| 摘 要: 针对针织圆纬机[1]喷雾加油装置[2]无法精准预测和受控计量问题，设计了基于ARM平台的喷雾加油系统，实现信号采样、信号处理、人机交互及润滑控制等功能。由于圆纬机振动引起装置油桶液面波动，导致油位检测精度下降，虽已开发出基于最小二乘法的线性回归模型的油位检测算法，但为进一步提高精度，开发出基于三次样条拟合的油位检测算法。通过残差分析、测试对比，表明三次样条拟合模型优于线性回归模型，三次样条拟合优度达0.9921，较之线性回归模型，拟合优度同比提升0.674%，能够进一步提高喷雾加油装置的工作准确性及稳定性。 |
| 关键词: 喷雾加油装置；三次样条拟合；残差分析 |
|
中图分类号: TP311
文献标识码: A
|
|
| Improvement of Oil Level Detection Algorithm for Spray Fueling Device based on Cubic Spline Fitting Model |
|
CHEN Weizheng1, PENG Laihu1,2, Tang Qilin1
|
( 1.Key Laboratory of Modern Textile Equipment Technology, Zhejiang Sci-tech University, Hangzhou 310018, China ;
2.Hangzhou Xuren Automation Limited Company, Hangzhou 310018, China)
936753998@qq.com; 43233212@qq.com; 765834836@qq.com
|
| Abstract: Aiming at the problem that the spray refueling device[1] of the circular weft knitting machine[2] is unable to accurately predict and control the measurement, this paper proposes to design a spray refueling system based on ARM platform to realize functions of signal sampling, signal processing, human-computer interaction and lubrication control. Due to the vibration of the circular weft knitting machine, the oil tank liquid level of the device fluctuates, which leads to the decline of the oil level detection accuracy. Although an oil level detection algorithm based on the linear regression model of the least squares method has been developed, an oil level detection algorithm based on cubic spline fitting is developed in order to further improve the accuracy. The residual analysis and test comparison show that the cubic spline fitting model is better than the linear regression model, and the cubic spline fitting goodness reaches 0.9921. Compared with the linear regression model, the fitting degree has increased by 0.674% over the same period of last year, which can further improve the accuracy and stability of the spray refueling device. |
| Keywords: spray refueling device; cubic spline fitting; residual analysis |
