基于遗传神经网络的扭矩传感器设计

Design of torque sensor base on genetic neural network

分析了传统扭矩传感器存在的主要误差,利用遗传算法优化的BP神经网络对扭矩传感器误差进行了补偿,提高了系统的测量精度.利用MSP430单片机完成神经网络的回想,使设计的智能传感器体积小、耗电省、成本低.

The torque sensor errors were analyzed and the BP neural network optimized by the genetic algorithm was used to compensate the errors,results showed that the precision of measurement was improved.MSP430 MCU was also used to finish the recollection of neural network,making the intelligent sensor have such merits as small volume,low electricity-consumption and low cost.