Research on Multi-source Sensing Technology for Electric Internet of Things Based on Edge Computing

With the power Internet of Things as the research background, this paper analyzed the multi-sensor data perception requirements. Combined with the distributed solution ideas provided by edge computing, this paper proposed a layered distributed architecture. In order to improve the stability of the collected data, a distributed Kalman filter algorithm based on local optimization was proposed to solve the problems of noise interference and redundant data in the monitoring data of sensors. Aiming at the data fusion of heterogeneous sensors, an adaptive weighting method was proposed, which ensured the data quality and improved the fusion efficiency. Finally, in the case of dynamic external interference and environmental state change, the data fusion could effectively reduce data redundancy and fully mined the multi-source information to realize the comprehensive evaluation of equipment states, which verified the effectiveness of the method.