Nuclear material warehouse abnormity early warning system and method

Abstract

The invention relates to a nuclear material warehouse abnormity early warning system and method, and the system comprises a laser radar personnel tracking system, a nuclear radiation detection network system, and an ultra wide band wireless positioning system, each system employs a layered architecture, and is composed of a sensing layer, a network layer, a processing layer, and an application layer, in the sensing layer, the laser radar personnel tracking system is provided with a plurality of laser radars, the nuclear radiation detection network system is provided with radiation detectors distributed in a place, and the ultra-wideband wireless positioning system is provided with an ultra-wideband tag and an ultra-wideband base station antenna which are arranged on a nuclear material container; the network layer is provided with equipment responsible for data transmission and communication of each system sensor; the processing layer is provided with a processor for carrying out processing, feature extraction and information fusion on data of each system sensor; and the application layer realizes anomaly detection, early warning and visual display. According to the invention, identification of illegal behaviors of personnel and early warning and positioning of illegal transfer of nuclear materials based on multi-sensing information fusion are realized, and the intelligent detection and identification capability of nuclear security is significantly improved.

Description

A nuclear material library anomaly early warning system and method Technical Field

[0001] This invention relates to the field of nuclear security technology, specifically to an anomaly early warning system and method for a nuclear material library. Background Technology

[0002] Nuclear security is a crucial application of social security within the nuclear field. Its core objective is to protect nuclear materials and other radioactive substances from theft and illegal transfer, prevent deliberate sabotage of nuclear facilities, and ultimately prevent the illegal possession, smuggling, diversion, and ultimate use of nuclear materials for nuclear proliferation, or the release of radioactive materials. Physical protection, as a vital safeguard for the security of nuclear materials and facilities, refers to the protective measures and technical systems used to prevent the theft and illegal transfer of nuclear materials and the destruction of nuclear facilities.

[0003] With the development of nuclear technology and the expansion of the application scope of nuclear materials, traditional nuclear material storage security systems face severe challenges in terms of intelligence and digitalization. Current nuclear material storage security solutions are mainly derived from those of ordinary civilian facilities, and are mostly passive mechanisms that assume staff operate according to regulations and pre-set procedures. Therefore, their intelligence level is low, they cannot identify highly concealed illegal operations, and their detection capabilities are insufficient, leading to inaccurate judgments on complex anomalies.

[0004] The existing nuclear security and control systems for radioactive source storage / nuclear material storage facilities mainly have the following technical problems:

[0005] 1) The lack of identification methods for non-standard operations makes it impossible to proactively track and accurately locate staff, resulting in the inability to determine the compliance of operations in a timely and accurate manner;

[0006] 2) There is a lack of accurate detection methods for abnormal handling of radioactive sources / nuclear materials, making it impossible to proactively identify complex and highly concealed abnormal behaviors;

[0007] 3) There is a lack of accurate detection methods for handling nuclear material containers, making it impossible to achieve accurate indoor positioning and tracking;

[0008] 4) The system has a low level of intelligence and mostly operates in a passive manner, making it unable to proactively detect and warn of abnormal situations.

[0009] In response to the above problems, there is an urgent need to substantially improve the nuclear security and control system capabilities of existing radioactive source / nuclear material storage facilities. Summary of the Invention

[0010] The purpose of this invention is to address the problems existing in the prior art by providing a novel nuclear material library anomaly early warning system and method based on multi-sensor information fusion. This system enables the identification of illegal personnel behavior and the early warning and location of illegal transfer of nuclear materials based on multi-sensor information fusion, significantly improving the intelligent detection and identification capabilities of nuclear security.

[0011] To achieve the above objectives, in one aspect, embodiments of the present invention provide a nuclear material library anomaly early warning system, including a lidar personnel tracking system, a nuclear radiation detection network system, and an ultra-wideband wireless positioning system. Each system adopts a layered architecture design, consisting of a perception layer, a network layer, a processing layer, and an application layer. In the perception layer, the lidar personnel tracking system is equipped with multiple lidars capable of covering the entire site area; the nuclear radiation detection network system is equipped with radiation detectors distributed throughout the site; and the ultra-wideband wireless positioning system is equipped with ultra-wideband tags attached to the nuclear material containers and ultra-wideband base station antennas. The network layer is equipped with devices responsible for the transmission and communication of sensor data from each system. The processing layer is equipped with processors that process, extract features from, and fuse information from the sensor data of each system. The application layer realizes anomaly detection, early warning alarms, and visualization display.

[0012] Furthermore, in a specific embodiment, the nuclear material library anomaly early warning system described above includes, in the processing layer, a lidar personnel tracking system comprising a multi-lidar joint calibration module, a point cloud data processing module, and a personnel trajectory tracking module.

[0013] The multi-lidar joint calibration module uses one lidar as a reference system and performs coordinate transformation on the transformation matrices of multiple lidars relative to the reference system to achieve unified fusion of multi-lidar data.

[0014] The point cloud data processing module processes the 3D point cloud data acquired by the LiDAR based on a point cloud segmentation algorithm using deep learning.

[0015] The personnel trajectory tracking module uses a multi-target tracking algorithm to continuously track detected personnel.

[0016] Furthermore, in a specific embodiment, in the nuclear material library anomaly early warning system described above, the nuclear radiation detection network system in the processing layer is provided with a radioactive source location module, which uses a radioactive source location method based on an optimization algorithm to solve for the location of the radioactive source.

[0017] Furthermore, in a specific embodiment, the nuclear material library anomaly early warning system described above includes, in the processing layer, an ultra-wideband wireless positioning system comprising a container positioning module and a container handling detection and identification module.

[0018] The container positioning module uses a positioning technology based on the angle of arrival (AHA) of the signal to achieve positioning by receiving angle information of the ultra-wideband tag signal attached to the nuclear material container from multiple base stations.

[0019] The container handling detection and identification module, in conjunction with the work plan, performs anomaly judgment and identification on the handling behavior of nuclear material containers.

[0020] Furthermore, in a specific embodiment, the nuclear material library anomaly early warning system described above includes an application layer comprising a tracking camera, a video surveillance device, and an alarm device. The tracking camera can detect and track target personnel, the video surveillance device displays monitoring video images of the target personnel, and the alarm device is used to issue early warning alarm information.

[0021] On the other hand, embodiments of the present invention provide a method for early warning of nuclear material library anomalies based on the above system, including:

[0022] S1) Sensor data is collected in real time through the perception layer of the lidar personnel tracking system, nuclear radiation detection network system, and ultra-wideband wireless positioning system.

[0023] S2) Transmit sensor data from each system to the processing layer for data processing, feature extraction, and information fusion;

[0024] S3) Upon detecting abnormal operations, the target personnel will be monitored and tracked.

[0025] S4) Track target personnel through surveillance video and report them to trigger an early warning mechanism to notify security personnel to take action.

[0026] Furthermore, in a specific embodiment, in the nuclear material library anomaly early warning method described above, step S2) involves processing the three-dimensional point cloud data acquired by the lidar personnel tracking system using a point cloud segmentation algorithm based on deep learning, including:

[0027] Data preprocessing: Denoising, filtering, and downsampling are performed on point cloud data to improve data quality;

[0028] Feature extraction: Extracting geometric features, texture features, and contextual features from point clouds;

[0029] Personnel detection: Using a point cloud deep learning architecture, accurate detection of personnel targets is achieved;

[0030] Dataset construction: By collecting information and manually labeling it, a relevant dataset is constructed, and a neural network is used for feature learning to obtain a network model for detection.

[0031] Furthermore, in a specific embodiment, in the nuclear material library anomaly early warning method described above, in step S2), the lidar personnel tracking system employs a multi-target tracking algorithm to achieve continuous tracking of detected personnel, including:

[0032] Target initialization: In the initial stage, assign a unique identifier (ID) to each person's test result;

[0033] Position prediction: The position of the current target is predicted using the Kalman filter algorithm;

[0034] Object matching: Matching the predicted location with the new detection result;

[0035] Lifecycle Management: Employs a time-based lifecycle mechanism to dynamically manage target information and prevent redundancy of object information.

[0036] Furthermore, in a specific embodiment, in the nuclear material library anomaly early warning method described above, in step S2), the nuclear radiation detection network system adopts a radioactive source localization method based on an optimization algorithm, which achieves accurate estimation of the radioactive source location by solving for the minimum mean square error between the expected response and the actual response of the radioactive source parameters.

[0037] Furthermore, in a specific embodiment, in the nuclear material library anomaly early warning method described above, in step S2), the ultra-wideband wireless positioning system employs positioning technology based on signal angle of arrival, achieving positioning by receiving angle information of ultra-wideband tag signals attached to the nuclear material container through multiple base station antennas, including:

[0038] Unified coordinate system: All base station antenna nodes are unified to the same spatial coordinate system in advance;

[0039] Angle measurement: The tag node emits a signal, and different base station antennas receive the signal in different directions;

[0040] Angle difference calculation: Select the angle at which a base station antenna receives the signal as a reference, and calculate the angle difference between other base station antennas and the reference;

[0041] Hyperbolic positioning: A hyperbola is established based on the angle of arrival between the tag and the two base station antennas, and the tag position is determined by the intersection of multiple hyperbolas.

[0042] Furthermore, in a specific embodiment, in the nuclear material library anomaly early warning method described above, in step S2), the ultra-wideband wireless positioning system, in conjunction with the work plan, performs anomaly judgment and identification on the handling behavior of nuclear material containers, including:

[0043] Position change detection: Real-time monitoring of container position changes to identify handling behavior;

[0044] Work plan comparison: Compare actual handling activities with the pre-set plan;

[0045] Anomaly detection: When unauthorized container handling is detected, an alert is immediately triggered;

[0046] Tag identification: Identify whether a container has been moved and the specific container that has been moved using an ultra-wideband base station antenna.

[0047] The beneficial effects of this invention are as follows:

[0048] 1) Proactive identification capability: It changes the traditional passive working mechanism, can proactively track and accurately locate staff, and timely and accurately determine the compliance of operations, thereby proactively discovering and identifying complex and highly concealed abnormal behaviors;

[0049] 2) Significantly improved recognition accuracy: The LiDAR scanning and recognition accuracy reaches over 95%, far exceeding that of traditional video surveillance systems;

[0050] 3) Significantly improved positioning accuracy: Ultra-wideband wireless positioning error is controlled within 20cm, and the positioning accuracy of radiation sources reaches below 25cm;

[0051] 4) Multi-dimensional monitoring: Through multi-sensor fusion, comprehensive monitoring of personnel, radiation sources, and containers can be achieved;

[0052] 5) Significantly enhanced real-time performance: Parallel processing of multiple sensors and intelligent algorithm optimization ensure system response time is within seconds;

[0053] 6) High level of intelligence: Based on deep learning, the intelligent decision-making algorithm can autonomously identify complex abnormal behavior patterns;

[0054] 7) High reliability: The multi-sensor redundancy design and cross-validation mechanism improve the overall reliability of the system.

[0055] This invention is not only applicable to the security protection of nuclear material libraries, but can also be extended to other high-security locations, such as hazardous materials warehouses and important material storage facilities, and has broad application prospects and promotional value. Attached Figure Description

[0056] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0057] Figure 1 is a general framework diagram of the nuclear material library anomaly early warning system in a specific embodiment of the present invention;

[0058] Figure 2 is a schematic diagram of the processing layer algorithm flow of the lidar personnel tracking system in a specific embodiment of the present invention;

[0059] Figure 3 is a schematic diagram of the radioactive source positioning module of the nuclear radiation detection network system in a specific embodiment of the present invention;

[0060] Figure 4 is a general framework diagram of the personnel trajectory tracking module of the lidar personnel tracking system in a specific embodiment of the present invention;

[0061] Figure 5 is a schematic diagram of the three-dimensional point cloud data processing flow of the lidar personnel tracking system in a specific embodiment of the present invention. Detailed Implementation

[0062] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0063] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0064] The terms “comprising”, “including”, etc., as used herein indicate the presence of the steps, features, operations, or components, but do not preclude the addition of one or more other steps, features, operations, or components.

[0065] Current security solutions for nuclear material storage facilities are mainly derived from those for ordinary civilian facilities. They are mostly passive working mechanisms with low levels of intelligence, unable to identify highly concealed illegal operations, and lack sufficient detection capabilities, resulting in inaccurate judgment of complex anomalies.

[0066] This invention addresses the shortcomings of existing nuclear material storage security systems and aims to solve the following technical problems:

[0067] 1) Improve the accuracy of personnel identification in nuclear material storage facilities, and achieve high-precision positioning and trajectory tracking of operators;

[0068] 2) To achieve precise location of radioactive sources and improve the reliability and real-time nature of monitoring radioactive materials;

[0069] 3) Establish a nuclear material container handling detection mechanism to effectively identify illegal handling activities;

[0070] 4) Construct a multi-sensor information fusion platform to enable collaborative work and intelligent decision-making among various subsystems.

[0071] By addressing the aforementioned technical problems, this invention aims to provide a novel nuclear material library anomaly early warning system and method based on multi-sensor information fusion. By integrating technologies such as nuclear radiation detection networks, lidar positioning, and ultra-wideband wireless positioning, it achieves the identification of illegal personnel behavior and the early warning and location of illegal transfer of nuclear materials based on multi-sensor information fusion. This enables comprehensive monitoring and anomaly early warning of personnel behavior and nuclear material status within the nuclear material library, significantly improving the intelligent detection and identification capabilities for nuclear security.

[0072] In some embodiments, the present invention provides a nuclear material library anomaly early warning system, including a lidar personnel tracking system, a nuclear radiation detection network system, and an ultra-wideband wireless positioning system. Each system adopts a layered architecture design, consisting of a perception layer, a network layer, a processing layer, and an application layer. In the perception layer, the lidar personnel tracking system is equipped with multiple lidars capable of covering the entire site area; the nuclear radiation detection network system is equipped with radiation detectors distributed around the site; and the ultra-wideband wireless positioning system is equipped with ultra-wideband tags attached to the nuclear material containers and ultra-wideband base station antennas. The network layer is equipped with devices responsible for the transmission and communication of sensor data from each system. The processing layer is equipped with processors that process, extract features from, and fuse information from the sensor data of each system. The application layer realizes anomaly detection, early warning alarms, and visualization display.

[0073] The nuclear material library anomaly early warning method based on the above system includes the following steps:

[0074] S1) Sensor data is collected in real time through the perception layer of the lidar personnel tracking system, nuclear radiation detection network system, and ultra-wideband wireless positioning system.

[0075] S2) Transmit sensor data from each system to the processing layer for data processing, feature extraction, and information fusion;

[0076] S3) Upon detecting abnormal operations, the target personnel will be monitored and tracked.

[0077] S4) Track target personnel through surveillance video and report them to trigger an early warning mechanism to notify security personnel to take action.

[0078] As shown in Figure 1, multiple lidar units of the lidar personnel tracking system are distributed within the nuclear materials warehouse. The lidars scan personnel entering the warehouse, and the acquired point cloud data is transmitted to the processor via a network system.

[0079] The lidar personnel tracking system uses lidar 3D scanning technology and artificial intelligence algorithms to identify personnel. It then locates and tracks personnel trajectories based on a 3D position model, thereby supporting intelligent analysis of abnormal behavior of personnel in nuclear material warehouses.

[0080] In some embodiments, the processing layer of the lidar personnel tracking system includes a multi-lidar joint calibration module, a point cloud data processing module, and a personnel trajectory tracking module. The overall system flow of the processing layer is shown in Figure 2.

[0081] The multi-lidar joint calibration module achieves comprehensive environmental perception within a field of view greater than 180° through multi-lidar joint calibration technology. During calibration, the point cloud of radar 1 is P_lidar-1, and the transformation matrix from the origin is M_1; the point cloud of radar 2 is P_lidar-2, and the transformation matrix from the origin is M_2. Using radar 1 as the reference frame, i.e., M_1 as the identity matrix, coordinate transformation achieves unified fusion of multi-lidar data, filling in blind spots and improving the global coverage and accuracy of the perception system.

[0082] The point cloud data processing module uses a deep learning-based point cloud segmentation algorithm to process the 3D point cloud data acquired by the LiDAR. The specific processing flow is shown in Figure 5, and includes the following steps:

[0083] 1) Data preprocessing: Denoising, filtering, and downsampling are performed on the point cloud data to improve data quality;

[0084] 2) Feature extraction: Extracting geometric features, texture features, and contextual features from the point cloud;

[0085] 3) Personnel detection: Using an improved PointNet++ network architecture, accurate detection of personnel targets is achieved;

[0086] 4) Dataset construction: By collecting information and manually annotating it, a relevant dataset is constructed, and a neural network is used for feature learning to obtain a network model for detection.

[0087] The personnel trajectory tracking module employs a multi-target tracking algorithm to continuously track detected personnel. The module architecture is shown in Figure 4. The multi-target tracking method includes the following steps:

[0088] 1) Target initialization: In the initial stage, assign a unique identifier (ID) to each person's test result;

[0089] 2) Position prediction: The position of the current target is predicted using the Kalman filter algorithm;

[0090] 3) Object matching: Matching the predicted location with the new detection results;

[0091] 4) Lifecycle Management: Adopt a time-based lifecycle management mechanism to dynamically manage target information and prevent redundancy of object information.

[0092] As shown in Figure 1, the nuclear radiation detection network system includes several radiation detectors distributed throughout the site, forming a distributed nuclear radiation detection network. The system uses an optimization algorithm to achieve precise location of the radiation source.

[0093] In some embodiments, a scientifically designed deployment scheme for radiation detectors is employed based on the area and layout characteristics of the nuclear material warehouse. Within the nuclear material warehouse, radioactive sources are stored in sealed containers, and the radiation intensity within the warehouse is within normal standards. When personnel enter the warehouse to retrieve a radioactive source from a designated area, opening the sealed container causes a sudden increase in radiation intensity in that area. At this time, radiation detectors distributed around the warehouse will detect the abnormal radiation intensity and focus their observation on that area.

[0094] While ensuring coverage of the detection range, the number of detectors installed should be minimized to reduce system costs. Each detector node has independent data acquisition, processing, and communication capabilities, and can transmit detection data to the central processing system in real time. In some embodiments, the radiation detectors are arranged according to the radiation source as shown in Figure 3, with several radiation detectors g1-g8 arranged around radiation sources S1 and S2.

[0095] In some embodiments, the processing layer of the nuclear radiation detection network system is equipped with a radiation source localization module, which employs a radiation source localization method based on an optimization algorithm to achieve accurate estimation of the radiation source location by solving a system of nonlinear equations.

[0096] Specifically, let the parameters of the radioactive source be represented by S', including its location coordinates and intensity parameters. The goal of the system is to find the optimal parameters S' that minimize the mean square error between the expected and actual responses.

[0097] Ψ(S') = Σ[D_i - F(S', r_i)]², where D_i is the measurement value of the i-th detector, and F(S', r_i) is the theoretical calculation value based on the parameter S' and the position r_i.

[0098] To solve the above nonlinear equations, the system can employ various optimization algorithms, such as:

[0099] Gradient descent: simple to implement, suitable for fast convergence in the initial stage;

[0100] Newton's method: fast convergence speed, suitable for high-precision calculations;

[0101] The Levenberg-Marquardt method (LM method) combines the advantages of gradient descent and Newton's method, and has good convergence performance.

[0102] Trust region method: It ensures convergence by limiting the iteration step size and is suitable for complex objective functions.

[0103] In practical applications, it is necessary to conduct algorithm research on the convergence speed, sensitivity to local optima, and computational accuracy of the above methods, and finally determine an algorithm that can achieve higher accuracy in a shorter time to be suitable for the application of nuclear security intelligent detection, early warning and prevention system.

[0104] As shown in Figure 1, the ultra-wideband wireless positioning system includes an ultra-wideband tag attached to a nuclear material container and an ultra-wideband base station antenna. This system locates and identifies the container by attaching the ultra-wideband tag and using the ultra-wideband base station antenna. Combined with operational information obtained from the nuclear safety information system, it can determine unauthorized container movement.

[0105] In some embodiments, the processing layer of the ultra-wideband wireless positioning system includes a container positioning module and a container handling detection and identification module.

[0106] The container positioning module employs positioning technology based on angle of arrival (AOA). It achieves positioning by receiving angle information from ultra-wideband tags attached to the nuclear material container via multiple base stations. The specific positioning method includes the following steps:

[0107] 1) Unified coordinate system: All base station antenna nodes are unified to the same spatial coordinate system in advance;

[0108] 2) Angle measurement: The ultra-wideband tag node emits a signal, and different base station antennas receive the signal in different directions;

[0109] 3) Angle difference calculation: Select the angle of the received signal of a certain base station antenna as the reference, and calculate the angle difference between other base station antennas and the reference;

[0110] 4) Hyperbolic positioning: A hyperbola is established based on the angle of arrival between the tag and the two base station antennas, and the tag position is determined by the intersection of multiple hyperbolas.

[0111] The container handling detection and identification module, in conjunction with the work plan obtained from the nuclear safety information system, performs anomaly judgment and identification on nuclear material container handling behavior, specifically including the following steps:

[0112] 1) Position change detection: Real-time monitoring of container position changes to identify handling behavior;

[0113] 2) Work plan comparison: Compare the actual handling activities with the pre-set plan;

[0114] 3) Anomaly Detection: When unauthorized container handling is detected, an alert is immediately triggered;

[0115] 4) Tag identification: Identify whether a container has been moved using an ultra-wideband antenna and determine the specific container based on the location of the container tag.

[0116] This invention achieves the organic integration and intelligent decision-making of multi-source data, including lidar, radiation detection networks, and ultra-wideband positioning, by constructing a unified information fusion and anomaly early warning system. High-performance processing equipment is deployed in the nuclear security control room to support the identification of illegal transfer of nuclear materials using multiple sensors, and related functional hardware and software are deployed on-site. The system architecture adopts a layered design, including a perception layer, a network layer, a processing layer, and an application layer.

[0117] The perception layer includes sensors such as lidar, nuclear radiation detectors, ultra-wideband wireless locators, and visual cameras.

[0118] Network layer: This includes the nuclear security network responsible for the transmission and communication of sensor data;

[0119] Processing layer: realizes data processing, feature extraction and information fusion; including multi-lidar joint calibration module, point cloud data processing module and personnel trajectory tracking module of lidar personnel tracking system, radioactive source positioning module of nuclear radiation detection network system, container positioning module and container handling detection and identification module of ultra-wideband wireless positioning system;

[0120] Application layer: Includes tracking cameras, video surveillance equipment and alarm devices, providing anomaly detection, early warning alarms and visualization display functions.

[0121] The above system architecture forms a multi-sensor fusion system for identifying illegal transfer of nuclear materials. The system's workflow is as follows:

[0122] 1) Anomaly detection: LiDAR, nuclear radiation detection network, and ultra-wideband wireless locators deployed in nuclear material warehouses detect abnormal operations based on data such as personnel trajectory tracking, radioactive material detection signals, or handling behavior of nuclear material containers.

[0123] 2) Target tracking: When an illegal operation is detected, the tracking camera at the warehouse entrance will detect and track the person.

[0124] 3) Video surveillance: Tracking target personnel and reporting their movements via video surveillance;

[0125] 4) Early warning and alarm: The system automatically triggers the early warning mechanism to notify relevant security personnel to take action.

[0126] The application method and system configuration of the novel nuclear material library anomaly early warning system of the present invention are described below with reference to specific embodiments.

[0127] Example 1: Anomaly Early Warning System for Small and Medium-Sized Nuclear Material Libraries

[0128] For small to medium-sized nuclear material storage facilities with an area of ​​approximately 1000 square meters, the anomaly early warning system of this invention is deployed:

[0129] 1) System hardware configuration:

[0130] LiDAR: Four LiDARs are deployed to cover the entire warehouse area, with a scanning frequency of 10Hz and a ranging accuracy of ±2cm.

[0131] Radiation detectors: Eight gamma radiation detectors are installed, with a detection range of 0.1 μSv / h to 10 Sv / h;

[0132] Ultra-wideband equipment: Deployed with 6 positioning base station antennas and 20 container tags, with a positioning accuracy of ±15cm;

[0133] Processing Server: Configured with a high-performance GPU server to support real-time data processing and algorithm execution.

[0134] 2) Software system deployment:

[0135] Data acquisition layer: Enables real-time acquisition and preprocessing of data from various sensors;

[0136] Algorithm processing layer: Deploys core algorithms for personnel tracking and detection, radiation source localization, and container handling identification;

[0137] Decision analysis layer: Enables multi-source information fusion and intelligent decision-making;

[0138] Display interface layer: Provides a visual monitoring interface and displays early warning information.

[0139] 3) System performance indicators:

[0140] Personnel identification accuracy: ≥95%;

[0141] Radiation source positioning accuracy: ≤25cm;

[0142] Container positioning error: ≤20cm;

[0143] System response time: ≤1 second.

[0144] Example 2: Integrated Protection System for Large Nuclear Material Warehouse

[0145] For a large nuclear material warehouse with an area of ​​approximately 5,000 square meters, a multi-layered comprehensive protection system was constructed:

[0146] 1) Layered protection architecture:

[0147] Perimeter protection layer: Deploy a perimeter intrusion detection system and video surveillance;

[0148] Regional monitoring layer: Deploy lidar and radiation detectors in different functional areas;

[0149] Key protection layer: Dense deployment and intensive monitoring of critical storage areas;

[0150] Core protection layer: 24-hour uninterrupted monitoring of high-risk nuclear materials.

[0151] 2) Multi-sensor deployment solution:

[0152] LiDAR: 16 LiDAR units are deployed to achieve 360° full coverage;

[0153] Radiation detectors: 32 gamma radiation detectors are installed to form a dense detection network;

[0154] Ultra-wideband equipment: Deployed 16 positioning base station antennas and 100 container tags;

[0155] Visual surveillance: 64 high-definition cameras are deployed to achieve all-round video surveillance.

[0156] 3) Intelligent linkage mechanism:

[0157] Anomaly Response: When an anomaly is detected in a certain area, the monitoring equipment in the relevant areas will automatically strengthen monitoring;

[0158] Device linkage: Intelligent linkage between devices such as LiDAR, cameras, and access control systems;

[0159] Personnel coordination: The system automatically notifies relevant security personnel to handle the situation;

[0160] Emergency Response Coordination: Seamless integration with the emergency command system.

[0161] Example 3: Mobile Nuclear Material Transportation Monitoring System

[0162] Applying the technology of this invention to real-time monitoring of the nuclear material transportation process:

[0163] 1) Vehicle-mounted monitoring system:

[0164] Personnel monitoring: Monitoring the driver's status through vehicle-mounted cameras and LiDAR;

[0165] Cargo monitoring: Installing ultra-wideband tags and radiation sensors inside transport containers;

[0166] Environmental monitoring: Monitor environmental parameters such as temperature, humidity, and radiation levels inside the carriage;

[0167] Location tracking: Combining GPS and BeiDou positioning, real-time tracking of transport vehicles is achieved.

[0168] 2) Remote monitoring center:

[0169] Real-time monitoring: Remotely monitor the status of transport vehicles and goods in real time;

[0170] Route planning: Optimize transportation routes based on road conditions and safety requirements;

[0171] Abnormal situation handling: Promptly handle any abnormal situations during transportation;

[0172] Historical tracking: Completely records transportation process data, supporting post-event traceability analysis.

[0173] As can be seen from the above embodiments, the novel nuclear material library anomaly early warning system of the present invention has good adaptability and scalability, and can meet the nuclear security needs of different scales and application scenarios, providing strong technical support for the safe management of nuclear materials.

[0174] Those skilled in the art will understand that the specific order of steps in the disclosed process is an example of an exemplary method. Based on design preferences, it should be understood that the specific order of steps in the process can be rearranged without departing from the scope of the invention. The appended methods provide elements of various steps in an exemplary order and are not intended to be limited to the specific order or hierarchy described.

[0175] Other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. Thus, the invention also intends to include such variations and adaptations if they fall within the scope of the claims and their equivalents.

[0176] The above embodiments are merely illustrative examples of the present invention. The present invention may also be implemented in other specific ways or forms without departing from its spirit or essential characteristics. Therefore, the described embodiments should be considered illustrative rather than limiting in any respect. The scope of protection of the present invention should be defined by the claims, and any variations equivalent to the intent and scope of the claims should also be included within the scope of the present invention.

Claims

1. A nuclear material library anomaly early warning system, characterized in that, The system includes a lidar personnel tracking system, a nuclear radiation detection network system, and an ultra-wideband wireless positioning system. Each system adopts a layered architecture design, consisting of a perception layer, a network layer, a processing layer, and an application layer. In the perception layer, the lidar personnel tracking system is equipped with multiple lidar units capable of covering the entire site area; the nuclear radiation detection network system has radiation detectors distributed throughout the site; and the ultra-wideband wireless positioning system has ultra-wideband tags attached to nuclear material containers and ultra-wideband base station antennas. The network layer includes equipment responsible for transmitting and communicating sensor data from each system. The processing layer includes processors that process, extract features from, and fuse information from the sensor data of each system. The application layer enables anomaly detection, early warning alarms, and visualization.

2. The nuclear material library anomaly early warning system as described in claim 1, characterized in that, In the processing layer, the lidar personnel tracking system includes a multi-lidar joint calibration module, a point cloud data processing module, and a personnel trajectory tracking module. The multi-lidar joint calibration module uses one lidar as a reference frame and performs coordinate transformation on the transformation matrices of multiple lidars relative to the reference frame to achieve unified fusion of multi-lidar data. The point cloud data processing module processes the 3D point cloud data acquired by the lidar based on a deep learning-based point cloud segmentation algorithm. The personnel trajectory tracking module uses a multi-target tracking algorithm to continuously track detected personnel.

3. The nuclear material library anomaly early warning system as described in claim 1, characterized in that, In the processing layer, the nuclear radiation detection network system is equipped with a radiation source localization module, which uses a radiation source localization method based on an optimization algorithm to solve for the location of the radiation source.

4. The nuclear material library anomaly early warning system as described in claim 1, characterized in that, In the processing layer, the ultra-wideband wireless positioning system includes a container positioning module and a container handling detection and identification module. The container positioning module uses positioning technology based on the angle of arrival (AHA) of the signal to achieve positioning by receiving angle information of the ultra-wideband tag signal attached to the nuclear material container through multiple base stations. The container handling detection and identification module, in conjunction with the work plan, performs anomaly judgment and identification on the handling behavior of the nuclear material container.

5. The nuclear material library anomaly early warning system as described in claim 1, characterized in that, The application layer includes tracking cameras, video surveillance equipment, and alarm devices. The tracking cameras can detect and track target personnel, the video surveillance equipment displays surveillance video images of the target personnel, and the alarm devices are used to issue early warning alarm information.

6. A method for early warning of nuclear material library anomalies using the system described in any one of claims 1-5, characterized in that, include: S1) Sensor data is collected in real time through the perception layer of the lidar personnel tracking system, nuclear radiation detection network system, and ultra-wideband wireless positioning system. S2) Transmit sensor data from each system to the processing layer for data processing, feature extraction, and information fusion; S3) Detect and track target personnel upon detecting abnormal operations; S4) Track target personnel through surveillance video and report the incident, triggering an early warning mechanism to notify security personnel for handling.

7. The nuclear material library anomaly early warning method as described in claim 6, characterized in that, In step S2), a point cloud segmentation algorithm based on deep learning is used to process the 3D point cloud data acquired by the lidar personnel tracking system, including: data preprocessing: denoising, filtering, and downsampling the point cloud data to improve data quality; feature extraction: extracting geometric features, texture features, and contextual features of the point cloud; personnel detection: using a point cloud deep learning architecture to achieve accurate detection of personnel targets; and dataset construction: constructing a relevant dataset through information collection and manual annotation, and using a neural network for feature learning to obtain a network model for detection.

8. The nuclear material library anomaly early warning method as described in claim 6, characterized in that, In step S2), the lidar personnel tracking system uses a multi-target tracking algorithm to continuously track detected personnel, including: target initialization: in the initial stage, each personnel detection result is assigned a unique identifier ID; position prediction: the position of the current target is predicted using a Kalman filter algorithm; object matching: the predicted position is matched with the new detection result; lifecycle management: a time-based lifecycle mechanism is used to dynamically manage target information and prevent redundancy of object information.

9. The nuclear material library anomaly early warning method as described in claim 6, characterized in that, In step S2), the nuclear radiation detection network system adopts a radiation source localization method based on an optimization algorithm, which achieves accurate estimation of the radiation source location by solving for the minimum mean square error between the expected response and the actual response of the radiation source parameters.

10. The nuclear material library anomaly early warning method as described in claim 6, characterized in that, In step S2), the ultra-wideband wireless positioning system employs positioning technology based on the angle of arrival (AHA). Positioning is achieved by multiple base stations receiving angle information from ultra-wideband tags attached to the nuclear material container. This includes: unified coordinate system: all base station antenna nodes are pre-unified to the same spatial coordinate system; angle measurement: the tag node emits a signal, and different base station antennas receive the signal in different directions; angle difference calculation: the angle at which one base station antenna receives the signal is selected as a reference, and the angle difference between other base station antennas and the reference is calculated; hyperbolic positioning: a hyperbola is established based on the AHA direction between the tag and the two base station antennas, and the tag position is determined by the intersection points of multiple hyperbolas.

11. The nuclear material library anomaly early warning method as described in claim 6, characterized in that, In step S2), the ultra-wideband wireless positioning system, in conjunction with the work plan, performs anomaly judgment and identification on the handling behavior of nuclear material containers, including: position change detection: real-time monitoring of container position changes and identification of handling behavior; work plan comparison: comparing actual handling behavior with preset plan; anomaly judgment: when unauthorized container handling behavior is detected, an early warning is immediately triggered; tag identification: identifying whether a container has been moved and the specific container that has been moved through the ultra-wideband base station antenna.

Images