Nuclear accident source item inversion method of neural network adaptive Kalman filter

Abstract

The present invention provides a nuclear accident source item inversion method of neural network adaptive Kalman filter. The method comprises the following steps: establishing a nuclear accident source item inversion model based on neural network adaptive Kalman filter, and determining the input parameters and output parameters of a neural network structure; establishing an atmospheric pollutant diffusion model; and determining observation variables, system variables, an observation matrix, a state transfer matrix and a calculation flow of an adaptive Kalman filter module, and finally obtaining a neural network adaptive Kalman filter nuclear accident source item inversion model. The method selects optimal parameters of the neural network when the neural network combines the adaptive Kalman filter to rapidly screen accidents to replace a simple approximation model with self learning of an atmospheric diffusion process in different weather conditions through the neural network, an adaptive algorithm is selected to overcome the defects of larger filtering errors and filtering diffusion of a conventional Kalman filter algorithm in the nonlinear variation condition, perform real-time online inversion and provide basis for the nuclear accident consequence evaluation and emergency decision.

Description

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine

Login to view more

AI Technical Summary

Benefits of technology

This patented technology uses an artificial intelligence (AI) system called Neural Networks that has been trained by experts over time through experience or experiments with different scenarios where radioactive materials were released during reactor shutdown events. These systems use this knowledge to predict how likely it will be dangerous at some future times when they should take measures against them. They also have their own way of learning about these situations without actually measuring things like air pollution levels. By doing this, researchers aimed towards developing more efficient ways to prevent radiation exposure hazards caused by uncontrolled releases of radium compounds into our environment.

Problems solved by technology

This patented technical problem addressed during this patents text relates to accurately determining when certain events occur (such as terrorist attacks), without relying solely upon external factors such as weather patterns or physical measurements like temperature). Current approaches involve estimating uncertain event terms through various techniques including advanced analyte measurement technology and machine learning models. However these current solutions lack sufficient precision due to unknown variables affecting their effectiveness over different scenarios.

Images