5 results about "Sequence determination" patented technology

The invention relates to the technical field of environmental monitoring and water ecological safety, in particular to a eukaryotic algae outbreak early warning method and system based on genus-level specific recognition. The method comprises the following steps: collecting a water body sample at a monitoring position according to a preset sampling plan, collecting an environment measurement value, and respectively obtaining environment parameter data, a water sample sampling identifier and a sampling timestamp; extracting nucleic acid from the water sample at the monitoring position based on the water sample sampling identifier, performing targeted amplification, and performing high-throughput sequencing at the same time to obtain eDNA original sequencing data; therefore, by constructing the eukaryotic algae outbreak early warning process based on genus-level specific recognition, the problems that in a traditional method, sampling disturbance is uncontrollable, sequence judgment precision is insufficient, trend recognition is lagged, and an early warning link is not transparent are solved, and the accuracy, stability and traceability of early judgment of algae outbreak are improved.

The invention provides a self-adaptive target grabbing sequence determination method, and relates to the technical field of data processing. The method comprises the following steps: segmenting current three-dimensional point cloud data of a target unloading area to generate pose information of a to-be-grabbed object; performing multi-dimensional quantitative evaluation on each to-be-grabbed object based on the pose information to generate a feature score; fusing the feature score with the pose information of the to-be-grabbed object, and constructing a comprehensive state vector at the current moment; inputting the comprehensive state vector into a pre-trained reinforcement learning strategy network to obtain an accumulated return value of each to-be-grabbed object, and selecting the to-be-grabbed object with the maximum accumulated return value as a current grabbed object to be grabbed; and the steps are repeated until the object to be grabbed does not exist in the target unloading area, multi-dimensional physical and geometric characteristics can be comprehensively considered, and therefore the overall efficiency and safety of unloading operation are improved.

This application provides an antenna ordering determining method and a related apparatus, to improve accuracy of antenna ordering detection. The method includes: obtaining a plurality of channel coefficients, where the plurality of channel coefficients include N channel coefficients corresponding to a first terminal, the N channel coefficients include channel coefficients of LOS channels between one antenna of the first terminal and N antennas of a network device, and N is an integer greater than 1; determining a target coefficient combination in M coefficient combinations corresponding to the first terminal, where a utility value of the target coefficient combination is greater than a utility value of another coefficient combination in the M coefficient combinations, each coefficient combination includes the N channel coefficients, channel coefficients included in any two of the M coefficient combinations are arranged in different manners, M=N! or M=(N!) / (2k), N! represents a factorial of N, and M and k are positive integers; and determining, based on a sequence of channel coefficients in the target coefficient combination, whether an antenna ordering is correct.

Embodiments of the present application disclose a ZC sequence determination method, device, equipment and storage medium. The ZC sequence determination method comprises: determining a fixed-point query table according to the length of a ZC sequence; determining a to-be-queried index according to the fixed-point query table and sequence correlation data of the ZC sequence; querying a basic complex number table of the ZC sequence according to the to-be-queried index; and determining target ZC sequence data according to the basic complex number table of the ZC sequence. The technical scheme of the embodiments of the present application can reduce the error of the ZC sequence determined by the table lookup method and improve the accuracy of the ZC sequence determined by the table lookup method.

The invention relates to the technical field of semantic network reasoning, in particular to a knowledge base and map collaborative enhancement large model reasoning system and method, and the system comprises a structure guiding module, a semantic difference module, a stability extraction module, a path comparison module and a label binding module. According to the method, graph path node semantic vectors and channel numbers are extracted, response entity fitting degrees and sequences are combined, structured distribution information is constructed, semantic path recognition precision is enhanced, difference entities are screened on the basis of semantic distances, and stability parameters are generated in combination with context word frequencies, co-occurrence frequencies and document distribution; semantic conflicts are positioned, paragraph stability is quantified, alignment of semantic contents and structural paths is achieved, causal chain consistency is improved, the relation between response paragraphs and path nodes is marked, traceable chains are formed, reasoning transparency and result verifiability are enhanced, and semantic calculation, structure extraction, sequence judgment and other means are fused. And the reasoning accuracy, interpretability and credibility are improved.