Distributed detection fusion method for single-point information source

Abstract

The invention relates to a distributed detection fusion method for a single-point information source, and the method comprises the following steps: (1) carrying out the mathematical modeling of all alarm processes through employing a correlation coefficient method according to the alarm signal work flow of an alarm, and obtaining an alarm process model of the alarm in a normal environment; (2) aiming at the problems of large concentration difference, uncontrollable temperature and humidity conditions and the like of the environment, adopting a joint distribution function to carry out missing alarm rate influence analysis on the concentration and the humidity, and constructing a relation model between the missing alarm rate of the alarm and the concentration and the humidity of the current position; (3) constructing a single alarm false alarm rate model according to the interference influence of gasoline, diesel oil, kerosene and the like in the environment on alarm equipment; and (4) designing a distributed information fusion rule by taking two alarms as examples according to the alarm distribution rule, and constructing a Bayesian-based detection data fusion algorithm. According to the invention, the missed alarm rate and the false alarm rate are greatly reduced.

Description

technical field [0001] The invention relates to a distributed detection fusion method of a single point information source and the technical field of chemical hazard information system. Background technique [0002] When an explosion accident occurs, toxic and harmful substances are dispersed into vapor and aerosol into the atmosphere, forming toxic and harmful clouds. The toxic and harmful clouds will move with the airflow and spread to the surroundings at the same time, forming a hazardous area. Single-point information sources are an important part of the chemical hazard information system, usually referring to chemical aerosol alarms, which are deployed in large numbers and widely distributed, and can provide timely alarms for common hazards. Under the conditions of informatization, the online information collection channels are diversified, and the types of information sources and technical means are also rich and diverse. At the same time, the struggle between the inf...

AI Technical Summary

Problems solved by technology

[0004] The technical problem of the present invention is to overcome the high rate of missed alarm and false alarm rate of the existing single alarm, which leads to low accuracy of detection data fusion alarm, which is difficult to meet the needs of operations, and provides a distributed detection fusion method of single-point information source, which is effective for alarm The physical model and mathematical model of the device are described, and the alarm process is mathematically modeled. Based on the Bayesian rule, the fusion algorithm of the single point source is designed, and the numerical simulation is carried out for the four influencing factors. The rate can be reduced by about 2 orders of magnitude

Images