摘要
减摇水舱实验台是典型的液压位置伺服系统,其功能是模拟船舶运动以验证减摇水舱的性能。进行强制震荡实验时,水舱中流体晃动对台体产生不断变化的干扰力矩。通常PID控制可实现理想的控制精度和稳定性,但是负载的变化会引起系统固有特性的变化,固定的PID参数很难保证一贯的控制效果。为此,设计基于Lab VIEW的液压伺服系统神经网络PID测控系统。当系统参数变化时,先通过实时仿真整定控制器参数,仿真测试达到预期目标后,加载到实际系统验证。实验结果表明,神经网络PID控制能够在负载变化时保证良好的控制效果,参数整定过程效率显著提高,操作方便。
Test bench of anti-rolling tank is a typical hydraulic position servo system, and its function is to simulate the motion of ship to verify the performance of the anti-rolling tank. When forced oscillation tests are carried out, the fluid in tank produces constantly changing disturbance torques. PID control can usually achieve the desired control precision and stability. However, the variation of the load results in the system inherent characteristics change, and the fixed PID parameters are difficult to ensure consistent control effect. Therefore, the neural network PID control and monitor system based on LabVIEW is designed for the platform. When the system parameters change, the controller parameters are adjusted by real- time simulation. When the simulation achieves the desired objectives, the controller parameters are loaded into the actual system. The test results show that neural network PID control can guarantee a good control effect when load changes. The parameters setting process efficiency increased significantly and is easy to operate.
作者
谢广东
黄广明
张松涛
XIE Guang-dong HUANG Guang-ming ZHANG Song-tao(Shanghai Division, China Ship Development and Design Center, Shanghai 201102 China College of Automation, Harbin Engineering University, Harbin 150001 China)
出处
《自动化技术与应用》
2017年第9期14-17,39,共5页
Techniques of Automation and Applications
基金
国家自然科学基金项目(编号61304076)
哈尔滨市科技创新人才专项基金项目(编号RC2013QN009035)
哈尔滨工程大学实验教学改革与实验技术研究基金项目(编号SYJG20130411)
