利用光纤的光强内调制原理设计了一种用于水下机器人的滑觉传感器。详细介绍了传感器的原理和设计方法 ,对传感器的静态、动态性能进行了理论和实验研究。用基于遗传算法的神经网络对滑觉传感器的动态性能进行了建摸和仿真 ,从传感器应用的角度出发 ,依据“逆模型”思想 ,采用神经网络对传感器建立辨识模型 ,从而提高了机械手的控制精度和灵敏度 A kind of sliding sensor for underwater robot is designed by using the principle of optical fiber internal modulation. The principle and design method of the sensor are introduced in detail, and the static and dynamic performance of the sensor are theoretically and experimentally studied. Using the neural network based on genetic algorithm, the dynamic performance of the gliding sensor was built and simulated. From the perspective of sensor application, the neural network was used to establish the recognition model based on the “inverse model” idea, which improved the control of the manipulator Accuracy and sensitivity