24results about How to "Avoid low accuracy" patented technology

The invention relates to an accelerator chip instruction operation analysis method and device, equipment, a medium and a product. The method comprises the following steps: obtaining a plurality of data streams obtained by executing a binary file corresponding to a target kernel function by an accelerator chip, then obtaining a plurality of assembly instruction codes corresponding to the binary file according to the binary file, and then obtaining a plurality of assembly instruction codes corresponding to the binary file according to a time sequence of a clock cycle, performing matching processing on each assembly instruction code and a data stream corresponding to the assembly instruction code in sequence to obtain an instruction code analysis result corresponding to the assembly instruction code, the instruction code analysis result comprises at least one of the number of clock cycles required by the accelerator chip to execute the assembly instruction code and the number of execution times corresponding to the assembly instruction code. By adopting the method, the accuracy of analysis results can be improved.

The invention discloses a service analysis report generation method and device and a computer storage medium. In the scheme, report demand information is acquired; the report demand information comprises key information required for generating the service analysis report; mapping the report demand information into a plurality of service statistical indexes according to a preset index relation tree; the preset index relation tree is used for representing a corresponding relation between the business statistical indexes; obtaining index data corresponding to each service statistical index; generating a target cue word based on the index data and a preset cue word template; and inputting the target cue word into the large language model, and performing analysis and reasoning on the target cue word by the large language model to generate a business analysis report. According to the technical scheme, the problem that the accuracy of the generated service analysis report is low due to the fact that a large language model is difficult to accurately understand specific vocabularies in the report demand information is solved, and the accuracy of generation of the service analysis report is improved.

The invention discloses an electrolytic cell fault analysis method, device and equipment, a storage medium and a related product. The method comprises the following steps: acquiring a training data set; performing data supplementation on the training data set to generate a supplemental data set; the supplementary data set comprises a fault type set and multiple groups of basic information corresponding to each fault type in the fault type set; performing clustering processing on the supplementary data set, and establishing a corresponding relation between each fault type and response characteristics; the response features are common feature data of multiple groups of basic information corresponding to the fault type; and on the basis of the corresponding relation between each fault type and the response feature, a fault type and response feature matching database is constructed, electrolytic cell fault diagnosis is carried out on the basis of the fault type and response feature matching database, and fault diagnosis can be carried out on the electrolytic cell on the basis of a small number of training data sets.

The invention discloses a highway bridge safety state assessment method and system, and relates to the technical field of highway bridge state assessment, and the method comprises the steps: collecting the time sequence vibration data of a plurality of key nodes of a highway bridge and the overall surface image data; inputting each time sequence vibration data into the improved generative adversarial network, generating denoised time sequence vibration data, and performing initial state evaluation according to the denoised data to obtain an initial evaluation value; wherein a generator in the network adopts an encoder-decoder structure, and a channel attention mechanism is additionally arranged between feature encoding and feature decoding; identifying all cracks in the surface image data, performing region division on the cracks, calculating according to the crack of each region to obtain a secondary evaluation value, and using the secondary evaluation value to assist in adjusting the initial evaluation value; and for each adjusted evaluation value, integration is carried out based on the support degree between the data, and an overall evaluation result is obtained. The accuracy and efficiency of highway bridge safety state evaluation can be effectively improved.

The invention discloses a warehouse data intelligent inventory system and method, the system comprises a warehouse intelligent system and a background server, the background server is used for loading a program to execute the following steps: obtaining the occurrence frequency of a first abnormal condition according to a set data statistics time period, the first abnormal condition is determined according to monitoring management data obtained from a warehouse intelligent system, and the starting time node of the data statistics time period is the ending time node of the last check work; determining a first inventory prompt value according to the occurrence frequency of the first abnormal condition; and when the first inventory prompt value is greater than the first early warning threshold value, outputting a first inventory early warning signal. By adopting the scheme of the invention, the timeliness and rationality of inventory work execution can be ensured, so that the accuracy of inventory work and the work processing efficiency are improved, and meanwhile, the increase of invalid operation cost is also avoided.

The present disclosure provides a text detection method, electronic device, medium and computer program product, which relate to the field of data processing technology. The method includes: obtaining image features of a target object; inputting the image features into a first decoder of a text detection model to obtain a first feature, and obtaining a first text detection result based on the first feature; inputting the first feature into a second decoder of a text detection model to obtain a second feature, and obtaining a second text detection result based on the second feature; wherein the first text detection result and the second text detection result are text detection results of different levels, and the levels of the text detection results include: word level, line level, or paragraph level. The present invention can obtain multiple levels of text detection results at the same time, and does not need to perform text detection on another level based on the text detection result of one level, but obtains the text detection results of each level based on the first feature, and the text detection results are highly accurate.

This invention relates to a method, apparatus, medium, device, and product for intelligent mail delivery. The method includes: acquiring electronic mail from citizens to be delivered; extracting features from the electronic mail to obtain mail feature data; extracting keywords from the electronic mail; determining candidate delivery departments based on the mail feature data and the keywords; selecting a target delivery department from the candidate departments; and delivering the electronic mail to the target delivery department. This application can improve the accuracy and efficiency of mail delivery.

A modular pin needle visual detection method, a detection server, a medium and a product relate to the field of machine vision detection. The method comprises: acquiring a standard model of a connector; acquiring a product image through a visual detection module; calculating a first actual coordinate of an anchor point in the product image through an image recognition algorithm; calculating an actual offset angle of the placement position of the connector according to the first actual coordinate and the first standard coordinate of the anchor point; establishing a detection coordinate system according to the actual offset angle and the actual coordinate system; determining a scanning path of the visual detection module; acquiring scanning images of each pin needle through the visual detection module according to the scanning path; calculating second actual coordinates of each pin needle according to the detection coordinate system and the scanning images; and performing abnormal detection on each pin needle to obtain a detection result. The technical scheme of the present application effectively improves the detection accuracy of the pin needle and avoids the problem of reduced detection accuracy of the pin needle caused by the deviation of the placement position of the connector.

This application discloses a method, apparatus, and electronic device for placing goods using an AGV, to improve the accuracy of goods placement by the AGV. The method includes: determining, based on a captured image, the horizontal distance between the AGV and a first reference line, and a first angle between the AGV and the first reference line; wherein the captured image is an image captured by an image acquisition device; determining whether the horizontal distance is less than a first threshold and whether the first angle is less than a second threshold; if so, controlling the AGV to enter a correction area until the distance between the AGV and a stop indicator in the ground markings is less than a third threshold, and placing the goods in a target area within the correction area; wherein the correction area includes the ground markings.

The utility model belongs to biscuit production technical field, concretely relates to a kind of automatic picking device of broken biscuit, including conveying frame, the port of air inlet pipe is successively equipped with baffle ring and wind speed detector, the bottom of the blocking plate is equipped with multiple automatic baffle, the conveying frame is symmetrically equipped with two the material guide plate and two conveyer belts, the symmetry axis of the material guide plate and the conveyer belt is perpendicular, the material guide plate covers the top of one the conveyer belt and then tilts to another the conveyer belt;By setting suction plate, baffle ring, wind speed detector and negative pressure device, biscuit will cause hindrance to the airflow entering air inlet pipe, the ability of airflow hindering of complete biscuit and broken biscuit is different, whether biscuit is broken can be determined by comparing wind speed, biscuit is lifted by airflow during detection, after detection, biscuit falls down, secondary detection is carried out at this time, avoid the accuracy of device detection to reduce due to crack in biscuit.

The invention provides a vibration-based axle box fault detection method and device, electronic equipment and a storage medium, and the method comprises the steps that target vibration data of a to-be-detected axle box is obtained, the target vibration data comprises a plurality of vibration time domain signals, and the to-be-detected axle box is a main roller shaft box in a multi-wire cutting machine; the amplitude of each vibration time domain signal in the target vibration data is compared with an amplitude threshold value, a normal signal proportion is determined, and the amplitude threshold value is obtained according to a vibration time domain signal generated by vibration of a sample axle box in a non-fault state; whether the normal signal proportion reaches a set proportion or not is judged, if not, it is determined that the axle box to be detected is in a fault state, and the set proportion is determined according to the normal sample signal proportion of the sample axle box in the fault state. According to the scheme, fault detection is automatically realized based on the vibration condition of the axle box, so that the accuracy and timeliness of fault detection are improved.

The invention provides a search method and device, equipment and a storage medium, and relates to the technical field of computers. The search method comprises the steps of obtaining a to-be-searched text; based on a standard verbal skill formulated in advance, performing fuzzy search on the to-be-searched text to obtain a first target fragment; in response to the fact that the text similarity between the first target fragment and the standard verbal skill is smaller than a first threshold value, segmenting the to-be-searched text to obtain a plurality of text fragments; based on the feature vector of the standard verbal skill and the feature vectors of the text segments, performing semantic search on the plurality of text segments to obtain a second target segment; and in response to the fact that the similarity between the feature vector of the second target fragment and the feature vector of the standard verbal skill is greater than a second threshold value, determining the second target fragment as a search result corresponding to the to-be-searched text.

Methods according to embodiments of the present disclosure include receiving channel state information-reference signals (CSI-RS) and reporting channel state information (CSI). The CSI includes i) first information related to channel prediction and ii) second information related to interference prediction.

The invention provides a method and device for evaluating the control state of a pipeline overturning lining repairing system. The method comprises the steps that m evaluation indexes collected at n different moments are acquired respectively, the jth index at the ith moment is marked as # imgabs0 #, the weight of each evaluation index is calculated based on # imgabs1 #, and the weight of the jth evaluation index is wj; and calculating an evaluation result of the control state at each moment based on # imgabs2 # wj, and sorting the evaluation results according to a sequence from large to small. By adopting the improved linear normalization method, the problem of low accuracy of evaluation index standardization can be avoided, and the evaluation index value which is more practical can be obtained; the objective weight method is introduced, and weighting is performed according to the variation degree of the evaluation index, so that low accuracy of subjective weighting can be avoided, and a more objective and accurate complex system evaluation result can be obtained.

This invention relates to material testing, specifically to a method for detecting defects in steel pipe columns used for foundation reinforcement. The method includes: acquiring echo signals received after emitting acoustic waves onto the steel pipe column; dividing the echo signals into multiple periodic segments; determining the probability of anomalies in each periodic segment based on its own waveform characteristics and the waveform characteristic differences between adjacent periodic signals; determining the multipath difference value between each periodic segment and the first periodic signal based on the multipath characteristics of the signal; determining the degree of distortion in each periodic segment based on the probability of anomalies and the multipath difference value; and determining the defects in the steel pipe column based on the degree of distortion of each periodic signal. This method can improve the accuracy of defect detection in steel pipe columns used for foundation reinforcement.

The commodity information identification method, device, storage medium and computer equipment provided by the present application can perform feature analysis on the target text to determine at least one commodity feature corresponding to the commodity activity link or commodity password when monitoring the target user to copy and paste the target text containing the commodity activity link or commodity password from any e-commerce platform to the present platform; then, the present application can match the at least one commodity feature with a plurality of pre-configured processing strategies according to the preset matching rules, and determine the target processing strategy corresponding to the at least one commodity feature according to the matching results; thus, the commodity activity link or commodity password can be routed to the corresponding processor for parsing according to the target processing strategy, and the commodity or activity corresponding to the commodity activity link or commodity password can be determined according to the parsing results. This process can effectively improve the matching efficiency and reduce the time and cost of manual matching, thereby effectively improving the user experience.

The utility model discloses a power transmission line tree obstacle detection device, and relates to the technical field of tree obstacle detection, the power transmission line tree obstacle detection device comprises an unmanned aerial vehicle, a connecting part and a detection mechanism used for detecting the tree obstacle distance, the detection mechanism is connected with the unmanned aerial vehicle through the connecting part, and the gravity centers of the connecting part and the detection mechanism are located under the gravity center of the unmanned aerial vehicle; the unmanned aerial vehicle is used for carrying the detection mechanism to detect the tree barrier distance, the convenience is improved compared with a heavy theodolite, the problem that the precision is low when the tree barrier distance is manually judged can be avoided through the arrangement of the detection mechanism, and the detection effect is improved; in addition, the gravity centers of the connecting part and the detection mechanism are located under the gravity center of the unmanned aerial vehicle, the flight stability of the unmanned aerial vehicle can be guaranteed, and the detection effect of the detection mechanism can be improved along with improvement of the flight stability.

The invention relates to a method and device for determining a filling wedge gas path, equipment and a medium. The method comprises the following steps: acquiring basic parameter information of a filling wedge, wherein the basic parameter information comprises a cylinder working stroke; calculating according to the basic parameter information to obtain the cylinder load and the cylinder maximum speed of the filling wedge; based on the cylinder load and in combination with a cylinder selection list, determining the model and the number of cylinders used by the filling wedge gas circuit; based on the air cylinder model, the air cylinder number, the air cylinder working stroke and the air cylinder maximum speed, the air pipe diameter, the air path connector model and the reversing valve model related to the filling wedge air path, whether an air storage tank is designed or not and the air storage tank model are determined in combination with an air path original part selection list. According to the method, the cylinder load and the cylinder maximum speed of the filling wedge can be calculated according to the basic parameters of the filling wedge, then the filling wedge gas circuit design is carried out based on the cylinder load and the cylinder maximum speed, and the accuracy of the filling wedge gas circuit design is improved.

Embodiments of the present application provide a kind of server attack determination method, device, storage medium and electronic device, wherein the method comprises: when obtaining first data packet set, determine the first quantity of data packet in first data packet set, wherein first data packet set includes the data packet that front end acquisition equipment transmits to back-end server in first period;When the difference between first quantity and second quantity is greater than or equal to first predetermined threshold, determine the third quantity of data packet in first data packet subset in first data packet set, second quantity indicates the data packet quantity that front end acquisition equipment transmits to back-end server in second period, and the length of data packet in first data packet subset is within the preset length range;According to third quantity and second predetermined threshold, judge whether back-end server is attacked.Through the embodiments of the present application, the technical problem that the accuracy of determining whether the server is attacked in the related art is low is solved.

The invention provides a sound-based axle box fault detection method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining a sound time domain signal set of a to-be-detected axle box in the cutting process of a current knife; each sound time domain signal in the sound time domain signal set is compared with a sound reference value, the normal sound ratio of the current cutter is determined, and the sound reference value is obtained according to sound time domain signals generated by a sample axle box in a non-fault state; the normal sound proportion is compared with a standard proportion, if the normal sound proportion does not reach the standard proportion, it is determined that the axle box to be detected is in a fault state, and the standard proportion is determined according to the normal sound proportion of a sample axle box in the fault state. According to the scheme, automatic fault detection is performed on the to-be-detected axle box according to the sound time domain signal generated by operation of the to-be-detected axle box, so that the accuracy and timeliness of fault detection are improved.

The application discloses a clothes material information acquisition method and a washing and protecting device. The clothes material information acquisition method comprises the following steps: acquiring the near-infrared spectrum of a detection area in an empty cylinder state; and after clothes are put into the cylinder, acquiring the near-infrared spectrum of the detection area for multiple times. The near-infrared spectrum acquired after the clothes are put in is analyzed, the spectrum close to the near-infrared spectrum acquired in the empty cylinder state is excluded, and the remaining near-infrared spectrum is analyzed. The clothes material information acquisition method of the application excludes the interference of the near-infrared spectrum of the empty cylinder on the near-infrared spectrum of the clothes, and the clothes material information detected and acquired is high in accuracy, simple and convenient, and harmless to the fabric. The washing and protecting device has the characteristics of improved washing and protecting effect, reduced energy consumption and higher intelligence.

This application provides a target monitoring method and system based on a low-altitude economic drone, relating to the field of image detection technology. In this application, firstly, target area monitoring video frames are acquired from the low-altitude economic drone; then, the target area monitoring video frames are denoised using non-local means to obtain denoised region monitoring video frames. During the non-local means denoising process, noise perception truncation is performed, ensuring that the similarity weights of video frame pixels with Euclidean distances greater than a predetermined reference distance are set to 0, thus preventing dissimilar video frame pixels from participating in the mean processing; finally, anomaly detection is performed on the denoised region monitoring video frames to obtain region anomaly detection results, reflecting whether abnormal behavior exists in the target area. Based on the above, the relatively low accuracy of anomaly monitoring in existing technologies can be improved.

The utility model discloses an aluminum flat pipe fixed-length cutting-off device capable of reducing noise, which comprises a mounting frame used for supporting, a sound insulation frame fixedly connected to the upper end of the mounting frame, a first electric telescopic rod, a second electric telescopic rod, a servo motor, a baffle and a sleeve, and the upper end of the baffle is fixedly connected with the lower end of the second toothed plate, the upper end of the first electric telescopic rod is fixedly connected with the inner side wall of the upper end of the sound insulation frame, the first connecting block is fixedly installed at the tail end of an extension rod of the first electric telescopic rod, and the sleeve is installed on the outer side of the first connecting block. According to the aluminum flat pipe fixed-length cutting-off equipment capable of reducing noise, an aluminum flat pipe can be stably clamped and limited on the table top during cutting, the situation that the pipe is impacted to displace and reduce the cutting-off precision during cutting off is prevented, fixed-length cutting-off can be achieved, follow-up secondary machining is avoided, the workload is reduced, noise and sweeps generated during cutting can be treated, noise pollution is avoided, and the cutting-off efficiency is improved. And the working area is kept clean.

This invention relates to the field of evaluation management, specifically to a method and platform for standard performance management based on dynamic evaluation. The platform includes: a standard publishing system, a standard evaluation management system, and a standard performance work system. The standard publishing system is used by standard evaluation units to input standard content, standard evaluation requirements, standard dynamic evaluation requirements, and standard dynamic evaluation rules. The standard evaluation management system automatically determines the compliance of corresponding standard evaluation data, generates and stores the standard evaluation results, and sends it to the standard performance work system of the evaluation entity. Then, it generates an evaluation digital identifier QR code based on the standard evaluation results and sends it to the evaluation entity client. The standard performance work system is used by the evaluation entity client to submit dynamic performance execution data, automatically determines the compliance of the dynamic performance execution data, and then automatically updates the latest dynamic evaluation result information to the information recorded in the evaluation digital identifier QR code.