Papers by Keyword: Batch Processing

Abstract: This paper describes an overall, simulation-based optimization approach to control plant operations for manufacturing calcium silicate masonry units (CS), which is directed towards and thus immediately applicable to practical processes. Starting from an investigation and classification of the CS production in order to differentiate the properties of each sub-process, specific target criteria are derived. To enable the influencing of these targets, relevant parameters including their mutual interdependencies are identified. On this basis, the criticality of each process step is assessed in order to determine improvement potentials and to investigate possible adjustments to the parameters.The elementary production types indicate a mix of the discontinuous and continuous processing in CS plants. Particularly, this work shows that through interrupting the continuous material flow, the hardening process is the main criteria for a plant’s success in meeting its targets, especially concerning energy efficiency. To achieve a feasible approach, the work develops a solving method geared to an optimized hardening process.Therefore, a formulation of a measureable target system is established, which is the prerequisite for modeling the whole optimization problem. An expedient decomposition of this optimization model to smaller sub-problems provides an efficient solving of these complex job-scheduling problems, in order to direct the method towards an operative use. The paper concludes with the determination of potential solving procedures for the overall problem and appropriate algorithms for solving the sub-problems.

Abstract: Plasma immersion ion implantation (PIII) method is often used to cleaning and enhancing mechanical properties of the surface of materials. In this work, the AISI 304 was treated in a PIII system to improve tribological and wear resistance properties. The new HV pulser was prepared to reach high average power (10 kW) using solid-state technology and a pulse transformer rather than using a conventional one based on hard-tube tetrodes with HV storage capacitors. For preliminary tests, low-density nitrogen plasma and pulses of 10 kV, 30μs width, and 1 kHz were used. A larger vacuum chamber used (600 liters) is very important for treating large area components and for batch processing. This is necessary in industrial applications and in cases that require high quality processing as in spatial or medical components. Stainless steel support was used to hold the samples in our case. XRD, SEM, and pin-on-disk surface diagnostics were used for investigation and characterization of the treated surfaces.

Abstract: Most commercial banks have put their business and data processing into centralized IT&Data centers. Fast business developments require IT centers to handle increasing number and volumes of batch processing; new challenges are how to make full use of available IT resources and have flexible configurations. Cloud Computation is changing the manner how information system architecture is to be redesigned to meet higher service level and IT cost constraints. The core features of IT resource and application virtualization in cloud computation could increase IT center performance and optimize resource allocation algorithms for batch processing in banking, e_commerce and other tense and large-scaled data processing industries. A batch processing optimization scheme is presented which consists in creating a dynamic model by dividing business process into parallel and independent tasks to whom are allocated adequate IT resources. An experimental environment based on Hadoop/MapReduce framework is set up to simulate the performance of the proposed scheme, the simulation results show inspiring advantages which encourage further research and application work.

Abstract: Aerial hyperspectral data have the characteristic of big size and complicated process, but we often need to extract information from aerial hyperspectral data quickly in practical work. To solve the contradiction between huge hyperspectral data and the demand of extracting information quickly, this paper researches the batch processing which based on traditional technique flow of alteration information extraction. The experiment result indicates this batch processing is an efficient technique to increase information extraction speed.

Abstract: A properties database of working medium is developed based on the IAWPS-IF97 mathematical model to satisfy the demand of the accurate properties of full scope simulation of coal-fired power plant. The database provides the continuous relationship of the specific enthalpy, the density and the specific heat capacity as the function of the pressure and temperature. On this basis, a graphical user interface is designed to ensure the database can be used as a separate software kit. The error caused by regression algorithm can be avoided in the properties database. The properties database is suit for full scope simulation of coal-fired power plants and can also be used in the nuclear power plant. The results show that a single-step execution of matrix calculation of the properties database costs no more than 1ms CPU time in the Pentium IV 1.6GHz computer, fast enough for most real-time applications. Compared with the Chinese Standard, the mean relative error of the properties is less than 0.1% in full operation scope.

Abstract: There is often more than one kind of operating systems in enterprises’ and organizations’ local area network. Therefore, exchanging visits under different operating systems would be very meaningful and important. Especially in the field of bioinformatics, in order to meet the requirements of researchers from various disciplines, such as mathematics, biology and computer science, we develop a new, intuitive and convenient system for exchanging resources between Windows and Linux.

Abstract: In semiconductor wafer manufacturing, furnaces are mainly used for diffusion and deposition operations. During operations, wherever a furnace becomes available, scheduling the next batch deals with decisions on a batch which operation to process next and how many wafer lots to group the batch. To resolve the mentioned scheduling problem of the furnace where existing multi-products and different due dates, a heuristic dynamic scheduling algorithm called WCRHA (weighted cost rate heuristic algorithm) is presented to minimize the objective value of the waiting cost per unit time based on due date constraints. Simulation results show that the proposed algorithm is valid and feasible. Compared with the previous dynamic scheduling algorithms, it is more efficient in ensuring delivery and reducing completion time.

Abstract: Scaling of CMOS structures through the deep sub-micron range and into the nano-scale range (< 100 nm) has posed a number of difficult problems for processing technology. One main technological approach has been to improve the uniformity and conformality of deposited layers. As the Atomic Layer Deposition (ALD) has already demonstrated that it can overcome many of the limitations of current film deposition techniques, it seems to be the solution for very conformal layers of high quality on severe topography. The ALD method has been developed already in the 1970’s by Tumo Suntola and co-workers [1-4]. However, it has been in the past a rather unused method in the semiconductor industry. This has recently changed. During the last couple of years, the large semiconductor companies have spent a lot of effort in the utilization of ALD, but until now a production worthy ALD tool with low ‘Cost-of-Ownership’ (CoO) numbers was not available. One reason for the late introduction of ALD is that the method is rather slow compared with the state of art methods like CVD, PECVD and PVD. Nevertheless, due to the outstanding properties of the ALD technique, the drawback of slow deposition rate may be balanced by the parallel processing of many wafers in semiconductor furnaces, as described here.