one of claims 1 to 7.

Images