Intelligent positioning system, method and device for nuclear fuel assembly loading and unloading machine

Abstract

The invention belongs to the technical field of nuclear power, and particularly relates to an intelligent positioning system, method and device for a nuclear fuel assembly loading and unloading machine. According to the positioning result of the system range finder, displacement data of the PLC system of the loading and unloading machine, a target detection technology and a laser ranging technology based on AI deep learning, and a multivariate positioning data coupling algorithm, accurate real-time positioning of the loading and unloading machine is realized, so that coordinates of a single terminal point are synchronously moved, and real-time monitoring of the position of the loading and unloading machine is realized. If the position is not the target position, an alarm is triggered, the loading and unloading supervision efficiency and accuracy are effectively improved, and human errors are reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of nuclear power, and particularly relates to an intelligent positioning system, method and device for a nuclear fuel assembly handling machine. Background Art

[0002] In the related art, the manual positioning implementation steps of the nuclear fuel assembly handling machine are as follows: (1) The fuel operator manually controls the trolley and the carriage to the end coordinates of the paper movement sheet; (2) The loading and unloading supervisor and the fuel supervisor determine the actual position of the current loading and unloading machine by visually observing the scale of the loading and unloading machine manually; (3) After comparing and matching with the end coordinates of the movement sheet, the assembly is released or grabbed.

[0003] In the existing method, the on-site personnel have a large workload, are prone to fatigue and errors, resulting in incorrect coordinate recognition, leading to events such as misplacement and stacking of fuel. Therefore, it is urgent to improve the positioning efficiency and accuracy of the handling machine. Summary of the Invention

[0004] To overcome the problems existing in the related art, an intelligent positioning system, method and device for a nuclear fuel assembly handling machine are provided.

[0005] According to an aspect of the embodiments of the present disclosure, an intelligent positioning system for a nuclear fuel assembly handling machine is provided. The system includes: a server, a handling machine PLC system, a plurality of cameras, and a plurality of rangefinders;

[0006] The plurality of cameras are respectively installed at positions where the transverse scale and the longitudinal scale of the handling machine in the RX plant of the nuclear power plant face the wall of the RX plant, and are used to collect complete images of the transverse scale and the longitudinal scale, and transmit the collected data to the server;

[0007] The server uses a scale recognition model to process and recognize the acquired image to obtain the scale pointer and scale value in the image;

[0008] The plurality of rangefinders are respectively installed at appropriate positions on the opposite wall of the carriage of the handling vehicle and the carriage, and are used to measure the distances between the carriage and the carriage and the transverse wall and the longitudinal wall, and determine the real-time coordinates of the handling machine by transmitting the measurement data to the server and according to the distance data.

[0009] In a possible implementation manner, during actual operation, when the operator controls the movement of the trolley and the carriage of the handling machine, the rangefinder continuously emits light beams to measure the distances between the trolley and the carriage and the fixed points on the opposite wall. The server obtains the distances measured by the rangefinder and determines the actual coordinates of the handling machine;

[0010] When the server determines that the actual coordinates are consistent with the target coordinates of the movement work order of the handling machine, it controls the camera to collect the scale image, and determines the coordinate value indicated by the current pointer according to the collected image.

[0011] When the server determines that the coordinate value pointed to by the current pointer is consistent with the target coordinate of the moving work order, it determines that the loading and unloading machine has accurately reached the specified position;

[0012] When the server determines that the coordinate value pointed to by the current pointer is inconsistent with the target coordinate of the moving work order, it starts an abnormal alarm, performs an AI retest, or instructs auxiliary manual visual supervision; if the retest result still indicates inconsistency, it determines that the current position is incorrect and continues to move the loading and unloading machine until the target coordinate is reached.

[0013] In a possible implementation manner, the algorithm network of the scale recognition model includes a Backbone, a Neck, and a head. Among them, the Backbone is used for feature extraction, the Neck is used for feature enhancement, and the head is used for network prediction;

[0014] Among them, prediction modules Head are respectively built from the feature maps P2 to P5, and four groups of feature maps generated by P2 to P5 are used to predict the coordinate value pointed to by the scale pointer in the current image.

[0015] In a possible implementation manner, the algorithm network structure and the proposed improved structure of the scale recognition model mainly further include an algorithm input layer;

[0016] After the image data of the algorithm input layer is input into the algorithm model, it is adjusted to pixels of a preset size through methods such as scaling and padding;

[0017] In the model training stage of the scale recognition model, an image augmentation method of random flipping, random scaling, and random rotation is adopted to expand the training data.

[0018] In a possible implementation manner, the backbone network designs a module combining a convolutional layer, a normalization layer, and an activation layer, and adopts a skip connection with a residual structure to change the simple parallel max-pooling layer to a combination of serial and parallel forms.

[0019] In a possible implementation manner, the prediction network is intended to adopt an Anchor free prediction method to decouple the classification prediction branch and the position prediction branch of the network, and predict the on-image coordinates and categories of foreign object targets. The loss function includes a classification loss function and a regression loss function. The classification loss function adopts a cross-entropy loss function, the regression loss function is CIoU loss, and a positive and negative sample matching strategy is adopted to accelerate model training.

[0020] In a possible implementation manner, the PLC system of the loading and unloading machine determines the actual coordinates S of the trolley and the gantry crane in real time according to the obtained laser ranging value D l and Equation 3: p :

[0021] Sp =(D l -d α ) / d β

[0022] where d α is the distance from the large vehicle and the small vehicle to the fixed point when they are at the origin of coordinates, and d β is the distance represented by one minimum measurement unit of the scale.

[0023] According to another aspect of the embodiments of the present disclosure, there is provided an intelligent positioning method for a nuclear fuel assembly handling machine, and the method includes the following steps:

[0024] Step 1: The server obtains the actual coordinates of the large vehicle and the small vehicle from the PLC system, and determines whether the current actual coordinates are consistent with the target coordinates of the movement work order of the handling machine;

[0025] Step 2: When the server determines that the current actual coordinates of the large vehicle and the small vehicle are consistent with the target coordinates of the movement order, it controls the camera to collect images of the horizontal scale and the vertical scale. According to the collected scale images, a scale recognition model is used to determine the coordinate value indicated by the current pointer, and it is determined whether the coordinate value indicated by the current pointer is consistent with the target coordinates;

[0026] Step 3: When the server determines that the coordinate value indicated by the current pointer is consistent with the target coordinates, it determines that the handling machine has accurately reached the specified position;

[0027] Step 4: When the server determines that the coordinate value indicated by the current pointer is inconsistent with the target coordinates, it starts an abnormal alarm, and repeats Step 2 for AI retest, prompting for auxiliary manual visual supervision; if the retest result still indicates that the coordinate value indicated by the current pointer is inconsistent with the target coordinates, it is considered that the current position is incorrect, and the handling machine continues to move until it reaches the target coordinates.

[0028] According to another aspect of the embodiments of the present disclosure, there is provided an intelligent positioning device for a nuclear fuel assembly handling machine, and the device includes the following modules:

[0029] An actual coordinate acquisition module, configured to obtain the actual coordinates of the large vehicle and the small vehicle from the PLC system, and determine whether the current actual coordinates are consistent with the target coordinates of the movement work order of the handling machine;

[0030] An intelligent recognition module, configured to control the camera to collect images of the horizontal scale and the vertical scale when it is determined that the current actual coordinates of the large vehicle and the small vehicle are consistent with the target coordinates of the movement order. According to the collected scale images, a scale recognition model is used to determine the coordinate value indicated by the current pointer, and it is determined whether the coordinate value indicated by the current pointer is consistent with the target coordinates;

[0031] A judgment module, used to judge whether the loading and unloading machine has accurately arrived at the designated position when judging that the coordinate value pointed by the current pointer is consistent with the target coordinate;

[0032] The alarm remeasurement module is used to activate an abnormal alarm when it is determined that the coordinate value indicated by the current pointer is inconsistent with the target coordinate, and repeat step 2 to perform AI remeasurement, prompting auxiliary manual visual supervision; if the remeasurement result still indicates that the coordinate value indicated by the current pointer is inconsistent with the target coordinate, it is considered that the current position is wrong, and the loader and unloader continues to move until it reaches the target coordinate.

[0033] According to another aspect of an embodiment of the present disclosure, a non-volatile computer-readable storage medium is provided, on which computer program instructions are stored, and the computer program instructions implement the above method when executed by a processor.

[0034] The beneficial effects of the present disclosure are: the system rangefinder positioning results and the PLC system displacement data of the loader and unloader provided by the present disclosure, the target detection technology based on AI deep learning and the laser ranging technology, the multi-positioning data coupling algorithm, realize the accurate real-time positioning of the loader and unloader, thereby simultaneously and synchronously moving the single end point coordinates, and realizing the real-time monitoring of the position of the loader and unloader. If it is not the target position, an alarm is triggered, which effectively improves the efficiency and accuracy of the loading and unloading supervision and reduces human errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a schematic diagram of the arrangement of cameras in an intelligent positioning system for a nuclear fuel assembly loader and unloader shown in an embodiment of the present disclosure.

[0036] Figure 2 It is a schematic diagram of the arrangement of rangefinders in an intelligent positioning system for a nuclear fuel assembly loader and unloader shown in an embodiment of the present disclosure.

[0037] Figure 3 It is a schematic diagram of an algorithm network structure of a scale recognition model shown in an embodiment of the present disclosure.

[0038] Figure 4 It is a flow chart of an intelligent positioning method for a nuclear fuel assembly loading and unloading machine shown in an embodiment of the present disclosure.

[0039] Figure 5 It is a block diagram of an intelligent positioning device for a nuclear fuel assembly loading and unloading machine shown in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0040] The present disclosure is further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0041] Unless otherwise defined, the technical and scientific terms used in this disclosure have the same meanings as those commonly understood by those skilled in the art to which this disclosure pertains; the terms used in this disclosure are only for the purpose of describing specific embodiments and are not intended to limit this disclosure; the term "including" and any variations thereof in this disclosure are intended to cover non-exclusive inclusion. Obviously, the embodiments described in this disclosure are only some of the embodiments of this disclosure, rather than all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in this disclosure without creative efforts belong to the scope of protection of this disclosure.

[0042] The mention of "embodiment" in this disclosure means that the specific features, structures or characteristics described in connection with the embodiment may be included in at least one embodiment of this disclosure. The phrase appears at various positions in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art explicitly and implicitly understand that the embodiments described herein may be combined with other embodiments.

[0043] Figure 1 It is a schematic diagram showing the arrangement of cameras in an intelligent positioning system for a nuclear fuel assembly handling machine according to an embodiment of this disclosure. Figure 2 It is a schematic diagram showing the arrangement of rangefinders in an intelligent positioning system for a nuclear fuel assembly handling machine according to an embodiment of this disclosure. Refer to Figure 1 and Figure 2 The system of this disclosure includes: a server, a handling machine PLC system, multiple cameras, and multiple rangefinders.

[0044] As Figure 1 shown, multiple rangefinders (for example, the first rangefinder 1, the second rangefinder 2, the third rangefinder 3, and the fourth rangefinder 4) are respectively installed at appropriate positions on the opposite walls of the trolley 6 of the handling vehicle and the trolley 5 to measure the distances between the trolley and the transverse and longitudinal walls, and transmit the measurement data to the server through wireless 5G communication technology to calculate the real-time coordinates of the handling machine and achieve automatic positioning of the handling machine. The installation process of the laser rangefinder includes layout planning, instrument installation and configuration, data collection and calibration, etc.

[0045] As Figure 2As shown in the figure, at the positions where the horizontal scale and the vertical scale of the refueling machine in the RX building of the nuclear power plant face the wall of the RX building, multiple cameras (for example, the first camera 7, the second camera 8, the third camera 9, the fourth camera 10, and the fifth camera 11) are installed respectively. Each camera captures the complete images of the horizontal scale and the vertical scale, and the data collected by the multiple cameras can be transmitted to the server through 5G data. The server starts the scale recognition model, and based on the obtained image processing, recognizes the scale pointer and scale value in the image, realizing the automatic reading of the coordinates of the refueling machine.

[0046] As an example of this embodiment, the image data of the scale recognition model of the present disclosure is mainly sourced from the scene. After installing the cameras by means of on-site survey, the video and image data collected by the cameras are preliminarily cleaned and corrected through data preprocessing, and the key frames of the video are extracted to reduce the storage pressure and improve the professionalism and reliability of the data. For the preprocessed data, manual annotation is carried out under the guidance of experts. Image data annotation can use annotation tools such as LableImg to annotate the objects to be recognized in the image, and the annotation content is the coordinates of the rectangular frame enclosing the object and its category. The annotated images and annotation files are divided into a training set and a validation set in a ratio of 8:2, which are used for model training and evaluation respectively. And the annotated data is used to establish a refueling machine scale data set according to the standard data set structure.

[0047] In the application scenario of the present disclosure, the scale and pointer recognition scenario of the refueling machine not only requires high-precision recognition effect, but also needs to ensure the recognition speed to achieve real-time and accurate positioning and monitoring. The scale recognition model of the present disclosure is based on a convolutional neural network, designs a refueling machine scale detection algorithm, recognizes and locates the scale pointer and scale value in the scale image, and uses the pointer position and the scale value closest to it as the coordinates of the refueling machine at this time. As Figure 3As shown in the figure, the algorithm network of the scale recognition model includes Backbone, Neck, and Head. Among them, Backbone is used for feature extraction, Neck is used for feature enhancement, and Head is used for network prediction. Backbone is a Darknet network based on the lightweight network structure C2f module, which contains multiple convolutional modules, extracts different scale features of the input image layer by layer, and outputs feature maps P1 to P5 of different scales. Neck is a multi-scale feature aggregation network that combines the feature pyramid network and the PAN structure, fuses the multi-scale feature maps output by Backbone, and completes the aggregation of shallow information into deep features. Head is a prediction module that uses a decoupled head structure, which is divided into a class prediction branch and a location prediction branch, and generates the target class and location of different feature maps. To make the best use of the spatial information of the network and ensure that the generated prediction boxes can cover more small targets, the present disclosure establishes prediction modules Head from feature maps P2 to P5 respectively, and uses the four groups of feature maps generated by P2 to P5 to predict the coordinate values pointed by the scale pointer in the current image. For example, since the camera is installed far from the scale, the scale values and the pointer on the collected image are small in scale and belong to small target objects. The feature maps P2, P3, P4, and P5 generated by Backbone can be processed by convolution, fully connected, and splicing to establish a prediction module, thereby improving the network's recognition and positioning capabilities for small targets.

[0048] The scale recognition model is implemented using the Python programming language and the Pytorch deep learning framework, and can be converted into a model format suitable for domestic equipment after model training. The algorithm network structure includes an algorithm input layer, a backbone network Backbone, a feature enhancement network Neck, and a prediction network Head. After the image data of the algorithm input layer is input into the algorithm model, it is adjusted to 640×640 pixels through methods such as scaling and padding. In addition, during the model training stage, image augmentation methods such as random flipping, random scaling, and random rotation are used to expand the training data and improve the generalization ability of the network. The backbone network designs a module that combines convolutional layers, normalization layers, and activation layers, and uses skip connections with a residual structure to change the simple parallel max pooling layer into a combination of serial and parallel forms, reducing the number of model parameters and computational complexity, improving the model's computational efficiency, and effectively extracting the shallow features of foreign object images. The feature enhancement network uses the multi-scale feature layers of the backbone network to construct a feature pyramid network structure, extracts the deep features of foreign object images, identifies foreign objects of different sizes, and improves the model's ability to distinguish different consumable foreign objects. The prediction network plans to use an Anchor free prediction method to decouple the classification prediction branch and the location prediction branch of the network, and predict the coordinates and classes of foreign object targets on the image. The loss function includes a classification loss function and a regression loss function. The classification loss function uses a cross-entropy loss function, the regression loss function is CIoU loss, and a positive and negative sample matching strategy is used to accelerate model training.

[0049] Figure 4 is a flowchart of an intelligent positioning method for a nuclear fuel assembly handling machine shown in an embodiment of the present disclosure. Refer to Figure 2 and Figure 4 , a two-dimensional coordinate system is established with two perpendicular walls of the plant as a reference, and the position where the displacement distance of the handling machine is 0 is used as the coordinate origin. Then, in the rangefinder calibration stage, the handling machine PLC system obtains the distances D l (x l , y l ) measured by each rangefinder l between the trolley and the fixed point of the handling machine. For example, as Figure 4 shown, 4 rangefinders can be set. The first rangefinder 1 and the second rangefinder 2 are used to measure the distance between the trolley 5 and the starting point, and the third rangefinder 3 and the fourth rangefinder 4 are used to measure the distance between the small car and the starting point.

[0050] The handling machine PLC system determines the displacement distances D p (x p , y p ) of the handling machine trolley according to Equation 1:

[0051] D p = D l - d α Equation 1

[0052] where d α is the distance from the trolley and the small car to the fixed point when they are at the coordinate origin.

[0053] The handling machine PLC system determines the scale coordinates S p of the trolley and the small car according to the displacement distances D p and Equation 2:

[0054] S p = D p / d β Equation 2

[0055] where d β is the distance represented by one minimum measurement unit of the scale.

[0056] In this way, the calculation model between the laser ranging value D l and the coordinate S p is shown in Equation 3:

[0057] S p = (D l - d α ) / d β

[0058] During actual operation, when the operator controls the movement of the trolley and the gantry of the fuel handling machine, the rangefinder continuously emits light beams to measure the distances of the trolley and the gantry from the fixed points on the opposite wall respectively. The PLC system of the fuel handling machine determines the actual coordinates S of the trolley and the gantry in real time according to the obtained laser ranging value D l and Equation 3 p .

[0059] Figure 4 is a flowchart of an intelligent positioning method for a nuclear fuel assembly handling machine shown in an embodiment of the present disclosure. As Figure 4 shown, the method includes the following steps

[0060] Step 1: The server obtains the actual coordinates of the trolley and the gantry from the PLC system and determines whether the current actual coordinates are consistent with the target coordinates of the movement work order of the fuel handling machine

[0061] Step 2: When the server determines that the current actual coordinates of the trolley and the gantry are consistent with the target coordinates of the movement order, it controls the camera to collect images of the horizontal scale and the vertical scale. According to the collected scale images, it uses a scale recognition model to determine the coordinate value indicated by the current pointer and determines whether the coordinate value indicated by the current pointer is consistent with the target coordinates

[0062] Step 3: When the server determines that the coordinate value indicated by the current pointer is consistent with the target coordinates, it determines that the fuel handling machine has accurately reached the specified position

[0063] Step 4: When the server determines that the coordinate value indicated by the current pointer is inconsistent with the target coordinates, it activates an abnormal alarm and repeats Step 2 for AI remeasurement, prompts auxiliary manual visual supervision, and further improves the positioning accuracy through appropriate manual intervention; if the remeasurement result still indicates that the coordinate value indicated by the current pointer is inconsistent with the target coordinates, it is considered that the current position is incorrect, and the fuel handling machine continues to move until it reaches the target coordinates

[0064] The system ranging result provided by the present disclosure and the displacement data of the PLC system of the fuel handling machine, based on the object detection technology and laser ranging technology of AI deep learning, and the multi-source positioning data coupling algorithm, realize the precise real-time positioning of the fuel handling machine, thereby synchronizing the end coordinates of the movement order at the same time and realizing the real-time monitoring of the position of the fuel handling machine. If it is not the target position, an alarm is triggered, effectively improving the supervision efficiency and accuracy of the loading and unloading, and reducing human errors

[0065] In a possible implementation, an intelligent positioning device for a nuclear fuel assembly handling machine is provided. The device includes the following modules

[0066] An actual coordinate acquisition module, configured to acquire the actual coordinates of the trolley and the small vehicle from the PLC system, and determine whether the current actual coordinates are consistent with the target coordinates of the movement work order of the loading and unloading machine;

[0067] An intelligent recognition module, configured to control a camera to collect images of the horizontal scale and the vertical scale when it is determined that the current actual coordinates of the trolley and the small vehicle are consistent with the target coordinates of the movement order, and according to the collected scale images, use a scale recognition model to determine the coordinate value indicated by the current pointer, and determine whether the coordinate value indicated by the current pointer is consistent with the target coordinates;

[0068] A judgment module, configured to determine that the loading and unloading machine accurately reaches the specified position when it is determined that the coordinate value indicated by the current pointer is consistent with the target coordinates;

[0069] An alarm retest module, configured to start an abnormal alarm when it is determined that the coordinate value indicated by the current pointer is inconsistent with the target coordinates, and repeat step two for AI retest, prompting auxiliary manual visual supervision; if the retest result still indicates that the coordinate value indicated by the current pointer is inconsistent with the target coordinates, it is considered that the current position is incorrect, and the loading and unloading machine is continuously moved until the target coordinates are reached.

[0070] The description of the above device has been elaborated in detail in the description of the above method, and will not be repeated here.

[0071] Figure 5 It is a block diagram of an intelligent positioning device for a nuclear fuel assembly loading and unloading machine shown in an embodiment of the present disclosure. For example, device 1900 may be provided as a server. Refer to Figure 5 , device 1900 includes a processing component 1922, which further includes one or more processors, and memory resources represented by a memory 1932 for storing instructions executable by the processing component 1922, such as application programs. The application programs stored in the memory 1932 may include one or more modules each corresponding to a set of instructions. In addition, the processing component 1922 is configured to execute instructions to perform the above method.

[0072] Device 1900 may further include a power supply component 1926 configured to perform power management of device 1900, a wired or wireless network interface 1950 configured to connect device 1900 to a network, and an input / output (I / O) interface 1958. Device 1900 may operate based on an operating system stored in the memory 1932, such as Windows ServerTM, MacOS XTM, UnixTM, LinuxTM, FreeBSDTM or the like.

[0073] In an exemplary embodiment, a non-volatile computer-readable storage medium is also provided, such as a memory 1932 including computer program instructions, and the computer program instructions can be executed by a processing component 1922 of the device 1900 to complete the above method.

[0074] The present disclosure may be a system, a method, and / or a computer program product. The computer program product may include a computer-readable storage medium having thereon computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0075] A computer-readable storage medium may be a tangible device that can retain and store instructions for use by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disc (DVD), a memory stick, a floppy disk, a mechanically encoded device such as a punch card or raised structures in a groove having instructions stored thereon, and any suitable combination of the foregoing. The computer-readable storage medium as used herein is not construed as a transitory signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., an optical pulse through an optical fiber cable), or an electrical signal transmitted through a wire.

[0076] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to various computing / processing devices, or downloaded to an external computer or external storage device through a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network may include a copper transmission cable, an optical fiber transmission, a wireless transmission, a router, a firewall, a switch, a gateway computer, and / or an edge server. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in each computing / processing device.

[0077] The computer program instructions for performing the operations of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine - related instructions, microcode, firmware instructions, state - setting data, or source code or object code written in any combination of one or more programming languages, including object - oriented programming languages such as Smalltalk, C++, etc., and conventional procedural programming languages such as the "C" language or similar programming languages. The computer - readable program instructions may be executed entirely on the user's computer, partially on the user's computer, executed as a stand - alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider). In some embodiments, by using the state information of the computer - readable program instructions to customize an electronic circuit, such as a programmable logic circuit, a field - programmable gate array (FPGA), or a programmable logic array (PLA), the electronic circuit can execute the computer - readable program instructions to implement various aspects of the present disclosure.

[0078] Aspects of the present disclosure are described herein with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present disclosure. It should be understood that each block of the flowcharts and / or block diagrams, and combinations of blocks in the flowcharts and / or block diagrams, can be implemented by computer - readable program instructions.

[0079] These computer - readable program instructions can be provided to a processor of a general - purpose computer, a special - purpose computer, or other programmable data - processing apparatus to produce a machine such that the instructions, when executed by the processor of the computer or other programmable data - processing apparatus, create a means for implementing the functions / acts specified in one or more blocks of the flowchart and / or block diagram. These computer - readable program instructions can also be stored in a computer - readable storage medium, which causes a computer, a programmable data - processing apparatus, and / or other devices to operate in a particular manner, so that the computer - readable medium storing the instructions includes a manufacture, which includes instructions for implementing various aspects of the functions / acts specified in one or more blocks of the flowchart and / or block diagram.

[0080] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device, causing a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process such that the instructions executed on the computer, other programmable data processing apparatus, or other device implement the functions / acts specified in one or more boxes of the flowchart and / or block diagram.

[0081] The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a segment of code, or a portion of an instruction, which contains one or more executable instructions for implementing the specified logical function. In some alternative implementations, the functions noted in the blocks may occur out of the order noted in the figures. For example, two consecutive blocks may in fact be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending upon the functionality involved. It should also be noted that each block of the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented by special purpose hardware-based systems that perform the specified functions or acts, or by combinations of special purpose hardware and computer instructions.

[0082] The embodiments of the present disclosure have been described above. The above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The choice of terms used herein is intended to best explain the principles of the embodiments, the practical application, or improvements made to the technology in the market, or to enable other ordinary skilled persons in the art to understand the embodiments disclosed herein.

Claims

1. An intelligent positioning system for a nuclear fuel assembly handling machine, characterized in that, The system includes: a server, a loading and unloading machine PLC system, multiple cameras, and multiple rangefinders; The multiple cameras are respectively installed at positions where the horizontal scale and the vertical scale of the loading and unloading machine in the RX building of the nuclear power plant face the wall of the RX building, and are used to collect complete images of the horizontal scale and the vertical scale and transmit the collected data to the server; The server uses a scale recognition model to process and recognize the obtained images to obtain the scale pointer and scale value in the images; The multiple rangefinders are respectively installed at appropriate positions on the opposite walls of the large vehicle of the loading and unloading vehicle and the small vehicle, and are used to measure the distances between the large vehicle and the small vehicle and the horizontal wall and the vertical wall, and determine the real-time coordinates of the loading and unloading machine according to the distance data by transmitting the measurement data to the server; 2. The system according to claim 1, characterized in that, During actual operation, when the operator controls the movement of the large and small vehicles of the loading and unloading machine, the rangefinders continuously emit light beams to measure the distances between the large and small vehicles and the fixed points on the opposite walls. The server obtains the distances measured by the rangefinders and determines the actual coordinates of the loading and unloading machine; When the server determines that the actual coordinates are consistent with the target coordinates of the movement work order of the loading and unloading machine, it controls the camera to collect scale images, and determines the coordinate value indicated by the current pointer according to the collected images; When the server determines that the coordinate value indicated by the current pointer is consistent with the target coordinates of the movement work order, it determines that the loading and unloading machine has accurately reached the specified position; When the server determines that the coordinate value indicated by the current pointer is inconsistent with the target coordinates of the movement work order, it starts an abnormal alarm, performs an AI remeasurement, or instructs auxiliary manual visual supervision; If the remeasurement result still indicates inconsistency, it is determined that the current position is incorrect, and the loading and unloading machine continues to move until it reaches the target coordinates.

3. The system according to claim 1, wherein The algorithm network of the scale recognition model includes a Backbone, a Neck, and a head. Among them, the Backbone is used for feature extraction, the Neck is used for feature enhancement, and the head is used for network prediction; Among them, prediction modules Head are respectively built from the feature maps P2 to P5, and four groups of feature maps generated by P2 to P5 are used to predict the coordinate value indicated by the scale pointer in the current image.

4. The system according to claim 1, wherein The algorithm network structure and the proposed improved structure of the scale recognition model mainly further include an algorithm input layer; After the image data of the algorithm input layer is input into the algorithm model, it is adjusted to pixels of a preset size through methods such as scaling and padding; In the model training stage of the scale recognition model, an image augmentation method of random flipping, random scaling, and random rotation is used to expand the training data.

5. The system according to claim 4, characterized in that, The backbone network designs a module combining a convolutional layer, a normalization layer, and an activation layer, and adopts a skip connection with a residual structure to change the simple parallel maximum pooling layer into a combination of serial and parallel forms.

6. The system according to claim 4, wherein The prediction network is proposed to adopt an Anchor free prediction method to decouple the classification prediction branch and the position prediction branch of the network, and predict the on-image coordinates and categories of foreign object targets; the loss function includes a classification loss function and a regression loss function. The classification loss function adopts a cross-entropy loss function, the regression loss function is a CIoU loss, and a positive and negative sample matching strategy is adopted to accelerate the model training.

7. The system according to claim 4, wherein The PLC system of the charging and discharging machine determines the actual coordinates S of the trolley and the carriage in real time according to the obtained laser ranging value D l and Equation 3 p : S p = (D l - d α ) / d β where d α is the distance from the fixed point when the large vehicle and the small vehicle are at the origin of coordinates, and d β is the distance represented by a minimum measurement unit of the scale.

8. An intelligent positioning method for a nuclear fuel assembly handling machine, characterized in that, The method includes the following steps: Step 1: The server obtains the actual coordinates of the trolley and the cart from the PLC system and determines whether the current actual coordinates are consistent with the target coordinates of the movement work order of the loading and unloading machine; Step 2: When the server determines that the current actual coordinates of the trolley and the cart are consistent with the target coordinates of the movement order, it controls the camera to collect images of the horizontal scale and the vertical scale. Based on the collected scale images, it uses a scale recognition model to determine the coordinate value indicated by the current pointer and determines whether the coordinate value indicated by the current pointer is consistent with the target coordinates; Step 3: When the server determines that the coordinate value indicated by the current pointer is consistent with the target coordinates, it determines that the loading and unloading machine has accurately reached the specified position; Step 4: When the server determines that the coordinate value indicated by the current pointer is inconsistent with the target coordinates, it activates an exception alarm and repeats Step 2 for AI retesting, prompting for auxiliary manual visual supervision; if the retest result still indicates that the coordinate value indicated by the current pointer is inconsistent with the target coordinates, it is considered that the current position is incorrect, and the loading and unloading machine continues to move until it reaches the target coordinates.

9. An intelligent positioning device for a nuclear fuel assembly handling machine, characterized in that, The device includes the following modules: An actual coordinate acquisition module, which is used to obtain the actual coordinates of the trolley and the cart from the PLC system and determine whether the current actual coordinates are consistent with the target coordinates of the movement work order of the loading and unloading machine; An intelligent recognition module, which is used to control the camera to collect images of the horizontal scale and the vertical scale when it is determined that the current actual coordinates of the trolley and the cart are consistent with the target coordinates of the movement order. Based on the collected scale images, it uses a scale recognition model to determine the coordinate value indicated by the current pointer and determines whether the coordinate value indicated by the current pointer is consistent with the target coordinates; A judgment module, which is used to determine that the loading and unloading machine has accurately reached the specified position when it is determined that the coordinate value indicated by the current pointer is consistent with the target coordinates; An alarm retest module, which is used to activate an exception alarm when it is determined that the coordinate value indicated by the current pointer is inconsistent with the target coordinates, and repeat Step 2 for AI retesting, prompting for auxiliary manual visual supervision; if the retest result still indicates that the coordinate value indicated by the current pointer is inconsistent with the target coordinates, it is considered that the current position is incorrect, and the loading and unloading machine continues to move until it reaches the target coordinates.

10. A non-volatile computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, the method described in claim 9 is implemented.

Images