Integration of multiple sensor fusion in controller design

The main focus of this research is to reduce the risk of a catastrophic response of a feedback control system when some of the feedback data from the system sensors are not reliable, while maintaining a reasonable performance of the control system. In this paper a methodology for integrating multiple sensor fusion into the controller design is presented. The multiple sensor fusion algorithm produces, in addition to the estimate of the measurand, a parameter that measures the confidence in the estimated value. This confidence is integrated as a parameter into the controller to produce fast system response when the confidence in the estimate is high, and a slow response when the confidence in the estimate is low. Conditions for the stability of the system with the developed controller are discussed. This methodology is demonstrated on a cupola furnace model. The simulations illustrate the advantages of the new methodology.     

A production planning in highly automated manufacturing system considering multiple process plans with different energy requirements

This paper focuses on a production planning problem in a highly automated manufacturing system considering multiple process plans with different energy requirements. The system consists of several closely interconnected sub-systems such as the processing system, the material (part) handling system, the tool transport system and the auxiliary system responsible for a supply of cooling/lubricants and a waste disposal. We propose a methodology for an estimation of energy consumption and material flows that are incurred at a system level with respect to multiple process plans for a part type. In addition, this study focuses on a production planning problem with the objective to minimize the weighted sum of energy consumption, inventory holding cost and backorder cost on a FMS considering multiple process plans. The production planning model is developed as a linear programming model. The benefit coming from the adoption of suggested model has been addressed with reference to a real industrial use case study.     

Modelling and analysis of multiple cluster tools system with random failures using coloured Petri net

Semiconductor manufacturing is among the most complicated and expensive operation systems encountered today and multiple cluster tool system has become an important technology in semiconductor manufacturing systems with the advantages of higher yields and shorter cycle times. Modelling and simulation has become a standard methodology in order to understand and predict performance of semiconductor manufacturing systems. Resources random failures during process execution are one major concern regarding system performance measured in terms of throughput and cycle time. This study aims to models and analyse random failures of processing modules of multiple cluster tool systems using coloured Petri net method. The impact of different input factors on system throughput and cycle time is presented.     

Dynamic multi-response optimization using principal component analysis and multiple criteria evaluation of the grey relation model

The application of parameter design methodology has been considerable in recent years to make system performance robust over a wide range of input conditions. This notion has been referred to as a robust design with dynamic characteristics. Due to product complexity, multiple correlated characteristics must be simultaneously evaluated for improving product quality. Dynamic multi-response optimization is becoming an important issue to contemporary industry. This study developed a novel procedure of optimizing dynamic multi-responses using principal component analysis (PCA) and multiple criteria evaluation of the grey relation model. PCA can consider the correlations among multiple quality characteristics to obtain uncorrelated components. These components are then substituted into multiple criteria evaluation of the grey relation model to determine the optimal factor level combination. A case study demonstrates the effectiveness of the proposed procedure for optimizing dynamic multi-response processes.     

Quality chain design and optimization by multiple response surface methodology

To maintain optimal quality characteristics in the defined specification limits is a vital decision for any industry and service system. To avoid nonconformity in outputs, the stream of variations and their potential causes must be identified so that the response variables fall into desirable limits across the manufacturing or service chain. Response surface methodology is considered as a powerful technique to facilitate the analysis of the mentioned problem. This paper presents the general quality chain design problem as a mathematical program and also proposes a method to solve it using multiple response surface methodology. An example of multistage processes is analyzed by the proposed approach to show its efficacies numerically and analytically.     

A real-time intelligent multiple fault diagnostic system

Modern manufacturing systems and their failure modes are very complex, and efficient fault diagnosis is essential for higher productivity. However, traditional fault diagnostic systems that perform sequential fault diagnosis can fail during diagnosis when fault propagation is very fast. This paper describes a real-time intelligent multiple fault diagnostic system (RIMFDS). This system deals with multiple fault diagnosis, and is based on multiprocessing by using a strata hierarchical artificial neural network (SHANN). If another fault occurs while an existing symptom is being diagnosed, the corresponding diagnosis module is triggered, and the fault diagnosis module of the new faulty unit begins to diagnose the faults in real time. RIMFDS can diagnose multiple faults with fast fault propagation and complex physical phenomena. The system consists of two main parts. One is a personal computer for remote signal generation and transmission, and the other is a host system for multiple fault diagnosis. The signal generator generates various faulty signals and image information and sends them to the host. The host has various modules and agents for efficient multiple fault diagnosis. A SUN workstation is used as a host for multiple fault diagnosis and graphic representation of the results.     

A real-time intelligent multiple fault diagnostic system

Modern manufacturing systems and their failure modes are very complex, and efficient fault diagnosis is essential for higher productivity. However, traditional fault diagnostic systems that perform sequential fault diagnosis can fail during diagnosis when fault propagation is very fast. This paper describes a real-time intelligent multiple fault diagnostic system (RIMFDS). This system deals with multiple fault diagnosis, and is based on multiprocessing by using a strata hierarchical artificial neural network (SHANN). If another fault occurs while an existing symptom is being diagnosed, the corresponding diagnosis module is triggered, and the fault diagnosis module of the new faulty unit begins to diagnose the faults in real time. RIMFDS can diagnose multiple faults with fast fault propagation and complex physical phenomena. The system consists of two main parts. One is a personal computer for remote signal generation and transmission, and the other is a host system for multiple fault diagnosis. The signal generator generates various faulty signals and image information and sends them to the host. The host has various modules and agents for efficient multiple fault diagnosis. A SUN workstation is used as a host for multiple fault diagnosis and graphic representation of the results.     

Solving multiple response optimisation problems using adaptive neural networks

This paper describes an integrated methodology using experimental designs and neural networks technologies for solving multiple response problems. This new methodology consists of an experiment reference template for designing and collecting training data samples and a parallel distributed computational adaptive neural network system to provide a powerful tool for data modelling, guiding experimentation and empirical investigations. While the experiment reference template is for determining the measurements to adopt in order to extract maximum information within minimum experimental efforts, the adaptive neural network provides a nonlinear multivariate data-fitting algorithm for analysing the results of the experimental design and providing decision support. This integrated methodology is used to model and optimise a multiple response metal inert gas (MIG) welding process. The neural network is trained with optimum welding experimental data, tested and compared in an actual welding environment in terms of weld quality. The relevant data is established using experimental design methods and is highlighted in the case study. The implementation for this case study was carried out using a semi-automatic welding facility, to mass weld a 20 in.×0.438 in. pin/box onto a 20 in.×0.5 in.×37 ft pipe (tubular drilling products), in an actual workshop which makes oilfield equipment. The entire range of welding combination that the process might be subject to during actual welding operations is included to study the weld quality.     

Stability analysis of multiple time-delayed system

A general class of linear time-invariant systems with time delays is studied. A number of methodologies have been suggested to assess the stability in the parametric domain of time delay or coefficient. This study offers an exact, structured and robust methodology to determine the stability regions of uncertain parameters in both time-delay space and coefficient space. The Rekasius transformation is used as a connection between time-delay space and coefficient space. An explicit analytical expression in terms of the system parameters which reveals the stability regions(pockets) in the domain of time delay and coefficient is presented. The method starts with the determination of all possible values of uncertain parameters which result in purely imaginary characteristic roots. In addition, some special stability boundaries are also discussed. After generating stability boundaries in parametric space, the two-step determination procedure is proposed to determine the actual stability regions. Such an approach can be used to determine the stability regions of any uncertain parameters of any retarded time-delay system. A complete example case study is also provided.     

Optimal load allocation of multiple fuel boilers

This paper presents a new methodology for optimally allocating a set of multiple industrial boilers that each simultaneously consumes multiple fuel types. Unlike recent similar approaches in the utility industry that use soft computing techniques, this approach is based on a second-order gradient search method that is easy to implement without any specialized optimization software. The algorithm converges rapidly and the application yields significant savings benefits, up to 3% of the overall operating cost of industrial boiler systems in the examples given and potentially higher in other cases, depending on the plant circumstances. Given today’s energy prices, this can yield significant savings benefits to manufacturers that raise steam for plant operations.     

Development of an experiment-based robust design paradigm for multiple quality characteristics using physical programming

The well-known quality improvement methodology, robust design, is a powerful and cost-effective technique for building quality into the design of products and processes. Although several approaches to robust design have been proposed in the literature, little attention has been given to the development of a flexible robust design model. Specifically, flexibility is needed in order to consider multiple quality characteristics simultaneously, just as customers do when judging products, and to capture design preferences with a reasonable degree of accuracy. Physical programming, a relatively new optimization technique, is an effective tool that can be used to transform design preferences into specific weighted objectives. In this paper, we extend the basic concept of physical programming to robust design by establishing the links of experimental design and response surface methodology to address designers’ preferences in a multiresponse robust design paradigm. A numerical example is used to show the proposed procedure and the results obtained are validated through a sensitivity study.     

Robust design of multiple dynamic quality characteristics

The Taguchi method is a useful technique to improve the performance of products or processes at a lower cost and in less time. The traditional Taguchi method focused on a single characteristic to optimize the parameter conditions. In practice, most products and processes have more than one quality characteristic. Several approaches dealing with multiple static quality characteristic problems have been reported. However, little attention has been focused on optimizing the multiple dynamic quality characteristics. This research presents an approach to optimizing multiple dynamic problems based on quality loss. The objective is to minimize the total average quality loss for the multiple dynamic quality characteristics experiment.     

Selecting optimal set of tool sequences for machining of multiple pockets

In computer-aided manufacturing domain, the selection of cutters significantly affects the machining time. However, it is difficult for humans to determine the optimal or near-optimal set of cutting tools, in particular for complex parts with multiple features. Traditionally, every feature is machined one-by-one, and no cutter sharing is taken into consideration. To perform machining operations efficiently, both the number and size of cutters should be selected carefully. Therefore, this paper proposes a new approach to determine optimal cutting tool sequences for machining multiple features in a single setup, which has been implemented and integrated with a feature-based process planning system for the generation of a numeric control code. To seek a general approach to compute the coverable area of each cutter for machining arbitrary region shapes in the case of multiple cutters combination, a new algorithm is introduced by improving the scan line-based polygon filling algorithm. To do so, the features with the nonpolygon profiles have to be approximated as polygons according to the requirement of the tolerance. The proposed method has been verified by experiments.     

Simultaneous optimization of multiple performance characteristics of carbonitrided pellets: a case study

The performance of a product is generally characterized by more than one response variable. Hence, the management often faces the problem of simultaneous optimization of many response variables. In recent years, a lot of literature has been published on various methodologies for tackling the multi-response optimization problems. Among them, the approach based on Taguchi’s quality loss function is very popular. This paper discusses a case study on multiple response optimization in carbonitriding process. The surface hardness, case depth, and dimensional variation of carbonitrided pellets were simultaneously optimized using quality loss function methodology. The optimum obtained through loss function approach was found to be superior to the ones obtained through optimizing the response variables separately. The result obtained through the implementation of the solution is also presented in the study.     

Real-time monitoring of complex sensor data using wavelet-based multiresolution analysis

Advancements made in sensor technology have resulted in complex sensor data that captures multiple process events. Real-time monitoring of complex manufacturing processes, such as nano-scale semiconductor polishing, often requires analysis of such complex sensor data. The multiple events in a process occur at multiple scales or frequencies (also referred to as multiscale) and are localized at different points in time. Recent literature contains several wavelet decomposition based multiscale sensor data analysis techniques including those that are developed for process monitoring applications, such as tool-life monitoring, bearing defect monitoring, and monitoring of ultra-precision processes. However, most of the above mentioned wavelet-based sensor data analysis techniques are designed for offline implementation. In an offline method, one can perform wavelet decomposition of longer data lengths in order to capture information needed for monitoring. However, this is computationally involved and needs longer processing time, which becomes a serious challenge in online (real time) applications. This paper first presents a complete online multiscale process monitoring methodology. The methodology is designed to deal with real-time analysis and testing of very high rate of process data collected by sensors. This is particularly critical and becomes a challenge for high rate of data collection by the sensors which pose additional difficulty of matching data processing rate with the data acquisition rate. The methodology is capable of displaying the analysis results through real time graphs for ease of process supervisory decision making. The methodology is demonstrated via a nano-scale silicon wafer polishing application. Sufficient details of the application are provided to assist readers in implementing this methodology for other processes. The results show that the methodology has the ability to deal with high rate of data collection as well as multiscale event detection.     

A parallel multiple Markov chain simulated annealing for multi-period manufacturing cell formation problems

Simulated annealing (SA) is a general purpose optimization technique capable of finding optimal or near optimal solutions in various applications. The major disadvantage of this technique is its slow convergence making it not suitable for solving many complex optimization problems. This limitation may be alleviated by parallel computing using a multiprocessor computer or a cluster of workstations. In this paper, we present an integer programming model for solving a multi-period cell formation problem in cellular manufacturing system. In order to solve the mathematical model efficiently, we developed a multiple Markov chain simulated annealing algorithm which allows multiple search directions to be traced simultaneously. Our computational results on a single processor machine showed that multiple Markov chain SA is much more efficient than a conventional single Markov chain SA. The parallel implementation of the multiple Markov chain SA further improves its computational efficiency in terms of solution quality and execution time.     

A novel hybrid method for supply chain optimization with capacity constraint and shipping option

In a modern market, supply chain network design is considered as a strategic decision that provides the proper platform for cost management and increases the competitive edge of enterprise. In a five-tier supply chain, there are several facilities such as suppliers, manufacturers, warehouses, distribution centers, and retailers or customers. For product transportation from one facility to another, different types of options may be used. These options have variety price and the manager should do the best assignment to reduce the total cost. In this paper, we formulated an integer programming model for a five-tier supply chain with capacitated facility and multiple transportation option with fixed lead time. We also proposed a novel meta-heuristic solution methodology that combines the Taguchi's feature with artificial immune approach in order to solve the proposed model. The performance of the proposed solution methodology has been examined against a set of numeric instances and the obtained results are compared with those provided by hybrid genetic algorithm and Taguchi and artificial immune system. Results indicate that this methodology can make better results than previous solutions effectively.     

Image processed tracking system of multiple moving objects based on Kalman filter

This paper presents a development result for image processed tracking system of multiple moving objects based on Kalman filter and a simple window tracking method. The proposed algorithm of foreground detection and background adaptation (FDBA) is composed of three modules: a block checking module (BCM), an object movement prediction module (OMPM), and an adaptive background estimation module (ABEM). The BCM is processed for checking the existence of objects. To speed up the image processing time and to precisely track multiple objects under the object’s mergence, a concept of a simple window tracking method is adopted in the OMPM. The ABEM separates the foreground from the background in the reset simple tracking window in the OMPM. It is shown through experimental results that the proposed FDBA algorithm is robustly adaptable to the background variation in a short processing time. Furthermore, it is shown that the proposed method can solve the problems of mergence, cross and split that are brought up in the case of tracking multiple moving objects.     

Assurance of system service level robustness in complex supply chain networks

Despite studies that consider service level rates of individual factories in a supply chain system, the interactions between connected factories and the impact of these interactions on the overall supply chain service level rate have been rarely studied. Moreover, due to uncertainties induced by various sources such as transportation delay and manufacturing processes variability, ensuring the robustness of system service level rate while considering these uncertainties in individual factories is a highly complex task. This paper studies uncertainty effect introduced by factory service level rates on the robustness of overall supply chain network performance, and presents a novel robust design optimization methodology to derive designs of factory service level rates in order to satisfy the service level rate requirement of the system and ensure its robustness. A case study with a multi-level multiple factories supply chain network is used to demonstrate the efficacy of the proposed methodology.     

A basic theoretical study of the temperature rise in sliding contact with multiple contacts

The objectives of this paper are to develop a theoretical solution for the temperature rise due to sliding contact between surfaces with multiple, interacting asperities and to use this solution to examine the effects of the important contact area and system parameters. A solution based on the Green's function method is developed for the basic problem of two half-space regions in sliding contact with any arbitrarily specified arrangement of rectangular asperities.Studies are conducted to demonstrate the effects of the contact area parameters, namely the number, size, spacing and orientation of the contacts, as well as sliding velocity. Results indicate that the contact temperatures are extremely sensitive to the number and relative spacing between contacts, where subdivision of a single contact into separated pieces significantly reduces the contact temperature rises. The orientation of the contacts relative to the sliding direction is shown to have only a small influence on temperature. The shape of the contacts also has only a small influence, except in the case of contact patches with large aspect ratios where significantly lower surface temperatures can occur. Sliding speed is shown to be extremely important in that increased speed causes both higher temperature levels and greater interaction between contacts due to the convective effect.The current paper is intended to describe the basic solution methodology for calculating temperature rises due to multiple, interacting contacts and to show some fundamental trends for a selected set of regularly arranged contact area distributions.