Information Processing Of The Fire Detection Algorithm

In recent years, the fire detection systems have been improved in the accuracy sincevarious types of fire detection algorithm emerge one after another, but they still couldn’t meetthe requirements of high accuracy for indoor fire detection. The research for informationprocessing in the fire detection algorithm, not only significantly reduces the fire hazards, butalso optimizes the design of fire protection systems. Furthermore, it protects people’s livesand property, and provides a reference for the study of multi-sensor data fusion.The current fire detection systems exist problems that the common feature recognitionrate is not high, the system of decision-making and ability to adapt are not powerful. Inresponse to this phenomenon, this paper analysis and compares a wide range of existing firedetection algorithm and introduces the fire detection algorithm based on data fusion.According to the three levels in data fusion, the author uses a modular algorithm foridentification, hierarchical information fusion, and introduces neural network algorithms andfuzzy reasoning techniques to solve these problems. Specifically, in order to efficientlyidentify the characteristics of the fire signal, this paper utilizes the learning andassociate ability of neural network algorithm, and makes target recognition on light fire andsmoldering fire by the BP algorithm based on LM. The experiments verify that the proposedalgorithm greatly improved the recognition accuracy of the characteristics of fire; In order tostrengthen the system of decision-making and adaptation, the fuzzy theory algorithm isimproved. The algorithm uses the feature level to provide direct-criterion (light fireprobability and smoldering fire probability) and decision-making indirect criterion (duration,fire damage degree, fire risk degree) fusion, and then arrive at a final decision in accordancewith the expertise to develop the fuzzy control rules, by verifying that the proposed algorithmis better enhance the accuracy of the decision-making system, and enhance the adaptability ofthe system.In summary, under the framework of data fusion, the author efficiently mixes the two current emerging types of intelligent algorithm, and solves the existing difficult point in thefield of fire detection, providing a more efficient idea and approach for the research fields.