9results about How to "Reduce inaccuracy" patented technology

The invention discloses an intelligent closestool cleaner performance test system, an intelligent closestool cleaner comprises a spray rod and a nozzle, the nozzle is arranged on the spray rod, and the intelligent closestool cleaner performance test system comprises a controller, an industrial personal computer, a manipulator, an image acquisition module, a cleaning water consumption measuring module, a cleaning force measuring module, a cleaning area measuring module and a nozzle self-cleaning force measuring module; the nozzle stretching time measuring module is used for measuring the stretching time and the retracting time of the nozzle; the cleaning water consumption measuring module is used for measuring the water consumption of one-time cleaning of the closestool cleaner; the cleaning force measuring module is used for measuring the cleaning force of the closestool cleaner; the cleaning area measuring module is used for measuring the cleaning area of the closestool cleaner; according to the full-process test, image recording, test process semi-automation, test result program judgment and test report automatic generation are carried out. Inaccuracy of test data caused by human factors in the test process is reduced, the whole test process is traceable in the whole process, and the test efficiency is greatly improved.

The invention relates to the technical field of data processing, in particular to a data acquisition and processing method for Actiwatch, which comprises the following steps of: obtaining a sleep state level at each moment according to data in all index data sequences, and obtaining a sleep time period of each small fragment sequence; according to the fluctuation condition of data in the sleep time period, the sleep abnormality degree of each small fragment sequence is obtained, according to the data of each sleep quality parameter of each small fragment sequence, the user exercise amount of each small fragment sequence is obtained, a sleep curve is obtained through fitting, and a low-sleep-quality sequence and a high-sleep-quality sequence are obtained; obtaining a sleep quality index of each small fragment sequence; and scoring the sleep quality of each small fragment sequence according to the sleep quality index of each small fragment sequence, and obtaining an evaluation parameter most suitable for the user. According to the method, the most suitable evaluation parameters of the sleep quality are obtained in combination with the physical conditions of different users, and the degree of misjudgment is reduced.

The application discloses a motion control system of a robot flexible joint module based on double encoders, and the flexible joint module comprises a drive controller, a motor, a first encoder and a second encoder. The drive controller comprises a communication module, a control module and an intelligent drive module which are sequentially connected. The intelligent drive module is electrically connected with the motor. The first encoder and the second encoder are both connected with the control module through the communication module. The intelligent drive module detects three-phase current and bus voltage of the motor and sends them to the control module. Meanwhile, the intelligent drive module receives and amplifies the control signal output by the control module to drive the motor to move. The control module runs a flexible joint motion control algorithm according to the position information of the motor end and the load end, the three-phase current of the motor and the bus voltage, and outputs the control signal to the intelligent drive module. The application has the advantages of small structure size, easy use and high positioning accuracy of the joint.

The invention discloses an equipment fault detection method based on Internet of Things and data driving. The method comprises the following steps: deploying and installing a production line environment monitoring module, an equipment state monitoring module, a fault detection module, a control module, a plurality of environment sensors and a plurality of equipment state sensors in a target monitoring area; the production line environment monitoring module obtains environment data around the equipment; the equipment state monitoring module obtains equipment state data; the production line environment monitoring module and the equipment state monitoring module respectively pre-process the acquired data and then transmit the data to the fault detection module; the fault detection module performs fault detection on the received data to obtain a detection result, and transmits the detection result to the control module; the control module adjusts equipment on the production line in real time according to the detection result. According to the method, efficient production line equipment fault detection is realized through the synergistic effect of the sensor, the monitoring module, the fault detection module and the control module.

The application provides a fault detection method and device, electronic equipment and storage medium. The method comprises: obtaining running data corresponding to a plurality of detection points of a target device; inputting the plurality of running data into a pre-trained fault detection model, wherein the fault detection model comprises at least an input layer, a multi-head attention mechanism layer, a plurality of feature extraction layers, a full connection layer and an output layer according to the transmission order of the running data; calculating the weight of each running data through the multi-head attention mechanism layer, and fusing the plurality of running data based on the weight; and extracting and classifying the fused running data through the plurality of feature extraction layers and the full connection layer, and outputting the fault probability of the target device through the output layer. The application solves the technical problem of excessive model parameters of the fault detection model in the prior art, and reduces the complexity of the fault detection model.

This invention discloses a method and system for determining the spatial relationship of potential hazards based on hierarchical early warning, belonging to the field of computer data processing technology. It includes acquiring historical risk data and real-time environmental data, delineating virtual electronic warning zones, and generating dynamic virtual warning zone data; collecting video stream data and identifying and locating potential hazard targets, generating hazard target identification data, and performing spatial matching analysis in conjunction with the dynamic virtual warning zone data to generate an intrusion determination result; if an intrusion occurs, calculating the three-dimensional spatial clearance distance between the hazard target and the live conductor, and outputting a hierarchical alarm result based on multiple key risk factors; if no intrusion occurs, analyzing and predicting the target movement status and location based on the hazard target identification data to generate an out-of-zone early warning result. This invention employs a data processing method that combines dynamic zoning with quantitative analysis and trend prediction of potential hazard targets both inside and outside the zones, enabling intelligent, precise, and forward-looking risk early warning.

The invention belongs to the technical field of big language model reasoning, and particularly relates to a big language model reasoning enhancement method and system based on logic chain verification. The method comprises the steps that historical question and answer pairs are classified into progressive question and answer pairs, multiple logic chains are constructed based on the progressive question and answer pairs, key nodes and secondary nodes are distinguished according to the distribution situation of the nodes in the multiple logic chains, and then a key node evaluation score and a secondary node evaluation score are calculated according to historical evaluation data; and when to-be-verified inquiry content is received, constructing a to-be-verified logic chain for response content and generating a reasoning answer, and then performing weighted summation on evaluation scores corresponding to key nodes and secondary nodes in the chain to calculate a verification score. According to the method, double verification is carried out on the structural importance and historical performance of the nodes in the logic chain, low-credibility results are filtered, and the accuracy and reliability of reasoning answers are improved.

A blade pitch control system for a cycloidal marine propulsion system is provided. A cycloidal marine vessel propulsion system includes a rotating wheel and a plurality of rotating blades attached to the rotating wheel and individually rotatable relative to the rotating wheel. The blade pitch control system includes: at least one rotating wheel position sensor configured to provide a sensor signal indicative of a measured position of a rotating wheel; a wheel position estimation unit configured to determine an estimated position of the rotating wheel based on a sensor signal of the at least one rotating wheel position sensor; and a blade pitch determination unit configured to determine a blade pitch set point of at least one of the plurality of rotating blades based on the estimated position of the rotating wheel. Determining the estimated position of the rotating wheel includes data processing of the sensor signals.

This application discloses a video decoding method, a video encoding method, and related equipment, belonging to the field of video processing technology. The video decoding method of this application includes: when the decoding end determines that the target information does not meet a first preset condition, using the Position Correlated Prediction Combination (PDPC) method to obtain the first pixel prediction value of each pixel point within the encoding unit to be decoded, wherein the target information includes information for determining whether to use the PDPC method to obtain the pixel prediction value; the decoding end decodes the encoding unit based on the first pixel prediction value.