Sensor Network Interface And Multisensor Fusion Technology; Sensor Network Interfaceand Multisensor Fusion Technology

With the development of network technology, remote data acquisition and control technology of high demand shows the trend. The existing RS232/RS485 protocol for the main transmission of multisensor system cannot satisfy modern industrial control requirements, a system to realize remote monitoring and control of network, and can be compatible with the original RS232/RS485 standard equipment is imperative. In addition, in a complex system, in order to reduce the amount of data transmission through network, it is necessary to fuse multisensor information; traditional fusion is registration and fusion division to consider. In fact they are interrelated and interact. In view of the above questions, this topic given the corresponding solving methods mainly includes the following content.(1)This design realized a network serial adapter; it is a generic network intermediate device.This adapter is suitable for any device with RS232 interface which is desired to be connected to network. As the realization of virtual serial port driver, conversion between network data and serial port data is left to virtual serial port driver allowing end users only need to do serial programming to achieve the goal of interconnecting PC and serial port device via network, greatly reducing the design difficulty of the client software. The design of the adapter consists of hardware and software. Based on the 8-bit MCU ATmega128L added the special network protocol stack W5100 chip, hardware design realized a 10/100Mbps Ethernet embedded device. Software design is made of virtual serial port driver and firmware, which makes communication between virtual serial port drive and firmware available. There is a manage program designed in firmware. User can update the parameters of the adapter through RS232 port.(2)This paper also studied on multisensor information fusion technology. A multisensor fusion and registration method which based on EM algorithm is given out, the sensor calibration and fusion in one algorithm can use at the same time, in E step is given system state estimate, M step is given registration. Finally, a tracking system of two sensors tracking a target simulation experiment is done using Matlab.