Control Chart Patterns Research Articles

A neural network approach to characterize pattern parameters in process control charts

Abnormal patterns on manufacturing process control charts can reveal potential quality problems due to assignable causes at an early stage, helping to prevent defects and improve quality performance. In recent years, neural networks have been applied to the pattern recognition task for control charts. The emphasis has been on pattern detection and identification rather than more detailed pattern parameter information, such as shift magnitude, trend slope, etc., which is vital for effective assignable cause analysis. Moreover, the identification of concurrent patterns (where two or more patterns exist together) which are commonly encountered in practical manufacturing processes has not been reported. This paper proposes a neural network-based approach to recognize typical abnormal patterns and in addition to accurately identify key parameters of the specific patterns involved. Both single and concurrent patterns can be characterized using this approach. A sequential pattern analysis (SPA) design was adopted to tackle complexity and prevent interference between pattern categories. The performance of the model has been evaluated using a simulation approach, and numerical and graphical results are presented which demonstrate that the approach performs effectively in control chart pattern recognition and accurately identifies the key parameters of the recognized pattern(s) in both single and concurrent pattern circumstances.

Simulation of weight prunning process in backpropagation neural network for pattern classification: A self-running threshold approach

Neural network minimization is a process in which non-affecting elements (non-contributing processing elements and weights) are deleted. This results in the network with a minimum configuration and positively contributing elements. In general, the scope of the minimization process is to achieve a network which provides optimal solution for a given task without loss of performance. Therefore, study into network minimization becomes essential for complicated tasks. The present study investigates ways of developing automatic thresholding methods for weight deleting process to minimize the network resulting in enhanced performances for classification of six control chart patterns. To achieve this aim, two weight deleting processes are developed including dynamic and static weight deleting processes. The automatic weight deleting mechanism introduced in the present study is based on the optimal RMS error variation. In this case, once the RMS error drops to the desired level, weight deleting process starts, which then enables one to select the optimal threshold automatically, in this case, RMS error becomes minimum. The study is extended to include two sets of data structures, namely first and second data kinds. In the first data kind, random variables and coefficients in equations governing the chart patterns are varied while in the second data kind, only random variables are changed. These enable one to compare the generalization capacity of the resulting network which uses the first and second data kinds separately. It is found that dynamic deleting process results improved performances and the late start of deleting process is fruitful for improved learning rate. In addition, the use of first data kind provides the network with better generalization capacity.

Control chart pattern recognition using a new type of self-organizing neural network

Control charts as used in statistical process control can exhibit six principal types of patterns: normal, cyclic, increasing trend, decreasing trend, upward shift and downward shift. Apart from normal patterns, all the other patterns indicate abnormalities in the process that must be corrected. Accurate and speedy detection of such patterns is important to achieving tight control of the process and ensuring good product quality. This paper describes a new type of neural network for control chart pattern recognition. The neural network is self-organizing and can learn to recognize new patterns in an on-line incremental manner. The key feature of the proposed neural network is the criterion employed to select the firing neuron, i.e. the neuron indicating the pattern class. The paper gives a comparison of the results obtained using the proposed network and those for other self-organizing networks employing a different firing criterion.

Feature-based control chart pattern recognition

This paper describes a new approach for the recognition of control chart patterns (CCPs). The approach uses features extracted from a CCP instead of the unprocessed CCP data or its statistical properties for the recognition task. These features represent the shape of the CCP explicitly. The approach has two main steps: (1) extraction of features and (2) recognition of patterns. A set of CCP feature extraction procedures are described in the paper. The extracted features are recognized using heuristics, induction and neural network techniques. The paper presents the results of analysing several hundred control chart patterns and gives a comparison with those reported in previous work.

AUTOMATED UNNATURAL PATTERN RECOGNITION ON CONTROL CHARTS USING CORRELATION ANALYSIS TECHNIQUES

Pattern recognition techniques are currently pursued to identify unnatural patterns on quality control charts. This approach has been shown to enhance the ability to utilize the information of the chart more effectively than conventional run rules. This paper presents analysis and development of a pattern recognition system for identifying unnatural patterns on quality control charts. The system is based on correlation analysis, where a set of optimal matched filters are generated. To illustrate the design methodology and operation of the system, a set of commonly encountered patterns is utilized, such as the trend, the systematic, and the cyclic patterns. A training algorithm that minimizes the probabilities of Type I and Type II errors i presented. To evaluate the system performance, a testing algorithm as well as a set of newly-defined performance measures are introduced. The obtained results, based on extensive simulation runs, have proved the effectiveness of correlation analysis for control chart pattern recognition. © 1997 Elsevier Science Ltd

A neural network approach for the analysis of control chart patterns

Control chart pattern recognition is an important aspect of statistical process control (SPC). The presence of unnatural patterns indicates that a process is affected by assignable causes, and corrective actions should be taken. This paper describes two types of pattern recognizers based on different neural network architectures: a multilayer perceptron trained by back-propagation and a modular neural network. The pattern recognizers were developed to take the advantage of the fact that a particular unnatural pattern is often associated with a set of assignable causes. Identification of unnatural patterns can greatly narrow the set of possible causes that must be investigated, and thus the diagnostic search could be reduced in length. The performances of the proposed pattern recognizers were evaluated through Monte Carlo simulations on the basis of appropriate performance measures. An extensive evaluation indicates that the proposed pattern recognizers could recognize multiple unnatural patterns for which they were trained. The modular neural network provides a better recognition accuracy than back-propagation when there will be strong interference effects. The notable features of the proposed pattern recognizers include the compactness of input/output representation, the ability to detect unnatural patterns when the signal-to-noise ratio is low, the outstanding capability to detect unnatural patterns starting anywhere in the sequence of data and the directional invariance property in the sense that patterns of different orientations could be detected equally well.

A neural network approach to identifying cyclic behaviour on control charts: a comparative study

The pattern recognition approach to expanding the usefulness and effectiveness of traditional control charts has been proposed and studied. The approaches adopted vary from statistical-based or artificial-intelligence-based (e.g. expert systems), to computational-intelligence-based (e.g. neural networks), or a mixture. Although general-purpose control chart pattern recognition systems have been shown to be useful in identifying a variety of non-random patterns, this was achieved at the price of losing the ability to identify certain details in individual pattern classes. Therefore, a special-purpose system is desirable to compensate for the limitation of a general-purpose system. In this research, a number of special-purpose (for cyclic data) control chart pattern recognizers based on several neural-network paradigms—namely back-propagation, ART1, ARTMAP, and fuzzy ARTMAP—were developed and their performance was carefully studied and compared. Extensive simulation was conducted to study: the use of proper training data and training strategies, the effect of complement coding, the use of binary or analogue input, and the proper range of pattern parameter values. The performance was measured by type I and type II errors as well as by average run length. In general, a special-purpose system shows significant improvement over its corresponding general-purpose system. Although back-propagation systems are more tolerant of a high level of noise, the family of ART-based systems performs more homogeneously across the range of the cycle amplitude studied. Within the ART family, fuzzy ARTMAP with complement coding outperforms other ART-based systems.

Self-adjusting diagnostic system for the manufacture of crystal resonators

In the development of a diagnostic system to monitor the condition of the frequency tuning process, fuzzy logic can be applied in recognition of unnatural statistical control chart patterns. The heuristics for reasoning is based on the principles behind statistical process control. With the experience of an expert in troubleshooting, one can derive the necessary knowledge to associate the pattern of a control chart to a set of known physical causes. Since these causes related to the unnatural statistical patterns are not independent, it is difficult to precisely associate the chart distribution patterns to the known causes. Furthermore, the causes of problems dealt with will also vary with time. Hence, by means of neural networks, it is possible to learn the association of fuzzily deduced chart patterns to plausible causes in order to achieve an optimum operating condition.

Unnatural pattern recognition on control charts using correlation analysis techniques

This paper presents analysis and development of a pattern recognition system for identifying unnatural patterns on quality control charts. The system is based on correlation analysis, where a set of optimal matched filters are generated. To illustrate the design methodology and operation of the system, a set of commonly encountered patterns is utilized, such as the trend, the systematic, and the cyclic patterns. A training algorithm that minimizes the probabilities of Type I and Type II errors is presented. To evaluate the system performance, a testing algorithm as well as a set of newly-defined performance measures are introduced. The obtained results have shown the effectiveness of correlation analysis for control chart pattern recognition.

Detecting process non-randomness through a fast and cumulative learning ART-based pattern recognizer

An adaptive resonance theory (ART) based, general-purpose control chart pattern recognizer (CCPR) which is capable of fast and cumulative learning is presented. The implementation of this ART-based CCPR was made possible by introducing two key alternatives, that is, incorporating a synthesis layer in addition to the existing two-layer architecture and adopting a quasi-supervised training strategy. i A detailed algorithm with the training and the testing modes was presented. Extensive simulations and performance evaluations were conducted and proved that this ART-based CCPR indeed possesses the capability of fast and cumulative learning. When compared with a back-propagation pattern recognizer (BPPR), the ART-based CCPR is superior on cyclic patterns, inferior on mixture patterns, and comparable on other patterns. Furthermore, an ART-based CCPR is easier to develop since it needs fewer training templates and takes less time to learn. This study not only provides a basis for understanding the capabilities o...

An Integrated Neural Network and Expert System Tool for Statistical Process Control

Control charts are a basic means for monitoring the quality characteristics of a manufacturing process to ensure the required quality level. They are used to track product and process variations through graphical representation of the quality variable of interest. A control chart shows the state of control of a process and can exhibit different types of patterns which are indicative of long-term trends in it. This paper describes the integration of an expert system and a neural-network-based pattern recognizer for analysing and interpreting control charts. The expert system has an on-line process monitoring package to detect general out-of-control situations and a diagnosis module to suggest corrective actions. The pattern recognizer is an on-line system comprising two neural networks and an heuristics module designed to identify incipient process abnormalities from control chart patterns. The paper also compares neural networks and expert systems and provides the rationale for the integration process.

Monitoring Projects Using Cash Flow Control Charts

ABSTRACT Post Completion Audits (PCA). an important area in capital budgeting have not received much attention in the past. In a competitive environment, however, PCA plays an important role. Uncertain initial forecasts and current estimates can be revised using post audit information under a Bayesian approach at virtually no additional cost, thus enhancing future decisions. Prueitt and Park [2]. developed the Cash Flow Control Chart (CFCC) model for a special class of investment problems with multiple identical units with uncertain cash flows, typically found in Advanced Manufacturing Systems (AMS) and fleet replacement. This paper provides meaningful interpretations to modified Cash Row Control Chart (M-CFCC) patterns.

Control chart pattern recognition using learning vector quantization networks

Pattern recognition systems using neural networks for discriminating between different types of control chart patterns are discussed. A class of pattern recognizers based on the Learning Vector Quantization (LVQ) network is described. A procedure to increase the classification accuracy and decrease the learning time for LVQ networks is presented. The results of control chart pattern recognition experiments using both existing LVQ networks and an LVQ network implementing the proposed procedure are given.

Control Chart Pattern Recognition Using Combinations of Multi-Layer Perceptrons and Learning-Vector-Quantization Neural Networks

Pattern recognition systems made up of independent multi-layer perceptrons and learning-vector-quantization neural network modules have been developed for classifying control chart patterns. These composite pattern recognition systems have better classification capabilities than their individual modules. The paper describes the structures of these pattern recognition systems and the results obtained on using them.

Back-propagation pattern recognizers for [formula omitted] control charts: Methodology and performance

The pattern recognition algorithm presented here is based on the perception that, as automated data collection becomes more widespread in manufacturing processes, the monitoring of control charts will be performed by computer-based algorithms. These algorithms will have to detect unnatural patterns to assist in the correction of assignable causes. The work currently being performed in addressing the application of pattern recognition to control charts is directed toward answering this need. In this paper, a control chart pattern recognition methodology based on the back-propagation algorithm, a neural computing theory, is presented. This classification algorithm, suitable for real-time statistical process control, evaluates observations routinely collected for control charting to determine whether a pattern, such as a trend or cycle, exists in the data. The foundation of the algorithm is based on the neural network concepts of constructing and training a network in the types of patterns to be detected. These concepts mimic the trained operator's ability to detect patterns. Here, the pattern recognizer is trained and tested with the consideration of Type I error (finding a pattern where none existed) as well as Type II error (failing to detect a known pattern). Performance measures sensitive to these types of errors are used to evaluate the algorithm's performance on an extensive series of simulated patterns of control chart data. This approach is promising because of its flexible training and high-speed computation.

Boltzmann machines that learn to recognize patterns on control charts

Boltzmann machines (BM), a type of neural networking algorithm, have been proven to be useful in pattern recognition. Patterns on quality control charts have long been recognized as providing useful information for correcting process performance problems. In computer-integrated manufacturing environments, where the control charts are monitored by computer algorithms, the potential for using pattern-recognition algorithms is considerable. The main purpose of this paper is to formulate a Boltzmann machine pattern recognizer (BMPR) and demonstrate its utility in control chart pattern recognition. It is not the intent of this paper to make comparisons between existing related algorithms. A factorial design of experiments was conducted to study the effects of numerous factors on the convergence behavior and performance of these BMPRs. These factors include the number of hidden nodes used in the network and the annealing schedule. Simulations indicate that the temperature level of the annealing schedule significantly affects the convergence behavior of the training process and that, to achieve a balanced performance of these BMPRs, a medium to high level of annealing temperatures is recommended. Numerical results for cyclical and stratification patterns illustrate that the classification capability of these BMPRs is quite powerful.

Interpretation of Quality-Control Charts for Concrete Production

The interpretation of quality-control charts is examined. Both moving average and moving average for range charts (R) are discussed. Control chart patterns are illustrated and their causes are identified. This paper will provide the users of control charts interpretation and guidance on their use and practical application for eliminating out-of-control problems.

A Novel Firing Rule For Training Kohonen SelforganisingMaps

Statistical Process Control Charts can exhibit six principal types of patterns: Normal, Cyclic, Increasing Trend, Decreasing Trend, Upward Shift and Downward Shift. All except Normal patterns indicate abnormalities in the process that must be corrected. Accurate and speedy detection of such patterns is important to achieving tight control of the process and ensuring good product quality. This paper describes an implementation of the Kohonen self-organising map which employs the Euclidean Distance as the firing rule for control chart pattern recognition. First, the structure of the network is outlined and the equations which govern its dynamics are given. Then the learning mechanism of the network is explained. The effects of different combinations of network parameters on classificatio n accuracy are discussed. A novel firing rule for the Kohonen self-organising map is proposed. This rule involves component-bycomponent comparison between the input pattern and the established class templates. When an input vector is presented, it is compared with the class templates in all the neurons in turn. The neuron containing the class template that best matches the input vector will subsequently fire. This approach is intended to enhance the generalisation capability and accuracy of the Kohonen selforganising map. The paper gives a comparison of the results obtained using the Euclidean Distance and the proposed firing rule.