32results about How to "Strong engineering application value" patented technology

The invention provides a semantic perception collaborative intrusion detection method for multiple vehicle-mounted networks, and relates to the field of vehicle-mounted network security, and the method comprises the steps: collecting communication data from a plurality of vehicle-mounted network domains such as a controller local area network, a vehicle-mounted Ethernet, a host system and the like; analyzing business features of each domain based on semantic information, and extracting feature representation from the time sequence dependency relationship; and a cross-domain attention mechanism is utilized to realize feature association among multiple network domains so as to identify potential multi-stage intrusion behaviors. Through multi-domain semantic fusion and cooperative detection, the method can effectively recognize complex attack behaviors which only show weak abnormal features, improves the intrusion detection precision and robustness of a vehicle-mounted network, and is suitable for a safety protection scene of an intelligent connected vehicle.

This invention provides a vehicle simulation speed correction method and system, relating to the field of vehicle intelligent dynamics modeling technology. The correction method includes the following steps: S1: Input and process simulation data output from the vehicle dynamics simulation model and corresponding real vehicle test data; S2: Construct a dynamic graph structure representing the interaction between state variables based on a preset physical coupling relationship of the vehicle powertrain; S3: Input the simulation data into a GCN and perform graph convolution operations under the constraints of the dynamic graph structure to extract graph embedding feature sequences representing the spatial dependencies between state variables; S4: Input the graph embedding feature sequences into a TCN and perform temporal convolution operations to learn the dynamic evolution law of state variables in the time dimension and output the correction amount of the vehicle simulation speed; S5: Output the result. Based on this, this invention solves the problem that existing correction methods have various limitations in practical applications.

This invention discloses an airport photovoltaic design method and system based on three-dimensional airspace constraints, which relates to the field of airport photovoltaic engineering technology. S1: Collect legal airspace control data of the target airport, site topography data, flight parameters, and regional annual meteorological and solar radiation data. Based on the collected multi-source data, construct a three-dimensional airspace constraint surface model of the airport, fully integrating four types of core airspace constraint boundaries within the airport: horizontal plane, conical surface, approach surface, and takeoff / climb surface, achieving a three-dimensional quantitative definition of the airspace constraint range. This airport photovoltaic design method and system based on three-dimensional airspace constraints is the first to deeply couple the airport's three-dimensional airspace constraint surface with photovoltaic design, abandoning the traditional two-dimensional uniform height restriction model. It constructs a full-space quantitative model based on the four types of core airspace constraint surfaces, representing a novel technical solution not disclosed in existing technologies, with significant design characteristics specific to airport scenarios.

This invention discloses a short-window gamma-gamma turbulence parameter prediction method for space-to-ground laser communication, belonging to the technical field of space-to-ground laser communication and atmospheric turbulence channel parameter prediction. This method addresses the problems of traditional methods, such as strong dependence on long observation windows, significant degradation in prediction accuracy under short-window scenarios, and insufficient robustness under low signal-to-noise ratio conditions. It establishes a short-window observation model for the space-to-ground optical link and a gamma-gamma channel statistical model, constructs time-dependent short-window training data, and designs a short-window gamma-gamma network. Temporal features are extracted through a convolutional backbone, and a scintillation exponential physical regularization auxiliary head is used to achieve joint parameter prediction under physical constraints, outputting predicted gamma-gamma distributed parameters. This method can achieve high accuracy and good stability in parameter prediction under short observation windows and low signal-to-noise ratio conditions, and can be used for turbulence channel state characterization at the space-to-ground laser communication receiver.

The application discloses a spine-pelvis joint segmentation method based on frequency-space cooperation and adaptive fusion, and belongs to the technical field of medical image processing. A joint segmentation dataset containing spine and pelvis structures is constructed, and standardization preprocessing is completed by supplementing pelvis annotation to public data and combining private clinical data; a multi-scale three-dimensional segmentation network is built, a frequency-space cooperation feature modeling HFSM module is introduced in the coding stage, local details and global semantic features are extracted through joint extraction in the spatial domain and the frequency domain; a selective cross adaptive fusion SCAF module is connected between the encoder and the decoder, boundary enhancement, structure perception and gate weight modulation are performed on different levels of coding features and decoding features, adaptive feature aggregation is realized, the network adopts an end-to-end training mode, a weighted combination loss function is used to optimize model parameters, and the joint segmentation result of the spine and the pelvis is output. The application can effectively improve the segmentation accuracy and integrity of the connection region, complex edge and small structure of the spine-pelvis.

This invention provides a method for optimizing BUCK circuit control based on adaptive reward weighting (DQN), comprising the following steps: Step 1, constructing a reinforcement learning control environment for the BUCK circuit; Step 2, designing a reward function based on output voltage error; Step 3, optimizing the reward weight parameters in the reward function using a particle swarm optimization (PSO) algorithm to obtain the optimal reward weight parameters and applying them to the reward function; Step 4, introducing external disturbances into the BUCK circuit control environment and training the DQN agent using the optimized reward function. This invention does not require a precise mathematical model and can achieve adaptive optimization control of the BUCK circuit under different operating conditions, demonstrating significant engineering application value.

The invention discloses a Beidou / 5G fusion positioning terminal for indoor and outdoor seamless switching, and belongs to the technical field of satellite navigation and wireless communication fusion positioning. Comprising a main control module, a Beidou positioning module, a 5G communication positioning module, an IMU inertial measurement module, a storage module, a power supply module and a positioning result output module. The master control module comprises a data preprocessing unit, a three-level vertical feature extraction unit, a fusion positioning decision model unit and an indoor and outdoor seamless switching execution unit. Through a depth feature extraction system of a three-level vertical processing architecture, progressive mining from a shallow single-source feature to a middle scene association feature and then to a deep switching decision feature is realized, and a vertical processing rule that a second-level feature is based on a first level and a third-level feature is based on a second level is strictly followed, so that cross-level feature interference is avoided, and the processing efficiency is improved. The depth and accuracy of multi-source data feature mining are greatly improved, and high-reliability feature support is provided for scene recognition and switching decision.

The application discloses a kind of gene variation evidence mining method and system based on multi-stage trusted inference, the method includes: obtaining academic literature to be parsed, literature is structured and semantically blocked;Respectively extract gene variation experimental evidence in table data and text text, construct dual-source evidence set;Based on gene variation identification, complete heterogeneous evidence fusion, and evidence is standardized processing;Combined with original text accurate matching and semantic checking mechanism, evidence is carried out traceability checking, abnormal identification and no illusion accurate correction;Further, cross-fragment evidence is globally aligned, deduplicated and fused, to generate standardized, high-precision, traceable gene variation functional evidence dataset;The application can effectively solve the problems of unstable long text analysis, evidence missed detection, data distortion and model illusion in the prior art, and improve the accuracy and engineering application capability of gene variation evidence mining.

This invention relates to the field of sliding bearing simulation and optimization technology, and more specifically to a high-precision simulation and optimization design method for hydrostatic thrust sliding bearings. It establishes an efficient decoupling model of the fluid-solid-thermal three-physics field of the hydrostatic thrust sliding bearing, decomposing the overall thermodynamic field into two subfields, which are coupled with the fluid force field and the solid force field, respectively. Then, it constructs a gap oil film flow / thermal field model and a friction pair deformation solid / thermal field model, achieving equivalent coupling through data transfer. The oil film control equations are iteratively solved to obtain oil film pressure, temperature, and flow velocity; the friction pair equations, including elastic constitutive and thermal strain equations, are solved to obtain their elastic and thermal deformations, with bidirectional data transfer achieving coupled iteration. Based on the above method, simulations are performed, establishing sub-models of oil film flow and friction pair deformation, using local interpolation and dynamic meshing to process the interface; the oil cavity is optimized based on the simulation results, parametric analysis is used to establish the mapping between structural parameters and lubrication performance, the objective function and constraints are defined, and the optimal oil cavity structure is obtained through algorithms.

The application discloses a rotor unbalance state evaluation method based on a rotor free stop phase angle, and is characterized in that the method comprises the following steps: step S1: rotor phase marking based on an engine assembly process; step S2: monitoring an engine operation state, and continuously recording rotor free stop phases of multiple marking points based on the marked rotor phase; step S3: analyzing and counting a characteristic parameter alpha value of the multiple marking points based on the recorded rotor free stop phases; and step S4: evaluating a rotor unbalance state according to the analyzed characteristic parameter alpha value, and outputting an evaluation result.

This invention relates to the field of energy-saving technology for marine propulsion systems, specifically to a high-efficiency energy-saving device for a refrigerated ship propeller hub cap fin with tilt angle adjustment. It aims to address the industry pain points of existing hub cap fins, such as poor adaptability to various operating conditions, easy failure of dynamic seals, and insufficient reliability. The device includes a propeller hub body and a hub cap fixed coaxially. Multiple sets of rotating shafts with fins are arranged in a ring array on the outer side of the hub cap, and rotating bevel teeth are fixed at the inner ends of the rotating shafts. The propeller hub body houses a magnetically coupled resonant wireless power supply unit, a telescopic device, and a rotating mechanism. The rotating mechanism drives a drive bevel tooth that meshes with the rotating bevel tooth, and the drive bevel tooth is coaxially connected and locked via a connecting shaft. This invention achieves synchronous adjustment and reliable locking of the fin angle through dual-position interlocking switching. It eliminates the risk of through-type dynamic seal failure by eliminating the need for penetrating motion transmission components, making it suitable for all operating conditions of refrigerated ships and the harsh marine environment of the open ocean. It offers a long maintenance-free period and high navigation safety.

The application discloses a kind of corn raw grain procurement processing and finished product grain transportation collaborative optimization methods under carbon constraint, the method includes: corn processing enterprise supply chain cost and different transportation mode carbon emission accounting mode;Comprehensively consider the seasonality fluctuation of raw grain supply, inventory loss, and different transportation mode and its packaging form in direct transport and reconditioning link timeliness, cost, damage, carbon footprint and other factors, construct mathematical model with corn processing enterprise supply chain cost minimization as target;To the constraint rules of purchase, processing, inventory, shortage, transportation and carbon emission management in target mathematical model, and linearization is handled to constraint condition;Based on the branch and price algorithm of fusion dynamic programming, obtain the collaborative optimization scheme of corn raw grain procurement processing and finished product grain transportation under carbon constraint.The present application solves the technical problems that the existing method is difficult to coordinate for the inventory loss of finished product grain, the transportation loss of multimodal transport and the carbon emission in transportation link.

This application discloses a suitability testing method for ultrasonic machining of honeycomb core materials, belonging to the field of honeycomb core material machining technology. The suitability testing method includes: determining the critical conditions of the process based on material properties and quality requirements; establishing a mapping relationship between material parameters and cutting forces and inversely calculating the critical cutting force; subsequently calculating the motion state and impact force of the tool under ultrasonic vibration; and finally, determining whether ultrasonic machining can be used based on the comparison between tool displacement and material fracture limit, and impact force and critical cutting force. This application can quickly determine the suitability of ultrasonic machining for new material honeycomb core parts without requiring extensive trial cutting experiments, effectively saving time and costs, and improving the application efficiency and accuracy of parameter determination of ultrasonic processes.

The invention discloses a preparation method of an aluminum foil blank for a high-strength 8021 battery soft package, and belongs to the technical field of aluminum alloy processing. According to the method, an 8021 alloy cast ingot serves as a raw material, hot rolling, segmented cold rolling, medium and low temperature intermediate annealing, finish rolling and other procedures are carried out, and the high-strength aluminum foil blank is prepared by optimizing rolling reduction distribution and roll surface roughness. The tensile strength of the obtained blank is larger than or equal to 170 MPa, the plasticity and the surface quality are excellent, the process energy consumption is lower, equipment does not need to be newly added, and the method is suitable for industrial production of the aluminum foil blank for the high-end soft package lithium battery aluminum-plastic film.

The application belongs to the technical field of electronic power, and particularly relates to a multi-target hybrid modulation method, system and device of a T-type three-level inverter. The method firstly judges whether the inverter is operated at an extremely low modulation degree; if yes, the theoretical midpoint currents of DPWM1 and DPWM3 modes are calculated in parallel, and the extended midpoint currents corresponding to four hybrid modulation modes generated after the introduction of VSVPWM sequences in the two modes are calculated; then, according to the numerical relationship between the real-time calculated midpoint compensation current and the midpoint currents, the optimal hybrid modulation mode is dynamically selected or the basic DPWM mode is returned to; finally, three-phase double modulation waves are generated, and PWM signals for driving all power switches are output after carrier comparison. The application realizes multi-target optimization of the midpoint potential, switching loss and output waveform under extreme working conditions, and significantly improves the system stability and performance.

This invention relates to the field of rolling bearing testing technology, specifically to a method for measuring bearing radial clearance. By acquiring three-dimensional coordinates at multiple points in the non-stressed area of ​​the bearing outer ring, the center position of the bearing outer ring is accurately determined using the least squares method. This effectively avoids measurement errors A caused by the measuring force of the measuring instrument itself leading to local elastic deformation of the bearing outer ring, and measurement errors B caused by radial loading leading to local elastic deformation of the bearing outer ring. Simultaneously, a zero-radial-clearance bearing is introduced as a calibration benchmark. Combined with measured data, the method identifies radial clearance changes caused by local deformation that are undetectable by traditional methods, and compensates and corrects the radial clearance measurement results accordingly. This not only avoids measurement errors caused by local deformation due to the measuring force of the measuring instrument itself in principle, but also significantly improves the accuracy and reliability of bearing radial clearance measurement, achieving high-precision and repeatable testing.

The invention discloses an LCL integrated filter based on transformer leakage inductance multiplexing. The LCL integrated filter comprises a discrete inductor, a transformer side inductor and a filtering branch, the transformer side inductor is the inherent leakage inductor of the transformer, and the filter branch comprises a filter capacitor C. According to the LCL filter, the inherent leakage inductor of the transformer is multiplexed to serve as topological optimization of the power grid side inductor and the filtering branch, and the technical problems that a traditional LCL filter needs to be independently provided with the power grid side inductor, so that the size is large, cost is high, the inherent leakage inductor of the transformer is idle and wasted, instability is likely to happen under a weak power grid, and harmonic suppression cannot meet the grid connection standard easily are solved.

This invention discloses a tracked mobile carrier based on highway guardrails, comprising a carrier body, a walking drive mechanism, and a support and guiding mechanism. The walking drive mechanism includes an outer drive-clamping track assembly and an upper drive-clamping track assembly located on both sides of the corrugated guardrail, forming a bidirectional clamping drive structure. The two tracks are in surface contact with the corresponding surfaces of the corrugated guardrail to provide walking drive force. Both track assemblies are equipped with adaptive adjusting slides with pre-tension springs, enabling adaptive inward and outward floating of the driven wheel assembly to continuously maintain the track's contact with the corrugated guardrail. The support and guiding mechanism includes a top support roller and a lower limiting roller with a buffer spring, achieving vertical load-bearing and lateral limiting and buffering. This invention enables highly stable and reliable all-weather autonomous movement along highway corrugated guardrails, adapting to various complex working conditions of the guardrails and meeting diverse application needs such as highway maintenance and inspection, and traffic control.

The invention relates to the technical field of material surface treatment, in particular to a method for plating molybdenum on the surface of diamond powder through salt bath. By optimizing the components of the molten salt, the molten salt forms low-melting-point eutectic in the salt bath molybdenum plating process, molybdenum powder and carbon atoms on the surface of diamond are subjected to a chemical reaction at the low plating temperature to generate a uniformly-distributed molybdenum carbide layer, and heat damage to the diamond can be effectively avoided. Besides, the molten salt can form a high degree of superheat, diffusion of molybdenum atoms and chemical reaction of the molybdenum atoms and carbon atoms are facilitated, it can be guaranteed that the coating completely covers the surface of the diamond and is firmly combined, and it can be guaranteed that the thickness of the plating layer is controllable by regulating and controlling the plating time; the chlorate powder is non-toxic, harmful substances are not generated in the salt bath molybdenum plating process, the environment-friendly property is good, the chlorate powder can be used for industrial production, and the chlorate powder has wide application prospects and high engineering application value.

The application discloses a gear NVH anchor point calibration and duty cycle switching cooperative optimization system and method, comprising: under selected torque, parameter value exhaustive combination is carried out on selected micro parameters of a gear transmission system, a plurality of micro parameter groups composed of micro parameters with different values are obtained, and a corresponding full-condition weighted peak-to-peak transmission error comprehensive evaluation value is obtained according to a peak-to-peak transmission error simulation curve and a peak-to-peak transmission error reference curve of the gear transmission system under each micro parameter combination, and the micro parameter group with the minimum full-condition weighted peak-to-peak transmission error comprehensive evaluation value is selected as the optimal anchor point reference parameter group for the gear transmission system to run; when the demand torque of the gear transmission system is in a preset optimization torque interval when the gear transmission system runs with the optimal anchor point reference parameter group, a two-point high-frequency duty cycle switching mode is executed.

The invention discloses an urban water supply system anti-seismic performance evaluation method considering function decline of a water treatment plant, and belongs to the technical field of urban water supply pipe network anti-seismic analysis. The method comprises the following steps that Monte Carlo simulation sampling is utilized, and earthquake damage scenes of various pipelines in the water supply network under the earthquake action are generated; quantitatively determining the residual function state of the water treatment plant after the earthquake through the vulnerability curve of the water treatment plant; establishing a coupling analysis model of the water treatment plant and the water supply pipe network to realize coupling of function changes between the water treatment plant and the pipe network; and calculating the anti-seismic performance index of the urban water supply system by integrating the pipe network damage and the water plant function attenuation. According to the method, the water plant vulnerability curve is introduced to quantitatively reflect the function attenuation of the water plant after the earthquake, and the water plant-pipe network coupling model is combined, so that the overall anti-seismic performance of the urban water supply system under the earthquake disaster can be more scientifically and comprehensively evaluated; the method has important theoretical and practical significance for guiding seismic design, toughness improvement and emergency decision-making of the urban water supply system.

This invention proposes a gated time-series graph fusion-based learning signal modulation recognition method, relating to the fields of pattern recognition and intelligent information processing. The method uses the input IQ signal as the basic input, introduces an amplitude-phase adaptive filter, and adaptively emphasizes frequency band features relevant to discrimination under the supervision of the modulation recognition task through parameterized modeling of the spectral amplitude and phase. In the discrimination stage, a classification framework combining temporal and structural modeling is adopted to mine complementary discrimination information of the modulation signal in both the temporal and structural dimensions. A long short-term memory network is used to perform temporal modeling of the signal frame sequence, capturing the dynamic evolution characteristics of the modulation signal. Based on the fusion features, a dynamic graph adjacency matrix that simultaneously reflects local time-delay correlation and global similarity is constructed, and inter-frame correlation information is propagated through a graph neural network. Finally, a hierarchical discrimination fusion module adaptively integrates embedding vectors from different layers of the temporal and graph branches.

The invention relates to the technical field of robot navigation, and discloses a robot navigation method and device based on multi-source data fusion, and the method comprises the steps: carrying out the fusion processing of the multi-source data of a robot, obtaining the environment map of the current environment of the robot and the pose of the robot, and obtaining the position of the robot according to the visual image data and the laser point cloud data; obstacle information in the current environment is recognized, the optimal path of the robot from the current position to the target end point position is determined according to the environment map, the robot pose and the obstacle information, a motion control instruction is generated according to the optimal path, and the robot is controlled to run according to the motion control instruction. Therefore, by implementing the method, the determination accuracy of the map and the pose of the robot can be improved through multi-source data fusion, the obstacle recognition accuracy is improved, and the stability, high efficiency and safety of autonomous navigation of the robot are further improved.

This invention provides a spatiotemporal collaborative decomposition method for berth allocation and quay crane scheduling problems under non-stationary conditions, comprising: Step 1, preprocessing historical port vessel data; Step 2, constructing an equivalent service capacity model under continuous coastline conditions; Step 3, modeling the port system as a time-varying queuing system; Step 4, extracting system state indicators to characterize operational features; Step 5, constructing a time decomposition mechanism; Step 6, constructing an adaptive spatial decomposition model to divide the continuous coastline and allocate quay crane resources; Step 7, obtaining a global scheduling scheme by integrating the solutions to sub-problems. This invention effectively alleviates the "curse of dimensionality" caused by the rapid expansion of the state space and decision space under non-stationary conditions, significantly reduces the difficulty of model solving, and shows varying degrees of reduction in the average time in port, while improving overall computational efficiency and making the algorithm convergence process more stable.

The invention provides a topological optimization design method of a multi-physics field multi-performance integrated lattice structure, and belongs to the field of aerospace lightweight and multifunctional structure design. According to the method, a dot matrix unit cell design domain is taken as an object, and a multi-objective optimization model including mechanical properties, heat-conducting properties, electrical conductivity and mass transfer and transportation capabilities is established based on normalization and a weight method for the comprehensive requirements of an aerospace structure under multi-field coupling service conditions such as bearing and heat management; and multi-performance collaborative improvement is realized under a multi-constraint condition, so that the engineering adaptation problems of function splitting, difficulty in coordination of performance conflicts and difficulty in meeting complex service conditions caused by design only based on a single physical field or a single performance index are avoided. The problem of function splitting caused by a traditional single physical field design is solved, the multi-performance integrated design of bearing / heat / electricity / mass transfer and the like is achieved, and the method is suitable for engineering implementation of high-performance lattice structures in the fields of aerospace and the like.

The invention relates to a method for positioning a 5G NR base station based on AOA and a least square method, and the method comprises the following steps: carrying out the selection of a test point, and collecting data; coordinate transformation is carried out, and an azimuth line is constructed; performing initial estimation to obtain an initial estimation value or an initial estimation area of the position of the target base station; and fitting the accurate position of the target base station by adopting a least square method. By adopting the method and the device for positioning the 5G NR base station based on the AOA and the least square method, the processor and the computer readable storage medium, the method and the device have initiative, do not depend on an engineering parameter table of an operator, and can actively discover and position an unknown base station; the method has high precision, adopts AOA direction finding and least square method optimal estimation, effectively balances the error of each measurement point, has practicability, is clear in process, can be realized by depending on the existing professional equipment such as a frequency sweep instrument, a directional antenna and a GPS antenna, and has relatively high engineering application value.

The invention provides a power prediction method for a solar cell array of an all-electric propulsion satellite. The power prediction method comprises the following steps: establishing a ground output power prediction model; the method comprises the following steps: calculating an output current predicted value of a solar cell array by using a ground model in combination with related telemetry data of the solar cell array during satellite electric propulsion orbital transfer and predicted ground radiation dose and sun-ground factors, and calculating a current predicted deviation coefficient reflecting a radiation dose predicted deviation through an output current telemetry value and the predicted value; the method comprises the following steps: periodically collecting telemetry data of a plurality of working points, and calculating to obtain an average current predicted deviation coefficient; and acquiring a preliminary predicted value of the output current at a future moment by using a ground model, correcting by using the average deviation coefficient to obtain a corrected output current predicted value, and multiplying the corrected output current predicted value by the fully adjusted bus voltage to obtain the total predicted output power of the solar cell array. In this way, the ground model is dynamically corrected by fusing the in-orbit measured data, and the power prediction precision of the all-electric propulsion satellite in the complex orbital transfer process is improved.

This invention relates to the field of optical fiber geometric reconstruction technology, specifically to a one-dimensional reconstruction analysis method for optical fiber ring stress, comprising: reading and processing finite metadata of the optical fiber ring to obtain Cartesian coordinate system node information; converting the Cartesian coordinate system node information to the cylindrical coordinate system; determining the forward and reverse optical fiber paths to which each node belongs; calibrating the length of each node based on the optical fiber winding process parameters and the optical fiber path to obtain the one-dimensional unfolded length of each node; interpolating based on the one-dimensional unfolded length to obtain a discrete point stress value sequence; conducting an actual optical fiber ring experiment based on the finite metadata to obtain a polarization coupling curve; comparing the polarization coupling curve with the discrete point stress value sequence to obtain the one-dimensional reconstruction accuracy result; this invention can improve the accuracy of optical fiber ring stress analysis results along the optical path.

The invention belongs to the technical field of image processing, and particularly relates to a rotor wing flight simulator wide area visual channel edge overlapping fusion method, system and device. The method comprises the following steps: acquiring single-channel image data of a multi-view channel; identifying and marking the independent display area and the edge overlapping area through comparison; defining internal and external transition zones at the boundary of the overlapping region; extracting brightness data of each area and establishing a brightness database; and performing brightness adaptation processing on the edge overlapping region and the transition zone based on the database, adjusting the brightness by calculating an adaptation coefficient based on the brightness mean value of adjacent regions, so that the edge overlapping region and the transition zone are in smooth transition with the independent display region, and finally generating a seamless fusion wide-area visual image. According to the method, hardware adjustment is replaced by a software algorithm, and the problems of uneven brightness and unnatural fusion of multi-channel projection are effectively solved.

The invention discloses an LNG loading valve identification method and system oriented to complex conditions. The invention provides a light weight multi-scale detection network LitePAN (Personal Area Network)-YOLO (Young Line Occurrence). The network is constructed based on a YOLOv8 framework, an improved Vanilla module is introduced to enhance the nonlinear expression ability of features, a C2f-ParNet module is designed to enhance cross-layer gradient flow and channel interaction, efficient multi-scale feature fusion is achieved, and a GSConv-based LitePAN structure is adopted in a neck structure to reduce the calculated amount and maintain the feature aggregation efficiency. Meanwhile, the detection head adopts an anchor-free decoupling design, and a WIoU v3 loss function is combined to improve bounding box regression and high IoU positioning precision. Compared with YOLOv8-n, the LitePAN-YOLO has the advantages that the parameter quantity of the LitePAN-YOLO is reduced by 7.7%, the calculated quantity is reduced by 5.9%, and mAP50-95 and mAP95 are respectively improved by 0.59% and 1.30%. The method shows higher robustness and detection precision under the scenes of dense small targets, large attitude change and complex illumination, and provides reliable visual perception support for loading and unloading operation of the LNG collaborative robot.