| With the emergence and development of new computer technologies,the demand for data collection of various intelligent terminal equipment is also increasing.How to use intelligent terminal collection equipment to solve the increasingly prominent electrical fire detection and monitoring problem has also begun to attract more and more attention.The design of electrical fire monitoring equipment should not only focus on obvious fire characteristics,but also be able to detect and judge whether electrical fires will occur in multiple dimensions and at multiple levels.It is necessary to achieve early warning and early cut-off as much as possible to minimize electrical fires,harm and loss.First of all,this article constructs an electrical fire detection model.After analyzing the current research status of electrical fire monitoring system and the principle of electrical faults that cause electrical fires,it further analyzes the defects and shortcomings of the main detection technologies used in electrical fire detection equipment at present.Through the analysis of the multi-source data fusion technology The advantages confirm the electrical fire detection model and research method based on multi-source data fusion technology.The research is based on the fusion processing and calculation of the historical data collected by multiple sensors,and the unsupervised learning clustering algorithm K-Means in machine learning is used to cluster the historical data of the equipment according to the industry to obtain the characteristics of industry user electricity consumption;The clustering feature of BP is used as the standard input set of BP neural network,and the BP neural network prediction model is obtained after corresponding training.This model makes full use of the historical data signals collected by the front-end sensor equipment for a long time.In principle,any signal will show a certain regularity and periodicity.These periodic laws can help us better understand the working characteristics of the equipment,and It is easy to find abnormal working conditions that do not conform to the law.By comparing the difference between the predicted feature and the monitored feature at a certain moment,it can be judged whether the electrical equipment is working abnormally.Using the system monitoring value obtained above and the comparison result of the model estimated value,preliminary judgment of the electrical working status.If the comparison result has a large difference value,at this time,it is necessary to combine the parameter change rate of the device at that moment(such as temperature change rate,current change rate)and the smoke sensor signal deployed on the scene to perform a higher level of data fusion,and finally Determine whether an electrical fire will occur and output an alarm signal.In this way,we use multi-source data fusion technology to build the multi-level and multi-dimensional electrical fire prediction model of this article,and after comparing the prediction results for a certain period of time,we conclude that the model can effectively improve the accuracy and accuracy of the system’s decisionmaking.The conclusion of the degree.Finally,a detailed circuit diagram design is given for the hardware design of the system.On the other hand,the software design of the system is also explained. |
