基于PID神经元网络的水面舰艇减摇方法

摘要

文章研究水面舰艇减摇问题,采用PID神经元网络控制方法.舰艇在大风浪条件下产生剧烈横摇,减摇鳍是目前应用最为广泛的减摇装置之一.鳍角与升力矩的水动力特性主要依靠静态实验获取,实际使用中,鳍控制力矩与鳍角呈复杂非线性关系,使水动力特性存在较大误差.为解决升力系数及航速等反馈中的重要参数不确定性而导致鳍角产生的力矩难以确定的问题,提出一种基于PID神经元网络的减摇方法.构造三层神经网络模型,将比例、积分、微分分别作为网络的隐含层单元,在减摇控制过程中动态调整鳍控制参数,仿真结果表明,PID神经元网络控制横摇系统,实时性能好,稳定性高,有较好的稳定舰艇横摇效果.%Surface vessels rolling reduction problem with PID neural networks control method is researched. Ships vibrate violently under the severe sea condition, stabilize fins is the most widely used device to reduce ship's rolling. The acquisition of fins' water dynamic torque characteristics rely mainly on static experiment In actual application moment, the fin angle and lift torque shows complicated nonlinear relation, and it made water dynamic characteristic variate with severe error. To solve the problem that important feedback parameters are uncertain, such as lift coefficient and the speed, this paper present a stabilize method based on PID neural networks. A three-layer neural network model is constructed, proportion, integral and differential units are used as network hide layer in the control process, adjust fins control parameters dynamically. Simulation result shows: the PID neural networks stabilizing system performs quickly, smoothly and effectively.