60results about How to "Guaranteed robustness" patented technology

This invention relates to a constrained automated optimization method for semiconductor measurement, comprising: step 3) optimizing auxiliary parameters using a training set; step 4) selecting one or more second candidate parameter combinations to form a candidate parameter combination pool, and saving the regression coefficient set corresponding to each second candidate parameter combination; step 5) using the auxiliary parameters in the optimal second candidate parameter combination as independent variables, using the corresponding regression coefficient set as the initial coefficients of the equation, and the key parameters as dependent variables to establish a second multiple linear regression equation; performing the first-stage optimization and the second-stage optimization to select the optimal predicted values ​​and optimal regression coefficients of the key parameters; and step 6) using a test set to perform the first-stage optimization and the second-stage optimization to verify the optimal candidate parameter combination and its corresponding optimal regression coefficients. The two-stage optimization strategy fundamentally solves industry pain points such as difficulty in convergence, difficulty in reproducibility, strong subjectivity, and long processing time.

The application discloses a lightweight indoor 3D laser ground extraction algorithm method, system and medium. The method comprises the following steps: acquiring indoor 3D laser radar data; obtaining coordinate values based on a laser radar coordinate system according to the indoor 3D laser radar data; sending the coordinate values to a preset plane equation to obtain a ground plane normal vector; sending the ground plane normal vector to a first preset coordinate system to establish a constraint to obtain a comprehensive error value; and accumulating squares of values in the comprehensive error value to obtain a comprehensive error sum of squares; extracting a minimum value in the comprehensive error sum of squares, and setting a coordinate point corresponding to the minimum value as an optimal current pose. The application is suitable for various lasers, and by giving a preset threshold, whether it is 16 lines, 32 lines or 128 lines, the algorithm accuracy and robustness are ensured, and the accuracy of recognized ground points is improved.

The application relates to the technical field of hybrid solid state disks, and discloses a hybrid solid state disk dynamic optimization method and device based on a large language model, which systematically classifies the parameter space of a hybrid solid state disk, constructs a parameter sensitivity matrix based on the classified parameters, determines the influence difference parameters of the parameters on performance under different workloads, collects system state information of the hybrid solid state disk in real time, generates an optimized parameter configuration scheme based on a large language model according to the system state information, the influence difference parameters and historical configuration information. The application can realize real-time perception of work load changes and dynamic focusing on the most sensitive parameter combination under the current load, realize collaborative reasoning and adaptive adjustment among multiple parameters, significantly reduce the write amplification rate and operation delay, and solve the problem that the response efficiency, system stability and energy consumption performance of the hybrid SSD are poor under a dynamic work load in the prior art.

This application relates to an event-driven method and computing device for recognizing revenue from intra-group transactions. The method includes: acquiring a transaction order and determining its type; if the transaction order is an intra-group transaction order, acquiring its triggering event; if the triggering event is a key business completion event, generating a cost recognition voucher for the purchaser of the transaction order and a revenue recognition voucher for the seller of the transaction order; wherein the key business completion event represents a business operation event indicating that the relevant performance obligations have been fulfilled. This achieves a fundamental shift from employee-based identification and triggering to system-based perception and response, improving the timeliness and accuracy of financial processing, thereby reducing operating costs and increasing processing efficiency.

This invention discloses an adaptive inertial navigation solution method and system based on motion feature recognition, comprising: extracting multidimensional features reflecting the motion state of the carrier based on angular velocity and specific force data collected by the inertial measurement unit, and constructing a feature matrix; performing dimensionality reduction processing on the feature matrix to obtain low-dimensional feature vectors; performing cluster analysis based on a clustering algorithm that initializes cluster centers using prior samples to identify the motion state of the carrier; and adaptively selecting an appropriate combined navigation filtering strategy for navigation solution. This invention effectively suppresses the accumulation of inertial navigation errors, improving the autonomous navigation accuracy and long-term stability of the system in scenarios of satellite signal rejection or visual failure; it can finely distinguish different motion modes such as hovering, uniform motion, and transitional states, avoiding the shortcomings of traditional threshold methods that are susceptible to noise interference and have poor adaptability; by introducing continuous frame mode voting and physical anti-jump mechanisms, it further eliminates single-frame misjudgments, ensuring the continuity and robustness of state recognition.

The application discloses an internet hospital hierarchical diagnosis and treatment and resource scheduling method and system, the method comprises the following steps: in response to the diagnosis and treatment request sent by the user, obtaining the diagnosis and treatment data containing the patient's complaint information; inputting the diagnosis and treatment data into the disease assessment model for feature extraction, and outputting a feature vector representing the patient's disease state; the feature vector at least includes a first dimension parameter representing the emergency degree of the disease and a second dimension parameter representing the complexity of the disease; based on the feature vector, determine the target resource node in the hierarchical medical resource database; according to the target resource node, generate a resource scheduling instruction, and establish a communication link between the user terminal and the corresponding server of the target resource node. The application introduces a two-dimensional evaluation system of emergency degree and complexity, combines a dynamic threshold adjustment mechanism based on supply and demand prediction, and significantly improves the utilization efficiency of medical resources and the accuracy of diagnosis matching.

The application discloses a frame structure design method of an AFDM system based on adaptive asymmetric guard bands, solves the problem that the frame structure design of the AFDM system under a high dynamic channel is difficult to simultaneously consider the bit error rate of a receiving end and spectral efficiency, and belongs to the technical field of information and communication.The application comprises the following steps: performing frame-level reliability determination according to a demodulation result vector of a current frame, when a corresponding threshold is satisfied, the current frame accesses an adaptive updating process; on the current frame that accesses, candidate pairs of a time delay index and a Doppler index are obtained, an effective path set of the current frame is screened out, a maximum Doppler spread priori and a Doppler energy directionality priori are extracted from the effective path set, a total guard band width of a next frame and a smooth direction factor are determined, asymmetric distribution of the lengths of guard bands on the left and right sides of a pilot is carried out according to the total guard band width of the next frame, the guard bands are relatively widened on the side of a Doppler energy concentration and are relatively shortened on the side of energy sparseness, and a next frame adaptive asymmetric guard band and a data area layout are generated.

This invention discloses an intelligent adaptive dose verification calculation method and system. The intelligent adaptive dose verification calculation method includes: obtaining preliminary dose distribution data for a patient based on a deep learning-based dose prediction model; dynamically assigning variance reduction parameters to particles in a Monte Carlo simulation based on the high and low dose distribution information in the preliminary dose distribution data, wherein the variance reduction parameters include simulation weights, with simulation weights assigned to particles predicted to be in high-dose regions being less than those assigned to particles predicted to be in low-dose regions; and using a Monte Carlo dose calculation engine, simulating the planned beam based on the dynamically assigned variance reduction parameters to obtain high-precision verification dose distribution data. This invention's intelligent adaptive dose verification calculation method ensures the physical authenticity of the final verification dose distribution and the credibility of the "gold standard."

This invention belongs to the field of vehicle automation technology, specifically relating to a machine vision-based method for locating target objects in vehicles. A camera on the vehicle acquires scene images of the target object, a neural network is used to obtain the center pixel coordinates of the target object in the scene image, and the distance between the vehicle and the target object is calculated using the camera's height from the ground and the camera's intrinsic parameter matrix. This invention is mainly applied to target object location during the retrofitting of old vehicles. By combining target object identification in a neural network with camera homography calibration, the relative position of the target object in 3D space is accurately calculated, enabling precise target object location even in complex scenes.

The motion quality evaluation method based on motion perception and double-flow space-time graph convolution relates to the motion quality evaluation field, solves the problems that the existing rehabilitation motion quality evaluation method is difficult to accurately model the cooperation between joints, the fixed joint grouping strategy is not flexible enough, the model has insufficient recognition of the detail differences of complex rehabilitation motions, and the reliability of the evaluation result is affected, a dynamic joint grouping strategy based on motion amplitude driving is provided, a double-flow STGCN architecture is designed, the position and orientation information of the joints are processed respectively and then fused, a two-stage self-attention module SADG is developed, the inter-group time sequence mode and the intra-group spatial relationship are modeled in stages, and multi-granularity feature interaction fusion is realized. The attention mechanism introduces a constraint guided by the motion amplitude, so that the attention weight is more interpretable and has physical meaning.

The application provides a low-altitude terrain following flight path guiding method for unmanned helicopters, and is based on an enhanced vector field histogram algorithm and on-board laser radar spatial gridding data, and proposes a low-altitude terrain following flight path guiding method for unmanned helicopters in a three-dimensional space, introduces unmanned aerial vehicle flight performance constraints and safe flight standards, constructs a safe flight pipeline model for unmanned helicopters, designs a laser radar gridding detection information partial loss filling model based on a first-order model, fully utilizes the detection information of the on-board laser radar, guarantees the robustness of algorithm solving and the safety of calculation results, and meets the obstacle avoidance requirements in a three-dimensional space in a complex and severe flight environment. According to the on-board laser radar detection data and the safe flight constraints of the unmanned helicopter, the application can realize the calculation of optimal flight waypoints for the low-altitude terrain following flight path guiding of the unmanned helicopter.

The application discloses a multi-agent dynamic real-time path planning method and device based on signal timing logic, and relates to the technical field of control methods. The method comprises the following steps: acquiring multi-robot path planning task information and a preset space-time constraint task demand; constructing a multi-robot task specification based on signal timing logic, and defining a corresponding robustness evaluation function; constructing a comprehensive path cost function based on the robustness evaluation function; introducing the path cost function into a path expansion and reconnection process based on a real-time rapid expansion random tree algorithm, updating a search tree structure, and generating a candidate path meeting environment constraints and timing constraints; planning paths for robots in sequence according to a preset multi-robot cooperative planning strategy, taking the path information of the robots that have been planned as dynamic obstacles and performing progressive cooperative obstacle avoidance, and outputting a conflict-free trajectory adaptive to a dynamic environment. The application can solve the path planning problem in a multi-agent scene.

This invention discloses a global collaborative prediction method for traffic signals based on quantum entanglement states. The method includes: establishing a quantumized model of the dynamic evolution path of the target traffic network within a future time window; constructing a spatiotemporal graph model; mapping the signal phase states of intersections to spatiotemporally integrated quantum states using a hierarchical quantum encoding scheme; constructing a Hamiltonian containing spatial coupling, temporal evolution, and cost terms based on a preset traffic optimization objective; solving for its ground state using a hybrid quantum-classical computation method to determine the optimal dynamic evolution path; encoding the current actual state of the traffic network as an initial state; decoding the collaborative signal control strategies for future time points using the optimal evolution operator and sequential measurements; and executing these strategies through a rolling time-domain control framework. This invention utilizes quantum parallelism for global optimization, enabling the acquisition of the optimal collaborative strategy across the entire spatiotemporal range in a single step, thereby improving the overall operational efficiency, predictability, and robustness of the traffic network.

The invention relates to a high-power green laser synthesis system based on a rotary frequency doubling crystal and multiple beams of infrared pulse laser, and belongs to the technical field of laser. The system comprises a multi-beam pulse infrared laser device, a frequency doubling crystal module, an annular rotating mechanism, a beam combining device, a continuous infrared laser device, a laser frequency regulation and control device and a rotating driving control system. The four beams of pulse infrared laser sources are symmetrically arranged along the circumferential edge, and periodically staggered output is realized through frequency regulation and control; the multiple frequency doubling crystals are distributed along the rotating circular ring at equal intervals and are sequentially aligned with laser pulses under driving of the rotating mechanism, and efficient frequency doubling is achieved. The center continuous infrared laser generates stable green light through frequency doubling of the fixed crystal. Finally, the green light generated by the multiple frequency doubling crystals and the central green light are integrated into a single light beam through a beam combining device to be output. According to the system, the power upper limit and the stability of green laser are remarkably improved through energy sharing of polycrystals, rotary dynamic matching and annular cooling, and the problems that existing green laser is limited in power and serious in heat effect are solved.

The invention relates to the technical field of power transaction strategy generation, and discloses a power transaction strategy generation method based on load prediction and profit fusion, which comprises the steps of inputting specified historical data and real-time data, and forming a feature set after completing data preprocessing through data cleaning, feature engineering, dimension reduction and standardization; a two-stage algorithm of improved LSTM basic prediction and real-time signal correction is adopted, and a dynamic load prediction value, a confidence interval and prediction error probability distribution are output; constructing a dynamic profit formula by taking profit maximization as a target, optimizing transaction electric quantity and opportunity through DQN reinforcement learning, and outputting an optimal transaction strategy and a price threshold value; pushing a strategy and generating a visual report, collecting actual transaction data, adjusting parameters according to deviation and profit conditions, and finely adjusting the model every day to guarantee long-term stability. Dynamic cooperation of load prediction and a profit target is realized, and the real-time performance, the economical efficiency and the robustness of a power transaction strategy are improved.

This invention provides a calibration-free path reconstruction method for buried cables based on Topo-NeRF and active sensing. Addressing the issues of traditional magnetic field inversion relying on calibration and being prone to mismatch in environments with strong interference such as reinforced concrete, the method preprocesses raw data to obtain a spatiotemporal feature matrix. This matrix undergoes feature extraction, spatial alignment, and drift compensation to eliminate extrinsic parameter calibration, resulting in multi-source fused data. This multi-source fused data is combined to characterize the cable's magnetic field using a continuous three-dimensional topological manifold. An implicit neural radiation field is constructed and coupled with differentiable electromagnetic rendering. The cable body and branch nodes are separated within a persistently cohomologically constrained topological bottleneck layer. The initial reconstruction results are combined with conditional diffusion and ground-penetrating radar dielectric priors to enhance low signal-to-noise ratio weak fields, ensuring consistency between the reconstruction results and the underground physical structure. Based on reconstruction uncertainties, topological entropy is calculated, and online acquisition trajectories are planned to form a closed-loop active sensing system. Finally, a three-dimensional path model is output, enabling rapid detection of underground cables.

The application belongs to the technical field of intelligent breeding, and provides a sow leg automatic scoring method and device based on three-dimensional point cloud. The method comprises the following steps: acquiring a target three-dimensional point cloud model of a pig; extracting a leg point cloud set corresponding to each leg of the pig from the target three-dimensional point cloud model; determining leg key point data of each leg from the leg point cloud set corresponding to each leg; determining trunk key point data from the target three-dimensional point cloud model according to the leg key point data; determining a posture determination result of the pig according to the target three-dimensional point cloud model, the leg key point data of each leg and the trunk key point data; and in the case that the posture determination result indicates that the pig is in a standard standing posture, determining a leg structure type and a leg structure score of the pig according to the leg key point data. The application can stably identify the leg structure characteristics of the pig and realize consistent automatic scoring, and has good application potential in terms of accuracy, robustness and interpretability.

The application relates to the technical field of blockchain security and discloses a data security encryption method and system based on a blockchain. The method comprises the following steps: obtaining to-be-transmitted plaintext data of a source chain and performing digital signature to obtain initial lattice basis ciphertext; obtaining target lattice parameters of a target chain, performing security level adaptation analysis on the source chain lattice parameters and the target lattice parameters, and obtaining a cross-chain conversion strategy; performing homomorphic key switching operation on the initial lattice basis ciphertext by using a cross-chain conversion key according to the cross-chain conversion strategy, performing modulus switching operation based on target modulus, and obtaining target compatible ciphertext; generating zero-knowledge correctness proof based on a conversion path, encapsulating the zero-knowledge correctness proof and the target compatible ciphertext, and obtaining cross-chain transmission data packets; and performing decryption and integrity check after the zero-knowledge correctness proof is verified to be correct on the target chain, and obtaining cross-chain interaction confirmation results. The method can realize low-noise cross-chain secure data transmission resisting quantum privacy.

This invention relates to a blockchain-based trusted collaborative control method and system for vehicle platoons, belonging to the field of intelligent connected transportation technology. The method includes steps such as vehicle state perception and data broadcasting, RSU node data verification and blockchain uploading, vehicle comprehensive trust value calculation and updating, predictive control compensation mechanism, abnormal behavior detection and malicious node isolation, and dynamic optimization and adjustment of the platoon structure. The system executes this method, ensuring data consistency and immutability through blockchain technology to enhance information security; utilizing trust value calculation and anomaly detection mechanisms to guarantee platoon security; and dynamically adjusting the platooning order based on priority scoring to improve collaborative control performance. This invention achieves high security, high real-time performance, and high robustness in collaborative vehicle platoon control under complex conditions, and has significant application value.

This invention proposes a health management method for multi-stack fuel cell power generation systems based on hierarchical collaborative optimization, belonging to the field of fuel cell technology. The method includes the following steps: Based on the operating parameters of the multi-stack fuel cell power generation system, a degradation model for the fuel cell stack and a capacity decay model for the battery energy storage system are established to obtain the lifespan cost of the fuel cell stack and the battery energy storage system; at a first time scale, based on the hydrogen consumption cost, the lifespan cost of the fuel cell stack, and the lifespan cost of the battery energy storage system, a rolling optimization problem is modeled to obtain a power reference trajectory and a state-of-charge reference trajectory for the battery energy storage system; at a second time scale, a Tube-model predictive control is used to construct an invariant tubular set and tighten the constraints, and an actual control law is generated based on the load prediction error for real-time rolling optimization to correct the power reference trajectory and the state-of-charge reference trajectory.

This invention relates to the fields of digital twin, chemical process control, and industrial intelligent management, specifically to a digital twin-based MOC intelligent identification and process analysis management system. The system includes a data acquisition module, an intent parsing and injection module, a topology map construction module, a spatiotemporal transmission simulation module, and a health assessment and feedback module. It acquires change intent data, digital twin base data, and real-time operational data; extracts the change content as parameterized features and generates micro-perturbation vectors, injects them into target topology nodes, constructs a cross-domain process topology map, iteratively generates predicted values ​​for each process parameter according to a preset time step, compares these values ​​with safety critical thresholds, and outputs a global health assessment result. When there is a risk of exceeding limits, it calculates the adjustment amount in reverse and generates a process compensation scheme; when safe, it outputs normal operation confirmation information, achieving early identification and compensation guidance for future risks of local changes.

This application provides an optimization method and related equipment for new energy vehicle charging scenarios on highways, belonging to the field of strategy optimization technology. The method includes: acquiring static basic data and dynamic real-time data of highway service areas; generating multiple candidate charging guidance strategies in a discretized decision space based on the static basic data and dynamic real-time data; rapidly evaluating each candidate charging guidance strategy to obtain a three-dimensional objective vector corresponding to each strategy; the three-dimensional objective vector includes a power grid load stability objective, a user waiting time objective, and a traffic congestion objective; selecting the optimal solution set from all candidate charging guidance strategies based on a non-dominated sorting algorithm, and determining the final charging guidance strategy from the optimal solution set using the ideal point method; executing the final charging guidance strategy and performing rolling optimization based on real-time feedback data. This application embodiment can ensure the high efficiency of new energy vehicle charging guidance operation in highway service areas.

This invention provides a low-code platform script generation method and system based on LLM+RAG, belonging to the field of script generation technology. The method includes constructing a semantic graph by analyzing user requirements, retrieving and filtering chainable component combinations, generating a raw script using a large language model, refining it through control flow analysis and variable initialization, and finally generating an executable script for deployment and execution. This invention achieves automated generation from natural language requirements to executable scripts, improving development efficiency and code reliability.

An automatic fixed format archive registration method, device, medium and product. In the method, an image preprocessing operation is performed on a digital image of an archive to be registered to obtain a standardized image; the standardized image is input into a field positioning model to obtain a plurality of candidate positioning boxes and confidence scores corresponding to a plurality of target field regions respectively; based on the confidence scores, the candidate positioning boxes are determined, and corresponding spatial position coordinates are obtained; a field sub-image is cropped from the standardized image according to the spatial position coordinates; optical character recognition processing is performed on the field sub-image to obtain preliminary recognized text; format verification and format conversion processing are performed on the preliminary recognized text to generate a field recognition result; and based on a field mapping relationship, structured registration data is generated. The technical scheme provided in the application guarantees the overall accuracy and reliability of the archive registration data.

This invention discloses a collaborative control method for unmanned trucks at intersections in open-pit mines, capable of identifying dynamic operating conditions. The onboard edge computing terminal includes a data preprocessing module, a parameter identification and state estimation module, a collaborative decision-making main control module, and a dual-loop safety circuit breaker module. The data preprocessing module is connected to the onboard sensor terminal, the collaborative decision-making main control module, the parameter identification and state estimation module, and the cloud training and control center. The parameter identification and state estimation module and the cloud training and control center are both connected to the collaborative decision-making main control module, which is connected to the dual-loop safety circuit breaker module. This invention solves the problem of static road surface model failure through online parameter identification, ensuring the robustness of intersection passage; it introduces a dual-loop safety circuit breaker mechanism to prevent safety accidents caused by AI algorithm "illusions"; and it incorporates energy management into the control closed loop, optimizing the speed curve for heavily loaded downhill vehicles to maximize the utilization of potential energy recovery.

This application provides a method, electronic device, storage medium, and product for updating parameters of a quantum circuit. The method includes: updating a first trainable parameter of a current noisy quantum circuit based on a current policy network to obtain at least one variation path of the first trainable parameter; determining a reward value based on each variation path and a reward function; the accuracy of the reward function and the output of the noisy quantum circuit is related to the degree of influence of noise on the output; updating a second trainable parameter of the policy network based on the reward value; and repeating the above steps until the policy network reaches a preset performance requirement. In this way, the policy network, through reinforcement learning, automatically adjusts the parameterized quantum gates in the noisy quantum circuit to achieve dynamic adaptation of the quantum circuit to noise. By dynamically adjusting the trainable parameters of the quantum circuit through a unified policy network, it is not necessary to model and train separately for each type of noise, thus improving processing efficiency.

This invention discloses a method and system for detecting life jacket wearing at port berths. The method includes: constructing a monitoring area and deploying computing nodes; training a YOLOv8 model for initial and secondary detection, a ResNet model for illumination classification, a Uformer model for rain and fog recovery, and a YOLOv8 model for semantic segmentation; performing illumination scene recognition on the acquired images and adaptively selecting the corresponding preprocessing method; using the initial detection results, for personnel targets with insufficient pixel size, predicting their position by tracking and calculating the magnification factor, controlling the PTZ camera to magnify and perform secondary detection for confirmation; using the semantic segmentation results to determine whether the personnel are located in the land-based work area; and combining target tracking, triggering an alarm only when the personnel are not wearing life jackets, are located on land, and their status continuously exceeds a threshold. The system includes an intelligent PTZ camera, a management platform, and a processing unit. This invention solves the problems of difficult identification of small targets at long distances, poor adaptability to complex illumination, and high false alarm rate due to rule-based methods.

This application provides a smoldering fire monitoring method based on wind field spatiotemporal constraints and multimodal fusion, relating to the field of multi-sensor information fusion. The method includes: acquiring thermal imaging, visible light data, gas concentration, environmental parameters, and wind field data; constructing a state-space model to filter gas concentration and extract rate-of-change features; establishing a gas transport model based on wind speed, wind direction, and the platform's relative position, calculating prior time delays and fine-tuning based on correlation analysis to align gas and image data; constructing a weighting function based on wind field geometric relationships to fuse and generate unified gas evidence; and finally constructing a temporal probability model including four states: normal, interference, smoldering, and open flame, using gas evidence, thermal evidence, and smoke evidence as observation inputs to infer the fire state. This invention solves the problems of spatiotemporal misalignment, weak anti-interference ability, and poor environmental adaptability in existing monitoring methods, enabling early, accurate, and robust monitoring of smoldering fires.

The application discloses a multi-model reasoning method and device based on a graph structure, a medium and a program product, wherein the method comprises the following steps: querying a task target node in a graph database and starting from the task target node, and performing reasoning path expansion in the graph database according to a corresponding driving model, wherein the graph database comprises a plurality of directed acyclic graphs, and each directed acyclic graph is obtained by mapping metadata of a registered analysis model; generating a task topology execution graph according to model nodes involved in the expanded reasoning path and the corresponding driving model; and sequentially scheduling actual execution of each model node according to the task topology execution graph, and obtaining an explanation result corresponding to the task target node. The application improves the reasoning execution efficiency by explicitly attributing the analysis model and optimizing the reasoning path by using graph calculation, and improves the intelligence, performance and interpretability of multi-model collaborative analysis.