A comparative evaluation of neural networks and hidden Markov models for monitoring turning tool wear

Condition monitoring of machine tool inserts is important for increasing the reliability and quality of machining operations. Various methods have been proposed for effective tool condition monitoring (TCM), and currently it is generally accepted that the indirect sensor-based approach is the best practical solution to reliable TCM. Furthermore, in recent years, neural networks (NNs) have been shown to model successfully, the complex relationships between input feature sets of sensor signals and tool wear data. NNs have several properties that make them ideal for effectively handling noisy and even incomplete data sets. There are several NN paradigms which can be combined to model static and dynamic systems. Another powerful method of modeling noisy dynamic systems is by using hidden Markov models (HMMs), which are commonly employed in modern speech-recognition systems. The use of HMMs for TCM was recently proposed in the literature. Though the results of these studies were quite promising, no comparative results of competing methods such as NNs are currently available. This paper is aimed at presenting a comparative evaluation of the performance of NNs and HMMs for a TCM application. The methods are employed on exactly the same data sets obtained from an industrial turning operation. The advantages and disadvantages of both methods are described, which will assist the condition-monitoring community to choose a modeling method for other applications.     

Process monitoring using auto-associative,feed-forward artificial neural networks

The potential of using artificially simulated neural networks as intelligent, adaptive process-monitoring devices is discussed. The investigation is considered as a method for automatic, intelligent exception reporting for quality control applications. The technique is also compared with the conventional statistical approaches of principal component analysis and Kohonen's feature map. The applications of the technique in aerospace and manufacturing environments are presented and a possible extension of the method to incorporate a diagnostic function is discussed.Seconded from Cheltenham and Gloucester College of Higher Education as a Royal Society/SERC Research Fellow at Smith's Industries Aerospace and Defence Systems, Bishop's Cleeve, Cheltenham, UK.     

Feature-filtered fuzzy clustering for condition monitoring of tool wear

Condition monitoring is of vital importance in order to assess the state of tool wear in unattended manufacturing. Various methods have been attempted, and it is considered that fuzzy clustering techniques may provide a realistic solution to the classification of tool wear states. Unlike fuzzy clustering methods used previously, which postulate cutting condition parameters as constants and define clustering centres subjectively, this paper presents a fuzzy clustering method based on filtered features for the monitoring of tool wear under different cutting conditions. The method uses partial factorial experimental design and regression analysis for the determination of coefficients of a filter, then calculates clustering centres for filtering the effect of various cutting conditions, and finally uses a developed mathematical model of membership functions for fuzzy classification. The validity and reliability of the method are experimentally illustrated using a CNC machining centre for milling.     

On-line tool wear monitoring in turning using neural networks

The on-line supervision of a tool's wear is the most difficult task in the context of tool monitoring. Based on an in-process acquisition of signals with multi-sensor systems, it is possible to estimate or classify wear parameters by means of neural networks. This article demonstrates that solutions can be improved significantly by using available secondary information about physical models of the cutting process and about the temporal development of wear. Process models describing the influence of process parameters are used for a dedicated pre-processing of the sensor signals. The essential signal behaviour in a certain time window is described by means of polynomial coefficients. These coefficients are used as inputs for feedforward networks considering the temporal development of wear (multilayer perceptrons with a sliding window technique and time-delay neural networks). With a combination of the proposed measures it is possible to obtain remarkable improvements of both tool wear estimation and classification.     

Force-torque based on-line tool wear estimation system for CNC milling of Inconel 718 using neural networks

In a modern machining system, tool condition monitoring systems are needed to get higher quality production and to prevent the downtime of machine tools due to catastrophic tool failures. Also, in precision machining processes surface quality of the manufactured part can be related to the conditions of the cutting tools. This increases industrial interest for in-process tool condition monitoring (TCM) systems. TCM supported modern unmanned manufacturing process is an integrated system composed of sensors, signal processing interface and intelligent decision making strategies. This study includes key considerations for development of an online TCM system for milling of Inconel 718 superalloy. An effective and efficient strategy based on artificial neural networks (ANN) is presented to estimate tool flank wear. ANN based decision making model was trained by using real time acquired three axis (F_x, F_y, F_z) cutting force and torque (M_z) signals and also with cutting conditions and time. The presented ANN model demonstrated a very good statistical performance with a high correlation and extremely low error ratio between the actual and predicted values of flank wear.     

Design of neural network-based estimator for tool wear modeling in hard turning

Hard turning with cubic boron nitride (CBN) tools has been proven to be more effective and efficient than traditional grinding operations in machining hardened steels. However, rapid tool wear is still one of the major hurdles affecting the wide implementation of hard turning in industry. Better prediction of the CBN tool wear progression helps to optimize cutting conditions and/or tool geometry to reduce tool wear, which further helps to make hard turning a viable technology. The objective of this study is to design a novel but simple neural network-based generalized optimal estimator for CBN tool wear prediction in hard turning. The proposed estimator is based on a fully forward connected neural network with cutting conditions and machining time as the inputs and tool flank wear as the output. Extended Kalman filter algorithm is utilized as the network training algorithm to speed up the learning convergence. Network neuron connection is optimized using a destructive optimization algorithm. Besides performance comparisons with the CBN tool wear measurements in hard turning, the proposed tool wear estimator is also evaluated against a multilayer perceptron neural network modeling approach and/or an analytical modeling approach, and it has been proven to be faster, more accurate, and more robust. Although this neural network-based estimator is designed for CBN tool wear modeling in this study, it is expected to be applicable to other tool wear modeling applications.     

Diagnosis of tool wear based on cutting forces and acoustic emission measures as inputs to a neural network

Cutting forces and acoustic emission measures as a function of tool wear are presented for different cutting parameters and their applicability for tool condition monitoring is evaluated. The best of them, together with cutting parameters, were chosen as inputs to a feedforward, back propagation (FFBP) neural network; some training techniques were applied and their effectiveness is also evaluated. Conventional training of FFBP neural networks very soon leads to overtraining, hence to deterioration in the net response. Training of these nets depends very much on the initial weight values. A good way of finding satisfactory results is to introduce random distortions to the weight system, which efficiently push the net out of a local minimum of testing errors. An even more effective method may be to employ temporary shifts in the weights, alternately negative and positive. This has two advantages: (1) it brings the net to balance between training and testing errors and (2) it enables a great reduction in the number of hidden nodes.     

Controlling fast spring-legged locomotion with artificial neural networks

Controlling the model of an one-legged robot is investigated. The model consists merely of a mass less spring attached to a point mass. The motion of this system is characterised by repeated changes between ground contact and flight phases. It can be kept in motion by active control only. Robots that are suited for fast legged locomotion require different hardware layouts and control approaches in contrast to slow moving ones. The spring mass system is a simple model that describes this principle movement of a spring-legged robot. Multi-Layer-Perceptrons (MLPs), Radial Basis Functions (RBFs) and Self-Organising Motoric Maps (SOMMs) were used to implement neurocontrollers for such a movement system. They all prove to be suitable for control of the movement. This is also shown by an experiment where the environment of the spring-mass system is changed from even to uneven ground. The neurocontroller is performing well with this additional complexity without being trained for it.     

In-process regressions and adaptive multicriteria neural networks for monitoring and supervising machining operations

The authors develop a monitoring and supervising system for machining operations using in-process regressions (for monitoring) and adaptive feedforward artificial neural networks (for supervising). The system is designed for: (1) in-process tool life measurement and prediction; (2) supervision of machining operations in terms of the best machining setup; and (3) catastrophic tool failure monitoring. The monitoring system predicts tool life by using different sensors for gathering information based on a regression model that allows for the variations between tools and different machine setups. The regression model makes its prediction by using the history of other tools and combining it with the information obtained about the tool under consideration. The supervision system identifies the best parameters for the machine setup problem within the framework of multiple criteria decision making. The decision maker (operator) considers several criteria, such as cutting quality, production rate and tool life. To make the optimal decision with several criteria, an adaptive feedforward artificial neural network is used to assess the decision maker's preferences. The authors' neural network approach learns from the decision maker's complex behavior and hence, in automatic mode, can make decisions for the decision maker. The approach is not computationally demanding, and experiments demonstrate that its predictions are accurate.     

Connectivity oriented fast Hough transform for tool wear monitoring

Tool wear monitoring can be achieved by analyzing the texture of machined surfaces. In this paper, we present the new connectivity oriented fast Hough transform, which easily detects all line segments in binary edge images of textures of machined surfaces. The features extracted from line segments are found to be highly correlated to the level of tool wear. A multilayer perceptron neural network is applied to estimate the flank wear in various machining processes. Our experiments show that this Hough transform based approach is effective in analyzing the quality of machined surfaces and could be used to monitor tool wear. A performance analysis of our Hough transform is also provided.     

Prediction of vehicle reliability performance using artificial neural networks

Product development is an important but also dynamic, lengthy and risky phase in the life of a new product. The optimisation of the product development phase through extensive knowledge of the involved procedures is believed to reduce the risks and improve the final product quality. Artificial intelligence and expert systems have been used successfully in optimising the development phase of some new products as it will be demonstrated by the first sections of this publication. This paper presents the first module of an expert system, a neural network architecture that could predict the reliability performance of a vehicle at later stages of its life by using only information from a first inspection after the vehicle’s prototype production. The paper demonstrates how a tool like neural networks can be designed and optimised for use in reliability performance predictions. Also, this paper presents an optimisation methodology that enabled the neural network to deal with the limited amount of available training data, common during new product development, and to finally achieve acceptable prediction performance with small error. A case example is presented to demonstrate the methodology.     

Modeling workpiece vibrations with neural networks

The entire workpiece on a lathe vibrates when it is excited at a single point. Frequency and time-domain/time-series techniques can estimate the force-displacement relationships between excitation and the individual points on the workpiece. In this paper, the use of single neural network is proposed to represent the force-displacement relationship between the applied excitation force and the vibration of the whole workpiece. The accuracy of the proposed approach is evaluated on the experimental data. Also, another neural network is used to store the frequency response characteristics of the workpiece.     

A novel solution for maze traversal problems using artificial neural networks

In this paper we have addressed the problem of finding a path through a maze of a given size. The traditional ways of finding a path through a maze employ recursive algorithms in which unwanted or non-paths are eliminated in a recursive manner. Neural networks with their parallel and distributed nature of processing seem to provide a natural solution to this problem. We present a biologically inspired solution using a two level hierarchical neural network for the mapping of the maze as also the generation of the path if it exists. For a maze of size S the amount of time it takes would be a function of S (O(S)) and a shortest path (if more than one path exists) could be found in around S cycles where each cycle involves all the neurons doing their processing in a parallel manner. The solution presented in this paper finds all valid paths and a simple technique for finding the shortest path amongst them is also given. The results are very encouraging and more applications of the network setup used in this report are currently being investigated. These include synthetic modeling of biological neural mechanisms, traversal of decision trees, modeling of associative neural networks (as in relating visual and auditory stimuli of a given phenomenon) and surgical micro-robot trajectory planning and execution.     

The training of neural networks to model manufacturing processes

Neural networks have been increasingly used in various areas of manufacturing. Modelling of manufacturing processes, to allow experimentation on the model, is one of the areas in which successful applications have been reported. Most literature in this area is focused on network results. This paper concentrates on methods for training neural networks to model complex manufacturing processes. It summarises the use of neural network for process modelling in the past decade and provides some detailed guidelines for network training. A case study of a complex forming process is used to demonstrate a real implementation case in industry, and the issues arising from this case are discussed.     

SR-30 turbojet engine real-time sensor health monitoring using neural networks,and Bayesian belief networks

This paper describes the use of artificial intelligence-based techniques for detecting and isolating sensor failures in a turbojet engine. Specifically, three artificial intelligence (AI) techniques are employed: artificial neural networks (NNs), statistical expectations, and Bayesian belief networks (BBNs). These techniques are combined into an overall system that is capable of distinguishing between sensor failure and engine failure—a critical capability in the operation of turbojet engines. The turbojet engine used in this study is an SR-30 developed by Turbine Technologies. Initially, NNs were designed and trained to recognize sensor failure in the engine. The increased random noise output from failing sensors was used as the key indicator. Next, a Bayesian statistical method was used to recognize sensor failure based on the bias error occurring in the sensors. Finally, a BBN was developed to interpret the results of the NN and statistical evaluations. The BBN determines whether single or multiple sensor failures signify engine failure, or whether sensor failures represent separate, unrelated incidences. The BBN algorithm is also used to distinguish between bias and noise errors on sensors used to monitor turbojet performance. The overall system is demonstrated to work equally well during start-up and main-stage operation of the engine. Results show that the method can efficiently detect and isolate single or multiple sensor failures within this dynamic environment.     

A comparison of genetic programming and artificial neural networks in metamodeling of discrete-event simulation models

Genetic programming (GP) and artificial neural networks (ANNs) can be used in the development of surrogate models of complex systems. The purpose of this paper is to provide a comparative analysis of GP and ANNs for metamodeling of discrete-event simulation (DES) models. Three stochastic industrial systems are empirically studied: an automated material handling system (AMHS) in semiconductor manufacturing, an (s,S) inventory model and a serial production line. The results of the study show that GP provides greater accuracy in validation tests, demonstrating a better generalization capability than ANN. However, GP when compared to ANN requires more computation in metamodel development. Even given this increased computational requirement, the results presented indicate that GP is very competitive in metamodeling of DES models.     

Distributed memory systems for simulating artificial neural networks

In executing tasks involving intelligent information processing, the human brain performs better than the digital computer. The human brain derives its power from a large number [O(10¹¹)] of neurons which are interconnected by a dense interconnection network [O(10⁵) connections per neuron]. Artificial neural network (ANN) paradigms adopt the structure of the brain to try to emulate the intelligent information processing methods of the brain. ANN techniques are being employed to solve problems in areas such as pattern recognition, and robotic processing. Simulation of ANNs involves implementation of large number of neurons and a massive interconnection network. In this paper, we discuss various simulation models of ANNs and their implementation on distributed memory systems. Our investigations reveal that communication-efficient networks of distributed memory systems perform better than other topologies in implementing ANNs.     

Machine grouping using sequence-based similarity coefficients and neural networks

Most neural network approaches to the cell formation problem do not use information on the sequence of operations on part types. They only use as input the binary part-machine incidence matrix. In this paper we investigate two sequence-based neural network approaches for cell formation. The objective function considered is the minimization of transportation costs (including both intracellular and intercellular movements). Constraints on the minimum and maximum number of machines per cell can be imposed. The problem is formulated mathematically and shown to be equivalent to a quadratic programming integer program that uses symmetric, sequence-based similarity coefficients between each pair of machines. Of the two energy-based neural network approaches investigated, namely Hopfield model and Potts Mean Field Annealing, the latter seems to give better and faster solutions, although not as good as a Tabu Search algorithm used for benchmarking.     

Developing empirical models from observational data using artificial neural networks

Computer-integrated manufacturing requires models of manufacturing processes to be implemented on the computer. Process models are required for designing adaptive control systems and selecting optimal parameters during process planning. Mechanistic models developed from the principles of machining science are useful for implementing on a computer. However, in spite of the progress being made in mechanistic process modeling, accurate models are not yet available for many manufacturing processes. Empirical models derived from experimental data still play a major role in manufacturing process modeling. Generally, statistical regression techniques are used for developing such models. However, these techniques suffer from several disadvantages. The structure (the significant terms) of the regression model needs to be decided a priori. These techniques cannot be used for incrementally improving models as new data becomes available. This limitation is particularly crucial in light of the advances in sensor technology that allow economical on-line collection of manufacturing data. In this paper, we explore the use of artificial neural networks (ANN) for developing empirical models from experimental data for a machining process. These models are compared with polynomial regression models to assess the applicability of ANN as a model-building tool for computer-integrated manufacturing.Operated for the United States Department of Energy under contract No. DE-AC04-76-DP00613.