FTU pole-climbing-free debugging communication device and method

Abstract

The invention discloses an FTU (feeder terminal unit) pole-climbing-free debugging communication device and method.The device comprises an image acquisition module, a direction deflection module, a lifting module and a horizontal moving module which are sequentially arranged from top to bottom, and a cable clamping mechanism is arranged on the image acquisition module; the image acquisition module, the direction deflection module, the lifting module and the horizontal moving module are all connected with the processor module, the processor module controls the direction deflection module, the lifting module and the horizontal moving module to work according to image information acquired by the image acquisition module, and a cable clamped by the cable clamping mechanism is inserted into a corresponding interface of an FTU terminal. According to the method, an image recognition model is adopted to recognize a target interface and obtain center point pixel coordinates of a bounding box of a recognition result, and the confidence coefficient is calculated by considering the distance and shape similarity between a predicted bounding box and a real bounding box when the target interface is recognized, so that the accuracy of the bounding box is improved. According to the invention, rod-climbing-free debugging is realized, the safety risk is reduced, and the operation efficiency is improved.

Description

Technical Field

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

[0002] In the distribution automation system, FTU (feeder terminal unit) is a key device to realize data acquisition, control, communication and other functions, such as Figure 1 As shown, the device is generally 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 surface, as shown in Figure 2 When maintaining or debugging the FTU, workers are currently required to climb the pole, connect the relevant communication cables to the control panel at the bottom of the FTU terminal, and then use a dedicated debugging terminal (such as a laptop or tablet computer) and dedicated debugging software to read data and send parameter data. This is not only time-consuming and labor-intensive, but also poses a safety hazard. 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, in which the debugging personnel can connect the relevant communication cables to the FTU terminal by 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 FTU pole-free debugging communication device comprises 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, wherein the bottom of the horizontal movement module is supported by a support rod, and a cable clamping mechanism is provided on the image acquisition module. 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 work respectively according to image information collected by the image acquisition module, and inserts a cable clamped by the cable clamping mechanism into an interface corresponding to 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 a top seat through a cross lifting arm, the top seat is supported below the direction deflection module, 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, the control end of the second motor is connected to the processor module, so that the second motor rotates under control 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 which are arranged in sequence from top to bottom, the third base is supported below the lifting module, the fifth base is detachably connected to the support rod, the fourth base and the fifth base are cross-arranged, 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 also 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 a FTU pole-free debugging communication method, which is applied to any of the FTU pole-free debugging communication devices described above, and comprises the following steps:

[0012] The image information collected by the image acquisition module is obtained, and the target interface is identified by the image recognition model and the pixel coordinates (u, v) of the center point of the bounding box of the recognition result are obtained. When identifying the target interface, the distance and shape similarity between the predicted bounding box and the real bounding box are considered to calculate the confidence to improve the accuracy of the bounding box.

[0013] Convert the center pixel coordinates (u, v) to the world coordinate system coordinates (X b , Yb , Z b ), and convert the image coordinate system into the camera coordinate system and then calculate 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 ) calculates the distance difference (X) between the cable clamped by the cable clamping mechanism (102) and the target interface b -X n |, Y b -Y n |, Z b -Z n |);

[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 |X b -X n |Difference in distance from Y direction|Y b -Y n |, control the lifting module movement to eliminate the vertical Z direction distance difference |Z b -Z n |, 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 getting a response, debugging work 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] Among them, λ c 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 whether the jth bounding box in the i-th grid 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;

[0021] The confidence error expression is as follows:

[0022]

[0023] Among them, λ p is the weight used to balance the ratio of confidence error to other errors, C i 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 whether the i-th grid contains the target. If it does, the value is 1, otherwise it is 0. i (c) is the probability distribution of the true category, is the probability distribution of the predicted category.

[0027] Furthermore, the target interface is a network port. 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] Among them, ρ is the center point of the predicted bounding box The Euclidean distance between the center point b of the true bounding box, d is the diagonal distance of the minimum enclosing rectangle containing the two bounding boxes, and μ reflects the aspect ratio of the predicted bounding box. The similarity with the aspect ratio η of the real bounding box, P(net port) represents the probability of the existence of a net 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, when calculating the angular difference θ between the cable clamped by the cable clamping mechanism (102) and the target interface after converting the image coordinate system into the camera coordinate system, the following steps are included:

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

[0032] A cuboid model of the target interface is established, wherein one face of the cuboid model is parallel to the camera imaging plane, and the corresponding normalized coordinates (x1, y1, z1) and (x2, y2, z2) in the camera coordinate system are obtained according to the coordinates (u1, v1) and (u2, v2) of the diagonal endpoints of the cuboid model in the image coordinate system;

[0033] The angle difference θ is calculated according to the normalized coordinates of the diagonal endpoints in the camera coordinate system, and the expression is as follows:

[0034]

[0035] Among them, x1, y1 and x2, y2 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, thereby avoiding the operator from climbing the pole.

[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, thereby improving the accuracy of the bounding box, and 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 the installation diagram of FTU on outdoor poles.

[0040] Figure 2 This is a schematic diagram of the FTU control panel.

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

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

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

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

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

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

[0047] Fig. 9 The figure is a flow chart of a method according to an embodiment of the present invention.

[0048] Legend: 1-image acquisition module, 2-direction 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 is further described below in conjunction with the accompanying drawings and specific preferred embodiments, but the protection scope of the present invention is not limited thereby.

[0050] Embodiment 1

[0051] In order to improve work efficiency and reduce operation risks, this embodiment proposes a FTU free pole debugging communication device, such as Figure 3 As shown, it includes an image acquisition module 1, a direction deflection module 2, a lifting module 3 and a horizontal movement module 4 which are arranged at the end of a support rod 101 and are 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 working. 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. 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 respectively 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, thereby 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 fixed to the bottom of the FTU terminal by itself during debugging, without the need for the operator to always hold the device, thus 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 command 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, so as to adjust the position of the cable 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 supply 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 5 As shown, the image acquisition module 1 of this embodiment includes an outer shell 11 and a camera 12 disposed in the outer shell 11 , the camera 12 specifically adopts an Omnivision OV2710 camera, and the cable clamping mechanism 102 includes a clamping claw 13 and a servo 14 disposed 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 clamp 13 is installed on one side of the step, and a gap for clamping the cable is formed between the side wall of the clamp 13 and the side wall of the step of the outer shell 11. A cable passing hole is provided at the bottom of the clamp 13, and the cable passing hole is connected to the gap, so that the cable is clamped and fixed after passing through the cable passing hole to reach the gap. The bottom of the clamp 13 is connected to the shaft of the servo 14, so that the clamp 13 opens and closes with the rotation of the shaft of the servo 14 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 clamp 13. For example, the servo 14 is controlled to rotate in a clockwise direction so that the gap between the side wall of the clamp 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 clamp 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 clamp 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 clamp 13 and the side wall of the step of the outer shell 11, the terminal of the cable (such as the crystal head 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 clamp 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 clamp 13 and the side wall of the step of the outer shell 11.

[0059] like Figure 6 As 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 rotation axis 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 is controlled to rotate and drive 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 7As 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 a 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 the second motor 31 drives the cross lifting arm 34 to rise and fall when it is controlled to rotate, 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 respectively provided on both sides of the lifting module 3, and the cross lifting arms 34 include two lifting arms that cross each other, and the intersection of the two lifting arms is hinged by a bolt 34a, and 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 shaft 36 at one end and a second pin shaft 37 that moves in the slide groove at the other end, and the lower end of one lifting arm in the cross lifting arms 34 is connected to the first pin shaft 36 of the second base 32 through a nut 34b, and the upper end is connected to the top base 3 through a nut 34b. The second lifting arm 34 is connected to the second pin shaft 37 of the second base 32, and the lower end of the other lifting arm in the cross lifting arm 34 is connected to the second pin shaft 37 of the second base 32 through a nut 34b, and the upper end is connected to the first pin shaft 36 of the top seat 33 through a nut 34b. The driving mechanism 35 includes a screw rod 35a, and the second pin shaft 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 another 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 8As shown, the horizontal moving module 4 of this embodiment includes a third base 41, a fourth base 42 and a fifth base 43 which are 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 a 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. 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 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 the 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 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.

[0064] Specifically, in this embodiment, the first driving mechanism 46 and the second driving mechanism 49 both use screw rods, the rotating shaft of the third motor 44 is connected to the screw rod of the first driving 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 that is threadedly matched with the screw rod of the first driving mechanism 46 is also provided. Similarly, the rotating shaft of the fourth motor 47 is connected to the screw rod of the second driving 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 that is threadedly matched with the screw rod of the second driving mechanism 49, so that when the third motor 44 and / or the fourth motor 47 moves to one side, 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 moves 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 horizontal plane X-axis and / or the Y-axis. 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 moves 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 horizontal plane X-axis and / or the Y-axis.

[0065] like Figure 1As shown, in this embodiment, in order to facilitate the 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, a first sleeve is provided at the bottom end of the support cover 103, and a second sleeve is provided at the bottom end of the fifth base 143 of the horizontal movement module 4. The end of the support rod 101 is inserted in the second sleeve, and the second sleeve is inserted in 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 a 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] Embodiment 2

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

[0069] S101) Preparation for debugging:

[0070] The operator stands on the ground and touches the upper end surface of the support cover 103 to the lower end surface of the FTU by lifting the support rod 101, then obtains the image information collected by the image acquisition module 1 through the information terminal of the processor module, and moves the entire device forward, backward, left and right for rough adjustment, so that the communication network port on the FTU panel is basically located at the center of the monitoring field of view. Finally, the operator controls the electromagnet 104 on the support cover 103 to be adsorbed to the base of the FTU through the information terminal, and then pulls the outer rod of the support rod 101 down to contact the ground and locks the buckle to complete the rough fixation of the whole set of equipment;

[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 (u, v) of the center point of the bounding box of the recognition result, and considers the distance and shape similarity between the predicted bounding box and the real bounding box to calculate the confidence when identifying the target interface, so as to improve the accuracy of the bounding box;

[0073] S103) Deviation calculation:

[0074] The information terminal converts the center point pixel coordinates (u, v) into world coordinate system 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 ) calculates the distance difference (X) between the cable clamped by the cable clamping mechanism 102 and the target interface b -X n |, Y b -Y n |, Z b -Z n |);

[0076] S104) Position adjustment:

[0077] The information terminal controls the direction deflection module 2 to move to eliminate the angle difference θ, and controls the horizontal movement module 4 to move to eliminate the distance difference in the horizontal X direction |X b -X n |Difference in distance from Y direction|Y b -Y n |, control the lifting module 3 to eliminate the distance difference in the vertical Z direction |Z b -Z n |, 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 after waiting for the operator's confirmation, the electromagnet 104 on the support cover 103 is powered off and detached 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 detached 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, when an image recognition model is used to identify the target interface and obtain the pixel coordinates (u, v) of the center point of the bounding box of the recognition result, the YOLO target detection algorithm is specifically 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 as a training data set for the YOLO algorithm. Then the input FTU image is divided into fixed grids, and multiple bounding boxes are predicted for each grid. The 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.

[0085] Typically, the confidence of each bounding box It is used to reflect whether this grid contains the network port target and the accuracy of the bounding box. If there is no network port in this grid, the probability P(network port) = 0 and the confidence is 0; otherwise, the confidence is the intersection-over-union ratio (IOU) 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 of ​​the predicted bounding box and the true bounding box. On this basis, the embodiment simultaneously 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]

[0087] Among them, ρ is the center point of the predicted bounding box The Euclidean distance between the center point b of the true bounding box, d is the diagonal distance of the minimum enclosing rectangle containing the two bounding boxes, and μ reflects the aspect ratio of the predicted bounding box. The similarity with the aspect ratio η of the real bounding box, P(net port) represents the probability of the existence of a net 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]

[0091] Among them, λc 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 whether the jth bounding box in the i-th grid 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 does not include the network port. Both are confidences after comprehensively considering the overlapping area, distance, and shape similarity, which can make the subsequent coordinate distance calculation more accurate. The expression is as follows:

[0093]

[0094] Among them, λ p is the weight used to balance the ratio of confidence error to other errors, C i 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]

[0097] in, Indicates whether the i-th grid contains the target. If it does, the value is 1, otherwise it is 0. i (c) 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 center pixel coordinates (u, v) of the bounding box of the target interface. Correspondingly, in step S103, the center pixel coordinates (u, v) are converted into the world coordinate system coordinates (x b , Y b , Z b ) is calculated as:

[0099]

[0100] Among them, Z cIt 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 debugging the FTU is a rectangular parallelepiped with a length and width of 12.5mm*9mm, and the two diagonal endpoints of the rectangular parallelepiped in the image can be detected by the monocular camera, and due to the positioning effect 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 taken as the origin of the world coordinate system, and the camera is calibrated using the Zhang Zhengyou calibration method to obtain the intrinsic parameter matrix and the extrinsic parameter matrix. The camera intrinsic parameter matrix is ​​usually expressed as:

[0103]

[0104] Among them, f x and f y is the focal length along the width and height of the image, c x and c y 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]

[0106] Wherein, (u, v) represents the image coordinates, (x, y, z) represents the normalized coordinates in the camera coordinate system, and in this embodiment, it is assumed that f x =f y =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 cable jack rectangular parallelepiped is W, the height is H, and one face of the network cable jack rectangular parallelepiped is parallel to the camera imaging plane, the position of the network cable jack rectangular parallelepiped in the camera coordinate system can be determined by the detected diagonal endpoints. Specifically, assuming that the center of the network cable jack rectangular parallelepiped is located at the origin of the camera coordinate system (0,0,0), and one side of the network cable jack rectangular parallelepiped is parallel to the x-axis of the camera, the rotation of the network cable jack rectangular parallelepiped can be calculated by the detected diagonal endpoints. Assuming that the coordinates of the diagonal endpoints in the image coordinate system are (u1,v1) and (u2,v2), the corresponding normalized coordinates (x1,y1,z1) and (x2,y2,z2) in the camera coordinate system can be obtained according to the coordinates (u1,v1) and (u2,v2) of the diagonal endpoints of the rectangular parallelepiped model in the image coordinate system in step S201;

[0108] S203) Calculate the angle difference θ according to 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 z1=z2=0. Then, the angle difference θ is calculated by the following formula:

[0109]

[0110] Among them, x1, y1 and x2, y2 are the normalized coordinate values ​​of the diagonal endpoints on the x-axis and y-axis in the camera coordinate system, and (u1, v1) and (u2, v2) are the coordinate values ​​of the diagonal endpoints in the image coordinate system.

[0111] In step S103 of this embodiment, the coordinates of the center point of the known network port line (X n , Y n , Z n ) is the coordinate of the reference center point, then the world coordinate system coordinate (X b , Y b , Z b ) and the target interface reference center coordinates (X n , Y n , Z n ) Calculate the distance difference (X) between the cable clamped by the cable clamping mechanism 102 and the FTU network port b -X n |, Y b -Y n |, Z b -Z n |). In step S104, the information terminal can b -X n |, Y b -Y n |, Z b -Zn |) calculates the distance to be moved in the horizontal X-axis direction, the Y-axis direction and the vertical Z-axis direction, and sends corresponding control instructions to the data processing module to control the horizontal movement module 4 and / or the lifting module 3. Further, based on the calculated angular difference θ, the information terminal can calculate the angle to be rotated, and send corresponding control instructions to the data processing module to control the direction deflection module 2. Specifically, the corresponding motors in the direction deflection module 2, the lifting module 3 and the horizontal movement module 4 are controlled to cooperate with the action, so as to realize the automatic adjustment of the crystal head inserted into the RJ45 network port on the FTU base to carry out debugging work.

[0112] In summary, the present invention proposes an FTU pole-free debugging communication device, and proposes an 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:

[0113] 1. The number of operators has been significantly reduced, from 4 to 2.

[0114] 2. The operation time has been greatly shortened. After testing, the operation time has been shortened from 30 minutes to 4-5 minutes;

[0115] 3. Transform pole climbing operations into ground operations to effectively reduce safety risks.

[0116] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A FTU pole-free debugging communication device, characterized in that: The invention 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 arranged 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 move according to image information collected by the image acquisition module (1), and inserts a cable clamped by the cable clamping mechanism (102) into an interface corresponding to an FTU terminal.

2. The FTU pole-free debugging communication device according to claim 1 is characterized in that: The image acquisition module (1) comprises an outer shell (11) and a camera (12) arranged in the outer shell (11); the cable clamping mechanism (102) comprises a clamping claw (13) and a steering gear (14) arranged 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 steering gear (14), so that the clamping claw (13) opens and closes with the rotation of the rotating shaft of the steering gear (14) to release or clamp the cable; the output end of the camera (12) and the control end of the steering gear (14) are respectively connected to the processor module.

3. The FTU pole-free debugging communication device according to claim 1 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); the rotation axis of the first motor (21) is connected to the image acquisition module (1); and the control end of the first motor (21) is connected to the processor module, so that the first motor (21) is controlled to rotate and drives the cable clamped by the cable clamping mechanism (102) on the image acquisition module (1) to adjust the angle.

4. The FTU pole-free debugging communication device according to claim 1 is characterized in that: The lifting module (3) comprises 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 rise and fall via a driving mechanism (35); a control end of the second motor (31) is connected to a processor module, so that the second motor (31) is controlled to rotate and drives the cable clamped by the cable clamping mechanism (102) on the image acquisition module (1) to rise and fall.

5. The FTU pole-free debugging communication device according to claim 1 is characterized in that: The horizontal moving module (4) comprises a third base (41), a fourth base (42) and a fifth base (43) which are 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 a 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 a 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.

6. The FTU pole-free debugging communication device according to claim 1 is characterized in that: A support cover (103) is also provided at the end of the support rod (101), and 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 an electromagnet (104) is provided at the top of the support cover (103), 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.

7. A FTU pole-free debugging communication method, applied to the FTU pole-free debugging communication device according to any one of claims 1 to 6, characterized in that: The following steps are involved: Acquire image information acquired by the image acquisition module (1), use an image recognition model to identify the target interface and obtain the pixel coordinates (u, v) of the center point of the bounding box of the recognition result, and calculate the confidence by considering the distance and shape similarity between the predicted bounding box and the real bounding box when identifying the target interface, so as to improve the accuracy of the bounding box; Convert the center pixel coordinates (u, v) to the world coordinate system 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 ) calculates the distance difference (|X) between the cable clamped by the cable clamping mechanism (102) and the target interface b -X n |,|Y b -Y n |,|Z b -Z n |); The direction deflection module (2) is controlled to move to eliminate the angle difference θ, and the horizontal movement module (4) is controlled to move to eliminate the distance difference in the horizontal X direction |X b -X n |Difference in distance from Y direction|Y b -Y n |, control the lifting module (3) to eliminate the distance difference in the vertical Z direction |Z b -Z n |, 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 getting a response, debugging work begins.

8. The FTU pole-free debugging communication method according to claim 7 is characterized in that: The loss function during the training of the image recognition model includes coordinate error, confidence error and classification error, where: The coordinate error expression is as follows: Among them, λ c 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 whether the jth bounding box in the i-th grid 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: Among them, λ p is the weight used to balance the ratio of confidence error to other errors, C i 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 whether the i-th grid contains the target. If it does, the value is 1, otherwise it is 0. i (c) is the probability distribution of the true category, is the probability distribution of the predicted category.

9. The FTU pole-free debugging communication method according to claim 7 is characterized in that: The target interface is a network port. 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: Among them, ρ is the center point of the predicted bounding box The Euclidean distance between the center point b of the true bounding box, d is the diagonal distance of the minimum enclosing rectangle containing the two bounding boxes, and μ reflects the aspect ratio of the predicted bounding box. The similarity with the aspect ratio η of the real bounding box, P(net port) represents the probability of the existence of a net port in the bounding box, It represents the intersection-over-union ratio of the predicted bounding box and the true bounding box.

10. The FTU pole-free debugging communication method according to claim 7 is characterized in that: When the image coordinate system is converted into the camera coordinate system and the angular difference θ between the cable clamped by the cable clamping mechanism (102) and the target interface is calculated, the following steps are included: Use the camera intrinsic parameter matrix to transform the image coordinates into normalized coordinates in the camera coordinate system; A cuboid model of the target interface is established, wherein one face of the cuboid model is parallel to the camera imaging plane, and the corresponding normalized coordinates (x1, y1, z1) and (x2, y2, z2) in the camera coordinate system are obtained according to the coordinates (u1, v1) and (u2, v2) of the diagonal endpoints of the cuboid model in the image coordinate system; The angle difference θ is calculated according to the normalized coordinates of the diagonal endpoints in the camera coordinate system, and the expression is as follows: Among them, x1, y1 and x2, y2 are the normalized coordinate values ​​of the x-axis and y-axis in the camera coordinate system respectively.

Images