Application of neural network data fusion algorithm in measurement circuit

Multi-sensor data fusion is a technology to fuse data from multiple sensors in order to make a more accurate estimation of the environment through measurement and detection. In this article a smart measurement circuit which embeds with neural network data fusion algorithm is designed with FPGA-based System-on-Chip (SoC) architecture. The circuit can operate in power down mode and its parameters are adjusted on-line. Influence of structure parameters to the network performance are analyzed and a partition succeed method is presented to shorten the time of network training on-line.