79 results about "Information science" patented technology

The invention provides a GIS industry large language model parameter fine tuning method considering parameter adaptability difference, and relates to the field of geographic information science and deep learning, the method comprises the following steps: giving a GIS downstream task training data set and a pre-training large language model, and calculating the adaptability of each parameter of the model to the data set; calculating the adaptability of each level of the large language model according to the calculated adaptability of each parameter; based on the adaptability of each layer, distributing different trainable parameters for the model layers with different adaptability; and based on the distributed trainable parameters, performing fine tuning training on the pre-trained large language model to obtain a GIS industry large language model, and completing efficient fine tuning of the large language model parameters. According to the technical scheme, the heterogeneity of the large language model parameters in the GIS professional knowledge adaptation process is considered, and the characteristic is utilized to optimize the LoRA fine adjustment process so as to improve the performance of the large language model on GIS professional tasks.

The invention discloses an electroencephalogram signal clustering method based on a Mama framework and comparative learning, and belongs to the crossing field of engineering application and information science, and the method comprises the following steps: collecting and preprocessing label-free electroencephalogram data; the method comprises the following steps of: constructing a Mama-based feature extractor, dividing an input electroencephalogram signal into a plurality of slices by adopting an electroencephalogram signal slice embedding strategy so as to obtain fine-grained local information, and modeling a potential context relationship by utilizing a slice perception scanning mechanism; the method comprises the following steps: constructing an original data view and an enhanced data view, and respectively inputting samples of the two data views into two Mamba feature extractors with shared network parameters for feature embedding; processing the feature pairs of the two views using an instance projector, constructing weighted instance level contrast learning wherein the distance in the original data space provides weight information; processing feature pairs of the two views by using a clustering projector, and constructing a clustering distribution discrimination contrast learning branch and a semantic perception contrast learning branch; and finally, performing joint training by using the comparative learning branches, and outputting a high-quality clustering result by ensuring feature, clustering distribution and semantic consistency.

The invention discloses a GeoJSON-oriented geospatial analysis execution method and system, and relates to the technical field of geographic information science and artificial intelligence. One embodiment of the method comprises the following steps: acquiring a GeoJSON file to be processed; acquiring a natural language task instruction input by a user, performing semantic understanding on the task instruction by using a first language model, extracting an operation target and a constraint condition, and generating an overall task; disassembling the whole task into a plurality of logically ordered sub-tasks to form a sub-task sequence; calling a function library API or generating a script code, executing the subtask, and generating a subtask result; continuously receiving and evaluating the generated sub-task results, and updating the sub-tasks which are not executed completely according to the evaluation result of the sub-task results until all the sub-tasks are completed; and integrating the subtask results to generate an output result. In this way, efficient, reliable and controllable automated geospatial analysis driven by natural languages is achieved.

The invention belongs to the technical field of geographic information science, and particularly relates to a Markov-FLUS model-based feedback land utilization change simulation method and system, and the method comprises the steps: firstly predicting an initial area target vector based on historical data and a Markov chain, and inputting the initial area target vector into a FLUS model to carry out the first space distribution simulation; by calculating the achievement degree of the simulation result and the area target, the space conflict type which cannot be achieved due to the space constraint is identified, the area target vector is dynamically corrected according to the space conflict type, and the corrected target is fed back to the FLUS model for iterative simulation until the result is converged. According to the method, the technical problem of disjunction of area prediction and spatial layout in a traditional coupling model is effectively solved, and the scientificity and accuracy of land utilization change simulation are remarkably improved.

The invention relates to the technical field of crossing of hydraulic engineering informatics, computational fluid mechanics and artificial intelligence. The invention discloses a river channel flow field reconstruction method, system and device and a storage medium, and the method comprises the steps: a data preprocessing module carries out the rasterization of unstructured hydrodynamic data, and extracts dynamic hydraulic boundary parameters and static geometric features containing a terrain and sign distance function; performing deep semantic coding on the boundary conditions by using an attention mechanism, and performing multi-modal splicing fusion on the boundary conditions and the static geometric features to generate a mixed feature map; based on the feature map, recovering flow field details step by step through cascade space up-sampling and convolution operation to generate a high-resolution prediction flow field; and further performing reverse normalization, and using a symbol distance function as a physical mask to identify a land invalid value. According to the method, the advantages of global attention and local convolution are fused, the proxy model is constructed, and high-precision and physical consistency rapid prediction of the complex river channel flow field is realized.

The invention belongs to the technical field of medical care informatics, and discloses a remote central monitoring system for community medical treatment, which comprises a data acquisition and integration module, a physiological state analysis engine and a monitoring information presentation module, and can construct a state-dependent baseline model cluster based on the activity state information of a user, according to the real-time activity state of the user in the monitoring period, the corresponding sub-model is dynamically selected to serve as a judgment reference, and then a continuously-changing physiological stability index is generated by comparing the multi-dimensional statistical characteristics of the real-time physiological data with the deviation degree of the reference; according to the method, by establishing an individualized physiological steady state judgment benchmark which is adaptively adjusted along with the activity situation, the capability of a monitoring system for continuously and quantitatively evaluating the overall stability of a human body as a complex system is improved, and the transformation from passive detection of discrete clinical events to state insight of a risk evolution process is realized.

The invention belongs to the technical field of geographic information science space optimization, and particularly discloses a low-altitude vertical take-off and landing field site selection method fusing mobile phone signaling big data dynamic maximum coverage, which comprises the steps of fusing multi-source geographic space big data, and constructing an urban low-altitude vertical take-off and landing field candidate position set; defining a multi-time-sequence take-off and landing field maximum coverage site selection model, and determining an optimization objective function, constraint conditions and a decision-making method; a genetic algorithm is used as a basic algorithm model, and three limiting conditions of a multi-time-sequence influence factor, total quantity constraint and integer programming are added to form a multi-time-sequence integer constraint spatial evolution algorithm; based on MP-ICSEA, an MP-MCLP problem is solved, an optimal solution of service facility configuration is obtained, the effectiveness of vertical take-off and landing field site selection and vertical take-off and landing aircraft redeployment along with the change of the flow of people is verified, and therefore the reasonability of vertical take-off and landing field site selection and the efficiency of vertical take-off and landing aircraft redeployment according to the time sequence are greatly improved.

The embodiment of the invention provides a medical progressive diagnosis method based on dynamic visual focusing, and belongs to the technical field of medical care informatics, and the method specifically comprises the steps: 1, obtaining medical image data and clinical query data; 2, constructing a hybrid reinforcement learning framework, and training a network model through a bimodal exploration mechanism and a double-flow optimization mechanism based on medical image data and clinical query data to obtain a trained diagnosis model; and step 3, inputting the to-be-diagnosed medical image and the corresponding clinical query into the diagnosis model, and outputting a diagnosis report containing the progressive visual focusing trajectory and the final diagnosis result. Through the scheme disclosed by the invention, the expert-level diagnosis thinking can be simulated to realize accurate diagnosis without relying on large-scale marking reasoning data, the problems of strategy collapse and visual illusion of pure online learning are effectively solved, and the diagnosis accuracy and interpretability are improved.

The application relates to the technical field of medical care information science, and discloses an old-age health prediction method and system, which comprises an information collection module, a phase-locked analysis module and a trend determination module, the information collection module synchronously acquires a behavior perturbation sequence and a physiological response sequence; the phase-locked analysis module calculates dynamic phase offset characteristics between the behavior perturbation sequence and the physiological response sequence, and determines a logical deviation residual of a measured phase offset relative to a reference mapping track by using a logic transfer check unit; and the trend determination module monitors a second-order cumulative change rate of the logical deviation residual in a sliding window, and outputs a warning signal when the second-order cumulative change rate exceeds a stability threshold value; the application can perceive the deterioration of body feedback regulation flexibility before a physiological parameter appears abnormal, and avoids prediction failure caused by a physiological compensation mechanism of an old individual by quantifying the information transmission efficiency of a feedback loop.

The application relates to an Internet autonomous domain type mapping method related to the dynamic characteristics of a topology network and belongs to the field of computer and information science. First, autonomous domain registration information is obtained through a whois database; second, BERT is used to extract semantic features of the registration information, and meanwhile, the similarity between different autonomous domains is calculated according to specific autonomous domain attributes in the registration information; then, an autonomous domain topology network is generated in combination with the specific autonomous domain attributes and the similarity between the autonomous domains, and time attention mechanism is used to extract dynamic characteristics of the topology network; finally, semantic features and topology network dynamic characteristics are fused to perform Internet autonomous domain type mapping. In view of the problem that the autonomous domain registration information is incomplete, an autonomous domain topology network dynamic characteristic extraction method is proposed, and the accuracy of Internet autonomous domain type mapping is improved.

This invention relates to a multi-temporal radiation signal inversion method based on Conditional Transformer, belonging to the field of satellite remote sensing data processing / spatial information science and technology. The method includes: acquiring meteorological time-series data of the target area to establish a dataset; designing a Conditional Transformer deep learning network; combining the target time-series data and meteorological data encoded by time and meteorological code generators with the output of the Encoder layer, and inputting this combination into the Decoder layer to train the Conditional Transformer deep learning network to obtain a surface radiation signal inversion model; inverting surface radiation brightness temperature data and evaluating the inversion accuracy; efficiently capturing long-distance spatiotemporal dependencies of radiation signals; flexibly inverting radiation changes according to the target time and meteorological conditions; removing nonlinear interference from the atmosphere on radiation; and more accurately inverting the true radiation characteristics of the surface.

This invention discloses a method for probabilistic quantification of mixed states in urban functional areas by integrating multi-source heterogeneous data, belonging to the fields of geographic information science, smart cities, and artificial intelligence. The method includes: S1, constructing a traffic analysis area; S2, acquiring multi-source data; S3, preprocessing the multi-source data acquired in S2; S4, generating true probability distribution labels; S5, extracting socio-semantic features; S6, extracting dynamic activity features; S7, extracting physical scene features; S8, feature fusion and distribution prediction; S9, model training; and S10, determining the mixed state and outputting the results. This invention achieves refined quantification of the mixed states of urban functional areas, outputting a continuous probability distribution vector rather than a single category label, significantly improving the information density and application value of the recognition results.

The invention relates to the field of medical care informatics, and discloses a lupus nephritis condition monitoring system based on TWEAK and CD163 dynamic change, which comprises a data adaptation gateway module, a data rationality arbitration engine, a time sequence regularization engine module, a high-order feature extraction engine module and a risk layering module, the time sequence regularization engine module switches an interpolation or smoothing algorithm based on a qualitative result of the data rationality arbitration engine and a time domain where a data point is located so as to generate a continuous time function. Through the information processing architecture, the technical problem that sparse asynchronous original medical data is prone to being affected by outlier pollution and end point distortion is solved, and the accuracy of data processing is improved. The system reconstructs discrete data snapshots into continuous mathematical objects, provides a stable and analyzable technical basis for subsequent extraction of interpretable high-order dynamic features, and improves the credibility of clinical decision support.

The invention provides a construction method of a life-like robot or intelligent system based on a life entropy minimization principle. The method aims at solving the problem that in the prior art, unified description and quantification of vital activity essence are lacked from the aspects of mathematics and information science, and a similar life intelligence system is not built accordingly. According to the method, a sensing and internal steady state system of a machine is constructed by defining ternary state coding (positive surprising, negative surprising and ground state) of a self-information unit (); constructing each function module topology of the machine by using a self-information complex (); and taking the minimum life entropy value as a core objective function of a machine system. A machine system eliminates a surprising state with high information content in real time through a PID feedback mechanism, so that the surprising state spontaneously returns to a low-entropy ordered ground state. According to the invention, the machine is endowed with the instinct of'benefiting and avoiding harm 'similar to organisms, and the method is suitable for constructing life-like robots or intelligent systems with high adaptivity.

The application discloses a kind of perfect preset area history flood disaster database method and system, belong to flood disaster information science and big data technology field;Including the following steps: obtaining preset area history hourly precipitation grid data, generates hourly grade grid map according to rainstorm threshold classification, connects the rainstorm area of corresponding moment of adjacent same grade grid;Combined with the process of spatial and temporal splitting and merging, complete rainstorm process is aggregated;Matching the corresponding city of preset area rainstorm flood related social media data, associate rainstorm process with city, extract corresponding moment data to generate pseudo document to determine city flood event;Linkage disaster city, build cross-city transmission chain;Based on flood propagation chain, flood event number caused by rainstorm, the number of affected cities, the start and end time of each city flood and cross-city chain are written into history flood disaster event database;The application can accurately identify and track flood events in space and time.

1. Name of the product in this design: Multifunctional Signal Analyzer. 2. Application of this design: This design is intended for use in information science electronic measuring instruments. 3. The key design feature of this product is its shape. 4. The image or photograph that best illustrates the design's key points: a 3D model.

The present application belongs to the technical field of medical care information science, and discloses a community medical remote central monitoring system, which comprises a data acquisition and integration module, a physiological state analysis engine and a monitoring information presentation module.The system can construct a state-dependent baseline model cluster based on the activity state information of a user, dynamically select a corresponding sub-model as a judgment benchmark according to the real-time activity state of the user in a monitoring period, and then generate a continuously changing physiological stability index by comparing the deviation of the multi-dimensional statistical characteristics of real-time physiological data from the benchmark.The present application establishes an individualized physiological steady state evaluation benchmark that is self-adapted to activity contexts, improves the ability of the monitoring system to continuously quantitatively evaluate the overall stability of the human body as a complex system, and realizes the transition from passive detection of discrete clinical events to state insight into the risk evolution process.

The invention discloses a man-machine collaborative geographical science popularization narrative generation and interactive teaching system and method, and relates to the technical field of wisdom education technology and geographical information science. The system comprises a user interaction module, an AIGIS analysis engine, a narrative generation engine, a story state manager and a multimedia presentation module. The user interaction module receives user input and interaction instructions, identifies user requirements and selects a corresponding interaction mode. And the AIGIS analysis engine executes teaching adaptive spatial analysis according to user input and outputs a structured result. And the narrative generation engine generates a scenario type narrative script and a guiding problem based on the structured analysis result in combination with a preset narrative generation logic. And the story state manager monitors events through rules, maintains task states and dynamically drives narrative flow adjustment. And the multimedia presentation module integrates the scenario type narrative script, the guiding problem and the visual content and multimedia resources generated based on the structured analysis result to perform immersive presentation.

This invention relates to a model theft detection method that combines training set data distribution and W-distance, belonging to the field of computer and information science and technology. First, the invention utilizes the VAE method to reduce the dimensionality of the training and query sets. Second, it uses maximum likelihood estimation to calculate the probability distribution of the query set, and samples multiple sets of samples to be detected according to this probability distribution. Then, for each set of samples to be detected, the same number of reference samples are randomly sampled from the training set, and the W-distance between each set of samples to be detected and the reference samples is calculated. Finally, the ratio of the number of categories in the reference samples to the total number of categories is used as a weight to calculate all W-distances. When the weighted calculation result is greater than a detection threshold, model theft is detected. This invention proposes a model theft detection method that correlates the training set data distribution, considers the characteristics of both the query set and training set sample distributions, and improves the W-distance calculation method, effectively enhancing the accuracy of model theft detection.

The application provides a fine-grained geographic spatial knowledge graph construction and query method and system, relates to the technical field of geographic information science, and comprises the following steps: expressing and calculating spatial relationship through a DE-9IM model; defining a fine-grained geographic spatial knowledge graph and a strong geographic spatial relationship; extracting target information in the data set to obtain a knowledge graph ontology structure; using a spatial interconnection algorithm to obtain SGRs and construct a knowledge graph; defining SGR query, dividing the SGR query into accurate query and approximate query; accurately querying the fine-grained geographic spatial knowledge graph FineGeoKG through an accuracy query algorithm; and approximately querying the fine-grained geographic spatial knowledge graph FineGeoKG through an approximate query algorithm. The fine-grained geographic spatial knowledge graph (FineGeoKG) and the related algorithm provided in the application adopt new technical means and methods, improve the efficiency of solving the spatial relationship related problems in the geographic space, and provide a reference solution for similar problems.

The invention discloses an AI-GIS system based on a multi-agent fine-tuning language model, and belongs to the field of artificial intelligence and geographic information science. According to the system, firstly, a task disassembling corpus and a tool matching corpus are constructed, and corresponding base models are finely adjusted according to the task disassembling corpus and the tool matching corpus, so that a task disassembling agent and a tool matching agent are obtained; designing a tree-shaped multi-agent framework, and associating an external application program interface of the GIS tool; when the AI-GIS system runs, a complex GIS task is automatically decomposed into sub-tasks, a proper GIS tool is matched for each sub-task, and finally the function of intelligent construction of the GIS workflow is achieved. According to the method, the use threshold is greatly reduced, operation can be carried out without professional knowledge, the processing efficiency and accuracy of the method exceed those of a mainstream scheme, the system is light in weight, easy to expand and low in resource requirement, a complete closed loop from task input to result output can be finally achieved, and available achievements can be directly delivered.

The invention discloses a DEM (Digital Elevation Model) sunshine duration method and system based on GPU (Graphics Processing Unit) and terrain entropy optimization, and belongs to the technical field of geographic information science and space calculation. The method comprises the following steps: firstly, calculating topographic information entropy by using digital elevation model data to quantify topographic complexity of different regions; and the terrain information entropy is used as a self-adaptive driving factor to dynamically adjust the search radius and the search step length of the sun visible range, so that the condition that no shielding exists in a terrain gentle region is quickly judged, and the terrain shielding relationship is finely simulated in a mountain region or a complex terrain region. Besides, the method is realized by adopting a CUDA (Compute Unified Device Architecture) parallel architecture, and is combined with a GPU (Graphics Processing Unit) parallel strategy of partitioning and partitioning, so that the calculation load of high-resolution data is effectively reduced, and the overall operation efficiency and the memory utilization rate are improved.

This invention relates to an automated penetration testing method based on reinforcement learning, belonging to the field of cyberspace security in computer and information science and technology. First, this invention utilizes environment-related parameters to construct the state space, action space, and reward function of the reinforcement learning model to generate training data. Second, it applies the training data and vulnerability exploitation components to train and generate a reinforcement learning model. Finally, based on a vulnerability exploitation component library, the reinforcement learning model recommends optimal vulnerability exploitation information, thereby automating the entire penetration testing process. This invention addresses the problems of manual reliance in penetration testing and the rigid application of vulnerability exploitation components by employing artificial intelligence technology to effectively improve the accuracy of vulnerability exploitation component recommendations, significantly enhancing the efficiency of penetration testing.

This invention belongs to the field of computer and information science and technology, and relates to a method, apparatus, equipment, and medium for extracting key features of power network assets. The method employs an improved k-means clustering algorithm to perform cluster analysis on power network asset data to obtain key features of the power network assets. The improvement of the k-means clustering algorithm lies in: calculating the Mahalanobis distance between each data point and each cluster center during the clustering process, and assigning the data point to the cluster center with the shortest distance; updating the cluster centers based on dynamic weight updates and adaptive learning rates; and using an attention mechanism to perform weighted feature summation and nonlinear transformation processing on the key features obtained from clustering to obtain an optimized feature vector matrix. This provides strong support for subsequent asset management, fault diagnosis, and predictive maintenance.

The invention relates to the technical field of medical care informatics, and discloses a medical health information processing system for blood test data, comprising: a time sequence data gating unit for acquiring test data and calculating a time interval, and calculating a unit time change rate of an index when the interval falls into an effective physiological response period; the polarity shunting locking unit is used for locking a positive or negative driving data set as a reference source according to the positive and negative polarities of the change rate; the collaborative drift calculation unit is used for calculating the instantaneous collaborative slope and the deviation degree of the instantaneous collaborative slope relative to the reference collaborative slope; the state closed-loop output unit generates a state signal when the deviation degree exceeds a relative steady-state threshold, through a bidirectional reference locking and timeliness gating mechanism, the problem of data analysis distortion caused by a physiological hysteresis effect and sampling discreteness is solved, and the signal-to-noise ratio and accuracy of physiological state evaluation under the non-equidistant sampling condition are effectively improved.

The utility model belongs to the technical field of information technology experiment boxes, and particularly relates to an information technology experiment box which comprises a bottom shell, the top of the bottom shell is provided with a face shell, one side of the face shell is rotatably provided with a buckle, and one side of the bottom shell is fixedly connected with a buckle seat. A placing mechanism is arranged in the bottom shell; wherein the placing mechanism comprises a large support and a clamping groove, a placing groove and a buckling groove are formed in the surface of the large support, and a splicing block is fixedly installed at the bottom of the large support. According to the information science and technology experiment box, customized design is adopted, each module is provided with a fixed placement position in the experiment box under the action of the placement grooves, so that storage management is facilitated, equipment can be quickly counted and recycled after class, most of the equipment is assembled through buckles and bolts, the assembly cost is reduced, and the assembly efficiency is improved. Meanwhile, the length and the width of the whole size are A4 paper, so that students do not occupy too much space during desktop operation.

The invention relates to an intelligent address resolution method based on word segmentation engine antagonism error correction, and belongs to the technical field of natural language processing and geographic information science. The method comprises the following steps: collecting multi-source heterogeneous address data, carrying out format standardization processing to obtain a standardized address, constructing an address corpus, storing the address corpus in a structured database, and synchronously establishing an incremental updating mechanism to guarantee the timeliness of corpora; a real-time input layer generates a noisy training set; a feature engineering layer outputs a fusion feature vector; a convolution layer and an LSTM layer capture local and global features respectively; an output layer is subjected to preliminary error correction by a pre-training language method and secondary verification of adversarial training error correction; outputting a structured result with a confidence coefficient and an error correction record; and carrying out batch verification on the analysis result and marking errors. Precise processing of Chinese address diversity and non-standardization problems is realized, manual dependence is reduced, and analysis robustness, accuracy and scene adaptability are improved.

1. Name of the product in this design: Bookshelf (Library and Information Science). 2. Purpose of this design: For placing books. 3. The key design feature of this product is its shape. 4. The image or photograph that best illustrates the design's key points: 3D view 1.

The invention discloses a geographic entity name derivation relation identification method, device and equipment, a medium and a product, and relates to the field of geographic information science, and the method comprises the steps: obtaining a target general name, a geographic entity name and derivation element attributes; identifying whether the geographic entity name is a derived place name of the target general name or not by adopting a derivation identification model; if yes, obtaining a plurality of candidate native place names according to the derived geometric parameters; performing feature engineering processing on the geographic entity name, the derived element attribute, any candidate native place name and the native element attribute to obtain a fusion feature corresponding to the candidate native place name; according to the fusion features, adopting a native probability model to obtain the probability that the candidate native place name is the native place name of the geographic entity name; according to the method, the neural network and the machine learning model are adopted, the space-time distribution features and the semantic features are fused, and the automation degree and accuracy of identification of the derived geographical names and the derived relations of the derived geographical names are improved.

The embodiment of the invention provides an aortic dissection risk prediction method, and belongs to the technical field of medical care informatics, and the method specifically comprises the steps: obtaining the Fourier transform infrared spectrum, Raman spectrum and clinical data of a serum sample of a to-be-detected object; preprocessing the Fourier transform infrared spectrum and the Raman spectrum, and standardizing clinical data; respectively extracting multi-scale features of the infrared spectrum and the Raman spectrum through a multi-scale feature extraction module; inputting the infrared and Raman spectrum features of the same scale into a cross-modal fusion module to obtain enhanced cross-modal features; performing information interaction through a multi-scale interaction module to obtain unified feature representation; and outputting a disease risk prediction result or a prognosis risk prediction result by using a classification and discrimination module based on the unified feature representation and the clinical data. According to the scheme, the stability and accuracy of disease risk prediction and prognosis risk prediction are improved.