21results about How to "Promote deep integration" patented technology

The invention provides a power load control method based on a multifunctional flexible adjusting unit, and belongs to the technical field of power load control. The power load control method based on the multifunctional flexible adjusting unit comprises the following steps: collecting current data and voltage data, monitoring the load condition of a special transformer user, and carrying out load identification by utilizing edge calculation so as to distinguish the power load type and the power load working state; constructing a load prediction model; acquiring historical load data stored in the terminal and the master station software; generating a prediction result in the load prediction model based on the historical load data, the power load type and the power load working state; and dynamically adjusting the operation state of the electric equipment based on the prediction result. According to the power load control method based on the multifunctional flexible adjustment unit, the refinement degree of load management and the accuracy of load prediction can be improved, and dynamic optimization scheduling can be realized.

The utility model relates to the technical field of new energy automobile charging, and discloses a mobile energy storage charging robot, which comprises a robot body, a charging pile and a platform dispatching system, and is characterized in that the robot body is provided with a high-capacity energy storage battery pack, a motor driving unit and an electric drive chassis; a charger, a charging and discharging switching controller and an anti-backflow terminal are arranged in the charging pile, the robot body is used for charging a new energy automobile, the charging pile is used for automatically switching to charge the robot body and receiving discharging of the robot body, and the platform dispatching system is used for managing resource distribution of the robot body and the charging pile of each city; according to the invention, the charging convenience and user experience are greatly improved, the problem of insufficient charging facilities is effectively relieved, the utilization rate of electric power resources is improved, deep fusion of new energy vehicles and smart grids is promoted, and an innovative solution is provided for promoting green travel and constructing a smart energy ecological system.

The invention provides a multi-source software-defined science and education resource recommendation method and system based on a knowledge graph, and the method comprises the steps: constructing an artificial intelligence knowledge graph based on the relation between artificial intelligence knowledge entities; obtaining science and education resource data of a science and education institution, extracting knowledge entities in the science and education resource data, and connecting the knowledge entities to the artificial intelligence knowledge graph to obtain a science and education resource knowledge graph; extracting local resource features / user features based on a graph neural network to obtain local resource features and user features of each client; constructing a local knowledge graph based on the science and education resource knowledge graph, and performing model training by taking the local resource features of each client as initial features until the model converges; and inputting the users and the resources into the graph neural network for prediction, and recommending the resources with high scores to each user. According to the scheme, the relationship between knowledge entities in the field of artificial intelligence is effectively extracted, a structured knowledge graph is formed, and the recommendation precision is improved.

The invention discloses a three-dimensional molecular text understanding method based on a hybrid expert model and a large language model, and the method comprises the steps: building a new MoE layer of the hybrid expert model, and enabling the MoE layer to comprise a plurality of experts, coding the task information text and the text context through a cross-modal projector, and inputting the coded task information text and text context into a route of a hybrid expert model; and training and processing molecular encoders of different tasks, and integrating the trained molecular encoders into a unified 3D molecular encoder. According to the method, task information is injected into routing input by introducing a text-assisted MoE layer, the capability of processing specific tasks by the model is enhanced, and a universal model capable of simultaneously understanding 2D and 3D molecules and various different molecular properties is provided. By combining a plurality of molecular encoders, the comprehensive understanding of molecular characteristics is improved. By designing a molecule-text mapper, deep fusion between molecular features and texts is promoted, and the understanding ability of the model for three-dimensional molecular texts is improved.

The invention provides a long-term action prediction method based on a multi-modal large language model and a related device, and the method comprises the steps: carrying out the frame extraction of a video according to a fixed sampling rate, and predicting text action labels corresponding to all video frames; inputting the text action tag into a multi-modal large language model to obtain a text feature, a visual feature and a preset action query feature; constructing a cross-modal interaction network architecture, and taking the text features, the visual features and preset action query features as input to obtain feature expressions after interaction of different modals; and obtaining an output result of action tuning based on the feature expression after interaction of different modalities and an action tuning module, then training a multi-modal large language model, and performing optimization solution on the network by using a loss function to obtain a prediction result.

The invention provides a Chinese medical report generation method and system based on a pre-trained large model and a terminal, and relates to the technical field of artificial intelligence, and the method comprises the steps: S1, obtaining an X-ray medical image of a to-be-generated report; s2, obtaining a Chinese medical report generation requirement input by a user; s3, importing the obtained X-ray medical image and a Chinese medical report generation requirement input by a user into a pre-trained large model to obtain a Chinese medical report of the X-ray medical image; the large model comprises a medical visual encoder, a medical large language model and a similar report retrieval module. According to the method, complex contents and lesion features in the medical image can be fully understood based on the pre-trained large model, the accuracy of the generated medical report in lesion description and diagnosis suggestion aspects is improved, the illusion problem when the model generates the report is relieved, and the quality and efficiency of the medical report are effectively improved.

The present application relates to the technical field of traditional Chinese medicine preparation, and particularly relates to a method and system for optimizing a frying process of traditional Chinese medicine decoction pieces. Formula information and process requirements of target traditional Chinese medicine decoction pieces are obtained; a frying process scheme of similar traditional Chinese medicine decoction pieces is retrieved from a traditional Chinese medicine frying expert knowledge base; initial parameters are generated by fusing the frying process scheme; first-round frying is performed; process data are collected by a multi-modal sensor; adjustment suggestions are proposed by experts; parameter adjustment schemes are generated in combination with the expert suggestions and the process data; interactive optimization is performed for multiple rounds; and optimal frying process parameters are recorded and stored in the knowledge base. By constructing a formula information-driven knowledge retrieval mechanism, a multi-modal sensing monitoring system, a parameter adjustment strategy combining expert evaluation and data analysis, and an interactive iterative optimization framework, a closed-loop optimization system is formed, and technical problems such as difficulty in knowledge reuse, single process monitoring, lack of systematic support for parameter adjustment, and insufficient human-machine cooperation in traditional Chinese medicine frying processes are solved.

The invention relates to a river flow prediction method, which comprises the following steps of: 1, dividing a daily runoff data set, and performing multi-stage decomposition on a training set by using Haar wavelets to obtain de-noising parameters; 2, multiplexing the verification set and the test set to obtain data; step 3, carrying out first-order differential transformation, and zooming to an interval of [0, 1] by using MinMaxScaler to obtain a normalized value of a differential sequence; 4, mapping the data to a high-dimensional space through a linear embedding layer, and then carrying out position coding to obtain data; step 5, inputting the data obtained in the step 4 into a DTransformer encoder, so that a catastrophe point obtains a higher attention weight so as to strengthen difference characteristics of adjacent time steps and obtain encoder output; 6, inputting the data obtained in the step 5 into an LSTM decoder, and performing feature recombination to obtain encoder output; 7, mapping the data obtained in the step 6 to 7-dimensional output through a full connection layer; and 8, performing inverse normalization and inverse difference transformation on the data obtained in the step 7 to obtain final output.

The invention provides a large-model multi-language enhancement model fusion method based on mixed experts, which comprises the following steps of: 1, selecting a to-be-fused layer from an anchor point model and a language expert model; step 2, constructing a fusion module; 3, fusing each to-be-fused layer in the anchor point model and the language expert model by using a fusion module to obtain a fused model; and step 4, training the fused model to obtain a final fusion result, and completing large-model multi-language enhancement model fusion based on the hybrid experts.

The application discloses a highway pavement disease detection method, system and equipment, and relates to the technical field of highway pavement image processing. The technical scheme points are: obtaining a to-be-detected pavement image of a highway; inputting the to-be-detected pavement image into a pre-trained disease area segmentation model for segmentation and extraction to obtain a disease area image of the highway pavement; wherein, a first discriminator and a first generator are used to form a first generative adversarial network, the first generative adversarial network is trained, and when the number of training reaches a maximum number, a disease detection model is obtained; the disease area image is input into the pre-trained disease detection model for detection to obtain a disease detection result of the highway pavement; wherein, a YOLOv8 neural network is trained, and when a first loss function converges, the disease detection model is obtained. The problems that the prior art cannot capture detailed information and has false and missed detection of diseases when detecting pavement diseases are solved.

The application discloses a floating photovoltaic power generation device suitable for fish culture, which comprises a corrosion-resistant floating body, a power generation module, a sun tracking module and a self-cleaning module integrated on the floating body to form an integrated independent power supply unit, and photovoltaic panels replace high-voltage power transmission to supply power to the culture device; the sun tracking module realizes sun tracking of the photovoltaic panels through a two-degree-of-freedom deflection structure combined with intelligent control and takes into account underwater culture illumination; the self-cleaning module realizes automatic cleaning of the photovoltaic panels through cooperation of a brush / roller brush and spraying; the floating body can be used alone or composed of a power generation group, and is matched with structures such as buoyancy reinforcement, gravity center lowering and wind and wave resistance linkage, and is suitable for different scales and types of culture water areas. The application realizes on-site power consumption, eliminates hidden dangers of high-voltage power transmission, improves photovoltaic power generation efficiency, reduces operation and maintenance cost, takes into account power generation and culture ecology, and has high practicability.

The present invention relates to the field of wastewater treatment, and in particular to a hydrogen peroxide wastewater treatment device combining aeration and sludge, comprising a tank body, wherein a stirring mechanism is arranged in the tank body, wherein the stirring mechanism comprises a rotating shaft, and a lower stirring rod and an upper stirring rod arranged parallel to each other along the axial direction thereof, and a counter-aeration component is arranged between the lower stirring rod and the upper stirring rod, wherein the counter-aeration component comprises upper aeration nozzles distributed at equal intervals along the axial direction of the lower stirring rod, and lower aeration nozzles distributed at equal intervals along the axial direction of the upper stirring rod, and an anti-blocking structure is arranged between the upper aeration nozzle and the lower aeration nozzle. The present invention ensures uniform mixing of wastewater and sludge through synchronous movement of the upper stirring rod and the lower stirring rod, and simultaneously utilizes the lower aeration nozzle and the upper aeration nozzle to inject oxygen-rich gas into the water body, thereby facilitating contact between suspended solids and bubbles and reducing accumulation of solid particles in the aeration gap.

The application belongs to the technical field of spaceflight, and discloses a satellite electromagnetic permanent magnet integrated connection separation and recovery device and a control method thereof. The satellite electromagnetic permanent magnet integrated connection separation and recovery device is provided with a curved surface permanent magnet interface on a micro-nano satellite. The curved surface permanent magnet interface is connected with a curved surface electromagnetic interface. The curved surface electromagnetic interface is arranged on a satellite ring. The application improves the carrying capacity of spaceflight equipment. The satellite ring can provide multiple micro-nano satellite connection separation and recovery interfaces, and can simultaneously release, separate and dock and recover multiple micro-nano satellites. The satellite ring is composed of a circular ring. The surface of the circular ring is provided with electromagnetic connection separation and recovery interfaces. The micro-nano satellite connection and recovery docking are realized by the butt joint of the micro-nano satellite curved surface permanent magnet interface and the curved surface electromagnetic interface. The locking of the connection interface is realized by adopting a bag belt fixing mode during the execution of a launch task.

The present invention belongs to the technical field of vehicle environment perception, and specifically discloses a vehicle environment perception method based on multi-view and multi-scale feature fusion. This method aims at the problem of information loss in the projection process of the projection-based point cloud segmentation method, and builds a vehicle environment perception network based on multi-view and multi-scale feature fusion. In terms of network input, the present invention uses multi-layer distance view representation and bird's-eye view feature dual encoding to design a multi-level multi-projection fusion deep learning network, which effectively improves the information extraction accuracy of targets in point clouds with prominent vertical information and long distances, and reduces the information loss of separated projections. In terms of network structure, the present invention proposes a multi-view fusion module, which combines MFB and MMViT to calculate the high-order correlation between BEV features and RV features, and deeply fuses multi-view features. The present invention is conducive to avoiding losses incurred during the projection process and improving the vehicle environment perception capability.

The invention provides a voice and voiceprint analysis and numeralization method and system for reflecting a phlegm-dampness state of a human body. The method comprises the steps of determining voice global feature parameters related to the phlegm-dampness state based on preset theoretical definition information of the phlegm-dampness state; voice data of normal people and phlegm-dampness people are collected, the collected voice data are marked, and corresponding physique categories are determined; feature extraction is carried out on the marked voice data, voice global feature parameters are obtained, and voice feature data sets of the normal crowd and the phlegm-dampness crowd are formed; constructing a normal voiceprint group and a phlegm-dampness state feature voiceprint group based on the voice feature data set by using a regression calculation method; performing matching analysis on the voice global feature parameter value of the to-be-detected voice and the normal voiceprint group and the phlegm-dampness state feature voiceprint group, and calculating the difference degree between the to-be-detected voice and the normal voiceprint group and the phlegm-dampness state feature voiceprint group; and quantitatively generating a diagnostic report of the phlegm-dampness state according to the calculated difference degree.

This disclosure relates to the field of digital rock physics, and particularly to a method and apparatus for adjusting the tortuosity of the pore space in a digital core sample. The method includes: acquiring a digital core model of the core sample and a pre-constructed tortuosity adjustment model; wherein the tortuosity adjustment model is an inverse mapping based on the wave principle; and using the tortuosity adjustment model to adjust the tortuosity of the pore space in the digital core model to construct an extended digital core model of the core sample.

The invention provides a fire risk assessment and management system of a data center. Comprising a fire-fighting digital twinborn digital body and entity mapping module, a fire risk information sensing module based on a data center, a fire risk information intelligent evaluation and fusion module and a high-rise building fire risk early warning, study and judgment module of the data center. The fire-fighting digital twin digital body and entity mapping module and the fire risk information sensing module based on the data center are connected with the fire risk information intelligent evaluation and fusion module; and the fire risk information intelligent evaluation and fusion module is connected with the high-rise building fire risk early warning and research and judgment module of the data center.

The present invention belongs to the technical field of processing and manufacturing of polytetrafluoroethylene tubes, and specifically relates to a transparent polytetrafluoroethylene tube, a preparation method and an application thereof. The present invention adopts high compression ratio polytetrafluoroethylene and PFA to blend and extrude to form a transparent polytetrafluoroethylene tube, thereby improving the transparency of the transparent polytetrafluoroethylene tube; at the same time, a secondary sintering method is adopted to promote the fusion of high compression ratio polytetrafluoroethylene and PFA, which can not only improve the transparency of the transparent polytetrafluoroethylene tube, but also improve the strength and elongation of the transparent polytetrafluoroethylene tube. The transparent polytetrafluoroethylene tube provided by the present invention has an outer diameter of less than 5.0 mm and has a high transparency, and is suitable for use in medical devices.

The invention discloses a new energy vehicle and power grid fusion interaction potential assessment method and system, and the method comprises the steps: obtaining a first parameter of a new energy vehicle, defining a user utility function and an agent profit function according to the first parameter, and constructing an interaction strategy behavior between an electric vehicle user and a charging station agent; based on the interactive strategy behavior, designing an evolutionary game model between the agent and the user, and searching strategy equilibrium between the user and the agent through an iterative algorithm; based on strategy equalization, calculating the shortest arrival time of the user responding to the charging and discharging demand by using a Wardrop first balance principle; and establishing a scheduling potential evaluation model under the condition of comprehensively considering the user activity type and the historical credit influence. The method optimizes the operation efficiency of the power grid, promotes the deep fusion of the new energy vehicle and the power system, improves the overall stability and economic benefits of the system, can provide a scientific basis for an electric power company to formulate reasonable scheduling measures, and enhances the capability of the power grid to cope with the access of large-scale electric vehicles.

The invention provides a mobility management method and device, a storage medium and a program product, relates to the technical field of communication, and is beneficial to realizing integration of communication and perception. The method comprises the following steps: after establishing a connection with a terminal, constructing an association relationship between a communication identifier of the terminal and a first sensing identifier; configuring a corresponding general inductance identifier for the association relationship; and performing mobility management on the terminal based on the sensing identifier.

The invention belongs to the field of artificial intelligence, and relates to a poster generation model training method comprising the following steps: obtaining a poster image with a label, the poster image comprising position label information of layout elements; extracting global image feature information of the poster image by adopting an image network; obtaining feature representation information of the layout elements, and performing self-attention coding on the feature representation information by adopting a layout network to obtain initial position information; performing feature interaction on the global image feature information and the initial position information by adopting an interaction layer to obtain layout element position information; and based on the difference between the layout element position information and the position labeling information, respectively training parameters of the image network, the layout network and the interaction layer so as to construct a poster generation model. The invention further provides a device, equipment and a medium. In addition, the invention also relates to a block chain technology, and the poster image and the target parameter can be stored in a block chain. According to the invention, automation of poster layout can be realized.