3.外文翻译（原版）.pdf

journal of materials processing technology 2 0 9 ( 2 0 0 9 ) 23022313 journal homepage: Intelligent adaptive control and monitoring of band sawing using a neural-fuzzy system Ilhan Asilt urka, Ali Un uvarb a Faculty of Technical Education, Selcuk University, Konya 42250, Turkey b Faculty of Mechanical Engineering, Selcuk University, Konya 42250, Turkey a r t i c l ei n f o Article history: Received 27 June 2007 Received in revised form 11 March 2008 Accepted 14 May 2008 Keywords: Intelligent manufacturing Fuzzy control Neural-fuzzy controller Adaptive control of band sawing Band sawing a b s t r a c t In bandsaw machines, it is desired to feed the bandsaw blade into the workpiece with an appropriate feeding force in order to perform an effi cient cutting operation. This can be accomplishedbycontrollingthefeedrateandthrustforcebyaccuratelydetectingthecutting resistance against the bandsaw blade during cutting operation. In this study, a neural-fuzzy- based force model for controlling band sawing process was established. Cutting parameters werecontinuouslyupdatedbyasecondaryneuralnetwork,tocompensatetheeffectofenvi- ronmental disturbances. Required feed rate and cutting speed were adjusted by developed fuzzy logic controller. Results of cutting experiments using several steel specimens show that the developed neural-fuzzy system performs well in real time in controlling cutting speed and feed rate during band sawing. A material identifi cation system was developed by using the measured cutting forces. Materials were identifi ed at the beginning of the cutting operation and cutting force model was updated by using the detected material type. Con- sequently, cutting speed and feed rate were adjusted by using the updated model. The new methodology is found to be easily integrable to existing production systems. 2008 Elsevier B.V. All rights reserved. 1.Introduction In band sawing, the power rating of the machine limits the thickness and hardness of the metal to be cut. In band sawing process, metal removal is accomplished by forcing a multi- toothed tool against the workpiece. The depth of cut in sawing cannotbepresetlikeothermetalcuttingprocessesandcontrol can only be exercised over the thrust load applied between the blade and workpiece material. The amount of metal removed by each tooth is dependent primarily on how well the blade transmits the applied pressure to the workpiece and also on the penetration ability of the cutting teeth. Machining forces generated during sawing process are therefore found to have greater signifi cance than in other chip removal pro- cesses. It has been found that thrust and cutting loads per tooth per unit thickness reduced with an increase in cutting Corresponding author. Tel.: +90 332 223 3344; fax: +90 332 241 2179. E-mail address: .tr (I. Asilt urk). speed. A reduction in the thrust force will cause a reduction in the depth of cut taken by the engaged teeth. An increase in the feed rate causes a substantial increase in both cut- ting and thrust loads per tooth. Geometry of the workpiece does also have a considerable infl uence on cutting perfor- mance. In band sawing, the thrust load is normally constant alongtheworkpiecebreadth.Whensawingroundsectionsthe width of the workpiece changes within the cut, the cutwidth increasesastheblademovestowardsthecentreanddecreases as the cut is being fi nished. Band saw machines that operate on a pressure feed principle maintain a constant chip load per tooth as described while the blade saws through varying sections. Artifi cial intelligence (AI) methods are widely used in solu- tion of complex engineering problems. Some of the most commonlyusedAItechniquesareneuralnetworks(NN),fuzzy 0924-0136/$ see front matter 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jmatprotec.2008.05.031 journal of materials processing technology 2 0 9 ( 2 0 0 9 ) 230223132303 Nomenclature Ddiameter eoutput error ffeed rate fiinstantaneous feed rate Foutprocess output Fmeasmeasured force Frefreference force Gmaterial group no Hhardness Iintegral of error IAM 1intelligent adaptive module Linstantaneous height MRRmaterial removal rate PIperformance index TWRtool wear rate yprocplant output yrefreference model output viinstantaneous cutting speed logics (FL), expert systems (ES) and models using hybrids of these. Artifi cial intelligence methods are used in every stage of manufacturing. Machining is one of the basic manufacturing methods used in the industry. Manufacturers must minimize costandprocesstime,andadditionallytheproductmustcom- ply with the required dimensions and quality criteria for a better competition. Increasing the productivity of metal cutting machine tools is a principal concern for manufacturing industry. In tra- ditional machining systems, cutting parameters are usually selected prior to machining according to machining hand- books or the users experience. The selected machining parameters are usually conservative to avoid machining fail- ure. To ensure the quality of machining products, to reduce the machining costs and increase the machining effi ciency, it is necessary to adjust the machining parameters in real- timeandtooptimizemachiningprocessatthattime.Adaptive control of the machining process is preferable to solve above problems. Since band sawing process is non-linear and time-varying, it is diffi cult for traditional identifi cation methods to provide an accurate model. Adaptive control methods provide on-line adjustment of the operating conditions. Therefore, parame- ter adaptive control techniques for machining processes were developed to adjust the feed rate automatically to maintain a constant cutting force. Applications of these techniques suc- cessfully increased both the metal removal rate and tool life. In this paper, an intelligent neural-fuzzy adaptive control scheme is proposed for band sawing process. The proposed adaptive control system can be applied effectively in various cutting situations. 2.Literature survey Therearealotofworksexistingintheliteratureonmonitoring and controlling of the machining operation. Groover pointed out that conventional control theory could be ineffi cient and unstable due to disturbing variations in the machining conditions. It is stated that fi xed cutting forces wouldbeausefulapproachforincreasingtoollifeandmaterial removal rate (Groover, 1987). The conventional PID feedback control system has been usedincontrollingmachiningprocessesbynumerous researchers (Masory and Koren, 1980, 1985; Lauderbaugh and Ulsoy, 1989; Koren, 1988). The main problem with the fi xed gain Adaptive Control Constraint (ACC) system is the one that produce poor performance and may become unstable during the time-varying machining process. The use of various forms of adaptive control in an ACC system has been examined by adjusting the gain of the controller. Model reference adaptive control-based ACC systems (MRAC)havebeendevelopedbysomeresearchers(Masoryand Koren, 1980, 1985; Lauderbaugh and Ulsoy, 1989). These stud- ies found that MRAC perform control duties better than fi xed gain controllers. A typical MRAC incorporates the parameter estimation of the cutting process. Recently, many studies have been devoted to the theory of fuzzy control and its application to machining processes. Tarng et al. developed a fuzzy logic-based controller (FLC) for adaptive control of turning operations. The developed FLC can adjust feed rate on-line so as to reduce machining time and maintain constant force (Tarng and Cheng, 1993; Tarng and Wang, 1993). IntheexperimentalstudiesofZhangand Khanchustambham (1993), it is shown that process opti- mization is possible by online monitoring and controlling of the machining process. This eliminates the effect of disturbances caused by operator. An online monitoring system was designed by Ordonez et al. (1997) by using artifi cial intelligence based on sensors. Signals which were taken from sensors are used in AI deci- sion making during the cutting process. The real time signals obtained through force transducers and estimated cutting forces obtained by using NN were compared. Consequently, estimated model was implemented to surface roughness, tool wear and geometric tolerances. Feed forward and back propagation algorithms were used as architecture and train- ing algorithm of NN model, respectively. Direct and indirect adaptive fuzzy techniques and simulations of conventional controls were compared. An adaptive control approach was suggested by Rodolfo et al. (1998) for maintaining the cutting force constant, in the milling process. The constant force feed rate was investigated without delay time. Tsai et al. (1999) observed that, surface roughness can experimentally be determined by one or more quantita- tive measurements. Estimated surface roughness model was based on relative vibrations between the tool and the work piece. Estimated surface roughness was improved by using signals that are taken from vibration and proximity sensors. System accuracy was observed as 9699%. An adaptive controller with optimization was designed based on two kinds of NN by Liu and Wang (1999) for milling process. A modifi ed back propagation NN was pro- posed adjusting its learning rate and adding dynamic factor in the learning process, and was used for the online modeling 2304journal of materials processing technology 2 0 9 ( 2 0 0 9 ) 23022313 of the milling system. A modifi ed Augmented Lagrange Multi- plier (ALM) neural network model was proposed adjusting its iteration step, and was used for the real time optimal control of the milling process. In this study, the simulation and exper- imental results show that not only does the milling system with the designed controller have high robustness and global stability, but also the machining effi ciency of the milling sys- tem with the adaptive controller is much higher than for the traditional CNC milling system. Adaptive control constraint is one of the methods used in controlling machining processes. Force control algorithms have been developed and evaluated by numerous researchers. Among the most common is the fi xed gain proportional inte- gral (PI) controller, originally proposed for milling by Kim et al. (1999). The gain of the controller is adjusted in response to variations in cutting conditions in the proposed controller. The essential aim of the neural network-based controller is toconstructareversefunctionforthemachiningsystemusing theNNsothattheoutputofthemachiningsystemapproaches to the desired output. Machining process can usually be con- trolled by adjusting the feed rate or spindle speed. The neural network-based ACC system has been applied to machining process control by (Liu et al., 1999; Hang and Chiou, 1996). In a study by Liu et al. (2001), the major adaptive control constraintsystemswerediscussedbasedonthefeedbackcon- trol, parameter adaptive control/self-tuning control, model reference adaptive control, variable structure control/sliding modecontrol,neuralnetworkcontrol,andfuzzycontrol.Their typical applications to constant cutting force control system are also described, and some recent experiments results were presented. Online method of achieving optimal settings of a fuzzy- neural network has been developed by Sandak and Tanaka. Results of the cutting experiments using several wood species show that the fuzzy-neural system developed performs well in online feed rate optimization during band sawing, while maintainingsawdeviationwithinspecifi edlimits(Sandakand Tanaka, 2003). Zuperl et al. (2005) discussed the application of fuzzy adap- tive control strategy to the problem of cutting force control in highendmillingoperations.IntheirACsystem,thefeedrateis adjusted on-line in order to maintain a constant cutting force in spite of variations in cutting conditions. They developed a simplefuzzycontrolstrategyintheintelligentsystemandcar- riedoutsomeexperimentalsimulationswiththefuzzycontrol strategy. The effect of cutting speed, feed rate and work piece geom- etry in band sawing were investigated by Ahmad et al. (1987). In the experimental studies, reduction in the thrust force and cutting force per teeth for unit thickness were observed, as the cutting speed increased. Ko and Kim observed that, in order to create a mechanis- tic model of cutting force, specifi c cutting pressure should be obtained through cutting experiments. The band sawing pro- cess is similar to milling in that it involves multi-point cutting, so it is not an easy matter to evaluate specifi c cutting pres- sure. The cutting force is predicted by analyzing the geometric shape of a saw tooth. They stated that the predicted cutting force coincided well with those measured in validation exper- iment. Therefore, the predicted cutting forces in band sawing can be used for the adaptive control of saw-engaging feed rate in band sawing (Ko and Kim, 1999). Anderson et al. (2001) pointed out a mechanical cutting force model for band sawing. The model describes the vari- ation in cutting force between individual teeth and relates it to initial positional errors, tool dynamics and edge wear. Band sawing is a multi-tooth cutting process, and the terminology of the cutting action is discussed and compared with other cutting processes. 3.Adaptive control of machining processes Intelligent machining system applications include monitor- ing and control technologies. This system also improves the machining operations. In this system, process related data is acquired, and then process is controlled. Manufacturing industries are affected by computer technologies. Recently, automation works were made on the material handling, quality monitoring, motion control, source planning, pro- cess control, etc. The sensing of the machining process is much more comprehensive and complex. Many papers have been prepared about monitoring and control of the machine tool. Researchers and industrialists concerned with tool mon- itoring and adaptive control. One of the most important functions of the intelligent control is the provision of required action in the unknown or indefi nite ambient processes. Machine tool and cutting tools are protected by the moni- toring system. Tool changing cost, scrap rate and production cost is reduced by real time tool wear measurement. Thus, full capacities of the machine tools are maintained. In the machining processes, feed rate is continuously adjusted for keeping on the process with constant reference force in the adaptive control systems. Thanks to adaptive control, since it aims to minimize the production time to adjust the feed rate to optimal values in the high capacity working conditions. Consequently, tool life increases with the restricted load application. 3.1.Necessities of the adaptive controls in the machining operation In the case of depth or width of cut, feed rate are usually adjusted to compensate for the variability. This type of vari- ability is often encountered in profi le milling or contouring operations (Groover, 1987). When hard spots or the areas of diffi culty to machine are encountered in the workpiece, either speed or feed is reduced to avoid premature failure of the tool. If the work piece defl ects as a result of insuffi cient rigidity inthesetup,thefeedratemustbereducedinordertomaintain accuracy in the process. As the tool begins to dull, the cutting forces increase. The adaptive controller will typically respond to tool dulling by reducing the feed rate. The workpiece geometry may contain shaped sections where no machining needs to be performed. If the tool were to continue feeding through these so-called air gaps at the same rate, time would be lost. Accordingly, the typical procedure is journal of materials processing technology 2 0 9 ( 2 0 0 9 ) 230223132305 Fig. 1 Adaptive control optimization in milling system. to increase the feed rate, by a factor of two or three, when air gaps are encountered. These sources of variability present themselves as time- varying and, for the most part, unpredictable changes in the machining process. It should be examined how adaptive con- trol can be used to compensate for these changes. 3.2.Adaptive control methods in the machining process The methods that are mentioned below are used in machining operations. These are namely Adaptive Control with Opti- mization, Geometric Adaptive Control and Adaptive Control Constraint. 3.2.1.Adaptive control optimization (ACO) In this type of adaptive control, a performance index is spec- ifi ed for the system. This performance index is a measure of overall process performance, such as production rate or cost per volume of metal removed. The objective is to opti- mize the index of performance by controlling speeds and/or feeds. A system with the adaptive controller for machining pro- cess can be constructed based on NNs as shown in Fig. 1. The system is modeled on-line by the modifi ed BP learning algo- rithm. The feed rate is adjusted and the process is optimized in real-time by the modifi ed ALM NN. In the process the differ- ence between measured cutting force and estimated cutting force is (e), which is used as back propagation NN and is to adjust the weights of the NN. Feed rate is adjusted in the sense of object function constraints (Liu et al., 1999). 3.2.2.Geometric adaptive control (GAC) Geometricadaptivecontrolisusuallyusedinfi nishmachining operations, where the objective is to achieve a desired surface quality and/or accurate part dimensions despite tool wear or tool defl ection (Liang et al., 2004). Owing to the relationship between feed rate and surface quality, surface roughness or dimensional accuracy of the part is continuously measured by the sensor by means of feed-back. Therefore in most GAC systems, the cutting speed is con- stant and the machining feed is manipulated to achieve the desired surface quality (Masory and Koren, 1980). The