30results about How to "Effective forecast" patented technology

The invention discloses a method for monitoring tire pressure of a running vehicle and is characterized in that: the method can analyze and compare pulse signals output from wheel velocity sensors and then judge the working conditions such as vehicle turning, driving wheel slippage and braking conditions and simultaneously send alarm information when determining that the tire pressure is abnormal. By collecting, sorting, analyzing, comparing and judging the pulse signals output from each wheel velocity sensor of the running vehicle, the method can determine the working conditions such as turning, braking and driving wheel slippage and provide the comparative analysis of detected parameters (pulse numbers). The method is designed with the alarm mode in which the abnormal tire pressure is determined when the monitored parameter exceeds the preset threshold value for several times. The method can accurately and efficiently predict the abnormal tire pressure of the running vehicle.

The invention discloses a TSP-based comprehensive advance forecast method for harmful gases in tunnels, which comprises the following steps: S1: detecting the rock mass in the tunnel, and recording the integrity of the rock mass in the tunnel; S2: according to the obtained completeness of the rock mass in the tunnel, Geological sketch; S3: Collect data for geological prediction of tunnel seismic wave method, and obtain the collected data; S4: Detect harmful gases in the rock mass, and obtain harmful gas detection results; S5: According to the data collected in S3 Harmful gas forecasting is carried out based on the harmful gas detection results in S4, and the forecast results are obtained; S6: according to the forecast results in S5, harmful gas protection is carried out; through the intervention of TSP technology, the accuracy of harmful gas prediction for non-coal-measure tunnels is effectively enhanced Degree, can more effectively complete the prediction of harmful gases.

The invention discloses a flash flood early warning method based on deep learning. The invention adopts the RNN model with LTSM function, and adding the long short-term memory (LTSM) method to the model algorithm can effectively solve the problems of gradient disappearance and gradient explosion in RNN, and more Effectively forecast small watersheds or mountain torrent areas, realize reliable mountain torrent disaster early warning and forecast, greatly improve the forecast time, the model only needs to pay attention to the input data, has the characteristics of easy operation, fast calculation speed, and easy transfer to other areas, and With the actual application of the model and the continuous accumulation of data, there is no need to manually adjust the parameters twice. The model will actively optimize and adjust the parameters to make its application easier and more convenient.

The invention relates to a method for estimating spherical equivalent change degree before and after cycloplegia. The spherical equivalent change degree is estimated by measuring and adjusting an accuracy numerical value; in order to further improve the prediction accuracy, the eye axis length and the subject age can be increased to serve as reference factors, and a prediction result is obtained through a calculation model. The beneficial effects of the invention are that the method can achieve the effective estimation of the spherical equivalent change degree before and after cycloplegia; themethod is higher in predictability and can provide a new clinical standard for clinical ophthalmologists to evaluate whether children patients need to use a cycloplegia agent or not, so the diagnosisand treatment efficiency is improved, medical resources are saved, and related drug side effects borne by the patients due to unnecessary cycloplegia are reduced.

The invention relates to a calculation method of heat conduction flux between riverbed and water body in a data-free area, including: fitting the interannual temperature change, calculating dimensionless air temperature value and water temperature value, calculating the heat content of the river bed, calculating the heat content difference at different times, and calculating The heat conduction flux of the river bed, and the establishment of the relationship curve between the air temperature and the heat conduction flux of the river bed. The present invention does not need to measure or calculate the temperature of the soil at different depths of the riverbed, and can evaluate the heat conduction flux of the riverbed in the data-free area according to the air temperature process, and the calculation is accurate and efficient; at the same time, the present invention provides the relationship curve between water temperature and air temperature and the heat conduction of the riverbed The relationship curve between flux and air temperature can effectively predict the change process of water temperature or riverbed heat conduction flux at different times.

The invention discloses a method for detecting a semiconductor device electrical property failure. The method for detecting the semiconductor device electrical property failure comprises the following steps of: S1, collecting the gray scale: a. taking the minimum region which can be scanned by an electron beam defect scanner as a scanning unit region, setting as a defect through a program form and detecting the defect, and shooting an electron microscope atlas; and b. taking a semiconductor device as a scanning object, scanning under different conditions, and analyzing the gray scale; S2, screening scanning conditions, and testing the electrical property of the scanning conditions; acquiring a similarity numerical value; and defining the optimal scanning condition; S3, carrying out fixed point gray scale analysis so as to obtain a gray scale data collecting interval with an effective electrical property; and S4, forecasting an electrical property failure region. According to the invention, the gray scale collected data obtained by the electron beam defect scanner and the electrical property testing data are compared, the electrical property failure region and an electrical property effective region can be forecast effectively, which provides a basis for monitoring the advantages and disadvantages of an ion implantation technology, provides a technical support for to-be-adopted improving measures, and provides a powerful guarantee for shortening the research cycle of the semiconductor device.

The invention relates to a method for judging abnormal gas by using the characteristics of gas desorption amount, belonging to the technical field of coal mine safety, including acquiring the minute average value data of gas monitoring of the jth shift; calculating the average value of the gas monitoring number of the jth shift; calculating the i minute gas K-minute moving average of the monitoring data; calculate the m-minute moving average of the i-th minute gas monitoring data; calculate the maximum value of the k-minute moving average in the jth shift; calculate the maximum value of the m-minute moving average in the jth shift; calculate The maximum gas desorption amount for k minutes in the jth shift; calculate the maximum gas desorption amount for m minutes in the jth shift; calculate the difference between the k or m minute gas desorption amount in the jth shift and the jth shift Ratio; calculate the ratio between the k-minute gas desorption amount and the m-minute gas desorption amount in the jth shift; calculate the ratio between the maximum value of the k-minute moving average and the m-minute moving average value in the jth shift; the final judgment knowledge.

The invention discloses a method for monitoring the air pressure of automobile tires while driving. An alarm message is issued when the air pressure is abnormal. The invention collects, sorts, analyzes, compares, and judges the pulse signals output by each wheel speed sensor of the car while driving, and can judge the driving conditions such as turning, braking drag, driving wheel slip, etc., and proposes detection parameters ( The comparative analysis method of the number of pulses) is designed, and the alarm mode of abnormal tire pressure is designed when the monitoring parameters exceed the threshold for many times. The invention can accurately and effectively predict the abnormal situation of the tire pressure of the automobile during driving.

The invention, which belongs to the earthquake prediction field, provides a geometric-graphic-method-based method for predicting an earthquake position and range with underwater temperature alarm points. The method comprises: many underwater temperature monitoring points (1) are marked on a corresponding regional map to form a network distribution graph (2) for earthquake prediction; when an earthquake will occur at a certain place, all alarm points (3) are marked correspondingly on the distribution graph, four low-temperature alarm points (4 and 5) and high-temperature alarm points (6 and 7) in different directions are selected, straight lines (8 and 9) in intersection and extending lines (10 and 11) are drawn on the four alarm points on the distribution graph (2); straight line centering points (23 and 35) of all straight lines are searched by using geometric graphic methods of an isosceles triangle and the like, so that prediction information of a central zone (45) position, a seismic wave, and a range (46) of an incoming earthquake can be obtained. Therefore, a prediction and prevention effect for an earthquake is realized; and thus losses caused by the earthquake disaster for human beings can be maximized.

A semiconductor device repeated defect detection method comprises a step S1: fixed point scanning, namely a fixed point scanning form is established, an internal zone of a repeated unit of a to-be-detected module is selected, a minimum zone which an electron beam defect scanner can scan is used as a scanning unit zone which is set to be a defect to be detected, and an electron microscope map is photographed; a step S2: gray scale sample collecting, namely the electron microscope map obtained in step S1 is analyzed in a gray scale mode, typical gray scale distribution and a gray scale distribution interval of the typical gray scale distribution are acquired, and a standard gray scale interval is defined; and a step S3: abnormal semiconductor device forecasting. Gray scale collecting data obtained through the electron beam defect scanner is compared with the standard gray scale interval of a normal semiconductor device, an abnormal semiconductor device can be forecast effectively, effective monitoring on a repeated defect generated in ion implantation technology is achieved, and accordingly a methodology is provided for process window optimizing and on-line monitoring, and on-line manufacturing of semiconductors and yield improvement are guaranteed.