| In recent years,with the continuous deterioration of the global environment and the continuous consumption of petroleum energy,people have begun to focus on nonpolluting natural environmental energy,such as solar energy,wind energy,mechanical energy,etc.Moreover,in order to extend the physical nodes of the cyber physical system and enable the wireless sensor network nodes to work for a long time,the power supply of the sensor nodes is one of the core elements that restrict WSN’s use cycle.In order to solve this,people began to research wireless sensor micro-energy self-powered technology.At present,the power supply of the wireless sensor nodes is mostly single energy device,such as solar battery.However,each energy device has its limitations.Therefore,the composite micro-energy power supply system can make wireless sensor nodes work faster and more smoothly.This paper conducted theoretical research on different micro-energy devices,analyzed their generator theory,and then designed a general micro-energy vector acquisition system to collect feature vectors of different micro-energy devices.The system consists of main control unit,input buffer unit,voltage acquisition unit,working current monitoring unit and wireless communication unit.The whole system adopts a separate design,which can detect the power consumption of the collection system while collecting vectors.After obtaining a large amount of vector data,by analyzing the correlation of vector groups and the complexity of the algorithm,BP neural network is selected as the recognition algorithm of the entire system among many recognition algorithms.In this paper,we designed and trained a BP neural network model.The accuracy of the entire model can reach more than 95%.Finally,in order to apply this recognition model to the composite micro-energy power supply system of wireless sensor nodes,this paper transplants the recognition model trained by the computer to the FPGA platform.The whole system is composed of a communication module and a recognition module.It has completed the interaction with the host computer,the acquisition system,and recognized the obtained vector.The entire system finally achieved has the characteristics of high accuracy and high response speed.It can accurately provide the existing energy types for the composite energy system composed of solar energy,mechanical energy,and radio frequency energy,and provide corresponding support for the composite energy supply system. |
