基于梯度型最小二乘滤波的船摇隔离度测试研究

Research on Testing Method of Ship Shaking Isolation Based on Gradient Descent and Least-squares Filtering Algorithm

摘要:

针对传统船摇隔离度测试方法效率低、误差大、操作复杂、且不能有效解决扰动跳变的不足,对船摇隔离的基本原理进行了分析,建立了梯度型最小二乘滤波模型,提出了基于梯度型最小二乘滤波算法对测量数据进行滤波处理,利用VC++编程实现对船摇隔离度的自动化测试,并在船舶生摇的条件下进行船摇隔离度的测试分析；对比传统方法测试结果表明,该方法能有效滤除测量数据中的随机误差,较大程度地降低了扰动跳变的抖动量,提高了伺服系统船摇隔离度的测试精度。

Abstract:

Traditional testing methods of ship shaking isolation need heavy workload, which has low efficiency and large error. Simultaneously the jitter caused by interference cannot be solved effectively. In order to solve these problems, the principle of ship shaking isolation is analyzed, and the model based on gradient descent and least-squares filtering algorithm is built, the filtering algorithm based on gradient descent and least squares is proposed to filter the measured data, the automated testing of ship shaking isolation is realized through VC++ programming, and the analysis of automated testing under the condition of ship shaking is carried out. Compared to the results tested by traditional methods, the simulation results show that the filtering algorithm which shortens testing time and improves the testing accuracy of ship shaking isolation could filter out random errors in the measurement data effectively, and the jitter caused by interference is significantly decreased.