Abstract
Feeding system modelling is the foundation for control strategy optimization, contour error compensation, etc., to improve the productivity and quality of a part. This paper proposes a novel residual learning approach for fitting the simulation error of the dynamics model of a machine tool feeding system. Then, the feeding system model consisting of the dynamics model and the residual model is constructed by integrating prior knowledge with statistical learning knowledge. The residual model is trained by using the training dataset generated from the dynamics model instead of using only the input and output data (i.e., the end-to-end data). In addition, a dynamic nonlinear correlate factor extraction method is proposed to extract the training dataset from the dynamics model and the reference data. Compared to the end-to-end data, the training dataset knows very well about the system’s nonlinear features owing to the internal prior knowledge of the dynamics model. Experiments conducted on a vertical milling centre confirm the effectiveness of the feeding system model in dynamic response prediction. Compared to the existing dynamics or data-driven modelling method, the proposed method can achieve higher prediction accuracy in nonlinear motion processes, such as the reverse process, and can obtain stable performance with respect to different feedrates owing to the residual learning approach.
Similar content being viewed by others
References
Li C, Li L, Tang Y, Zhu Y, Li L (2019) A comprehensive approach to parameters optimization of energy-aware CNC milling. J Intell Manuf 30(1):123–138. https://doi.org/10.1007/s10845-016-1233-y
Shao G, Brodsky A, Miller R (2018) Modeling and optimization of manufacturing process performance using Modelica graphical representation and process analytics formalism. J Intell Manuf 29(6):1287–1301. https://doi.org/10.1007/s10845-015-1178-6
Yang M, Yang J, Zhu L, Yu X (2020) A novel curvature circle iterative algorithm for contour error control of multi-axis CNC machine tools. Precis Eng 65:23–31. https://doi.org/10.1016/j.precisioneng.2020.05.005
Chen HR, Cheng MY, Wu CH, Su KH (2016) Real time parameter based contour error estimation algorithms for free form contour following. Int J Mach Tools Manuf 102:1–8. https://doi.org/10.1016/j.ijmachtools.2015.11.009
Chiang H-S, Chen M-Y, Huang Y-J (2019) Wavelet-based EEG processing for epilepsy detection using fuzzy entropy and associative petri net. IEEE Access 7:103255–103262. https://doi.org/10.1109/access.2019.2929266
Elias I et al. (2020) Genetic algorithm with radial basis mapping network for the electricity consumption modeling. Appl Sci 10(12). https://doi.org/10.3390/app10124239
Aquino G, Rubio JDJ, Pacheco J, Gutierrez GJ, Ochoa G, Balcazar R, Cruz DR, Garcia E, Novoa JF, Zacarias A (2020) Novel nonlinear hypothesis for the Delta parallel robot modeling. IEEE Access 8:46324–46334. https://doi.org/10.1109/ACCESS.2020.2979141
Erkorkmaz K, Altintas Y (2001) High speed CNC system design. Part II: modeling and identification of feed drives. Int J Mach Tools Manuf 41(10):1487–1509. https://doi.org/10.1016/S0890-6955(01)00003-7
Ansoategui I, Campa FJ (2017) Mechatronics of a ball screw drive using an N degrees of freedom dynamic model. Int J Adv Manuf Technol 93(1–4):1307–1318. https://doi.org/10.1007/s00170-017-0597-2
Pandilov Z, Milecki A, Nowak A, Grajewski D, Ciglar D, Mulc T (2015) Virtual Modelling and simulation of a CNC machine feed drive system. Trans FAMENA 4:37–54
Li X, Zhao H, Zhao X, Ding H (2016) Dual sliding mode contouring control with high accuracy contour error estimation for five-axis CNC machine tools. Int J Mach Tools Manuf 108:74–82. https://doi.org/10.1016/j.ijmachtools.2016.05.007
Huang HW, Tsai MS, Huang YC (2018) Modeling and elastic deformation compensation of flexural feed drive system. Int J Mach Tools Manuf 132:96–112. https://doi.org/10.1016/j.ijmachtools.2018.05.002
Altintas Y, Verl A, Brecher C, Uriarte L, Pritschow G (2011) Machine tool feed drives. CIRP Ann Manuf Technol 60(2):779–796. https://doi.org/10.1016/j.cirp.2011.05.010
Wu J, Yu G, Gao Y, Wang L (2018) Mechatronics modeling and vibration analysis of a 2-DOF parallel manipulator in a 5-DOF hybrid machine tool. Mech Mach Theory 121:1339–1351. https://doi.org/10.1016/j.mechmachtheory.2017.10.023
Yang M, Yang J, Ding H (2018) A two-stage friction model and its application in tracking error pre-compensation of CNC machine tools. Precis Eng 51:426–436. https://doi.org/10.1016/j.precisioneng.2017.09.014
Bui BD, Uchiyama N, Simba KR (2016) Contouring control for three-axis machine tools based on nonlinear friction compensation for lead screws. Int J Mach Tools Manuf 108:95–105. https://doi.org/10.1016/j.ijmachtools.2016.06.001
Rafan NA, Jamaludin Z, Chiew TH, Abdullah L, Maslan MN (2015) Contour error analysis of precise positioning for ball screw driven stage using friction model feedforward. Procedia CIRP 26:712–717. https://doi.org/10.1016/j.procir.2014.08.021
Pei M, Wu X, Guo Y, Fujita H (2017) Small bowel motility assessment based on fully convolutional networks and long short-term memory. Knowledge-Based Syst 121:163–172. https://doi.org/10.1016/j.knosys.2017.01.023
Huang K, Wen H, Zhou C, Yang C, Gui W (2020) Transfer dictionary learning method for cross-domain multimode process monitoring and fault isolation. IEEE Trans Instrum Meas 9456(c):1. https://doi.org/10.1109/tim.2020.2998875
Liu Y, Yang C, Huang K, Gui W (2020) Non-ferrous metals price forecasting based on variational mode decomposition and LSTM network. Knowledge-Based Syst 188:105006. https://doi.org/10.1016/j.knosys.2019.105006
Huo F, Poo AN (2013) Nonlinear autoregressive network with exogenous inputs based contour error reduction in CNC machines. Int J Mach Tools Manuf 67:45–52. https://doi.org/10.1016/j.ijmachtools.2012.12.007
Huo F, Xi XC, Poo AN (2012) Generalized Taylor series expansion for free-form two-dimensional contour error compensation. Int J Mach Tools Manuf 53(1):91–99. https://doi.org/10.1016/j.ijmachtools.2011.10.001
Erwinski K, Paprocki M, Wawrzak A, Grzesiak LM (2016) Neural network contour error predictor in CNC control systems. 2016 21st Int. Conf Methods Model Autom Robot MMAR 2016, 537–542. https://doi.org/10.1109/MMAR.2016.7575193
Jiang Y, Chen J, Zhou H, Yang J, Xu G (2020) Nonlinear time-series modeling of feed drive system based on motion states classification. J Intell Manuf 31:1935–1948. https://doi.org/10.1007/s10845-020-01546-5
Abu M, Wee H (2020) Hierarchical linear and nonlinear adaptive learning model for system identification and prediction. Appl Intell 50:1699–1710. https://doi.org/10.1007/s10489-019-01615-0
Agand P, Shoorehdeli MA, Khaki-Sedigh A (2017) Adaptive recurrent neural network with Lyapunov stability learning rules for robot dynamic terms identification. Eng Appl Artif Intell 65:1–11. https://doi.org/10.1016/j.engappai.2017.07.009
Hou Z, Gao H, Lewis FL (2017) Data-driven control and learning systems. IEEE Trans Ind Electron 64(5):4070–4075. https://doi.org/10.1109/TIE.2017.2653767
Xu W, Peng H, Zeng X, Zhou F, Tian X, Peng X (2019) A hybrid modelling method for time series forecasting based on a linear regression model and deep learning. Appl Intell 49(8):3002–3015. https://doi.org/10.1007/s10489-019-01426-3
Li F, Jiang Y, Li T, Du Y (2017) An improved dynamic model of preloaded ball screw drives considering torque transmission and its application to frequency analysis. Adv Mech Eng 9(7):1–11. https://doi.org/10.1177/1687814017710580
Guo S et al. (2017) Cable-driven interventional operation robot with Stribeck friction feedforward compensation. In: 2017 IEEE International Conference on Mechatronics and Automation, ICMA 2017, 1787–1791. https://doi.org/10.1109/ICMA.2017.8016088
Górecki T, Krzýsko M, Ratajczak W, Wolýnski W (2016) An extension of the classical distance correlation coefficient for multivariate functional data with applications. Stat Transit 17(3):449–466. https://doi.org/10.21307/stattrans-2016-032
Kundrata J, Fujimoto D, Hayashi Y (2020) Comparison of Pearson correlation coefficient and distance correlation in Correlation Power Analysis on Digital Multiplier. In: 2020 43rd International Convention on Information, Communication and Electronic Technology (MIPRO). IEEE, 146–151. https://doi.org/10.23919/MIPRO48935.2020.9245325
Banks HT, Joyner ML (2017) AIC under the framework of least squares estimation. Appl Math Lett 74(1):33–45. https://doi.org/10.1016/j.aml.2017.05.005
Acknowledgements
This research was supported by the National Science and Technology Major Project [grant number 2018ZX04035002-002] and the National Natural Science Foundation of China [grant numbers 51575210 and 51675204].
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix 1. Machine tool feeding system
Based on a part program, the CNC system decomposes the contour to each single-axis reference position, u, which is fed to the machine tool feeding system. As illustrated in Appendix Fig. 20, the feeding system, as the core component of a machine tool, consists of mechanical transmission elements, a servo motor and its controller. Compared to the output position,w′, detected by the rotary encoder, the actual position,w, representing the position of the worktable, can reflect the true size of a part because of the errors caused by the mechanical transmission elements [13] are included.
In the work described here, the model of the feeding system is used to predict the actual position, w. To build an accurate model of the feeding system, the key challenges are decreasing the contour error in the nonlinear motion process and achieving stable prediction performance with different feedrates.
Appendix 2. Parameters and their identification results
Owing to the residual model for modelling the simulation error of the dynamics model, the parameters of the dynamics model do not need to be accurately identified. Therefore, the models for the servo drive and servo motor were built based on their design parameters. The main parameters of the mechanical transmission elements identified by using the least squares estimation (LSE) method [33] for the X- and Y-axis are given in Appendix Tables 4 and 5, respectively. Otherwise, the other parameters were set up by using the default values.
Appendix 3. Associated parameters of the NURBS curves
Heart contour: Order: k = 4, knot vector: {0, 0, 0, 0.15, 0.5, 0.5, 0.85, 1, 1, 1}; control points (x, y): {(0.0, 0.0), (−20.0, 50.0), (40.0, 20.0), (75.0, 0.0), (40.0, 20.0), (−20.0, 50.0), (0.0, 0.0)}; weights: {1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0}.
Goggles contour: Order: k = 4, knot vector: {0, 0, 0, 0.15, 0.3, 0.45, 0.6, 0.75, 0.9,1,1, 1}; control points (x, y): {(0.0, 0.0), (10.0, −40.0), (40.0, −10.0), (70.0, −40.0), (80.0, 0.0), (70.0, 10.0), (40.0, 20.0), (10.0,10.0), (0.0, 0.0)}; weights: {1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0}.
Rights and permissions
About this article
Cite this article
Jiang, Y., Chen, J., Zhou, H. et al. Residual learning of the dynamics model for feeding system modelling based on dynamic nonlinear correlate factor analysis. Appl Intell 51, 5067–5080 (2021). https://doi.org/10.1007/s10489-020-02096-2
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10489-020-02096-2