A FTU pole-free debugging communication device and method

Abstract

The present invention discloses a communication device and method for FTU (Functional Transmitter Unit) (FTU) debugging without climbing a pole. The device includes an image acquisition module, a directional deflection module, a lifting module, and a horizontal movement module, arranged sequentially from top to bottom. The image acquisition module is provided with a cable clamping mechanism. The image acquisition module, the directional deflection module, the lifting module, and the horizontal movement module are all connected to a processor module. The processor module controls the operation of the directional deflection module, the lifting module, and the horizontal movement module based on image information collected by the image acquisition module. The cable clamped by the cable clamping mechanism is inserted into the corresponding interface of the FTU terminal. The method uses an image recognition model to identify the target interface and obtains the pixel coordinates of the center point of the bounding box of the recognition result. When identifying the target interface, the confidence level is calculated by considering the distance and shape similarity between the predicted bounding box and the actual bounding box to improve the accuracy of the bounding box. The present invention achieves debugging without climbing a pole, reduces safety risks, and improves work efficiency.

Description

Technical Field

[0001] The present invention relates to the fields of power distribution automation technology and robot visual recognition and control, and in particular to an FTU pole-free debugging communication device and method. Background Art

[0002] In the distribution automation system, FTU (Feeder Terminal Unit) is the key equipment to realize data acquisition, control, communication and other functions, such as Figure 1 As shown, the device is usually installed on an outdoor pole, 2-4 meters above the ground, and in order to prevent dust and rain from affecting the operation of the device, the control panel of the device is located on the bottom, as shown in Figure 2 Currently, maintenance or debugging of the FTU requires workers to climb a pole to connect the relevant communication cables to the control panel on the bottom of the FTU terminal. They then use a dedicated debugging terminal (such as a laptop or tablet) and dedicated debugging software to read data and send parameter data. This is not only time-consuming and labor-intensive, but also poses safety risks. Summary of the Invention

[0003] The technical problem to be solved by the present invention is as follows: In view of the above-mentioned problems in the prior art, a communication device and method for FTU pole-free debugging is provided. The debugging personnel can connect the relevant communication cables to the FTU terminal while standing on the ground, thereby realizing pole-free debugging, minimizing safety risks and improving work efficiency.

[0004] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0005] A communication device for debugging an FTU without climbing a pole includes an image acquisition module, a direction deflection module, a lifting module, and a horizontal movement module, which are arranged in sequence from top to bottom. The bottom of the horizontal movement module is supported by a support rod. The image acquisition module is provided with a cable clamping mechanism. The image acquisition module, the direction deflection module, the lifting module, and the horizontal movement module are all connected to a processor module, so that the processor module controls the direction deflection module, the lifting module, and the horizontal movement module to operate according to image information collected by the image acquisition module, and inserts the cable clamped by the cable clamping mechanism into the corresponding interface of the FTU terminal.

[0006] Optionally, the image acquisition module includes an outer shell and a camera arranged in the outer shell, the cable clamping mechanism includes a clamp and a servo arranged in the outer shell, a gap for clamping the cable is formed between the side wall of the clamp and the side wall of the outer shell, the bottom of the clamp is connected to the rotating shaft of the servo, so that the clamp opens and closes as the rotating shaft of the servo rotates to release or clamp the cable, and the output end of the camera and the control end of the servo are respectively connected to the processor module.

[0007] Optionally, the direction deflection module includes a first motor installed on a first base, the lifting module is supported under the first base, the rotating shaft of the first motor is connected to the image acquisition module, and the control end of the first motor is connected to the processor module, so that the first motor rotates in a controlled manner and drives the cable clamped by the cable clamping mechanism on the image acquisition module to adjust the angle.

[0008] Optionally, the lifting module includes a second motor installed on a second base, the second base is connected to the top seat through a cross lifting arm, the top seat is supported below the direction deflection module, and the horizontal movement module is supported below the second base. The second motor drives the cross lifting arm to lift and lower through a driving mechanism, and the control end of the second motor is connected to the processor module, so that the second motor rotates in a controlled manner and drives the cable clamped by the cable clamping mechanism on the image acquisition module to lift and lower.

[0009] Optionally, the horizontal movement module includes a third base, a fourth base and a fifth base arranged in sequence from top to bottom, the third base is supported below the lifting module, the fifth base and the support rod are detachably connected, the fourth base and the fifth base are arranged crosswise, the fourth base is provided with a third motor and a first guide rail arranged along the Y direction, the third motor drives the third base to move along the first guide rail through a first driving mechanism, the fifth base is provided with a fourth motor and a second guide rail arranged along the X direction, the fourth motor drives the fourth base to move along the second guide rail through a second driving mechanism, and the control ends of the third motor and the fourth motor are respectively connected to the processor module, so that the third motor and the fourth motor are controlled to rotate and drive the cable clamped by the cable clamping mechanism on the image acquisition module to move along the horizontal X direction or Y direction.

[0010] Optionally, a support cover is further provided at the end of the support rod, the main body is arranged in the support cover, and an electromagnet is provided at the top of the support cover, and the control end of the power switch of the electromagnet is connected to the processor module, so that the electromagnet is controlled to be magnetically connected to the FTU terminal.

[0011] The present invention also provides an FTU pole-free debugging communication method, which is applied to any of the FTU pole-free debugging communication devices described above, and includes the following steps:

[0012] Obtain the image information collected by the image acquisition module, use the image recognition model to identify the target interface and obtain the pixel coordinates of the center point of the bounding box of the recognition result ( u,v ), when identifying the target interface, the distance and shape similarity between the predicted bounding box and the true bounding box are considered to calculate the confidence to improve the accuracy of the bounding box;

[0013] The pixel coordinates of the center point ( u,v ) is converted to world coordinates (X b , Y b , Z b ), and converting the image coordinate system into the camera coordinate system, calculates the angular difference θ between the cable clamped by the cable clamping mechanism and the target interface;

[0014] According to the world coordinate system coordinates ( X b , Y b , Z b ) and the target interface reference center coordinates ( X n , Y n , Z n ) Calculate the distance difference between the cable clamped by the cable clamping mechanism (102) and the target interface ;

[0015] Control the direction deflection module to eliminate the angle difference θ, and control the horizontal movement module to eliminate the distance difference in the horizontal X direction. Distance difference in Y direction , control the lifting module movement to eliminate the vertical Z direction distance difference , so that the cable clamped by the cable clamping mechanism is inserted into the target interface;

[0016] After sending a request to the FTU terminal and receiving a response, debugging begins.

[0017] Furthermore, the loss function during the training of the image recognition model includes coordinate error, confidence error and classification error, where:

[0018] The coordinate error expression is as follows:

[0019]

[0020] in, is the weight used to balance the ratio of coordinate error to other errors, S is the grid size, B is the number of bounding boxes, Indicates the i The first j Whether the bounding box contains the target, if it does, the value is 1, otherwise it is 0, ( u i ，v i ) is the pixel coordinate of the center point of the real bounding box, w i and hi are the length and width of the ground-truth bounding box, ( ， ) is the pixel coordinate of the center point of the predicted bounding box, and To predict the length and width of the bounding box;

[0021] The confidence error expression is as follows:

[0022]

[0023] in, is the weight used to balance the ratio of confidence error to other errors, is the confidence of the true bounding box, is the confidence of the predicted bounding box;

[0024] The classification error expression is as follows:

[0025]

[0026] in, Indicates the i Whether the grid contains the target, if it does, the value is 1, otherwise it is 0. is the probability distribution of the true category, is the probability distribution of the predicted category.

[0027] Furthermore, the target interface is a network interface. When the confidence is calculated by considering the distance and shape similarity between the predicted bounding box and the true bounding box, the expression is as follows:

[0028]

[0029] in, ρ To predict the center point of the bounding box and the center point of the ground truth bounding box b The Euclidean distance between d is the diagonal distance of the minimum bounding rectangle containing the two bounding boxes, µ Reflects the aspect ratio of the predicted bounding box Aspect ratio with the true bounding box The similarity, Indicates the probability that there is a network port in the bounding box, It represents the ratio of the intersection of the predicted bounding box and the true bounding box to their union, that is, the intersection-union ratio.

[0030] Furthermore, after converting the image coordinate system into the camera coordinate system, the angular difference θ between the cable clamped by the cable clamping mechanism (102) and the target interface is calculated, including the following steps:

[0031] Use the camera intrinsic parameter matrix to convert the image coordinates into normalized coordinates in the camera coordinate system;

[0032] Establish a cuboid model of the target interface, one face of the cuboid model is parallel to the camera imaging plane, and according to the coordinates of the diagonal endpoints of the cuboid model in the image coordinate system ( )and( ), get the corresponding normalized coordinates in the camera coordinate system ( )and( );

[0033] The angular difference θ is calculated based on the normalized coordinates of the diagonal endpoints in the camera coordinate system. The expression is as follows:

[0034]

[0035] in, and They are the normalized coordinate values ​​of the x-axis and y-axis in the camera coordinate system respectively.

[0036] Compared with the prior art, the advantages of the present invention are:

[0037] The device body of the present invention includes an image acquisition module, a direction deflection module, a lifting module, and a horizontal movement module, which are arranged in sequence from top to bottom. The image acquisition module is provided with a cable clamping mechanism, which can control the direction deflection module, the lifting module, and the horizontal movement module to work together based on the image information collected by the image acquisition module, and insert the cable clamped by the cable clamping mechanism into the control panel at the bottom of the FTU terminal, avoiding the operator's climbing operation.

[0038] In the method of the present invention, when performing target recognition and positioning of the interface based on the image information collected by the image acquisition module, the distance and shape similarity between the predicted bounding box and the real bounding box are considered to calculate the confidence level, thereby improving the accuracy of the bounding box. Then, according to the difference between the positioned coordinates and the preset coordinates, the direction deflection module, the lifting module, and the horizontal movement module are controlled to move to the preset coordinates to ensure that the cable can be smoothly inserted into the corresponding interface. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 This is a schematic diagram of the installation of FTU on an outdoor pole.

[0040] Figure 2 Schematic diagram of the FTU control panel.

[0041] Figure 3 This is an exploded view of a device according to an embodiment of the present invention.

[0042] Figure 4 FIG. 2 is an electrical connection block diagram of an embodiment of the present invention.

[0043] Figure 5 This is an exploded view of the image acquisition module according to an embodiment of the present invention.

[0044] Figure 6 This is an exploded view of the direction deflection module according to an embodiment of the present invention.

[0045] Figure 7 This is an exploded view of the lifting module according to an embodiment of the present invention.

[0046] Figure 8 This is an exploded view of the horizontal movement module according to an embodiment of the present invention.

[0047] Figure 9 Flowchart of a method according to an embodiment of the present invention.

[0048] Legend: 1-image acquisition module, 2-directional deflection module, 3-lifting module, 4-horizontal movement module, 11-outer shell, 12-camera, 13-gripper, 14-servo, 21-first motor, 22-first base, 31-second motor, 32-second base, 33-top seat, 34-cross lifting arm, 34a-bolt, 34b-nut, 35-driving mechanism, 35a-screw, 35b-slider, 36-first pin, 37- Second pin shaft, 41-third base, 41a-first through hole, 41b-first slider, 42-fourth base, 42a-second through hole, 42b-second slider, 43-fifth base, 44-third motor, 45-first guide rail, 46-first driving mechanism, 47-fourth motor, 48-second guide rail, 49-second driving mechanism, 101-support rod, 102-cable clamping mechanism, 103-support cover, 104-electromagnet, 105-pin. DETAILED DESCRIPTION

[0049] The present invention will be further described below in conjunction with the accompanying drawings and specific preferred embodiments, but the scope of protection of the present invention is not limited thereby.

[0050] Example 1

[0051] In order to improve work efficiency and reduce operation risks, this embodiment proposes a FTU free-to-climb-pole debugging communication device, such as Figure 3As shown, it includes an image acquisition module 1, a direction deflection module 2, a lifting module 3 and a horizontal movement module 4 arranged at the end of a support rod 101 and arranged in sequence from top to bottom. The support rod 101 is supported at the bottom of the horizontal movement module 4. A cable clamping mechanism 102 is provided on the image acquisition module 1. A support cover 103 is provided on the outside of the body 1. The top of the support cover 103 is about 2 to 3 cm higher than the image acquisition module 11 to prevent the image acquisition module 1 from touching the bottom surface of the FTU before work. An electromagnet 104 is provided on the top of the support cover 103. The image acquisition module 1, the direction deflection module 2, the lifting module 3 and the horizontal movement module 4 are all connected to the processor module, and the control end of the power switch of the electromagnet 104 is connected to the processor module.

[0052] Through the above structure, the operator only needs to lift the device of this embodiment to the vicinity of the FTU terminal. The processor module controls the actions of the directional deflection module 2, the lifting module 3 and the horizontal movement module 4 according to the image information collected by the image acquisition module 1, and inserts the cable clamped by the cable clamping mechanism 102 into the corresponding interface of the FTU terminal, avoiding the need for personnel to perform high-altitude operations. At the same time, the processor module can control the electromagnet 104 to be magnetically connected to the FTU terminal, so that the device can be automatically fixed to the bottom of the FTU terminal during debugging, without the operator having to always hold the device, saving physical strength.

[0053] In this embodiment, a remote control method is adopted for the device, such as Figure 4 As shown, the processor module includes a data processing module, a communication module and an information terminal. The data processing module exchanges data with the information terminal through the communication module, and the content of the interaction includes data uploading and instruction receiving. The image acquisition module 1, the direction deflection module 2, the lifting module 3 and the horizontal movement module 4 are all connected to the data processing module of the processor module. The data processing module obtains the image data collected by the image acquisition module 1 and uploads it to the information terminal through the communication module. The information terminal can issue corresponding adjustment instructions according to the position of the target interface in the image, such as angle adjustment instructions, height adjustment instructions and horizontal adjustment instructions. After the data processing module obtains the adjustment instructions through the communication module, it converts them into different control signals and sends them to the corresponding functional modules in the direction deflection module 2, the lifting module 3 and the horizontal movement module 4 to achieve adjustment of the cable position clamped by the cable clamping mechanism 102.

[0054] like Figure 4 As shown, the processor module of this embodiment is also provided with a power supply module, which can provide power to the data processing module, the image acquisition module 1, the direction deflection module 2, the lifting module 3 and the horizontal movement module 4 to ensure that the device can work normally.

[0055] like Figure 5As shown, the image acquisition module 1 of this embodiment includes an outer shell 11 and a camera 12 arranged in the outer shell 11. The camera 12 specifically adopts an Omnivision OV2710 camera. The cable clamping mechanism 102 includes a clamping claw 13 and a servo 14 arranged in the outer shell 11.

[0056] In this embodiment, a hole is opened at the top of the outer shell 11 to facilitate the camera 12 to collect images from the bottom of the FTU terminal. The output end of the camera 12 is connected to the data processing module of the processor module, so that the data processing module can upload the image data collected by the camera 12 to the information terminal through the communication module in real time.

[0057] The top of the outer shell 11 of this embodiment is stepped, and the clamping jaw 13 is installed on one side of the step, and a gap for clamping the cable is formed between the side wall of the clamping jaw 13 and the side wall of the step of the outer shell 11. The bottom of the clamping jaw 13 is provided with a cable through hole, and the cable through hole is connected to the gap, so that the cable passes through the cable through hole and reaches the gap and is clamped and fixed. The bottom of the clamping jaw 13 is connected to the rotating shaft of the servo 14, so that the clamping jaw 13 opens and closes as the rotating shaft of the servo 14 rotates to release or clamp the cable. At the same time, the control end of the servo 14 is connected to the data processing module of the processor module, so that the data processing module can generate the corresponding motor rotation direction according to the opening and closing instructions of the information terminal and send it to the servo 14 to realize the control of the opening and closing of the clamping jaw 13. For example, the servo 14 is controlled to rotate in a clockwise direction so that the gap between the side wall of the clamping jaw 13 and the side wall of the step of the outer shell 11 is opened, thereby facilitating the passage of the cable through the cable through hole. The servo 14 is controlled to rotate in a clockwise direction so that the gap between the side wall of the clamping jaw 13 and the side wall of the step of the outer shell 11 is closed, thereby fixing the cable between the side wall of the clamping jaw 13 and the side wall of the step of the outer shell 11.

[0058] It should be noted that when the cable is clamped between the side wall of the clamping jaw 13 and the side wall of the step of the outer shell 11, the terminal of the cable (such as the crystal plug of the network cable, etc.) needs to be higher than the top of the image acquisition module 1 so that it can be smoothly inserted into the interface at the bottom of the FTU terminal. On this basis, in this embodiment, the end of the clamping jaw 13 and the top of the step of the outer shell 11 are also provided with a cavity for accommodating the root of the cable terminal, thereby ensuring that the cable terminal and the cable can be firmly clamped between the side wall of the clamping jaw 13 and the side wall of the step of the outer shell 11.

[0059] like Figure 6As shown, the direction deflection module 2 of this embodiment includes a first base 22 and a first motor 21 installed on the first base 22. The first motor 21 adopts an MG966R servo motor. The lifting module 3 is supported below the first base 22. The rotating shaft of the first motor 21 is connected to the bottom of the outer shell 11 of the image acquisition module 1. The control end of the first motor 21 is connected to the data processing module of the processor module. The data processing module can generate the corresponding motor rotation direction and number of rotations according to the angle value in the angle adjustment instruction of the information terminal, and send it to the first motor 21, so that the first motor 21 rotates in a controlled manner and drives the cable clamped by the cable clamping mechanism 102 on the outer shell 11 of the image acquisition module 1 to adjust the angle.

[0060] like Figure 7 As shown, the lifting module 3 of this embodiment includes a second motor 31 installed on a second base 32. The second motor 31 adopts a reduction motor. The second base 32 is connected to the top seat 33 through a cross lifting arm 34. The top seat 33 is supported below the direction deflection module 2. The horizontal movement module 4 is supported below the second base 32. The second motor 31 drives the cross lifting arm 34 to rise and fall through the driving mechanism 35. The control end of the second motor 31 is connected to the data processing module of the processor module. The data processing module can generate the corresponding motor rotation direction and number of rotations according to the height value in the height adjustment instruction of the information terminal and send it to the second motor 31, so that when the second motor 31 is controlled to rotate, it drives the cross lifting arm 34 to rise and fall, thereby driving the cable clamped by the cable clamping mechanism 102 on the image acquisition module 1 to rise and fall.

[0061] like Figure 7 As shown, in this embodiment, a pair of cross lifting arms 34 are provided on both sides of the lifting module 3. The cross lifting arms 34 include two cross lifting arms, and the intersection of the two lifting arms is hinged by a bolt 34a. The side wall of the second base 32 and the side wall of the top base 33 are both provided with a fixed first pin 36 at one end and a second pin 37 that moves in a slide groove at the other end. The lower end of one lifting arm in the cross lifting arms 34 is connected to the first pin 36 of the second base 32 by a nut 34b, and the upper end is connected to the top base 3 by a nut 34b. 3, while the lower end of the other lifting arm in the cross lifting arm 34 is connected to the second pin 37 of the second base 32 through a nut 34b, and the upper end is connected to the first pin 36 of the top seat 33 through a nut 34b. The driving mechanism 35 includes a screw rod 35a, and the second pin 37 of the second base 32 is provided with a slider 35b. The rotating shaft of the second motor 31 is connected to the screw rod 35a, the screw rod 35a passes through the slider 35b and cooperates with the thread of the slider 35b, and the slider 35b contacts the upper surface of the second base 32.

[0062] Through the above structure, when the second motor 31 rotates in one direction (such as clockwise), the slider 35b moves along the screw rod 35a toward the second motor 31, thereby driving the cross-lifting arm 34 to drop in height, thereby lowering the top seat 33. When the second motor 31 rotates in the other direction (such as counterclockwise), the slider 35b moves along the screw rod 35a away from the second motor 31, thereby driving the cross-lifting arm 34 to rise in height, thereby raising the top seat 33.

[0063] like Figure 8 As shown, the horizontal moving module 4 of this embodiment includes a third base 41, a fourth base 42 and a fifth base 43 arranged in sequence from top to bottom, the third base 41 is supported below the lifting module 3, the fifth base 43 and the support rod 101 are detachably connected, the fourth base 42 and the fifth base 43 are arranged crosswise, the fourth base 42 is provided with a third motor 44 and a first guide rail 45 arranged along the Y direction, the third motor 44 drives the third base 41 to move along the first guide rail 45 through the first driving mechanism 46, the fifth base 43 is provided with a fourth motor 47 and a second guide rail 48 arranged along the X direction, the fourth motor 48 is provided on the 47 drives the fourth base 42 to move along the second guide rail 48 through the second driving mechanism 49. The control ends of the third motor 44 and the fourth motor 47 are respectively connected to the data processing module of the processor module. The data processing module can generate the corresponding motor rotation direction and number of rotations according to the X-direction distance value and / or Y-direction distance value in the horizontal adjustment instruction of the information terminal and send them to the third motor 44 and / or the fourth motor 47, so that the third motor 44 and / or the fourth motor 47 rotates in a controlled manner and drives the cable clamped by the cable clamping mechanism 102 on the image acquisition module 1 to move along the horizontal X direction or Y direction.

[0064] Specifically, in this embodiment, the first drive mechanism 46 and the second drive mechanism 49 both use screw rods, the rotating shaft of the third motor 44 is connected to the screw rod of the first drive mechanism 46, the bottom of the third base 41 is provided with a first through hole 41a for the first guide rail 45 to pass through, and a first slider 41b is also provided to cooperate with the screw rod of the first drive mechanism 46. Similarly, the rotating shaft of the fourth motor 47 is connected to the screw rod of the second drive mechanism 49, the bottom of the fourth base 42 is provided with a second through hole 42a for the second guide rail 48 to pass through, and a second slider 42b is also provided to cooperate with the screw rod of the second drive mechanism 49, so that when the third motor 44 and / or the fourth motor 47 move to a certain direction, the third motor 44 and / or the fourth motor 47 can rotate. When the third motor 44 and / or the fourth motor 47 rotates in one direction (such as clockwise), the first slider 41b and / or the second slider 42b move along the corresponding screw rod toward the third motor 44 and / or the fourth motor 47, so that the third base 41 and / or the fourth base 42 and all the components above them move along the positive direction of the X-axis and / or the Y-axis of the horizontal plane. When the third motor 44 and / or the fourth motor 47 rotates in the other direction (such as counterclockwise), the first slider 41b and / or the second slider 42b move along the corresponding screw rod away from the third motor 44 and / or the fourth motor 47, so that the third base 41 and / or the fourth base 42 and all the components above them move along the opposite direction of the X-axis and / or the Y-axis of the horizontal plane.

[0065] like Figure 1 As shown, in this embodiment, in order to facilitate maintenance of the device, the device body composed of the image acquisition module 1, the direction deflection module 2, the lifting module 3, and the horizontal movement module 4 and the support rod 101 and the support cover 103 adopt a detachable connection method. Specifically, the bottom end of the support cover 103 is provided with a first sleeve, and the bottom end of the fifth base 143 of the horizontal movement module 4 is provided with a second sleeve. The end of the support rod 101 is inserted into the second sleeve, and the second sleeve is inserted into the first sleeve, and the end of the support rod 101 and the first sleeve and the second sleeve are all provided with corresponding pin holes, so that after the support rod 101, the body 1 and the support cover 103 are installed, the pin 105 can be inserted into the pin holes of the end of the support rod 101, the first sleeve and the second sleeve in sequence to achieve stable installation of the device.

[0066] In this embodiment, in order to adapt to FTU terminals of different heights, the support rod 101 adopts a telescopic rod, including an inner rod and an outer rod. The device body and the support cover 103 are installed at the end of the inner rod, and the outer rod is provided with a buckle to lock the inner rod.

[0067] Example 2

[0068] This embodiment proposes a FTU free pole debugging communication method, which is applied to the FTU free pole debugging communication device of embodiment 1. Figure 9 As shown, the following steps are included:

[0069] S101) Preparation for commissioning:

[0070] The operator stands on the ground and uses the lifting support rod 101 to touch the upper end surface of the support cover 103 to the lower end surface of the FTU. Then, the operator obtains the image information collected by the image acquisition module 1 through the information terminal of the processor module, and moves the entire device back and forth and left and right for rough adjustment, so that the communication network port on the FTU panel is basically in the center of the monitoring field of view. Finally, the operator controls the electromagnet 104 on the support cover 103 through the information terminal to attract it to the base of the FTU, then pulls the outer rod of the support rod 101 down to contact the ground and locks the buckle, completing the rough fixation of the entire device.

[0071] S102) Target Identification:

[0072] The information terminal obtains the image information collected by the image acquisition module 1, uses the image recognition model to identify the target interface and obtains the pixel coordinates of the center point of the boundary box of the recognition result ( u,v ), when identifying the target interface, the distance and shape similarity between the predicted bounding box and the true bounding box are considered to calculate the confidence to improve the accuracy of the bounding box;

[0073] S103) Deviation calculation:

[0074] The information terminal will take the pixel coordinates of the center point ( u,v ) is converted to world coordinates ( X b , Y b , Z b ), and converting the image coordinate system into the camera coordinate system, and then calculating the angular difference θ between the cable clamped by the cable clamping mechanism 102 and the target interface;

[0075] The information terminal is based on the world coordinate system coordinates ( X b , Y b , Z b ) and the target interface reference center coordinates ( X n , Y n , Z n ) Calculate the distance difference between the cable clamped by the cable clamping mechanism 102 and the target interface ;

[0076] S104) Position adjustment:

[0077] The information terminal controls the direction deflection module 2 to eliminate the angle difference θ and controls the horizontal movement module 4 to eliminate the distance difference in the horizontal X direction. Distance difference in Y direction , control the movement of lifting module 3 to eliminate the distance difference in the vertical Z direction , so that the cable clamped by the cable clamping mechanism 102 is inserted into the target interface;

[0078] S105) Equipment disengagement:

[0079] After the information terminal sends a request to the FTU terminal and receives a response, the debugging work begins;

[0080] After debugging is completed, the information terminal generates a prompt message and waits for the operator to confirm. The electromagnet 104 on the control support cover 103 is powered off and separated from the base of the FTU. The operator unlocks the outer rod buckle of the support rod 101 and retracts the inner rod, so that the cable on the device is disconnected from the corresponding interface of the FTU, and the work is completed.

[0081] The relevant steps are described below.

[0082] In step S102 of this embodiment, after the information terminal obtains the image information captured by the image acquisition module 1, it performs standardization processing, including size adjustment, grayscale, and denoising steps. The specific implementation process of these standardization processes is well known to those skilled in the art and will not be repeated in this embodiment.

[0083] In step S102 of this embodiment, the image recognition model is used to identify the target interface and obtain the pixel coordinates of the center point of the boundary box of the recognition result ( u,v ), the YOLO target detection algorithm is used to locate and identify the RJ45 network port in the image. By improving the confidence of the bounding box to optimize the loss function, the network port target can be accurately located.

[0084] Before step S102, this embodiment first collects a large number of FTU images and annotates the network port bounding boxes to serve as a training dataset for the YOLO algorithm. The input FTU images are then segmented into fixed grids, with multiple bounding boxes predicted for each grid. A convolutional neural network is used to extract image features. Based on the extracted features, the probability and location of each bounding box containing a network port are predicted. Within each grid, the bounding box with the highest confidence score is selected as the final prediction. Bounding boxes with an overlap with other bounding boxes exceeding a set threshold are discarded to ensure that network port targets are not detected repeatedly.

[0085] Typically, the confidence of each bounding box , used to reflect whether the grid contains the network port target and the accuracy of the bounding box. If there is no network port in this grid, the probability PIf (network port) = 0, the confidence is 0; otherwise, the confidence is the intersection-over-union (IOU) ratio between the predicted bounding box and the true bounding box (i.e., the ratio of the intersection of the predicted bounding box and the true bounding box to their union). However, IOU only considers the overlapping area between the predicted bounding box and the true bounding box. Based on this, this embodiment also considers the distance and shape similarity between the two to calculate the confidence, thereby improving the accuracy of the bounding box. The expression is as follows:

[0086] (1)

[0087] in, ρ To predict the center point of the bounding box and the center point of the ground truth bounding box b The Euclidean distance between d is the diagonal distance of the minimum bounding rectangle containing the two bounding boxes, µ Reflects the aspect ratio of the predicted bounding box Aspect ratio with the true bounding box The similarity, Indicates the probability that there is a network port in the bounding box, It represents the ratio of the intersection of the predicted bounding box and the true bounding box to their union, that is, the intersection-union ratio.

[0088] In this embodiment, the loss function during image recognition model training includes coordinate error, confidence error, and classification error, where:

[0089] The coordinate error is divided into the center point error of the bounding box and the length and width error of the bounding box. The expression is as follows:

[0090] (2)

[0091] in, is the weight used to balance the ratio of coordinate error to other errors, S is the grid size, B is the number of bounding boxes, Indicates the i The first j Whether the bounding box contains the target, if it does, the value is 1, otherwise it is 0, ( u i ，v i ) is the pixel coordinate of the center point of the real bounding box, w i and h i are the length and width of the ground-truth bounding box, ( ， ) is the pixel coordinate of the center point of the predicted bounding box, and To predict the length and width of the bounding box;

[0092] The confidence error is divided into the confidence error of the bounding box that includes the network port and the confidence error of the bounding box that excludes the network port. Both are confidence levels that comprehensively consider the overlapping area, distance, and shape similarity. This can make subsequent coordinate distance calculations more accurate. The expression is as follows:

[0093] (3)

[0094] in, is the weight used to balance the ratio of confidence error to other errors, is the confidence of the true bounding box, is the confidence of the predicted bounding box;

[0095] The classification error is the probability loss of being judged as a network port target, and the expression is as follows:

[0096] (4)

[0097] in, Indicates the i Whether the grid contains the target, if it contains the value is 1, otherwise it is 0. is the probability distribution of the true category, is the probability distribution of the predicted category.

[0098] In step S102 of this embodiment, the trained YOLO target detection algorithm can obtain the visualization or data format of the recognition and classification results, so as to obtain the pixel coordinates of the center point of the bounding box of the target interface ( u,v ), correspondingly, in step S103, the center pixel coordinates ( u,v ) is converted to world coordinates ( X b , Y b , Z b ) is calculated as:

[0099] (5)

[0100] in, Z c It is the distance between the intersection of the optical center along the optical axis and the imaging plane. In this embodiment, it is a fixed distance of 2-3 cm between the top of the image acquisition module 1 and the bottom of the FTU in the initial state, so it can be ignored when taking images.

[0101] In step S103 of this embodiment, after converting the image coordinate system into the camera coordinate system, when calculating the angular difference θ between the cable clamped by the cable clamping mechanism 102 and the target interface, since the network cable jack for FTU debugging is a rectangular parallelepiped with a length and width of 12.5 mm*9 mm, and the two diagonal endpoints of the rectangular parallelepiped in the image can be detected by the monocular camera, and due to the positioning function of the support cover 103, it can be ensured that the bottom plane of the FTU and one surface of the rectangular parallelepiped of the network cable jack are parallel to the imaging plane, the angular difference θ can be calculated by the following steps and formula:

[0102] S201) The image coordinates are converted into normalized coordinates in the camera coordinate system using the camera intrinsic parameter matrix. In this embodiment, the camera position is used as the origin of the world coordinate system, and the camera calibration is performed using the Zhang Zhengyou calibration method to obtain the intrinsic parameter matrix and the extrinsic parameter matrix. The camera intrinsic parameter matrix is ​​generally expressed as:

[0103] K= (6)

[0104] in, and is the focal length along the width and height of the image, and is the principal point coordinate of the image, and thus, the expression for converting the image coordinate into the normalized coordinate in the camera coordinate system is as follows:

[0105] x= , y= ,z= (7)

[0106] Where (u, v) represents the image coordinates, (x, y, z) represents the normalized coordinates in the camera coordinate system. In this embodiment, it is assumed that = =f, that is, the focal length is equal in both directions;

[0107] S202) Establish a rectangular model of the target interface. Assuming that the width of the network jack is W and the height is H, and one face of the network jack is parallel to the camera imaging plane, the position of the network jack in the camera coordinate system can be determined by the detected diagonal endpoints. Specifically, assuming that the center of the network jack is located at the origin of the camera coordinate system (0,0,0), and one side of the network jack is parallel to the x-axis of the camera, the rotation of the network jack can be calculated by the detected diagonal endpoints. Let the coordinates of the diagonal endpoints in the image coordinate system be ( )and( ), according to step S201, the coordinates of the diagonal endpoints of the rectangular parallelepiped model in the image coordinate system ( )and( ), get the corresponding normalized coordinates in the camera coordinate system ( )and( );

[0108] S203) Calculate the angle difference θ based on the normalized coordinates of the diagonal endpoints in the camera coordinate system. Since one face of the network cable jack-like cuboid is parallel to the imaging plane, it can be assumed that = =0, then the angle difference θ is calculated by the following formula:

[0109] (8)

[0110] (9)

[0111] in, and are the normalized coordinate values ​​of the x-axis and y-axis of the diagonal endpoints in the camera coordinate system, respectively. )and( ) are the coordinate values ​​of the diagonal endpoints in the image coordinate system.

[0112] In step S103 of this embodiment, the coordinates of the center point of the known network port line are set ( X n , Y n , Z n ) is the coordinate of the reference center point, then the world coordinate system coordinates ( X b , Y b , Z b ) and the target interface reference center coordinates ( X n , Y n , Z n ) Calculate the distance difference between the cable clamped by the cable clamping mechanism 102 and the FTU network port In step S104, the information terminal can The required distances in the horizontal X-axis, Y-axis, and vertical Z-axis directions are calculated, and corresponding control instructions are sent to the data processing module to control the movement of the horizontal movement module 4 and / or the movement of the lifting module 3. Furthermore, based on the calculated angular difference θ, the information terminal calculates the required rotation angle and sends corresponding control instructions to the data processing module to control the movement of the directional deflection module 2. Specifically, the corresponding motors in the directional deflection module 2, the lifting module 3, and the horizontal movement module 4 are controlled to coordinate their movements, thereby automatically adjusting the RJ45 network port on the FTU base for debugging.

[0113] In summary, the present invention proposes a FTU pole-free debugging communication device and a FTU pole-free debugging communication method for the device, which enables workers to quickly and safely maintain and debug the FTU on the ground, improves work efficiency, reduces safety risks, and has the following beneficial effects:

[0114] 1. The number of operators has been significantly reduced, from 4 to 2.

[0115] 2. The operation time is greatly shortened. After testing, the operation time is shortened from 30 minutes to 4-5 minutes;

[0116] 3. Convert pole climbing operations into ground operations to effectively reduce safety risks.

[0117] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiment. All technical solutions based on the concept of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications that do not depart from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A FTU pole-free debugging communication method, applied to an FTU pole-free debugging communication device, characterized in that: The FTU free-to-climb pole debugging communication device comprises an image acquisition module (1), a direction deflection module (2), a lifting module (3) and a horizontal movement module (4) which are arranged in sequence from top to bottom, wherein the bottom of the horizontal movement module (4) is supported by a support rod (101), and a cable clamping mechanism (102) is provided on the image acquisition module (1). The image acquisition module (1), the direction deflection module (2), the lifting module (3) and the horizontal movement module (4) are all connected to a processor module, so that the processor module controls the direction deflection module (2), the lifting module (3) and the horizontal movement module (4) to act according to image information collected by the image acquisition module (1), and inserts the cable clamped by the cable clamping mechanism (102) into the corresponding interface of the FTU terminal. The method comprises the following steps: Obtain the image information collected by the image acquisition module (1), use the image recognition model to identify the target interface and obtain the pixel coordinates of the center point of the boundary box of the recognition result ( u,v ), specifically, the YOLO target detection algorithm is used to locate and identify the RJ45 network port in the image. By improving the confidence of the bounding box to optimize the loss function, the network port target can be accurately located. When identifying the target interface, the distance and shape similarity between the predicted bounding box and the true bounding box are considered to calculate the confidence to improve the accuracy of the bounding box. The expression is as follows: in, ρ To predict the center point of the bounding box and the center point of the ground truth bounding box b The Euclidean distance between d is the diagonal distance of the minimum bounding rectangle containing the two bounding boxes, µ Reflects the aspect ratio of the predicted bounding box Aspect ratio with the true bounding box The similarity, Indicates the probability that there is a network port in the bounding box, represents the intersection-over-union ratio of the predicted bounding box to the true bounding box; The pixel coordinates of the center point ( u,v ) is converted to world coordinates ( X b , Y b , Z b ), and converting the image coordinate system into the camera coordinate system, and then calculating the angular difference θ between the cable clamped by the cable clamping mechanism (102) and the target interface; According to the world coordinate system coordinates ( X b , Y b , Z b ) and the target interface reference center coordinates ( X n , Y n , Z n ) Calculate the distance difference between the cable clamped by the cable clamping mechanism (102) and the target interface ; Control the direction deflection module (2) to eliminate the angle difference θ, and control the horizontal movement module (4) to eliminate the distance difference in the horizontal X direction. Distance difference in Y direction , control the lifting module (3) to eliminate the distance difference in the vertical Z direction , so that the cable clamped by the cable clamping mechanism (102) is inserted into the target interface; After sending a request to the FTU terminal and receiving a response, debugging begins.

2. The FTU pole-free debugging communication method according to claim 1 is characterized in that: Before obtaining the image information collected by the image acquisition module (1), the method further includes the step of training the image recognition model, specifically collecting a large number of FTU images, marking the network port bounding boxes, and using them as a training data set for the YOLO algorithm. The FTU images are then divided into fixed grids, and multiple bounding boxes are predicted for each grid. Image features are extracted using a convolutional neural network, and the probability and location information of each bounding box containing the network port are predicted based on the extracted features. In each grid, the bounding box with the highest confidence is selected as the final prediction result, and the bounding boxes with an overlap with other bounding boxes higher than the set threshold are eliminated to ensure that the network port target is not detected repeatedly. The loss function during the image recognition model training includes coordinate error, confidence error, and classification error, where: The coordinate error expression is as follows: in, is the weight used to balance the ratio of coordinate error to other errors, S is the grid size, B is the number of bounding boxes, Indicates the i The first j Whether the bounding box contains the target, if it does, the value is 1, otherwise it is 0, ( u i ，v i ) is the pixel coordinate of the center point of the real bounding box, w i and h i are the length and width of the ground-truth bounding box, ( ， ) is the pixel coordinate of the center point of the predicted bounding box, and To predict the length and width of the bounding box; The confidence error expression is as follows: in, is the weight used to balance the ratio of confidence error to other errors, is the confidence of the true bounding box, is the confidence of the predicted bounding box; The classification error expression is as follows: in, Indicates the i Whether the grid contains the target, if it contains the value is 1, otherwise it is 0. is the probability distribution of the true category, is the probability distribution of the predicted category.

3. The FTU pole-free debugging communication method according to claim 1 is characterized in that: After converting the image coordinate system into the camera coordinate system, the calculation of the angular difference θ between the cable clamped by the cable clamping mechanism (102) and the target interface includes the following steps: Use the camera intrinsic parameter matrix to convert the image coordinates into normalized coordinates in the camera coordinate system; Establish a cuboid model of the target interface, one face of the cuboid model is parallel to the camera imaging plane, and according to the coordinates of the diagonal endpoints of the cuboid model in the image coordinate system ( )and( ), get the corresponding normalized coordinates in the camera coordinate system ( )and( ); The angular difference θ is calculated based on the normalized coordinates of the diagonal endpoints in the camera coordinate system. The expression is as follows: in, and They are the normalized coordinate values ​​of the x-axis and y-axis in the camera coordinate system respectively.

4. A FTU pole-free debugging communication device, used to execute the FTU pole-free debugging communication method according to any one of claims 1 to 3, characterized in that: The image acquisition module (1) comprises an outer shell (11) and a camera (12) disposed in the outer shell (11); the cable clamping mechanism (102) comprises a clamping claw (13) and a servo (14) disposed in the outer shell (11); a gap for clamping the cable is formed between the side wall of the clamping claw (13) and the side wall of the outer shell (11); the bottom of the clamping claw (13) is connected to the rotating shaft of the servo (14), so that the clamping claw (13) opens and closes as the rotating shaft of the servo (14) rotates to release or clamp the cable; the output end of the camera (12) and the control end of the servo (14) are respectively connected to the processor module.

5. The FTU pole-free debugging communication device according to claim 4 is characterized in that: The direction deflection module (2) comprises a first motor (21) mounted on a first base (22); the lifting module (3) is supported below the first base (22); a rotating shaft of the first motor (21) is connected to the image acquisition module (1); and a control end of the first motor (21) is connected to a processor module, so that the first motor (21) rotates in a controlled manner and drives the cable clamped by the cable clamping mechanism (102) on the image acquisition module (1) to adjust the angle.

6. The FTU pole-free debugging communication device according to claim 4 is characterized in that: The lifting module (3) includes a second motor (31) mounted on a second base (32). The second base (32) is connected to a top base (33) via a cross lifting arm (34). The top base (33) is supported below the direction deflection module (2). The horizontal movement module (4) is supported below the second base (32). The second motor (31) drives the cross lifting arm (34) to move up and down via a driving mechanism (35). The control end of the second motor (31) is connected to a processor module, so that the second motor (31) rotates in a controlled manner and drives the cable clamped by the cable clamping mechanism (102) on the image acquisition module (1) to move up and down.

7. The FTU pole-free debugging communication device according to claim 4 is characterized in that: The horizontal moving module (4) includes a third base (41), a fourth base (42) and a fifth base (43) arranged in sequence from top to bottom, the third base (41) is supported below the lifting module (3), the fifth base (43) and the support rod (101) are detachably connected, the fourth base (42) and the fifth base (43) are arranged crosswise, the fourth base (42) is provided with a third motor (44) and a first guide rail (45) arranged along the Y direction, the third motor (44) drives the third base (41) along the first guide rail (45) through the first driving mechanism (46). The guide rail (45) moves, the fifth base (43) is provided with a fourth motor (47) and a second guide rail (48) arranged along the X direction, the fourth motor (47) drives the fourth base (42) to move along the second guide rail (48) through the second driving mechanism (49), and the control ends of the third motor (44) and the fourth motor (47) are respectively connected to the processor module, so that the third motor (44) and the fourth motor (47) are controlled to rotate and drive the cable clamped by the cable clamping mechanism (102) on the image acquisition module (1) to move along the horizontal X direction or Y direction.

8. The FTU pole-free debugging communication device according to claim 4 is characterized in that: The end of the support rod (101) is further provided with a support cover (103), the image acquisition module (1), the direction deflection module (2), the lifting module (3) and the horizontal movement module (4) are all arranged in the support cover (103), and the top of the support cover (103) is provided with an electromagnet (104), and the control end of the power switch of the electromagnet (104) is connected to the processor module, so that the electromagnet (104) is controlled to be magnetically connected to the FTU terminal.

Images