6results about How to "Guaranteed space" patented technology

This invention discloses a method and system for identifying abnormal driver behavior based on continuous learning, relating to the field of driving behavior detection technology. The method includes dividing a driver monitoring image into image blocks and projecting them into feature vectors, adding positional encoding, and combining extracted prompt features with segmentation into prompt vectors to achieve dual coverage of global and key local features, improving feature extraction accuracy. The two types of vectors are concatenated into a fused feature sequence, input to a pre-trained Vision Transformer model with frozen parameters, and enhanced features are extracted through a self-attention mechanism. Then, a global feature vector is obtained through average pooling. Continuous learning is achieved by outputting a vector of all seen task logic values, using a binary mask for filtering, and classifying using the argmax function. This allows for adaptation to behavior category expansion, balancing real-time performance with stability in complex environments, improving the accuracy of similar behavior recognition, and making it easily deployable in in-vehicle scenarios.

The application relates to the field of robot control, in particular to a multi-modal perception pipeline inspection robot control method and an inspection robot. The method comprises the following steps: based on the inspection robot, pipeline detection perception is carried out, and a real-time three-dimensional structure map is generated; the real-time three-dimensional structure map is compressed and uploaded to a cloud monitoring platform; when the cloud monitoring platform receives the real-time three-dimensional structure map, spatial topology analysis is carried out, an inspection instruction is generated, and the inspection instruction is sent to the pipeline internal inspection robot; the internal inspection robot receives the inspection instruction, carries out adaptive path planning, and generates a path execution sequence; the inspection robot carries out dynamic inspection operation based on the path execution sequence, synchronously collects data and returns the data to the cloud monitoring platform, and completes the inspection control operation. The application realizes high-precision and automatic buried oil pipeline inspection operation, and improves the detection efficiency and safety of the buried oil pipeline.

The application discloses a road segmentation method based on dual-modal feature uncertainty fusion, aiming at solving the problems of modal feature conflict, poor adaptability of fixed weight fusion and insufficient utilization of modal uncertainty in existing depth camera road segmentation; the method constructs an integrated framework of dual-modal feature extraction and uncertainty perception fusion: for color images, a pure thinking method module is adopted to extract global semantic features, and for depth data, a convolution network with fused asymmetric dilated convolution is adopted to enhance spatial geometric perception; through a multi-scale dilated attention module, feature interaction is optimized, and pixel-level classification evidence is generated by using evidence collection; based on the Dempster-Sheffer evidence theory, dual-modal uncertainty is quantified and dynamically fused, the fusion weight is adaptively adjusted, and feature conflict and information redundancy are effectively inhibited; the application significantly improves the road segmentation precision and robustness in complex scenes, and provides reliable drivable area perception information for an automatic driving system.

The invention relates to the technical field of compound fertilizers, in particular to a preparation process of a slow-release compound phosphate fertilizer based on multistage activation. The preparation process comprises the following steps: carrying out powder refining treatment on a phosphorus-based raw material to obtain a pretreated phosphorus material; mixing the pretreated phosphorus material with a weak acid or organic acid activation medium to obtain a dissolution activation material; s3, adding a salifying regulation component and a complexing regulation component into the dissolved activated material to obtain a secondary activated mixture; s4, granulating the secondarily activated mixture to obtain fertilizer core particles; s5, interface activation: performing interface activation treatment on the surfaces of the fertilizer core particles to form an interface activation layer; s6, sequentially constructing an inner quick-release layer, a middle reaction slow-release layer and an outer controlled-release layer on the surfaces of the fertilizer core particles; and S7, drying the coated particles, and screening and finishing the particles to obtain the slow-release composite phosphate fertilizer finished product. The problems that in the prior art, the available phosphorus activation rate is low, the coating binding force is poor, the release curve cannot be adjusted, and the batch consistency is poor are solved.

The invention relates to the technical field of concrete admixture preparation, and particularly discloses a cement concrete admixture and a preparation method and application thereof.According to the cement concrete admixture, through multi-synergistic copolymerization of a polyether macromonomer, acrylic acid, modified cellulose and sodium allysulfonate in a specific proportion, the cement concrete admixture is prepared; the comprehensive performance of the concrete is improved. Methylallyl alcohol polyoxyethylene ether and acrylic acid construct a main framework of the admixture, so that the steric hindrance and the initial dispersion force of a foundation are guaranteed; by sulfonating cellulose and introducing a negatively charged sulfonic acid group, the flocculation structure among cement particles can be broken, and the initial water-reducing rate and the flowing speed are remarkably improved; and by introducing a carbon-carbon double bond, the modified cellulose can be introduced into a polycarboxylic acid macromolecular chain through a copolymerization reaction, and a rigid cellulose skeleton and a flexible polyether side chain jointly construct a three-dimensional space network structure, so that the steric hindrance effect of the admixture is jointly improved.

The invention discloses an unmanned aerial vehicle aerial survey image real-time interpretation and ground object labeling method and system, and relates to the field of real-time interpretation and ground object labeling, and the system comprises an acquisition positioning module, a visual interference module, a spatio-temporal evolution module and a determination processing module. The method comprises the following steps: determining a pose calculation bumping event and an occurrence time window thereof by combining clustering and correlation analysis of a characteristic curve, identifying a causal interference domain causing bumping by analyzing a dense motion vector field and a characteristic loss region between images, reversely projecting the causal interference domain to a physical space, determining a risk analysis region, and determining a risk analysis result. Comprehensively calculating multiple risk factors in the region, and generating a drift risk spatio-temporal evolution prediction map; according to the method, the space-time alignment and the unified mapping are carried out, the comprehensive annotation risk index of each grid is decoded, and a hierarchical response mechanism is started, so that the final effectiveness judgment and processing of the surface feature annotation are obtained, and the space credibility and the time sequence availability of the annotation result are fundamentally guaranteed.