44results about How to "Improve analysis rate" patented technology

Light from an object such as a cell moving through an imaging system is collected and dispersed so that it is imaged onto a plurality of separate detectors. The light is spectrally dispersed by a plurality of spaced-apart dichroic reflectors, each detector receiving light from a different one or the dichroic reflectors. Each dichroic filter reflects light of a different predefined color, passing light or other colors. The output signal from each detector is indicative of a different characteristic of the object. In one configuration, each detector is provided with a separate imaging lens. In another configuration, the detectors are spaced at varying distances from the corresponding dichroic reflectors, so that separate imaging lenses are not required.

A new method for rapid and accurate determination of aromatics content in fuel oil, the specific process is as follows: first scan the ultraviolet spectrum of a certain number of fuel oil samples, and use conventional methods to test the content of the target value as a training set. Then, the empirical mode decomposition is performed on the spectral signals of the training set in turn, and a certain number of intrinsic mode functions (IMF) and residual items are obtained. Then, the previous high-frequency IMF is summed to form a high-frequency signal, and the remaining IMF and residual term are summed to form a low-frequency signal. The high-frequency signal and the low-frequency signal are expanded into an extended matrix in the variable direction, and a partial least squares regression model is established between the extended matrix and the target detection value. The same decomposition, addition and expansion are performed on the ultraviolet spectrum of the prediction set, and the content of unknown components can be predicted by substituting it into the model. The advantage of the present invention is that the self-adaptive Hilbert-Huang transformation method is introduced, and at the same time, the unfolding strategy is used for reference, which not only fully utilizes the data information, but also avoids the selection of model parameters.

The invention discloses an energy saving method applied to a mobile crowd sensing network node. A data transfer station is in communication connection with a cloud server; The GPS positioning module is arranged on an AP node; The dormancy module is connected with the AP node to control whether the mobile terminal sleeps or not; The mobile terminal collects node port information of each AP node around the terminal and GPS positioning information of the terminal; the data is fed back to the data transfer station; the data transfer station packages the node port information and the GPS positioning information and sends the packaged information and GPS positioning information to the cloud server; the cloud server compares the GPS positioning information; sending node port information of the APnode, closest to the mobile terminal, on the AP node to the mobile terminal; and the mobile terminal selects the AP nodes corresponding to the node port information for connection, the cloud server sends a dormancy instruction to other AP nodes 50 meters away from the AP nodes, and at the moment, each dormancy module controls the corresponding AP node to enter a dormancy mode.

The present invention is a user data processing method and server based on big data and business analysis. Before user portrait analysis, the user data analysis strategy will be determined based on business data analysis instructions, and then the business interaction objects and business interaction types related to the business interaction period data, so that the user data collection strategy can be determined. The server can collect user data of user business terminals in a targeted manner based on the user data collection strategy, thereby filtering out some useless data. In this way, when performing user portrait analysis, the user data that has data analysis and mining value can be directly processed The analysis can not only improve the analysis rate of user portraits, but also improve the analysis accuracy of user portraits, thereby improving the processing efficiency of user data. It can be understood that the above method can combine user data analysis strategies and user data collection strategies, thereby improving the intelligence of user portrait analysis.

The invention relates to an entrance guard record query screenshot linkage system based on fingerprint identification. The entrance guard record query screenshot linkage system comprises a fingerprintdetection unit, an entrance guard unit, a camera shooting unit, a central control unit, a reset unit, a recording unit and a central control display. According to the invention, the central control unit controls the reset unit to select the corresponding reset time according to a specified alert level; when the fingerprint detection unit transmits an alarm signal, the reset unit can cut off the power supply to the fingerprint detection unit when the specified reset time is up, thereby effectively preventing the buzzer from continuously buzzing, quickly judging the alert level of the linkage system according to the different buzzing durations of a buzzer, and improving the alert efficiency of the system. Meanwhile, the central control unit readjusts the alert level according to the triggering times of the alarm, and completes the alert level self-adjustment of the system through self-learning, thereby enabling the system to select the most suitable alert level for different environments, and further improving the alert efficiency of the system.

The invention provides a method for determining tris (2, 3-dibromopropyl) phosphate content of water, and belongs to the field of organic phosphate flame retardant. The method mainly comprises the following steps: (1) conducting solid-phase extraction on a sample; (2) eluting and collecting the extraction column; (3) concentrating the eluent obtained in the step (2) to a constant volume; and (4) determining the concentrated liquid with a high-performance liquid chromatograph-triple quadrupole mass spectrometer. The method provided by the invention improves determination method of tris (2,3-dibromopropyl) phosphate in water environment; and through ultra performance liquid chromatograph-triple quadrupole mass spectrometry combination, the method for detecting tris (2,3-dibromopropyl) phosphate method has more advantages than other methods. The present invention using ultra performance liquid chromatograph has larger separation speed than gas chromatograph and general high performance liquid chromatograp; triple quadrupole mass spectrometry for SRM scanning realizes higher selectivity, more accurate qualitative diagnosis, and improved sensitivity. Therefore, the method provided by the invention has obvious advantages compared with other gas-phase method and liquid mass chromatography method.

The invention relates to the technical field of petrochemical engineering development, and provides an online carbon determination system for catalysts in a fluidized bed device, which includes a first catalyst delivery pipe, a burnt measurement reactor, a burnt flue gas delivery pipe, and flue gas Analyzer. The carbon determination system can measure the carbon deposition amount of the catalyst in real time when the fluidized bed device is working continuously, and improve the analysis rate of the catalyst carbon determination, which is fast, convenient and practical. The embodiment of the present invention also provides a fluidized bed device, which includes the above-mentioned online carbon determination system, which can obtain the carbon deposition amount of the catalyst in real time to understand the active state of the catalyst, and improve the working efficiency of the fluidized bed device. In addition, the embodiments of the present invention also provide an online carbon determination evaluation method for catalysts in fluidized bed devices, which can quickly and real-time monitor the state of carbon deposition in catalysts.

The present application is a big data-based user data processing method and a big data server, which can perform user portrait analysis on the user data to be processed based on a user data analysis strategy, and obtain a user portrait analysis result. In this way, global and local analysis of user behavior data characteristics can be carried out, so as to consider the differences and correlations of user portraits at the global level and local level, so that the determined user portrait analysis results can reflect the actual level of users. image situation.

The present application discloses an instance processing method and its device. The method includes: acquiring the diagnosis data of the instance, wherein the diagnosis data includes the diagnosis results at multiple diagnosis time points corresponding to the diagnosis rules of the instance; The diagnostic data is input to the trained instance anomaly detection component to obtain diagnostic information on whether the instance is abnormal, wherein the instance anomaly detection component is a neural network model component that uses time series information modeling processing. By adopting the application, automatic diagnosis of instances can be realized.