5results about How to "Show superiority" patented technology

The invention provides a metasurface optimization design method based on reinforcement learning, relates to the technical field of electromagnetic metasurfaces, and solves the problems that a metasurface design method based on supervised learning highly depends on a data set and the manufacturing cost of the data set is high. The method comprises the following steps: constructing a reinforcement learning environment model corresponding to a metasurface, and constructing a value network corresponding to a reinforcement learning algorithm; based on the constructed reinforcement learning environment model and the value network, through an intelligent agent and environment interaction mode, an objective function of the reinforcement learning environment model is updated, decision making of an intelligent agent is optimized, after a preset optimization requirement is met, the interaction process is stopped, and a metasurface optimization design result is output; wherein the learning environment model is reinforced to give an agent reward by comparing the difference between a target function value of a new state and a given optimization value; the method provided by the invention has stronger solving capability for nonlinear and uncertain complex environments, has the characteristics of intelligence and rapidness, and reduces data set dependency and manufacturing cost.

The invention discloses a rainstorm disaster event argument extraction method and device based on event text enhancement, and the method comprises the steps: constructing a rainstorm disaster event template, and defining an event type and an argument role; extracting data related to the rainstorm disaster from the microblog platform; preprocessing the data; labeling the preprocessed data to obtain a data set; generating a problem template; performing event text enhancement; integrating the problem template and the enhanced event text as model input; encoding the integrated text by using a pre-trained Chine-BERT-WWM model, and obtaining a hidden layer representation; deep features are extracted through an intermediate linear layer, and starting and ending positions of an input sequence are predicted by using a full connection layer; according to the method, the rich MRC data set can be directly utilized to enhance event extraction, the problem of data scarcity is relieved, and possibility is provided for zero sample event extraction of invisible event types.

The invention discloses a tunnel two-parameter safety evaluation index construction method based on dislocation and strong earthquake coupling, and belongs to the technical field of tunnel safety evaluation. The method comprises the steps that tunnel characteristic parameters and stratum rock-soil mechanical characteristic parameters are obtained, the fault dislocation size is determined, and seismic waves conforming to a standard spectrum are selected; establishing a multi-feature coupled nonlinear dynamic finite element numerical analysis model; on the basis of the fault dislocation size and the selected seismic wave, the nonlinear dynamic finite element numerical analysis model is adopted to calculate an incremental dynamic time history, and a dynamic response curve of the tunnel lining structure under the action of multiple disasters is obtained; establishing a tunnel two-parameter safety index function considering the coupling effect based on the tunnel radial convergence rate and the plastic strain; and determining an index damage state and threshold division based on the tunnel two-parameter safety index function and the lining response-structure performance curve, and taking the index damage state and threshold division as a tunnel two-parameter safety evaluation index.

This invention discloses a method for segmenting colorectal cancer polyps based on cross-domain feature fusion. It relates to the field of medical image segmentation technology. The method includes: acquiring an image dataset with diverse polyp morphologies; constructing an FFNet network model, comprising: a pyramid visual encoder, a multi-scale feature enhancement module, an adaptive filtering module, and a similarity aggregation module connected sequentially; and fusing image domain and frequency domain features from the image dataset using the FFNet network model to achieve polyp segmentation. This invention achieves more accurate and comprehensive segmentation through the FFNet network model.

The invention provides a non-viral vector for treating phenylketonuria as well as a preparation method and application of the non-viral vector, and belongs to the technical field of biological pharmacy. The non-viral vector for treating phenylketonuria comprises a lipid vector and a nucleic acid molecule for editing PAH gene expression, the lipid carrier comprises a cationic lipid and an auxiliary lipid, and the auxiliary lipid comprises a pH sensitive lipid component and an immune escape promoting lipid component; the nucleic acid molecule for editing PAH gene expression comprises DNA (deoxyribonucleic acid), siRNA (small interfering ribonucleic acid), mRNA (messenger ribonucleic acid) or sgRNA. The non-viral vector is prepared by screening lipid components and proportions. The non-viral vector prepared by the invention is uniform in particle size and stable in system, can wrap or load a gene editing drug, is high in drug encapsulation efficiency, has relatively high drug delivery efficiency for hepatocytes, can obviously recover the phenylalanine hydroxylase level, and achieves the effect of treating phenylketonuria.