39results about How to "Performance understanding" patented technology

The invention provides a magnetic resonance quality control method and system, and a server. The method comprises the following steps: receiving a quality control report set sent by a test terminal, wherein the quality control report set at least carries a detection identifier of a to-be-detected magnetic resonance imager and a target parameter; determining a reference parameter according to the detection identifier; comparing the reference parameter with the target parameter to obtain a device quality report; and sending the device quality report to the corresponding test terminal according to a corresponding relation. According to the magnetic resonance quality control method and system provided by the invention, the quality control report can be automatically obtained by the test terminal without manual analysis or calculation, so the efficiency is improved; the server can analyze the quality control reports of a large number of magnetic resonance imagers to provide an evaluation report of the imager for a client, the performance of the imager can be comprehensively and accurately evaluated, and it is beneficial for the client to understand the performance of the imager more accurately and to quickly locate and maintain the fault of the imager; and a manufacturer can also download data from the server so as to provide support for the research, development and optimization of the imager.

The invention discloses a sugarcane pressing process prediction method based on deep feature recognition, which comprises the following steps: 1. collecting several sets of original data; 2. removing abnormal data and standardizing the original data collected in step 1 to Obtain standardized data; 3. Multi-level screening is performed on the standardized data obtained in step 2 to obtain feature vectors with high correlation with energy consumption and extraction rate and low redundancy; 4. Use the mixed chicken flock algorithm to obtain from step 3 Search the effect of different feature combinations and model parameters on a single data-driven model in the eigenvector candidate set, and obtain the parameter variables, energy consumption and extraction rate under the optimal performance of a single model; 5. Establish the first layer of deterministic prediction output; 6. , establish a multi-model combination model, and realize the deterministic and probabilistic prediction of extraction rate and energy consumption. The invention greatly improves the model fitting effect and prediction accuracy, and solves the problems that these indexes are difficult to measure online.

The invention discloses an antenna switching method, a multi-antenna terminal, and a computer-readable storage medium. The antenna switching method is applied to the multi-antenna terminal including aprimary transceiver path, a secondary receiver path and at least three groups of antennas. Whether antenna switching is required is determined by acquiring the communication parameters of respectivegroups of antennas and subjecting the performance each group of antennas to condition advanced decision with at least two decision stages. Since two or more decision phases with a timing sequence areused to evaluate the current performance of each group of antennas before it is determined whether the primary antenna or the secondary antenna is required to be reselected, the overall performance ofeach group of antennas can be objectively and comprehensively evaluated during the condition advanced decision and an influence of accidental factors on the antenna performance can be removed. Moreover, since the communication parameters are different in the at least two decision phases, the performance of the antennas in at least two communication parameter dimensions can be evaluated, thereby achieving a comprehensive antenna performance evaluation result and reliable and effective subsequent switching.

The present invention belongs to the field of computer graphics, and in particular relates to a GPU graphics processing performance monitoring and analysis method. The method comprises: (1) GPU resource statistics, (2) data relay, (3) graphics processor performance data generation, (4) data filtering, (5) online viewing, (6) data saving, and (7) offline viewing. According to the method provided bythe present invention, by detecting GPU performance data during application execution, online real-time viewing or offline playback are performed on the performance data that the user cares about, and tool support is provided for GPU design and application optimization.

The invention relates to the technical field of a floor drain, and particularly provides a floor drain display device. The floor drain display device includes a sewer display part and an anti-water returning display part. The sewer display part includes a water inlet pipe, a water tank, and a sewer pipe; a pipe mouth of the water inlet pipe is arranged above the water tank; the bottom of the water tank is provided with a sewer opening for installing the floor drain; the sewer pipe communicates with the sewer opening; the bottom end of the sewer pipe is provided with a sewer stop valve. The anti-water returning display part includes a water returning test tank and a water returning test pipe; a position of the bottom of the water returning test tank is higher than a position of the bottom surface of the water tank; one end of the water returning pipe communicates with the bottom of the water returning test tank, and the other end communicates with the sewer pipe, wherein a communication position is positioned between the sewer opening and the sewer stop valve; and the water returning test pipe is provided with a water returning test valve. Through the floor drain display device, product performance of the floor drain can be displayed from different angles.

The invention discloses a fuel cell parameter control system. The fuel cell parameter control system mainly comprises an air compressor, a pressure reducing valve, a pressure gauge, a filter, a gas supply control valve, a flow meter, a humidification path control valve, a humidifier, a water supplying control valve, a water replenisher, a manually operated valve, a dry gas path control vale, a pressure sensor, a temperature sensor, a fuel cell stack, a pressure sensor, a back pressure valve, a water tank, a PC end control and display device, a hydrogen cylinder and a nitrogen cylinder. The system can test a single parameter of a fuel cell stack system, has a safe use method and stable performances, can detect the performance parameters of the fuel cell under a single variable, and realizes accurate and comprehensive detection of cell stack performances. After a test reaction of the fuel cell, only water is produced and any pollutants are avoided so that zero discharge of pollutants is realized. The invention also provides a working method of the fuel cell parameter control system.

The invention discloses a method and system for effectively analyzing the performance of an SSD front-end NVMe module. The method includes the following steps: using a DRAM module to simulate a Flash module at the back end of an SSD; receiving and executing a command from a host; Perform data transmission between; obtain SSD performance data as NVME module performance data. This solution uses the DRAM module to simulate the Flash module at the back end of the SSD, and fully realizes the function of the NVMe module. Through the SSD performance test, the ultimate performance of the NVMe module hardware IP at the front end of the SSD can be obtained, and the hardware IP of the NVMe module can be accurately understood. Performance, and maintain or upgrade the SSD according to the performance of the NVMe module hardware IP, so as to improve the overall performance of the SSD and improve the product experience.