Error Optimization of Machine Vision based Tool Movement using a Hybrid CLONALG and PSO Algorithm

Error Optimization of Machine Vision based Tool Movement using a Hybrid CLONALG and PSO Algorithm

Prasant Kumar Mahapatra (Central Scientific Instruments Organization, Chandigarh, India), Anu Garg (Central Scientific Instruments Organization, Chandigarh, India) and Amod Kumar (Central Scientific Instruments Organization, Chandigarh, India)
Copyright: © 2016 |Pages: 14
DOI: 10.4018/IJAMC.2016010104
OnDemand PDF Download:
No Current Special Offers


A machine vision system with single monochrome CCD camera and backlight was developed for tool positioning and verification. While evaluating the performance of the developed vision system, the experiments showed that the output of machine vision system was not comparable to the output of sensors embedded in motion stages. Inherent factors like Imaging setup, camera calibration, environmental effects etc. are responsible for the error. These errors must be minimized to achieve maximum efficiency of developed vision system. In this paper, a novel hybrid algorithm is proposed to optimize the tool position error. The proposed algorithm comprises of CLONALG (one of the techniques of Artificial Immune System) and Particle Swarm Optimization (PSO) (a global optimization algorithm). Hybrid algorithm is tested on tool movement ranging from 0.020 mm to 7 mm. Performance of proposed algorithm is evaluated and also compared with CLONALG and PSO algorithms individually.
Article Preview

1. Introduction

Numerous types of Computer Numerically Controlled (CNC) machines like lathe, milling, drilling etc. are used for various manufacturing processes. All these machines possess different types of errors. Those errors are geometric, thermal and cutting force induced errors detected in the machine tools. In order to enhance the accuracy of machine tools and to improve quality of parts produced with them, the errors must be measured and compensated. A review of estimation and compensation of all three errors is presented in Ramesh et al. (2000). A unique single model has been proposed for geometric and cutting force induced error compensation in a 3-axis CNC milling machine (Raksiri and Parnichkun, 2004). A review of technologies available for measurement of geometry, numerical error compensation and use of Numerical Compensation (NC) for manufacturing machines and various challenges faced are explained in Schwenke et al. (2008).

These errors contribute to positional error of machine tools. Position errors come under the category of systematic error of geometric and kinematic errors (Mekid and Ogedengbe, 2010). Chen (1995) says geometric errors are the major contributors to the positioning errors. But thermally induced errors also cause positioning error of machine tools. Thermal generated is due to continual use of machine tools and this heat causes expansion of structural elements. This expansion causes inaccuracy in positioning of tools (Ramesh et al., 2000). Transducers used for measurement of tool position in CNC machines have been discussed by Bell (1963). Most prevalent method of detecting positional error is linear measurement. As described by Rashid (2005), linear positioning error is the difference between the real position measured by laser interferometer and the programmed position of tool.

An algorithm has been developed by Munlin (2002) to estimate and minimize the tool positional error in 5-axis milling machine. Fines et. al. (2008) has implemented a real position error compensation system based on Artificial Neural Network for machine tools in an industrial environment. In Barman et al. (2010), laser interferometry technique have been used to measure and compensate the linear positional errors of CNC machine tools. Numerous error compensation methods like direct sensor-based, indirect sensor-based, hybrid etc. have been discussed by Turek et. al. (2010).

Numerous machine vision systems for various applications like tool and flank wear monitoring and assessment (Kurada et al., 1997; Pfeifer et al., 2000; Bradley et al., 2001; Jurkovic et al., 2005; Su et al., 2006; Kerr et al., 2006; Atli et al., 2006; Castejon et al., 2007; Shahabi et al., 2008; 2009), measurement of surface roughness (Luk et al., 1989; Ho et al., 2002; Gadelmawla, 2004; Shahabi et al., 2010) etc. have been developed. Literature review shows that till date a machine vision system for lathe tool position estimation and compensation have not been developed. The authors developed a machine vision system for lathe tool positioning and verification without deploying any sensor. Motion sensors were initially embedded in motion controller to verify the performance of the developed system. A number of experiments during performance evaluation of the developed vision system showed that output of machine vision system was not accurately matching with output of motion sensors. These errors are due to camera calibration, noise in images, imaging setup, image measurement, environmental effects etc. In order to design a robust machine vision system, it was felt that these errors must be optimized. A hybrid (AIS-PSO) algorithm is proposed to optimize these errors. After applying the proposed algorithm, it was observed that lathe tool movement provided by a machine vision system and sensors in motion controller became comparable.

Complete Article List

Search this Journal:
Open Access Articles
Volume 13: 4 Issues (2022): 2 Released, 2 Forthcoming
Volume 12: 4 Issues (2021): 3 Released, 1 Forthcoming
Volume 11: 4 Issues (2020)
Volume 10: 4 Issues (2019)
Volume 9: 4 Issues (2018)
Volume 8: 4 Issues (2017)
Volume 7: 4 Issues (2016)
Volume 6: 4 Issues (2015)
Volume 5: 4 Issues (2014)
Volume 4: 4 Issues (2013)
Volume 3: 4 Issues (2012)
Volume 2: 4 Issues (2011)
Volume 1: 4 Issues (2010)
View Complete Journal Contents Listing