Information For Process Control Research Articles

Ｓｅｎｓｏｄｅｃ６Ｓランナビリティーモニタリングシステム

For a number of years on-line machinery condition and runnability monitoring systems have allowed maintenance departments and paper mill operating staff to measure machinery vibration, process pulsation and paper quality in order to detect mechanical deterioration or process stability problems at an early stage. With the analysis tools built into these systems, developing problems have been averted, thereby reducing the occurrence of unscheduled downtime and improving runnability. Preventive maintenance scheduling and papermachine fabric changes have been made more proactive and effective.Today, industry standards for network communications and emerging industry standards for operating systems and operator interfaces are facilitating the development of millwide networks. With the adoption of these standards, the seams between process control systems, quality control, production management and information systems are dissolving. Operations, quality control, and engineering personnel are now multiple users on the same network. Similarly, on-line machinery condition and runnability monitoring systems are becoming an integral part of the same network. Distributed architectures and high speed communications allow maintenance and operations personnel to analyze the same process and machinery condition information.While general information sharing is becoming commonplace, there is a continuing need for machinery condition and runnability monitoring systems to offer their specific users more insight into the complex, dynamic area of machinery condition, process pulsation and paper quality interactions. These task-oriented systems have recently introduced new signal processing techniques and user tools to make predictive maintenance diagnosis and operator interaction more insightful and effective.

A novel data reconciliation and gross error detection tool for the mineral processing industry

This paper describes a data reconciliation and gross error detection software (RAGE) to convert raw data into useful plant information for performance analysis, process control and circuit optimization. The conventional data reconcillation tools used in mineral processing industries generally focus on solid and component balances. RAGE uses Successive Quadratic Programming for data reconciliation and Generalized Likelihood Ratio Test with serial compensation technique for gross error detection. The software not only focuses on solid and component balances but also gives liquid balance across the entire circuit.

Solid electrolyte sensor for studying the behaviour of a partial oxidation catalyst

Oxygen-ion conducting solid electrolyte cells were prepared with Pt reference electrodes and Fe 2O 3–Sb 2O 4 working electrodes. The working electrode served as both electrode and catalyst during the partial oxidation of propene. The electrochemical cell allowed thermodynamic catalyst surface oxygen activities to be followed and acted as a sensor of catalyst state. It is clear that the catalyst was much more selective for acrolein formation at lower oxygen activities as indicated by the sensor signal. There is evidence to suggest that the oxygen activity measured by the sensor is as a result of the presence of adsorbed oxygen and not lattice oxygen. Based on the results it appears that potentiometric sensors could provide important additional information for catalytic partial oxidation process monitoring and control.

Use of supervisory information in process control

This article presents an approach to integrate supervision information directly into process control strategies. It focuses on one of the basic tasks of plant and process supervision: the validation of process information. A framework is suggested for how sophisticated process control functions with the knowledge of additional supervision information can react to abnormal process situations in order to keep the process running and stable. These tasks can be realised using function blocks, an established software approach for distributed control systems applicable to workstations as well as to microprocessors in remote I/O.

Applications of Intelligent Systems in Electronics Manufacturing

Fast growing production in rapidly changing environment has many challenges for decision-making, testing, inspection and control in electronics manufacturing. By combining and comparing the information collected from the production process it is possible to find reasons for some defects, and use this information in process control and development. Intelligent systems are useful in taking into account non-linearity especially in multivariable systems. Intelligent systems can successfully learn the knowledge from human inspectors by analysing on-line data and then use the knowledge to support decision making. These applications provide a good basis for wider use of intelligent systems together with other techniques in improving process design and control. This paper reviews the results gained in Control Engineering Laboratory, University of Oulu.

High speed fiber optics backbones

In late 1995, the data communications networks at the Fletcher Challenge Canada's Crofton pulp and paper mill had reached capacity and were suffering from overload. The large growth in process control information and personal computers over the past few years was straining the abilities of existing coaxial and fiber optic systems to reliably transfer data. In addition, increased network complexity made troubleshooting the networks extremely difficult and time-consuming. This article discusses the design methods used to create a new fiber optic backbone to provide the data carrying capacity and reliability needed for the pulp and paper mill's computers and control systems. High-speed network technologies such as ATM (asynchronous transfer mode), FDDI (fiber distributed data interface) and fast Ethernet were compared for their suitability in industrial environments. The design also incorporated Ethernet switches so that the existing Ethernet networks could attach at minimal cost. These switches subdivided the original single large network into approximately 24 sub-networks. This division greatly increased the network's allowable traffic capacity and prevented problems in any one area from propagating throughout the mill, and thereby reduced network and process downtime.

Temporal management using relative time in knowledge-based process control

In this paper, a knowledge-based approach is proposed for the management of temporal information in process control. A common-sense theory of temporal constraints over processes/events, allowing relative temporal knowledge, is employed here as the temporal basis for the system. This theory supports duration reasoning and consistency checking, and accepts relative temporal knowledge which is in a form normally used by human operators. An architecture for process control is proposed which centres on an historical database consisting of events and processes, together with the qualitative temporal relationships between their occurrences. The dynamics of the system is expressed by means of three types of rule: database updating rules, process control rules, and data deletion rules. An example is provided in the form of a life scheduler, to illustrate the database and the rule sets. The example demonstrates the transitions of the database over time, and identifies the procedure in terms of a state transition model for the application. The dividing instant problem for logical inference is discussed with reference to this process control example, and it is shown how the temporal theory employed can be used to deal with the problem.

Quantitative Analysis of Process Streams by On-Line FT-IR Spectrometry

Current efforts in the Process Analytical Chemistry and Control Technology group at Bell Laboratories, Lucent Technologies, have focused on the development of an on-line method for real-time characterization of organic-containing effluents (principally trimethyamine (TMA) and methanol) produced during a high-temperature processing. On-line analysis of the gas streams was performed by combining FT-IR spectrometry with partial least squares (PLS) to obtain quantitative information for process control. The gas-phase infrared spectra were measured at a resolution of 16 cm-1 under atmospheric conditions. The correlation coefficients for all the components including water were larger than 0.999, and the standard errors of prediction were much less than 1% by weight. The effect of CO2 on the predictability of other components is discussed. Results of our experiment showed that the predicted errors for TMA and methanol increased by as much as 10 and 15%, respectively, as the amount of CO2 increased to 100% of the...

Real-time automatic statistical control for the grinding process

The highly consistent axes control of current computer numerically controlled (CNC) grinding machines has ensured a high-quality output provided that the process parameters can be kept constant. However, there are many variations in the process that give rise to differences between workpieces. Typical process variations such as tool wear, materials and temperature changes are inevitable and cannot be compensated by advancements in axis servo-control. This paper addresses this problem by applying a high-level control methodology which makes use of real-time statistical process control (SPC) information to determine whether the process has deteriorated due to process variations. Adjustments to machine parameters to offset these changes can then be made. The methodology has been successfully incorporated in a control loop within the controller of the CNC grinders to achieve truly unattended grinding operations.

Spatial characterization of wafer state using principal component analysis of optical emission spectra in plasma etch

Optical emission spectroscopy (OES) is often used to obtain in-situ estimates of process parameters and conditions in plasma etch processes. Two barriers must be overcome to enable the use of such information for real-time process diagnosis and control. The first barrier is the large number of measurements in wide-spectrum scans, which hinders real-time processing. The second barrier is the need to understand and estimate not only process conditions, but also what is happening on the surface of wafer, particularly the spatial uniformity of the etch. This paper presents a diagnostic method that utilizes multivariable OES data collected during plasma etch to estimate spatial asymmetries in commercially available reactor technology. Key elements of this method are: first, the use of principal component analysis (PCA) for dimensionality reduction, and second, regression and function approximation to correlate observed spatial wafer information (i.e., line width reduction) with these reduced measurements. Here we compare principal component regression (PCR), partial least squares (PLS), and principal components combined with multilayer perceptron neural networks (PCA/MLP) for this in-situ estimation of spatial uniformity. This approach has been verified for a 0.35-/spl mu/m aluminum etch process using a Lam 9600 TCP etcher. Models of metal line width reduction across the wafer are constructed and compared: the root mean square prediction errors on a test set withheld from training are 0.0134 /spl mu/m for PCR, 0.014 /spl mu/m for PLS, and 0.016 /spl mu/m for PCA/MLP. These results demonstrate that in-situ spatially resolved OES in conjunction with principal component analysis and linear or nonlinear function approximation can be effective in predicting important product characteristics across the wafer.

Wafer level reliability: Process control for reliability

Wafer Level Reliability test techniques can be used to provide fast feedback process control information regarding the reliability of the product of a semiconductor process. The purpose of wafer level reliability (WLR) tests is the measurement of variation in the materials comprising the semiconductor device. They are not intended as modeling tools for the quantification of the effect of stress on these materials. As such, WLR tests must provide a repeatable stress, independent of normal process variation. The results of these tests will be a measurement of the “rate of degradation” of the basic circuit elements caused by a standard stress.

Real-time data management in a flexible manufacturing system (FMS)

The need for more flexible and efficient data management in manufacturing systems has become obvious. To secure the highest utilization rate and maximum productivity of manufacturing systems, it is necessary to be able to find the right information at the right time and at the right place: Just In Time. Non production and inefficiency in single operated, stand-along NC-machine tools and other NC-equipment are due largely to humanly caused elapsed times between the receipt of information, the formulation of a decision and the performance of the command. The advantages of Flexible Manufacturing systems (FMS) are obvious if they are compared to stand-along machines. FMS-concepts, hardware and software-state of the art, are designed mostly on the idea of centralized control. These types of systems are difficult and expensive to expand and maintain. Centralization makes the FMS dependent on a network and a main computer that might cause troubles with breakdowns. This paper presents solutions for decentralization and real-time data management of the control system of an FMS. An FMS has to be capable of controlling its resources and its state in real-time and take care of the connections to upper levels of enterprise network (CIM). This leads to the use of real time multi-tasking operating systems and suitable software for system developing. Relational Database Management systems (RDBMS) are very common in administrative use but in process-control information management solutions are still based on traditional ways of programming. Standard RDBMSs offer an easier way to integrate the manufacturing information system with other computer aided (CA-) subsystems through Local Area Networks (LANs). Dependency on the format in which the data is stored is not so crucial if the strategic decision of the company-wide standard for data management has been approved. The most useful approach to factory systems integration is the modularity of flexible components that are to be implemented to the system. Only in this way it is possible to design systems that are flexible and expandable also in the future. When the system is implemented the technology is at the certain level and the requirements for the products made in the system are well known. In the future, demands to expand the system with new facilities might become topical. If there is already a modular system it is not a great problem to install new equiptment in the system.

PVM-Parallel Virtual Machine: AUsers' Guide and Tutorial for Networked Parallel Computing

Part 1 Introduction: heterogeneous network computing trends in distributed computing PVM overview other packages. Part 2 The PVM system. Part 3 Using PVM: how to obtain the PVM software setup to use PVM setup summary starting PVM common startup problems running PVM programs PVM console details host file options. Part 4 Basic programming techniques: common parallel programming paradigms workload allocation porting existing applications to PVM. Part 5 PVM user interface: process control information dynamic configuration signalling setting and getting options message passing dynamic process groups. Part 6 Program examples: fork-join dot product failure matrix multiply one-dimensional heat equation. Part 7 How PVM works: components messages PVM daemon libpvm library protocols message routing task environment console program resource limitations multiprocessor systems. Part 8 Advanced topics: XPVM porting PVM to new architectures. Part 9 Troubleshooting: geting PVM installed getting PVM running compiling applications running applications debugging and tracing debugging the system. Appendices: history of PVM versions PVM 3 routines.

Scanning-probe microscope analysis of optical thin films: A new analytical tool in a manufacturing environment

OCLI is a leading manufacturer of thin films for optical and thermal control applications. The determination of thin film and substrate topography can be a powerful way to obtain information for deposition process design and control, and about the final thin film device properties. At OCLI we use a scanning probe microscope (SPM) in the analytical lab to obtain qualitative and quantitative data about thin film and substrate surfaces for applications in production and research and development. This manufacturing environment requires a rapid response, and a large degree of flexibility, which poses special challenges for this emerging technology. The types of information the SPM provides can be broken into three categories:(1)Imaging of surface topography for visualization purposes, especially for samples that are not SEM compatible due to size or material constraints;(2)Examination of sample surface features to make physical measurements such as surface roughness, lateral feature spacing, grain size, and surface area;(3)Determination of physical properties such as surface compliance, i.e. “hardness”, surface frictional forces, surface electrical properties.

X¯ CONTROL CHART PATTERN IDENTIFICATION THROUGH EFFICIENT OFF-LINE NEURAL NETWORK TRAINING

Back-propagation pattern recognizers (BPPR) are proposed to identify unnatural patterns exhibited on Shewhart control charts. These unnatural patterns, such as cycles and trends, can provide valuable information for real-time process control. In a computer-integrated manufacturing environment, the operator need not routinely monitor the control chart but, rather, can be alerted to patterns by a computer signal generated by the proposed algorithm. In this paper, an off-line analysis is performed to investigate the training and learning speed of these BPPRs on simulated X¯ data. The best configuration of the network is further tested to demonstrate the classification capability of the proposed BPPR.

Fault management in process control: eye movements and action.

Operators controlled a simulated thermal hydraulic system based on Crossman's waterbath task. They were required to keep the system at set points for temperature, level and flow rates. The system was subjected to disturbances and to failures, which the operators were required to manage. Behavioural data were collected in the form of keystroke commands, and in addition eye movements were measured. Evidence for 'cognitive lockup' and for a preference for serial fault management were found. Fixation durations do not change, but faulty subsystems are examined more frequently. During lockup other subsystems may be examined but action on them is delayed. Eye movements give considerable additional insight into the nature of process control information processing compared with traditional measures of operator performance.

Plutonium process control using an on-line gamma monitor for uranium, plutonium, and Americium

An On-Line Gamma Monitor profiles the concentration of uranium, plutonium, and americium in waste and product streams of the anion exchange process used to purify plutonium at Los Alamos. The Monitor employs passive gamma Spectrometry to measure the 59.5-KeV and 129-KeV gamma rays of241Am and239Pu, respectively. Because the uranium impurity in typical process streams has no gamma ray suitable for passive measurement, a novel radiotracer technique is used. Uranium-237, always present in plutonium processed at Los Alamos as a minor alpha-decay daughter of241Pu, has a 6.8-day half-life and 208-KeV gamma energy, which make it an ideal radiotracer for macro amounts of uranium in the process. The On-Line Gamma Monitor is used routinely to provide Los Alamos operators with continuous, real-time process control information.

Computer Communication for the Integration of Information in Industrial Production

Industrial production is based on multi-level information ranging from strategic considerations on the enterprise level to process control and product information on the production level. This information was very often processed locally on those different information levels; the interchange and interpretation of (digital) data carrying the information, was the task of information users and programmers. Today, computerized information processing facilitates an integration of all above mentioned levels through computer communication. In order to arrive at an open information exchange, standardized protocols are required considering the transmission of data as well as their semantic interpretation. The discussion of general demands of information integration is followed by the characterization of the currently available communication methods and networks (WAN, LAN, PABX). The ISO-OSI standards and their production-oriented extension towards MAP/TOP and the Field Bus are discussed in detail with special emphasis on application layer protocols. This leads to conformance and interoperability testing considerations and standards. The presentation of an example of information integration in steel production concludes the paper

ACOUSTIC EMISSION AND MACHINING - PROCESS ANALYSIS AND CONTROL

ABSTRACT A quantitative model considering signal attenuation and tool flank wear is proposed to describe the influence of machining process parameters on the level of acoustic emission (AE) generation. The AE sensitivities to width of cut, feed rate, and hardness of material in three-dimensional cutting are experimentally studied. The application of AE signal analysis to chip formation detection, providing information for process control in unattended machining, is described.

The valuing of management information

This article is the first in a series of four articles concerned with the valuing of management information. The first three articles examine the major approaches to valuation, the final one summarizes and links together these methods into a practi cal approach to the valuing of management information. This article considers the Bayesian approach to the valuing of management information by examining two examples. One of these examples is an unprogrammed strategic decision con cerned with valuing a piece of market research information, the other concerned with the valuing of process control informa tion.