5results about How to "Short generation cycle" patented technology

The application discloses an automatic quantitative glue coating and plugging system based on 3D vision, which comprises an industrial computer, a movable platform, a mechanical arm, a 3D camera, a plug stock bin assembly and a terminal device; the mechanical arm is provided with the 3D camera and the terminal device at the tail end; the terminal device comprises a pneumatic gripper assembly, a glue coating assembly and a glue scraping plate; when coating, the industrial computer controls the mechanical arm to drive the 3D camera to move to the top of a hole to be coated to obtain point cloud data of the hole to be coated, and processes the point cloud to obtain the center coordinates, the radius and the depth of the hole to be coated; the industrial computer calculates the glue coating amount according to the radius and the depth of the hole and the glue coating thickness; the plug stock bin assembly pops out a plug matched with the radius of the hole; after the industrial computer controls the pneumatic gripper to take out the popped plug, the industrial computer controls the glue coating assembly to move to the center coordinates to coat glue according to the calculated glue coating amount, and after coating, the industrial computer controls the pneumatic gripper to put the taken plug into the hole.

The application relates to an end-to-end petroleum business knowledge graph construction method based on a large language model, which comprises the following steps: acquiring multi-source data in a petroleum business field to form rich corpus information; constructing a knowledge extraction module, designing a petroleum business field special prompt template, extracting and generating input data under instruction prompts by using a large language model, and outputting entity relationship triple knowledge structure; constructing a knowledge consistency verification module, introducing a thinking chain reasoning technology and a self-reflection mechanism, and performing correction and optimization; constructing a knowledge storage and management module, automatically importing triple knowledge data into a graph data system, establishing a multi-dimensional index structure, and completing preliminary establishment of a knowledge graph; and constructing a knowledge updating and feedback learning module, realizing dynamic evolution and self-adaptive updating of the knowledge graph. The application can realize rapid construction and dynamic updating of a petroleum business field knowledge graph at a low cost.

The application belongs to the technical field of artificial intelligence, and discloses a standard response knowledge unit generation method, device, equipment and storage medium. The processing method comprises the following steps: inputting unstructured documents and historical interaction records of a business field into an initial generation model to obtain a standard response knowledge unit set; submitting the standard response knowledge unit set to a man-machine collaborative quality inspection interface to output structured defect annotations of the unit; forming an optimization sample pair of the standard response knowledge unit and the corresponding correct content of the structured defect annotation to form a model optimization data set; incrementally training the initial generation model based on the model optimization data set to obtain an optimized generation model; and iteratively optimizing the optimized generation model based on the annotation data until the optimized generation model meets a preset standard. The application can be applied to financial technology, medical health and other business management systems, and solves the technical problems of poor generation quality and low generation efficiency of question and answer dialogue sample pairs.

The invention belongs to the technical field of industrial process control, and discloses a jacquard weaving and dyeing intelligent cooperative control method, which comprises the following steps: generating a batch identifier and associating a loom identifier, obtaining a fabric length position and dividing a length section, and establishing a traceable mapping between the length section and a weaving time interval; weaving process data are collected, pattern plate indexes are synchronously extracted, and fabric structure indexes are calculated according to length sections. Before dyeing, a dyeing control track is calculated based on a structure index, a target chromatic value and equipment constraint and closed-loop execution is carried out, during dyeing, a suction state and a chromatic state are updated and an unexecuted set value sequence is corrected, after dyeing, an end point chromatic value corresponds to a length section, a weaving time interval is reversely positioned, and a weaving deviation correction target value or a dyeing parameter correction amount is generated; chromatic aberration fluctuation is reduced; and color fastness stability is improved.

The invention discloses a generation method and device of an optical proximity correction model, a correction method, a medium and equipment. The generation method comprises the following steps: acquiring sample light source information, preset pattern parameters corresponding to a plurality of mask test patterns and actual measurement key sizes; and taking the sample light source information, the plurality of preset pattern parameters and the plurality of actually measured key sizes as inputs of an initial correction model, taking an estimated key size corresponding to the threshold distribution of the photoresist image as an expected output, and optimizing the initial correction model based on an expected matching degree between the estimated key size and the actually measured key size to obtain an optical proximity correction model. The preset pattern density is introduced into the preset pattern parameters, the density effect can be well captured, the critical dimension deviation trend caused by the density effect is reduced, the prediction accuracy of the optical proximity correction model is improved, additional compensation is not needed, the generation step and the correction step of the optical proximity correction model are simplified, and the prediction accuracy of the optical proximity correction model is improved. And the generation period of the model is shortened.