基于BP神经网络的空中机器人传感控制研究

Experimental Study on Sensing Control of Air Robot in Complex Electromagnetic Environment

对空中机器人进行传感控制极易受到电磁的干扰,导致接收到的信息存在大量噪声,无法实现准确稳定地传感控制。由此,设计基于BP神经网络的复杂电磁环境中空中机器人传感控制系统,实现空中机器人在复杂电磁环境中的有效控制。设计系统架构包括机载飞控子系统和地面站子系统两部分;系统通过超声波测距模块确定障碍物同空中机器人的详细距离,利用核心是S3C2410芯片的主控模块对测距模块获取的信息实施处理,基于处理后的信息采用神经网络算法实现空中机器人避障传感控制。系统使用BP算法对神经网络进行训练,实现空中机器人正确判断障碍物的同时降低传感控制误差。实验结果表明,所设计系统能够在复杂电磁环境中实现对空中机器人准确的、快速稳定的传感控制。

The sensor control of aerial robot is easy to be interfered by electromagnetic,resulting in a large amount of noise in the received information,which cannot achieve accurate and stable sensing control. Therefore,the sensor control system of aerial robot in complex electromagnetic environment based on BP neural network is designed to realize the effective control of aerial robot in complex electromagnetic environment. The system architecture includes two parts： airborne flight control subsystem and ground station subsystem. The system determines the detailed distance between obstacle and aerial robot by ultrasonic ranging module. The main control module of the S3 C2410 chip is used to process the information obtained by the ranging module. Neural network algorithm is used to realize obstacle avoidance sensing control of aerial robot. The BP algorithm is used to train the neural network to realize the correct judgment of obstacles and the reduction of sensor control error. The experimental results show that the designed system can realize accurate,fast and stable sensing control of the aerial robot in complex electromagnetic environment.