31results about How to "Quick prediction" patented technology

The invention relates to an intelligent pre-judgment system for setting a fire isolation zone, including a wind speed and direction detection system, a soil detection system and a terminal control system; the wind speed and direction detection system is composed of several aerial intelligent bodies equipped with wind speed and direction detection equipment. It is used to collect wind speed and wind direction information around the fire; the soil detection system is composed of several land agents equipped with soil detection equipment, which is used to collect soil information around the fire; the terminal control system is used to control the air agents, And receive the wind speed and direction information from the air agent; the terminal control system includes the air agent control unit, the land agent control unit and the pre-judgment system, and the pre-judgment system has a deep neural network. Through intelligent control and data processing, the present invention uses the wind speed and direction information and soil information in the area around the fire to predict the direction of the fire, quickly and accurately predict the location of the isolation zone, and save the pre-judgment time and process of the isolation zone setting.

The present invention relates to a performance testing device, method and system of an IPTV set-top box. The testing device includes: a test request receiving module, which is used to obtain a test request, and analyze and obtain test content; a browser test module, when the test content is a browser Response during the test is used to test the browser performance of the set-top box to be tested; the player test module responds when the test content is a player test, and is used to test the player performance of the set-top box to be tested; the test result feedback module, For storing and feeding back test results; multiple test cases are stored in the browser test module, and the test results of browser performance are obtained based on the test cases; multiple video formats corresponding to the video format are stored in the player test module. The playback operation instance of is used to obtain the test result of the player performance based on the playback operation instance. Compared with the prior art, the present invention has the advantages of simple operation, rapid result feedback, high accuracy and the like.

The invention discloses an abnormal event processing system and an application method thereof, comprising a plurality of distributed processing terminals used for preliminary data acquisition and preprocessing; the distributed processing terminals are connected with a cloud server with autonomous processing capability, and the cloud server is used for The data uploaded by the distributed processing terminal is stored, analyzed and processed; the cloud server is connected with a manual processing terminal; the distributed processing terminal is also connected with a mobile terminal. The invention has the characteristics of strong data collection capability, capable of processing abnormal events quickly, and analyzing and predicting abnormal events that will occur according to the data.

The invention relates to a prediction method of appearance bubbles of hardened concrete. The method comprises the steps of dividing concrete mixture into three parts in average, wherein one of the three parts is used for the test of expansion degree Fm, another part is used for the test of air content Af, and the rest part is used for the test of bubble discharge resistance coefficient cdischarge; then, calculating according to the formula of k=Fm.Af.cdischarge; dividing k value into three level according to ranges being more than 10000, being between 5000 and 10000, and being less than 5000, wherein the appearance quality of the hardened concrete corresponding to the three level ranges is respectively poor, medium and excellent. According to the prediction method, the number of the appearance bubbles of the hardened concrete can be rapidly predicted, and the prediction result is high in accuracy; the problem that at present, the appearance bubbles of the hardened concrete can be estimated only after form removal can be solved; furthermore, According to the method, the cost is low, and simplicity is realized.

The invention discloses an intelligent analysis method and system for paroxysmal atrial fibrillation based on difference time scatter diagrams. The method includes a training stage and a detection stage; the training stage includes: A1) collecting original training data of electrocardiogram; A2) drawing The difference time scattergram in the training stage; A3) mark the atrial fibrillation attack label and the non-atrial fibrillation attack label on the difference time scattergram; A4) train the convolutional neural network model; the detection stage includes: B1) collect the original detection of the electrocardiogram Data; B2) draw the difference time scatter diagram in the detection stage; B3) input the difference time scatter diagram in the detection stage into the convolutional neural network model, and output the paroxysmal atrial fibrillation intelligent analysis result. The invention can realize the intelligent diagnosis of paroxysmal atrial fibrillation by using the method of classification of difference time scatter diagrams, thereby realizing the detection of subclinical atrial fibrillation patients.

The invention discloses a junk mail identification method and device, belonging to the technical field of networks. The method includes: obtaining the header information of the target mail, calling the mail classification rule base, and comparing the header information with a plurality of spam rules in the mail classification rule base; if it is determined according to the comparison result that the target mail is not spam, then Perform feature extraction on the text of the target email to obtain the text features of the target email; call the first email recognition model to identify the text features of the target email, and output the first recognition result; determine whether the target email is spam according to the first recognition result . The present invention first uses the header information of the mail and the mail classification rule base for pre-judgment, and then uses the text features of the mail and the mail recognition model for further identification, so that the spam mail can be identified more accurately, and the missed detection of the spam mail is avoided and false positives.

The invention relates to a magnetic resonance water detection signal detection apparatus and detection method based on instantaneous floating point amplification. A preamplifier circuit is connected to a secondary amplifier circuit and an analog switch through a narrow-band filtering circuit. The secondary amplifier circuit is connected to an A / D converting circuit through a high speed program control amplifier circuit. The secondary amplifier circuit is connected to a comparator circuit through an envelope circuit. The analog switch is in separate connection to the high speed program control amplifier circuit, the envelope circuit and a master control circuit. The master control circuit is connected to the A / D converting circuit and the analog switch through a PC 10. According to the invention, the secondary amplifier circuit and the high speed program control amplifier circuit realizes large dynamic range through cascade connection therebetween, which meets the requirement for the detection of large-magnitude varying noise; the apparatus uses the envelope circuit to predict the converting trend of an input signal and provide reference to gain adjustment; the envelope circuit and the comparator circuit which are based on instantaneous floating point amplification achieve the effects of fast pre-anticipation and in-time gain adjustment. Further, no manual operation is required, the work efficiency is increased, and amplifier saturation is effectively suppressed.

The present invention provides an automatic early warning method and an automatic early warning system for dangerous driving behaviors of vehicles. The automatic early warning method of the present invention includes the following steps: constructing an automatic discrimination cloud model for dangerous driving behaviors of vehicles; Prediction cloud model-Elman neural network; real-time collection and storage of vehicle motion attitude data; use the measured data to train the automatic prediction cloud model-Elman neural network to obtain automatic early warning cloud model-Elman neural network that meets the accuracy requirements; based on automatic early warning cloud Model-Elman neural network for automatic warning. The invention can accurately, quickly and reliably analyze and predict the dangerous driving behavior of the vehicle, realize active early warning of the dangerous driving behavior of the vehicle, and effectively reduce road traffic accidents.