NC end milling optimization using evolutionary computation

Typically, NC programmers generate tool paths for end milling using a computer-aided process planner but manually schedule “conservative” cutting conditions. In this paper, a new evolutionary computation technique, particle swarm optimization (PSO), is proposed and implemented to efficiently and robustly optimize multiple machining parameters simultaneously for the case of milling. An artificial neural networks (ANN) predictive model for critical process parameters is used to predict the cutting forces which in turn are used by the PSO developed algorithm to optimize the cutting conditions subject to a comprehensive set of constraints. Next, the algorithm is used to optimize both feed and speed for a typical case found in industry, namely, pocket-milling. Machining time reductions of up to 35% are observed. In addition, the new technique is found to be efficient and robust.     

An in-process surface recognition system based on neural networks in end milling cutting operations

An in-process based surface recognition system to predict the surface roughness of machined parts in the end milling process was developed in this research to assure product quality and increase production rate by predicting the surface finish parameters in real time. In this system, an accelerometer and a proximity sensor are employed as in-process surface recognition sensors during cutting to collect the vibration and rotation data, respectively. Using spindle speed, feed rate, depth of cut, and the vibration average per revolution (VAPR) as four input neurons, an artificial neural networks (ANN) model based on backpropagation was developed to predict the output neuron-surface roughness R_a values. The experimental results show that the proposed ANN surface recognition model has a high accuracy rate (96–99%) for predicting surface roughness under a variety of combinations of cutting conditions. This system is also economical, efficient, and able to be implemented to achieve the goal of in-process surface recognition by retrieving the weightings (which were generated from training and testing by the artificial neural networks), predicting the surface roughness R_a values while the part is being machined, and giving feedback to the operators when the necessary action has to be taken.     

A neural network approach for force and contour error control in multi-dimensional end milling operations

The problem of controlling the average resultant cutting force together with the contour error in multi-dimensional end milling operations is considered in this study. Two sets of neural networks are used in the control system. The first set is used to specify the feed rate to maintain a desired cutting force. This feed rate is resolved along the feed axes using a parametric interpolation algorithm so that the desired part shape is obtained. The second set is used to make corrections to the feed rate components specified by the parametric interpolation algorithm to minimize the contour error caused by the dynamic lag of the closed-loop servo systems controlling the feed drives. In addition, the control system includes a feedforward input to compensate for static friction effects. Experimental results are presented for machining two-dimensional circular slots and a three-dimensional spherical surface to show the validity of the proposed approach.     

Intelligent tool path generation for milling of free surfaces using neural networks

The presented paper has an intention to show how with the help of Artificial Neural Network (ANN), the prediction of milling tool-path strategy could be made in order to establish which milling path strategy or their sequence will show the best results (will be the most appropriate) at free surface machining, according to set technological aim. In our case the best possible surface quality of machined surface was taken as the primary technological aim. Configuration of used Neural Network (NN) is presented, and the whole procedure is shown on an example of mould, for producing light switches. The verification of machined surface quality, according to average mean roughness, R_a, is also being done, and compared with the NN predicted results.     

Direct adaptive control of end milling process

A direct adaptive control algorithm, which is spindle speed and drive dynamics independent, has been developed for machining operations. The combined dynamics of feed motion and cutting process are modelled as a third order system whose parameters may vary with spindle speed and part geometry changes during machining. The algorithm does not use any specific time interval, thus sampling time dependent discrete transfer functions and pole assignments are avoided. The adaptive controller is designed to have a closed loop characteristic function which behaves like an open loop regular and stable machining operation. The proposed direct adaptive controller is practical, can be used in any multi-axes machining, and can be combined with chatter suppression techniques which require spindle speed regulation. The algorithm is applied to the adaptive control of milling. Satisfactory results are obtained in constraining the maximum cutting forces and dimensional surface errors in milling experiments.     

Fuzzy control of feed rate in end milling operations

In order to improve productivity in end milling operations, a new adaptive control system based on fuzzy logics to maintain a constant cutting force is developed. It is shown, by experimental cutting tests, that the cutting tool travels in the air cut with fast feed rate, yet in the varying depths of cut, the tool travels with an adjustable feed rate to prevent the occurrence of tool breakage and maintain a high metal removal rate.     

Fuzzy control of spindle power in end milling processes

This paper reports a fuzzy control system for power regulation in end milling processes. This control system is capable of adjusting both feedrate and spindle speed simultaneously. Experiments have been carried out using both steel and aluminum workpieces of various cutting geometries. Different tools (HSS and carbide tools of different diameters and different number of teeth) have been used for aluminum workpieces. Both full immersion slotting and partial immersion cutting were tested. Our test results show that the system was in sensitive to workpiece and tool changes and cutting power was well regulated around the target levels for various types of variations in depth of cut. Our test results also show that as compared to single parameter (feedrate) adjustment, further savings in machining time can be achieved by adjusting both feedrate and spindle speed.     

基于人工神经网络和遗传算法的平面铣削加工参数自适应优化

机械加工最优自适应控制的关键在于自适应加工模型的建立和实时优化策略的制定。本文提出用人工神经网络方法建立加工过程模型 ,用遗传算法实现在线优化。基于以上算法 ,构造了平面铣削加工参数自适应优化系统 ,可使加工系统在不违反加工约束的前提下 ,总是获得最大材料去除率。     

Expert control strategy using neural networks for electrolytic zinc process

1　ＩＮＴＲＯＤＵＣＴＩＯＮＴｈｅｔｈｒｅｅｂａｓｉｃｓｔｅｐｓｉｎｚｉｎｃｈｙｄｒｏｍｅｔａｌｌｕｒｇｙａｒｅｌｅａｃｈｉｎｇ ,ｐｕｒｉｆｉｃａｔｉｏｎａｎｄｅｌｅｃｔｒｏｌｙｓｉｓ .Ｔｈｅｅｌｅｃｔｒｏｌｙｔ ｉｃ ｐｒｏｃｅｓｓｉｎｖｏｌｖｅｓ ｐａｓｓｉｎｇａｎｅｌｅｃｔｒｉｃａｌｃｕｒｒｅｎｔｔｈｒｏｕｇｈｉｎｓｏｌｕｂｌｅｅｌｅｃｔｒｏｄｅｓｔｏｃａｕｓｅｔｈｅｄｅｃｏｍｐｏｓｉ ｔｉｏｎｏｆａｎａｑｕｅｏｕｓｚｉｎｃｓｕｌｆａｔｅｅｌｅｃｔｒｏｌｙｔｅａｎｄｔｈｅｄｅ ｐｏｓｉｔｉｏｎｏ…     

Simulation of optimum control for electrodischarge dressing using neural networks

A neural identifier and a neural controller for optimum control of the electro-discharge dressing systems are proposed. A modeling of a system is obtained from a neural identifier and a neural control structure satisfying stability is proposed. Computer simulation results show that the proposed neural identifier not only gives accurate modeling results but also can find the relationship of the electro-discharge dressing system. In addition, the proposed neural controller gives very effective control according to gap increase by the learning process in spite of the nonlinear characteristics of electro-discharge conditions.     

Monitoring of tool fracture in end milling using induction motor current

This paper describes the use of induction motor current to monitor tool fracture in end milling operations. The principles of induction motors are studied in this paper to establish the relationship between the motor current and the motor torque. It is shown that the square of the stator current of induction motors is approximately proportional to the motor torque. Since the occurrence of tool fracture will cause variations in the motor torque, measurement of the stator current appears to be an indirect technique for monitoring tool fracture. A sensitivity analysis of the stator current to the occurrence of tool fracture is also reported. Finally, experimental results under varying cutting conditions have been presented to demonstrate the effectiveness of this approach for the detection of tool fracture in end milling operations.     

Dynamic response analysis in end milling using pre-twisted beam finite elements

This paper develops an analytical model for estimating the dynamic responses in end milling, i.e. dynamic milling cutter deflections and cutting forces, by using the finite-element method along with an adequate end milling-cutting force model. The whole cutting system includes the spindle, the bearings and the cutter. The spindle is modelled structurally with the Timoshenko-beam element, the milling cutter with the pre-twisted Timoshenko-beam element due to its special geometry, and the bearings with lumped springs and dampers. Because the damping matrix in the resulting finite-element equation of motion for the whole cutting system is not one of proportional damping due to the presence of bearing damping, the state-vector approach and the convolution integral is used to find the solution of the equation of motion. To assure the accuracy of prediction of the dynamic response, the associated cutting force model should be sufficiently precise. Since the dynamic cutting force is proportional to the chip thickness, a quite accurate alogorithm for the calculation of the variation of the chip thickness due to geometry, run-out and spindle-tool viration is developed. A number of dynamic cutting forces and tool deflections obtained from the present model for various cutting conditions are compared with the experimental and analytical results available in the literature, good agreement being demonstrated for these comparisons. The present model is useful, therefore, for the prediction of end milling instability. Also, the tool deflections obtained using the pre-twisted beam element are found to be smaller than those obtained using the straight beam element without pre-twist angle. Hence neglecting the pre-twist angle in the structural model of the milling cutter may overestimate the tool deflections.     

Geometric adaptive straightness control system for the peripheral end milling process

A relationship between the tool deflection and the feed rate is modelled by a modified Taylor's tool-life equation. An off-line Geometric Adaptive Control (GAC) system to compensate for machining straightness error in the finished surface due to tool deflection and guideway error generated by the peripheral end milling process is proposed.

Without a priori knowledge of the variations of the cutting parameters, the time-varying parameters are estimated by an exponentially windowed recursive least squares method with only post-process measurements of the straightness error through a gap sensor. The location error is compensated by moving the milling bed through a numerical control command before cutting. The waviness error is regulated by using optimal feed rate manipulation as obtained from the proposed GAC method during machining although the parameters do not converge to fixed values.

Experimental results show that the location error is controlled within the range of fixturing error of the milling bed on the guideway, the waviness accuracy can be increased to more than three times that of the case with no control action. A single-pass milling operation can become feasible through practical application of the proposed GAC system for finish cutting conditions.     

Hybrid adaptive control based on the characteristics of CNC end milling

In this paper, hybrid adaptive control algorithm that controls feedrate is proposed to regulate spindle current. For this purpose, variation of steady state spindle current, time constant and time delay were examined through experiments based on step end milling under various cutting condition. The developed hybrid adaptive control algorithm is composed of adjustable proportional feedback control (adjustable P control), fine control carries out detailed control, and entry feedrate control reduces the peak current produced when a tool makes contact with a workpiece. The adjustable P control was verified through comparisons between calculate P gains and experimentally obtained P gains. The hybrid adaptive control algorithm was applied to various cutting conditions and it showed global stability as well as excellent applicability behavior.     

铝基复合材料铣削加工过程参数的预测与优化(英文)

研究了LM25Al/SiCp复合材料的铣削加工特征。将获得的相关试验数据采用响应面方法建立了一个数学模型来描述各种加工参数对后刀面磨损率的影响。采用标准的响应面方法来设计实验。方差分析结果表明,在实验研究范围内,所建立的数学模型能够很好地描述铣削加工各参数的影响。采用优化组合参数得到了最小的后刀面磨损率。     

Performance-based optimization of multi-pass face milling operations using Tribes

The paper proposes a new optimization technique based on Tribes for determination of the cutting parameters in multi-pass milling operations such as plain milling and face milling by simultaneously considering multi-pass rough machining and finish machining. The optimum milling parameters are determined by minimizing the maximum production rate criterion subject to several practical technological constraints. The cutting model formulated is a nonlinear, constrained programming problem. Experimental results show that the proposed Tribes-based approach is both effective and efficient.     

Thermal modelling of end milling

Determination of the temperatures during machining is one of the most important challenges for accurate milling simulations. Coupled with excessive shearing, plastic deformation and friction in a small region of cutting, the temperatures in milling may have very significant impact on parts and tools such as dimensional errors, residual stresses and tool wear. Temperature exhibits a non-linear complex-modelling problem in milling process. In this article, for the first time, a novel thermal modelling is introduced for fast and accurate prediction of temperatures in end milling processes. A theoretical modelling approach and experimental validations are presented for various cutting conditions.     

Drill wear monitoring using neural networks

The primary objective of this research is to monitor drill wear on-line. In this paper, drill wear monitoring is carried out by measuring the thrust force and torque signals. In order to identify the tool wear conditions based on the signal measured, a neural network, using a cumulative back-propagation algorithm, is adopted. This paper also describes the experimental procedure used and presents the results obtained for establishing the neural network. The inputs to the neural network are the mean values of thrust force and torque, spindle rotational speed, feedrate and drill diameter. The neural network is trained to estimate the average drill wear. It is confirmed experimentally that the tool wear can be accurately estimated by the trained neural network. The accuracy of tool wear estimation using the neural network is superior to that using other regression models.     

Tracking control of two-wheel driven mobile robot using compound sine function neural networks

The purpose of this paper is to propose a compound sine function neural network (NN) with continuous learning algorithm for the velocity and orientation angle tracking control of a mobile robot. Herein, two NN controllers embedded in the closed-loop control system are capable of on-line continuous learning and do not require any knowledge of the dynamics model. The neuron function of the hidden layer in the three-layer feed-forward network structure is on the basis of combining a sine function with a unipolar sigmoid function. In the NN algorithm, the weight values are only adjusted between the nodes in hidden layer and the output nodes, while the weight values between the input layer and the hidden layer are one, that is, constant, without the weight adjustment. The developed NN controllers have simple algorithm and fast learning convergence. Therefore, the proposed NN controllers can be suitable for the real-time tracking control of the mobile robots. The simulation results show that the proposed NN controller has better control performance in the tracking control of the mobile robot. The compound sine function NN provides a new way to solve tracking control problems for a mobile robot.