35results about How to "Verify validity" patented technology

The application discloses a method for designing and optimizing a VLCP system assisted by light reflection RIS under low signal-to-noise ratio, wherein the VLCP system comprises an intelligent wall adopting a smart surface RIS with mirror reflection, a transmitter and a user terminal equipped with a receiver with a liquid crystal smart surface, communication and positioning signals are integrated into the transmitter, the transmitter comprises an LED, and the method comprises the following steps: under low signal-to-noise ratio, taking asymptotic capacity as a communication metric constraint condition, taking a Cramer-Rao lower bound of positioning error as a positioning metric constraint condition, combining a positioning constraint satisfaction process based on a two-stage greedy algorithm and a communication performance maximization process based on a greedy strategy convex relaxation algorithm, optimizing elements of an RIS allocation matrix to maximize the asymptotic capacity, obtaining an optimal RIS allocation matrix, and completing the design and optimization of the VLCP system. In a low signal-to-noise ratio scene, the VLCP system of the application can significantly improve system performance by applying RIS.

This invention belongs to the field of solid oxide fuel cell control and optimization technology, and particularly to a dual-input model predictive control method for solid oxide fuel cells with carbon deposition constraints. It includes the following steps: S1: Constructing a DIR-SOFC controlled object model with 11 state variables; S2: Solving for the optimal control increment through sequential quadratic programming; S3: Constructing a constraint system; S4: Introducing an adaptive setpoint for the output voltage that adjusts linearly with the load current density. The model dimension of this invention is expanded from the existing 3 states to 11 states, fully covering six gas phase partial pressures, bilinear coupling of carbon accumulation and catalyst activity, and three-zone thermal dynamics. The ODE white-box prediction engine accurately reflects MSR nonlinearity and positive feedback of carbon deposition. Compared with ARMAX black-box prediction, the error is reduced from approximately 3% to less than 0.8%, with an average single-step calculation time of 133ms, meeting the real-time constraint of a 1s sampling time.

This invention relates to the field of multimodal sentiment analysis technology, and in particular to a method and system for multimodal sentiment analysis based on sequential translation under uncertain and missing modalities. The method includes: acquiring multimodal sentiment data, including visual, textual, and auditory modalities; constructing a multimodal sentiment analysis model based on sequential translation, firstly by extracting self-attention encoding features of each modality using a Transformer encoder; and then designing a bidirectional sequential translation architecture centered on the textual modality. This invention improves the utilization of textual modal features in the feature fusion process, thereby enhancing the quality of multimodal fusion features and providing strong support for subsequent complex sentiment analysis.

This invention relates to the field of low-altitude emergency response technology and discloses a method for joint scheduling optimization of heterogeneous equipment for low-altitude emergency response tasks. The method includes: obtaining a task set and a set of heterogeneous equipment for the low-altitude emergency response task; constructing a dual-objective joint scheduling model for the low-altitude emergency response task based on the task set and the heterogeneous equipment set; performing non-dominated sorting optimization on the dual-objective joint scheduling model to generate a Pareto optimal solution set for the low-altitude emergency response task; executing the scheduling scheme in the Pareto optimal solution set; and monitoring the heterogeneous equipment of the low-altitude emergency response task in real time based on a preset dynamic rescheduling mechanism to obtain triggering events for the heterogeneous equipment; dynamically rescheduling the scheduling scheme based on the triggering events to generate a scheduling update scheme for the low-altitude emergency response task. This invention can improve the overall success rate, resource utilization efficiency, and system robustness of low-altitude emergency response tasks.

This invention belongs to the technical field of aircraft manufacturing engineering and intelligent scheduling optimization. It discloses an aircraft final assembly scheduling method and equipment based on an improved genetic algorithm. The steps are as follows: (1) Construct a scheduling model based on process and worker data with the optimization objectives of minimizing the final assembly completion time and minimizing the number of workers involved; (2) Encode the process priority and worker skill slots obtained from the scheduling model using an improved genetic algorithm in the encoding stage, and use the encoding as chromosomes to form an initial population; (3) Perform event-driven decoding on the individuals in the initial population to obtain an initial scheduling scheme; (4) Calculate the fitness value of each chromosome according to the total completion time and the number of workers involved corresponding to the initial scheduling scheme, and screen all chromosomes according to the fitness value to obtain a new population, and then finally obtain the aircraft final assembly scheduling scheme. This invention realizes the collaborative optimization of completion time and the number of workers involved.

This invention belongs to the field of wheat breeding, specifically relating to a wheat Fusarium head blight resistance locus and molecular breeding method. This invention identified five stable Fusarium head blight resistance-associated loci, Qfhb.jaas-2D, Qfhb.jaas-3A, Qfhb.jaas-3B.1, Qfhb.jaas-5A.2, and Qfhb.jaas-6D, on chromosomes 2A, 2D, 3A, 5A, and 6D. For the identified resistance loci, this invention designed KASP molecular markers. These markers have high specificity and sensitivity, enabling accurate and rapid detection of resistance genotypes in wheat materials, providing a powerful tool for marker-assisted selection (MAS). This invention verified the effectiveness of the screened resistance loci. The results showed that the above five loci were significantly associated with Fusarium head blight resistance, providing a reliable basis for the breeding of resistant varieties.

This invention belongs to the field of image processing and segmentation technology, specifically relating to a method for segmenting corrosion defects on the exterior of nuclear power plant equipment based on an attention mechanism and edge detection operators. The method includes the following steps: acquiring exterior images of key nuclear power plant equipment; performing semantic-level annotation on the acquired images to construct a nuclear power equipment corrosion segmentation dataset; using the TransUnett model as the basic network framework and obtaining pre-trained weights from the dataset; obtaining an improved TransUnet model; obtaining the final weight file suitable for nuclear power equipment corrosion defect segmentation; inputting the image of the nuclear power equipment to be detected into the trained model, which simultaneously outputs a semantic segmentation mask and edge response results; and generating the final corrosion defect segmentation result through post-processing. The beneficial effect is that it can simultaneously utilize a combination of semantic segmentation and edge detection to achieve defect segmentation, solving the shortcomings of existing model structures in edge blurring and inaccurate segmentation under complex environments.

The application discloses a kind of personalized federated learning method and system based on multi-dimensional adaptive aggregation and local dynamic representation alignment, the method in federated learning process, first by multi-dimensional adaptive aggregation mechanism, comprehensive historical local characteristics, round attenuation and model structure characteristics, to build personalized training starting point for client;Subsequently in the local training process, introduce local dynamic representation alignment mechanism, dynamically update class prototype and build contrast loss, to enhance feature discrimination ability;Further, the importance of each layer parameter of the model is evaluated by parameter importance fair quantization mechanism, and a layer importance mask is generated to guide the next round of refined aggregation.Based on the above technical solutions, the application effectively alleviates the personalized and generalization imbalance problem under the condition of data heterogeneity, while improving the model accuracy, effectively reducing the communication round and communication overhead.

This invention proposes a generalizable federated learning method and system based on perceptual heterogeneity features and consistency of the fundamental matrix. The system includes a central server and multiple clients. The central server has an unlabeled public image dataset, and each client holds a labeled private image dataset and deploys a local image classification model that can be decomposed into an image feature extractor and a classifier. Each client is equipped with a feature mapping layer that projects the image features output by the image feature extractor onto a unified high-dimensional feature space. This invention solves the heterogeneity and catastrophic forgetting problems through the FBCL federated learning framework.

This invention discloses a method for registering industrial printed images based on a lightweight network. The method includes: S1 Image acquisition and dataset construction, which uses a CCD industrial camera to acquire images of the front and back of printed materials and constructs a training dataset by combining multispectral images; S2 Constructing a DHSiam network, which includes two parallel feature extraction networks; S3 Proposing an attention heterogeneous mechanism, which includes a position regression unit and a lightweight unit to enhance data representation and reduce model parameters; S4 Constructing a cross-entropy loss function with additional margin to enhance network performance metrics. This invention builds a hardware platform with 8 parallel GPUs, uses only AMC-Loss to distinguish image similarity, utilizes the attention heterogeneous mechanism to extract important information from feature maps at different scales, and adjusts model parameters to enhance the algorithm's performance on mobile devices, thereby comprehensively improving the model's ability to register printed images with different perspectives and deformations.

This invention proposes a tracking control method and system for tower cranes with unknown disturbances, belonging to the technical field of anti-swaying for tower cranes. The method includes: acquiring the state information of the tower crane's boom, trolley, and load; establishing a dynamic model of the tower crane and decoupling the unknown disturbance and its nonlinear unknown coefficients in the dynamic model using Young's inequality; designing a coordinate transformation to eliminate the pure feedback structure in the system; introducing coordinate transformation into a virtual controller designed based on the backstepping method, constructing the virtual controller, and decoupling it from the disturbance; and, based on the virtual controller and the actual controller, controlling the position of the trolley on the tower crane to maintain synchronization with the designed desired trajectory under unknown external disturbances using a near-disturbance decoupling control method. This invention avoids the swaying of the tower crane load due to uncertain disturbances, which is of great significance for improving the working efficiency and safe construction of tower cranes.

A method for obtaining micro-vibration test conditions for aerospace vehicles involves: searching for micro-vibration disturbance sources during the vehicle's on-orbit operation; classifying the micro-vibration disturbance sources and identifying those that cannot be avoided; conducting micro-vibration environment measurement tests on the unavoidable disturbance sources to obtain the time-domain curve of the maximum vibration environment of the disturbance sources; establishing a finite element model for micro-vibration response analysis; conducting transient response analysis to determine the micro-vibration response of precision instruments; converting the time-domain curve of the micro-vibration response of the precision instruments into a frequency-domain curve and performing envelope design to obtain the micro-vibration test conditions of the precision instruments for subsequent experimental verification.

This invention belongs to the field of elevator light curtain technology and discloses an elevator light curtain with door opening and closing warning function. It includes a light curtain housing and positioning frames fixed at both ends. A groove is formed on the left side of the light curtain housing, and the light curtain body is connected inside the groove. Warning LEDs are installed on both sides of the light curtain body. A movable frame is fitted around the outside of the light curtain body. Limiting grooves are formed on the front and rear sides of the light curtain housing, and the movable frame slides in cooperation with the limiting grooves. This elevator light curtain with door opening and closing warning function completely changes the traditional "passive response" safety protection mode of light curtains through the design of multiple arrays of dual-color warning LEDs, constructing an active and efficient warning system. During the elevator door closing process, the high brightness and wide warning range of the dual-color flashing lights can clearly remind the elderly and children who move slowly, and can also effectively attract the attention of passengers focused on electronic devices, avoiding collisions or pinching accidents caused by delayed reaction.

The utility model discloses a kind of microwave heating devices for pipeline flow heating, it is related to microwave heating technical field, including metal shell, medium filled in the metal shell, coaxial port being arranged at one end of the metal shell, outer conductor being arranged at the coaxial port, inner conductor being extended into the metal shell by the outer conductor, mode converter being arranged in the metal shell and being connected with inner conductor, insulating medium is arranged between the outer conductor and the inner conductor, mode converter is matched network between coaxial feeder input impedance and waveguide wave impedance, the metal shell bottom is provided with medium extension area that is inserted heated object in heating process. The utility model inserts medium extension area into heated object, to realize the effect of plug-in heating, the efficient heating of material can be realized to dielectric constant real part between 5 to 80, loss tangent value between 0.1-1, while realizing the uniform heating to load.

The application relates to the technical field of communication, and discloses a verification method for a three-dimensional channel model, which comprises the following steps: determining a sampling region of the three-dimensional channel model; wherein the sampling region is related to a coordinate axis of a three-dimensional coordinate system established with a test region center as an origin; collecting channel data of sampling points located in the sampling region; and verifying the effectiveness of the three-dimensional channel model by using the channel data. The selection mode of the sampling points in the air interface test channel verification link of the three-dimensional channel model is improved, so that the verification scheme based on the channel data of the sampling points can better represent the modeling quality of the three-dimensional channel model, thereby more accurately verifying the effectiveness of the three-dimensional channel model. The application further discloses a verification device for the three-dimensional channel model and an electronic equipment.

The application discloses a design method and system of a V-shaped built-in high-speed permanent magnet rotor containing multiple magnetic isolation bridges, and comprises the following steps: obtaining stress analytical relations of a central magnetic isolation bridge, an intermediate magnetic isolation bridge and two side magnetic isolation bridges according to structural characteristics of the magnetic isolation bridges, geometric characteristics of pole shoe regions, force balance equations of the magnetic isolation bridges, geometric coordination conditions, material physical equations and load forms; performing calculation and analysis based on the stress analytical relations to obtain design parameters of the intermediate magnetic isolation bridge and stress variation laws of the central magnetic isolation bridge, the intermediate magnetic isolation bridge and the two side magnetic isolation bridges; and determining optimal design parameters of the intermediate magnetic isolation bridge based on the stress variation laws. The five-magnetic-isolation-bridge V-shaped permanent magnet rotor is adopted, the width direction of the intermediate magnetic isolation bridge is perpendicular to the pole shoe surface of the rotor at the position of the intermediate magnetic isolation bridge, the mechanical strength of the permanent magnet rotor structure is remarkably improved, and the maximum operating speed of the permanent magnet rotor is further improved.

The application belongs to the technical field of computer vision, and discloses a few-shot classification method based on an image-text pre-training model.For the optimal transport of the base element level, an image block reweighting mechanism based on prototype distance is proposed.The Euclidean distance between the prototype and the image block is converted into the weight of each image block, so that different weights are assigned to the image blocks.For the prompt level stage, a cascaded optimal transport module is proposed, which not only considers the optimal transport between the few-shot image-text features, but also considers the cross optimal transport between the zero-shot and few-shot image-text features.A prototype-based unbalanced consistency loss function is proposed to supervise the network.The loss function is divided into a prototype distance loss and an unbalanced consistency loss.The test is carried out on Caltech01, DTD and EuroSAT data sets, and the experimental results prove that the network model proposed in the application is superior to the most advanced algorithm at present, and the effectiveness of the algorithm is verified.

The application discloses a kind of under messy scene robot autonomous pickup and pose acquisition method of placement, the color image of work scene obtained by camera is as input, target information is obtained using semantic segmentation model and input information representation ability is strengthened, a kind of actor-critic form deep reinforcement learning method is used to remove obstacle autonomously and explore target object pose, then target placement pose is obtained using the method based on depth feature template matching, finally, pick and place object to specific pose, the scheme can explore the occluded target object in complex scene and place to specific pose.The scheme realizes high-precision positioning of target placement pose according to pickup pose, and the target pickup success rate in messy scene can reach more than 80%, which improves the success rate and stability of target placement pose acquisition task as a whole.

The application relates to a water surface unmanned ship improved GA-SA path planning method for offshore wind farm inspection, and relates to a water surface unmanned ship path planning method.The application aims to solve the problems of low efficiency and low precision of an existing algorithm.The process is as follows: one, initializing algorithm parameters;two, performing chromosome coding of feasible solutions, and setting a fitness value function;three, taking the highest fitness value as the optimal individual;four, judging whether the convergence condition is met, if yes, outputting the current optimal individual, otherwise, entering five;five, taking the current optimal value as an initial solution of the SA algorithm, and randomly obtaining an updated solution;judging the energy change of the updated solution;if the energy change is less than 0, the updated solution is taken as the latest solution;otherwise, the updated solution is reserved according to a probability;six, judging whether the convergence condition is met, if no, repeating five;if yes, returning to three to replace the optimal individual, and repeatedly executing four to six until the convergence condition is met.The application is used in the field of path planning of water surface unmanned ships.

This application discloses a circuit device and method for monitoring the status of a light-controlled thyristor and reporting overvoltage protection. The device includes a BOD (Browser Overvoltage) action and overvoltage event identification and reporting unit and a passive thyristor status monitoring unit connected in parallel across the two ends of the light-controlled thyristor pair within the switch. The BOD action and overvoltage event identification and reporting unit uses an external BOD mirror to reflect the internal protection threshold and utilizes back-end logic to intelligently identify signals, simultaneously achieving reliable reporting of BOD overvoltage protection actions and independent detection of unexpected system overvoltage events. The high-capacitive reactance circuit of the passive thyristor status monitoring unit enables low-power, passive online monitoring of the light-controlled thyristor status, directly indicating the health of the light-controlled thyristor component. This application integrates three major functions: real-time component status monitoring, BOD protection action reporting, and intelligent overvoltage event identification. It features full fiber optic isolation, passive high reliability, and is an innovative front-end sensing and intelligent monitoring and reporting solution specifically designed for railway automatic phase-splitting electronic switches.

The application provides a supervised contrast fine-tuning method and device for active few-shot learning, and relates to the technical field of text classification in natural language processing. Active few-shot learning (AFSL) is an effective paradigm to improve the performance of large language models under limited annotation budget. In view of the problem that the traditional fine-tuning target is low in efficiency in AFSL, the application proposes a supervised contrast fine-tuning framework specially for the text classification task of natural language processing. The framework combines supervised contrast learning and hard negative sample mining, optimizes the embedding space by using an enhanced hybrid loss function, and thus improves the utilization efficiency of labeled samples. A large number of experiments on five benchmark datasets show that, under the fixed SOTA query strategy, the text classification performance of the method of the application is superior to that of the baseline model. These findings prove that the improvement of the learning target to optimize "how to learn" provides a new direction complementary to the existing query strategy for promoting AFSL research.

The application discloses a kind of lithium ion battery distributed thermal process sensor fault estimation method, comprising: using two-dimensional partial differential equation to describe rectangular lithium ion battery distributed thermodynamic equation;Distributed thermal model of large size lithium ion battery is constructed;Using Chebyshev-Galerkin method, distributed thermodynamic equation and distributed thermal model are decomposed into reduced order model described by standard state space equation in time domain;State variable in reduced order model is constructed using sensor measurement output and a Hurwitz matrix, and enhanced reduced order model is constructed;Enhanced adaptive observer and error state space equation are constructed using enhanced reduced order model;Based on error state space equation, the fault strength of temperature sensor that fails is estimated using fast adaptive algorithm.The application can effectively estimate sensor fault, including time-invariant and time-varying fault, and single sensor and multiple sensor fault.

The present disclosure belongs to the technical field of geophysical information processing, and is a spatial anomaly weight determination method and system based on time series similarity, comprising: data clustering classification on CO surface concentration original data; obtaining CO anomaly data based on the clustering result; calculating the Benioff energy accumulation curve; performing anomaly accumulation processing on the extracted CO anomaly data to obtain an anomaly accumulation curve, and performing normalization processing on the anomaly accumulation curve and the corresponding Benioff energy accumulation curve; calculating the time series similarity between the normalized anomaly accumulation curve and the Benioff energy accumulation curve; and taking the time series similarity as a target function to obtain an optimized spatial weight parameter. The present disclosure can adaptively determine the spatial weight based on the time structure similarity, improve the correlation between the anomaly analysis result and the seismic activity, and is suitable for seismic related anomaly detection and analysis.

The invention belongs to the field of intelligent biological prediction, and particularly relates to an RNA-small molecule binding affinity prediction method based on multi-view learning. The method comprises three stages of data set up and down sampling processing, feature construction and extraction based on RNA and small molecule sequences, and binding affinity prediction. According to the method, data set distribution is optimized through an up-down sampling method, multi-view feature information is constructed through sequence information of RNA and small molecules, a deep network is established to extract and depth biological information features, and finally, an MLP regression device is adopted to predict an RNA-small molecule binding affinity value, so that drug discovery work is assisted to be achieved. In order to solve the problem of imbalance of the existing data set, the invention provides a method of combining up and down sampling for processing, and the problem of imbalance is effectively relieved.

This invention provides an aerial formation prediction method and system, belonging to the field of situational awareness. The method includes: using the extreme values ​​of the latitude and longitude coordinates of each target in the formation at the current moment as edges, and taking the movement direction of the lead aircraft in the formation as the formation movement direction, selecting the coordinates of the upper left and lower right vertices of the formation area to construct the formation area coordinates, thus completing the construction of the formation area; dividing the formation area into grids and encoding each grid according to whether a target exists, forming a formation grid code; concatenating the formation grid codes from multiple consecutive moments into feature vectors and inputting them into a trained LSTM model to predict the feature vector corresponding to the formation grid code at the next moment; inputting the feature vector corresponding to the formation grid code at the next moment into an SVM model to predict the formation formation at the next moment. This invention solves the problems of difficult feature extraction and template construction in formation prediction algorithms.

This invention discloses a spatiotemporal traffic flow prediction method based on dual-flow decoupling. The method first decomposes the temporal data of road network traffic flow into trend and periodic components using the exponential moving average method. Then, features are extracted and fused using a deep linear network and a local feature hybrid network with temporal block embedding to obtain the global temporal prediction component. After the traffic flow temporal data is encoded in the temporal domain by gated dilated convolution, multi-view spatial aggregation is performed using forward, backward, and normalized graph convolutions with adaptive adjacency matrices to obtain the local spatiotemporal prediction component. Finally, dynamic weights are generated by a gated network, and the two types of components are weighted and fused to obtain the final prediction result. This invention accurately separates long-term and short-term traffic flow features, adaptively balances global patterns and local details, improves the accuracy and robustness of long-term temporal prediction, and reduces computational complexity, making it suitable for intelligent traffic flow management scenarios.

This invention provides an autonomous tracking control method for unmanned surface vessels (USVs) that integrates Kalman filtering and deep reinforcement learning, belonging to the field of intelligent ship and autonomous navigation control technology. Addressing issues such as increased drift angle and frequent steering in real-world marine environments, it constructs a control framework that combines Kalman filtering-assisted line-of-sight guidance with deep reinforcement learning. By introducing a Kalman filtering-based line-of-sight guidance algorithm to predict and compensate for drift angle in real time, a target heading is generated, ensuring the USV's speed and direction align with the desired path. A deep reinforcement learning controller learns and outputs the optimal rudder angle control strategy based on state information such as the deviation between the target heading and the actual heading. A composite reward function is designed that considers heading error, drift angle changes, and rudder angle stability, significantly suppressing the frequent steering issues found in traditional control methods. This invention outperforms existing traditional control technologies in terms of track tracking accuracy, rudder angle adjustment smoothness, and robustness under complex sea conditions.

The application discloses a kind of image semantic transmission and multi-outlet DNN inference collaborative design method for star-ground fusion network, constructs an image semantic transmission and multi-outlet DNN inference collaborative processing model under the scene of star-ground fusion network, model includes the LEO satellite equipped with semantic encoder, at least one ground station equipped with semantic decoder and multi-outlet DNN inference module.Satellite is collected after remote sensing image data, by encoder, the semantic features extracted are transmitted to ground station, ground station reconstructs image by decoder, and executes multi-outlet DNN inference task, adopts early return mechanism in inference process, and when meeting inference accuracy threshold, inference result is output.On the basis of considering semantic recovery quality and inference accuracy, jointly satellite launch power, semantic compression rate, EE decision and computing resource allocation, total time delay of task completion is minimized, and image semantic transmission and multi-outlet DNN inference collaborative design are realized.