Massey Documents by Type
Permanent URI for this communityhttps://mro.massey.ac.nz/handle/10179/294
Browse
10 results
Search Results
Item Contributions to improve the power, efficiency and scope of control-chart methods : a thesis submitted in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Statistics at Massey University, Albany, New Zealand(Massey University, 2019) Adegoke, Nurudeen AdedayoDetection of outliers and other anomalies in multivariate datasets is a particularly difficult problem which spans across a range of systems, such as quality control in factories, microarrays or proteomic analyses, identification of features in image analysis, identifying unauthorized access in network traffic patterns, and detection of changes in ecosystems. Multivariate control charts (MCC) are popular and sophisticated statistical process control (SPC) methods for monitoring characteristics of interest and detecting changes in a multivariate process. These methods are divided into memory-less and memory-type charts which are used to monitor large and small-to-moderate shifts in the process, respectively. For example, the multivariate χ2 is a memory-less control chart that uses only the most current process information and disregards any previous observations; it is typically used where any shifts in the process mean are expected to be relatively large. To increase the sensitivity of the multivariate process control tool for the detection of small-to-moderate shifts in the process mean vector, different multivariate memory-type tools that use information from both the current and previous process observations have been proposed. These tools have proven very useful for multivariate independent normal or "nearly" normal distributed processes. Like most univariate control-chart methods, when the process parameters (i.e., the process mean vector or covariance parameters, or both) are unknown, then MCC methods are based on estimated parameters, and their implementation occurs in two phases. In Phase I (retrospective phase), a historical reference sample is studied to establish the characteristics of the in-control state and evaluate the stability of the process. Once the in-control reference sample has been deemed to be stable, the process parameters are estimated from Phase I, and control chart limits are obtained for use in Phase II. The Phase II aspect initiates ongoing regular monitoring of the process. If successive observed values obtained at the beginning of Phase II fall within specified desired in-control limits, the process is considered to be in control. In contrast, any observed values during Phase II which fall outside the specified control limits indicate that the process may be out of control, and remedial responses are then required. Although conventional MCC are well developed from a statistical point of view, they can be difficult to apply in modern, data-rich contexts. This serious drawback comes from the fact that classical MCC plotting statistics requires the inversion of the covariance matrix, which is typically assumed to be known. In practice, the covariance matrix is seldom known and often empirically estimated, using a sample covariance matrix from historical data. While the empirical estimate of the covariance matrix may be an unbiased and consistent estimator for a low-dimensional data matrix with an adequate prior sample size, it performs inconsistently in high-dimensional settings. In particular, the empirical estimate of the covariance matrix can lead to in ated false-alarm rates and decreased sensitivity of the chart to detect changes in the process. Also, the statistical properties of traditional MCC tools are accurate only if the assumption of multivariate normality is satisfied. However, in many cases, the underlying system is not multivariate normal, and as a result, the traditional charts can be adversely affected. The necessity of this assumption generally restricts the application of traditional control charts to monitoring industrial processes. Most MCC applications also typically focus on monitoring either the process mean vector or the process variability, and they require that the process mean vector be stable, and that the process variability be independent of the process mean. However, in many real-life processes, the process variability is dependent on the mean, and the mean is not necessarily constant. In such cases, it is more appropriate to monitor the coefficient of variation (CV). The univariate CV is the ratio of the standard deviation to the mean of a random variable. As a relative dispersion measure to the mean, it is useful for comparing the variability of populations having very different process means. More recently, MCC methods have been adapted for monitoring the multivariate coefficient of variation (CV). However, to date, studies of multivariate CV control charts have focused on power - the detection of out-of-control parameters in Phase II, while no study has investigated their in-control performance in Phase I. The Phase I data set can contain unusual observations, which are problematic as they can in uence the parameter estimates, resulting in Phase II control charts with reduced power. Relevant Phase I analysis will guide practitioners with the choice of appropriate multivariate CV estimation procedures when the Phase I data contain contaminated samples. In this thesis, we investigated the performance of the most widely adopted memory-type MCC methods: the multivariate cumulative sum (MCUSUM) and the multivariate exponentially weighted moving average (MEWMA) charts, for monitoring shifts in a process mean vector when the process parameters are unknown and estimated from Phase I (chapters 2 and 3). We demonstrate that using a shrinkage estimate of the covariance matrix improves the run-length performance of these methods, particularly when only a small Phase I sample size is available. In chapter 4, we investigate the Phase I performance of a variety of multivariate CV charts, considering both diffuse symmetric and localized CV disturbance scenarios, and using probability to signal (PTS) as a performance measure. We present a new memory-type control chart for monitoring the mean vector of a multivariate normally distributed process, namely, the multivariate homogeneously weighted moving average (MHWMA) control chart (chapter 5). We present the design procedure and compare the run length performance of the proposed MHWMA chart for the detection of small shifts in the process mean vector with a variety of other existing MCC methods. We also present a dissimilarity-based distribution-free control chart for monitoring changes in the centroid of a multivariate ecological community (chapter 6). The proposed chart may be used, for example, to discover when an impact may have occurred in a monitored ecosystem, and is based on a change-point method that does not require prior knowledge of the ecosystem's behaviour before the monitoring begins. A novel permutation procedure is employed to obtain the control-chart limits of the proposed charting test-statistic to obtain a suitable distance-based model of the target ecological community through time. Finally, we propose enhancements to some classical univariate control chart tools for monitoring small shifts in the process mean, for those scenarios where the process variable is observed along with a correlated auxiliary variable (chapters 7 through 9). We provide the design structure of the charts and examine their performance in terms of their run length properties. We compare the run length performance of the proposed charts with several existing charts for detecting a small shift in the process mean. We offer suggestions on the applications of the proposed charts (in chapters 7 and 8), for cases where the exact measurement of the process variable of interest or the auxiliary variable is diffcult or expensive to obtain, but where the rank ordering of its units can be obtained at a negligible cost. Thus, this thesis, in general, will aid practitioners in applying a wider variety of enhanced and novel control chart tools for more powerful and effcient monitoring of multivariate process. In particular, we develop and test alternative methods for estimating covariance matrices of some useful control-charts' tools (chapters 2 and 3), give recommendations on the choice of an appropriate multivariate CV chart in Phase I (chapter 4), present an efficient method for monitoring small shifts in the process mean vector (chapter 5), expand MCC analyses to cope with non-normally distributed datasets (chapter 6) and contribute to methods that allow efficient use of an auxiliary variable that is observed and correlated with the process variable of interest (chapters 7 through 9).Item A simulation of selected statistical process control methods :a thesis presented in partial fulfilment of the requirements for the degree of Master of Technology at Massey University(Massey University, 1989) Chanprasert, SireeA simulation program, SQC, was developed at the Production Technology Department, Massey University. The program was written in Vax Basic 3.0 which is structured programming language and is run on the Vax computer under the VAX/VMS operating system 4.5. SQC is a menu-driven program which was designed to simulate data from a variety of production processes subject to inherent random variation and predetermined changes; sample selection for statistical quality purposes. Such decisions were made via the available feature to allow for user interactive control of the process parameters and sample selection methods while the chart of selected method was plotted on the terminal screen as well as optionally on the printer. The exercise has been done to test and to observe how the program performed and produced the output on the screen and terminal-format files. Moreover, the program evaluation was carried out by comparing with a published article, which is satisfactorily acceptable. The SQC can be utilized as a teaching tool for students in practising how each statistical process control method performs and how to make a right decision at a right time and as a research tool to observe and use the simulated results to predict and to improve the production process in the future.Item Design of a monitor for the debugging and development of multiprocessing process control systems : a thesis presented in partial fulfilment of the requirements for the degree of Master of Technology in Computing Technology at Massey University(Massey University, 1987) Dobbie, GillianThis thesis describes the design of a general purpose tool for debugging and developing multimicroprocessor process control systems. With the decreasing pnce of computers, multimicroprocessors are increasingly being used for process control. However, the lack of published information on multiprocessing systems and distributed systems has meant that methodologies and tools for debugging and developing such systems have been slow to develop. The monitor designed here is system independent, a considerable advantage over other such tools that are currently available.Item The Pro-t-con project : the development of a system for continuous process improvement using Pro-t-con process optimisation software at AEP Filmpac Ltd : a thesis submitted to the Faculty of Technology and Engineering, Massey University, in partial fulfilment of the requirements for the degree of Master of Technology in Quality Systems(Massey University, 2000) Moynagh, Paul KennethThis project details the work done to develop a methodology for process improvement at AEP Filmpac in Auckland, New Zealand. The company had purchased a process optimisation software package called Pro-t-con which they intended to use to improve the operating conditions for each product on each machine in the plant. Early use of the Pro-t-con software produced a number of questions as to its ability to optimise processes as effectively as expected. Thus research was done to test the effectiveness of the package and analyse its strengths and weaknesses. The results of this work suggest that Pro-t-con although very easy to use is limited in its ability to effectively optimise processes. Statistically it lacks the rigor of Classical and Taguchi design of experiment methods and cannot resolve processes with interactions or non-linear factors. At the outset of the project the plant did not possess a system for suitably storing and retrieving machine setup information, thus any improvements made to the settings one day would not be available for use the next time that product was run. Consequently in order to longitudinally develop process settings it was also necessary to develop a setting sheet system to support the process improvement initiatives. The combination of a methodology for continuously improving processes and one for actually undertaking experiments to exploit such a process produced a coherent 10 step method for general process improvement This method was used successfully on a variety of processes at plants in Auckland and Sydney.Item A 'pinch' technology analysis of energy integration in the Huntly power station : a thesis presented in partial fulfilment of the requirements for the degree of Master of Technology(Massey University, 1990) Tioe, Jen FukMost processing industries use combinations of heat exchangers for heating and/or cooling process streams. A large amount of the energy used by the processing industries is used just for process heat. In the industrial nations of Western Europe approximately one third of the national energy use is for process heat (Smith, 1981). The design of most industrial processes is based on a long period of development with many evolutions and improvements leading to a current flowsheet. It is often assumed that these flowsheets are more or less optimal, with no significant "faults" left in them. This is of course not true. An analysis carried out by the New Zealand Dairy Research Institute (NZDRI) showed that theoretical minimum energy consumptions are 25 -30% lower than the actual energy consumption of the most efficient of New Zealand's dairy processing plants (Lovell-Smith and Baldwin,1988). For instance, the average energy use per tonne of casein powder produced in New Zealand is 16.9 GJ/tonne but the optimal energy usage is only 9.9 GJ/tonne. Recently, Linnhoff March Limited claimed that 2.5% energy saving in Huntly Power Station is possible but unproven. In terms of money this means about NZ$ 80 million of total saving over the station operating life( ≈ 30 years). [From Introduction]Item Characterisation & process control of pumping systems in the dairy industry : a thesis presented in partial fulfilment of the requirements for the degree of Masters of Technology at Massey University(Massey University, 2000) Dorsey, Mark RichardThe interaction between control of pumping systems in the dairy industry and the performance of the process has been investigated. Pumping in a precooling system at Massey University No.1 Dairy Unit was chosen as a case study. The requirements of the precooling system were determined from previous work done by dairy technologists. Part of these requirements were that: i) microbial damage to milk must be minimised by good temperature control, specifically by cooling the milk down to 18°C immediately after milking as specified in the New Zealand Dairy Industry Farm Dairy Code of Practice (COP). ii) handling should be gentle to minimise damage to the milk fat globule membrane by avoidance of cavitation and foaming. However controlled pumping, which minimises damage to the fat globule membrane, has been reported to decrease the cooling capacity of the plate heat exchanger (PHE). The precooling system at No.1 Dairy Unit was modified to allow continuous monitoring of key process variables (temperatures, flows and pressures). These were logged continuously and automatically to allow analyses to be carried out for whole milking sessions. The analysis shows that the releaser pump in the precooling system at No.1 Dairy Unit was oversized. This resulted in the pump only operating for 10 to 50% of the time and consequent inefficient usage of cooling water. In general the average temperature of the milk entering the vat complied with the COP requirement. However, as a consequence of the pump control system, the instantaneous temperature at times exceeded the COP recommended temperature. The analysis showed that cooling of the milk held up in the PHE during the pump-off phase contributed significantly to the cooling performance of the system. The present set up of the releaser pump pumping regime is based on a fixed pump-on phase of 6 seconds. The pump starts when the milk level reached a predetermined level in the milk receiver tank, which holds the milk coming from the cows. The duration of the pump on phase was set so that there would always be a milk fluid head in the receiver tank; which was decided by the relative size of the pump and receiver tank. The present pumping regime did not make best use of the ability of the system to cool the milk held up in the PHE during the pump-off phase. By simply changing the pump-on phase to 3s, more milk could be held up during the pump off phase in the PHE, giving a 10% increase in efficiency in the use of the cooling capacity of the water. This was achieved without changing the size of the PHE or any additional capital investment. Synchronising the water with milk flow rate resulted in further gains in efficiency of cooling water usage but this resulted in an increase in the temperature of milk exiting the PHE. This conflict of goals made evident that an improvement in efficiency could only be attained by using cooler water, which could be achieved by additional equipment such as a cooling tower. However it is recommended that any modification to the process must be accompanied by a reanalysis of the performance of the system in conjunction with an appropriate control system to optimise the performance of the system.Item Development of a headrig process information system : a thesis presented in partial fulfilment of the requirements for the degree of Masters of Technology in Computer Systems Engineering at Massey University(Massey University, 1999) Bayne, PeterA computer-based process information system was developed to gather operational information about the headrig handsaw at the Timber Technology Centre (TiTC) sawmill in the Waiariki Institute of Technology, store the data in a database, and display the information in various forms to the user. The project was the first part of an encompassing programme to instrument an entire commercial sawmill. This research programme aims to determine which variables are crucial to quantifying the sawing processes and to investigate the best techniques for measuring the variables. The system developed was extremely modular. Both analysis modules and sensor hardware can be added or removed without any need for restarting the system. A client-server architecture using networking communications was used to facilitate this. A central server gathers and stores the data, and individual clients analyse the data and display the information to the user. This enables analysis modules to be added and removed without even restarting the system. An experiment to determine the effect of wood density on the variables measured was used to test the viability of the completed system. The system successfully gathered all of the information required for the experiment and performed 70% of the data collation and analysis automatically. The remainder was performed using spreadsheets as this was deemed to be the most suitable method. The loosely coupled design of the system allows it to be up-scaled to a mill-wide program easily. Experiments performed to gather information about pivotal process variables are currently being planned, and should be underway as the expansion into other machine stations is being designed.Item Development of a real-time simulator for a three-effect falling film evaporator : a thesis presented in partial fulfilment of the requirements for the degree of Master of Technology at Institute of Technology and Engineering, Massey University(Massey University, 2003) Guo, Tong1.1 Background 1.1.1 A Falling-film evaporator Evaporation is an important process used in the chemical industry, which concentrates a solution containing dissolved or suspended solids by boiling off the solvent. A large proportion of the energy used in industry is given to this process. It is a very appropriate process to study due to its wide use and energy intensity [3]. In a modern evaporation system the evaporator acts as the main process unit in the evaporation process for drying a number of products. Good evaporator control is particularly important because of its widespread use, especially in the New Zealand dairy industry, and the direct effect of better control on energy efficiency and a more consistent product.Item The control of multivariable time-delayed processes and a generalized Smith predictor : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Production Technology at Massey University(Massey University, 1994) Crawford, Robert AndrewIn this thesis the description, analysis and control of time-delayed multivariable processes are investigated, particularly the descriptions of multivariable processes that facilitate a multivariable extension of the Smith predictor. Two new pseudo-commutativity results for matrix multiplication are presented. These results are used to show that a general time-delayed transfer function can be decomposed into three components representing input-delays, output-delays and the delay-free dynamics of the process. It is also shown that any such time-delayed transfer function can also be written in a form in which all the delays appear as output-delays. These time-delayed transfer functions are used in the development of a multivariable Smith predictor. It is also shown that the pseudo-commutativity results can be applied to non-delayed processes. In particular a new method, based on these results, for reformulating a transfer function description of a process as a state-space description is developed. A case study of a time-delayed process is investigated.Item Modelling, optimisation and control of a falling-film evaporator : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Production Technology at Institute of Technology and Engineering, Massey University, Palmerston North, New Zealand(Massey University, 2004) Paramalingam, ShabesheFalling-film evaporators in the dairy industry are key process units where most of the water is removed in the production of milk and whey powders. Evaporators of two to eight effects are common in the dairy industry. Mechanical vapour recompression is widely used to increase the energy efficiency of the evaporation process. A Thermal vapour recompression is used to control the final total solids concentration exiting the evaporator at Fonterra-Ingredients, Whareroa. Previous research into the performance of dairy evaporators has focused largely on milk evaporators (Choudhary, 1996; Runyon et al., 1991; Quaak & Gerritsen, 1990; Quaak et al, 1994; Winchester, 2000). The performance of the whey evaporator at whey products was completely different (see Chapter-9) to the performance with milk products (Winchester, 2000). This could be due to the difference in the physical and chemical properties (discussed in Chpter-3).The aim of the current study was to apply mathematical modelling to the optimisation and control of a two-effect thermal vapour recompression evaporator for whey products at Fonterra Ingredients-Whareroa, Fonterra Co-operative Ltd. The purpose of this study is to solve the problems (low throughput, fouling due to film break-up, increased energy consumption and unknown running conditions for each product) currently experienced with the evaporator at whey products. Figure 1.0 illustrates a general whey powder process at Whareroa (see section 2.2). The manufacture of protein ingredients from cheese and casein whey has evolved during the last fifty years into an established part of the world dairy industry (Bylund, 1995). Raw milk is variable in its composition (see Appendix A-3) and most dairy products can be produced in a variety of ways from this milk. Therefore, it is not surprising to find significant variations in reported values for the physical properties of dairy products (Bloore et al., 1981; Snoeren, 1982; Murakami and Okos, 1989; Fernandez-Martin, 1971; Jeurnink and Kruif, 1993, Antonio, 1983; Adam et al., 1994; Middleton, 1996). A major problem in the processing of whey protein ingredients is variation in final product properties (foamability and solubility index) due to factors relating to protein denaturation during processing and to the high variability of raw material composition. These interrelated factors include the source of raw material (whey), cheese manufacturing practices, heat treatment history, protein fractionation procedures and storage conditions (Bloore et al., 1981). The processing of whey powder, types of whey products and their properties are discussed in Chapter 2.