Agent learning in autonomic manufacturing execution systems for enterprise networking

In enterprise networks, companies interact on a temporal basis through client–server relationships between order agents (clients) and resource agents (servers) acting as autonomic managers. In this work, the autonomic MES (@MES) proposed by Rolón and Martinez (2012) has been extended to allow selfish behavior and adaptive decision-making in distributed execution control and emergent scheduling. Agent learning in the @MES is addressed by rewarding order agents in order to continuously optimize their processing routes based on cost and reliability of alternative resource agents (servers). Service providers are rewarded so as to learn the quality level corresponding to each task which is used to define the processing time and cost for each client request. Two reinforcement learning algorithms have been implemented to simulate learning curves of client–server relationships in the @MES. Emerging behaviors obtained through generative simulation in a case study show that despite selfish behavior and policy adaptation in order and resource agents, the autonomic MES is able to reject significant disturbances and handle unplanned events successfully.     

Systematic ultimate bound computation for sampled-data systems with quantization

We present a novel systematic method to obtain componentwise ultimate bounds in perturbed sampled-data systems, especially when the perturbations arise due to quantization. The proposed method exploits the system geometry as well as the perturbation structure, and takes intersample behavior into account. The main features of the method are its systematic nature, whereby it can be readily computer coded, without requiring adjustment of parameters for its application, and its suitability for dealing with highly structured perturbation schemes, whereby the information on the perturbation structure is directly taken into account. The latter feature distinguishes the method from other approaches that require a bound on the norm of the perturbation and thus disregard information on the perturbation structure. We apply the method to a numerical example taken from the literature to illustrate its simplicity and potential.     

Kinetics of the xylitol crystallization in hydro-alcoholic solution

Nývlt method was used to determine the kinetic parameters of commercial xylitol in ethanol:water (50:50 %w/w) solution by batch cooling crystallization. The kinetic exponents (n, g and m) and the system kinetic constant (B_N) were determined. Model experiments were carried out in order to verify the combined effects of saturation temperatures (40, 50 and 60 °C) and cooling rates (0.10, 0.25 and 0.50 °C/min) on these parameters. The fitting between experimental and calculated crystal sizes has 11.30% mean deviation.     

Efficiency of chain codes to represent binary objects

We present a study of compression efficiency for binary objects or bi-level images for different chain-code schemes. Chain-code techniques are used for compression of bi-level images because they preserve information and allow a considerable data reduction. Furthermore, chain codes are the standard input format for numerous shape-analysis algorithms. In this work we apply chain codes to represent object with holes and we compare their compression efficiency for seven chain codes. We have also compared all these chain codes with the JBIG standard for bi-level images.     

International differences of productivity in scholarly management knowledge

Using a dataset of refereed conference papers, this work explores the determinants of academic production in the field of management. The estimation of a count data model shows that the countries’ level of economic development and their economy size have a positive and highly significant effect on scholarly management knowledge production. The linguistic variable (English as official language), which has been cited by the literature as an important factor facilitating the participation in the international scientific arena, has also a positive and statistically significant effect.     

Performance analysis of multiprocessor DSPs: a stream-oriented component model

This article introduces a composable component model that can be used for performance analysis of complex digital signal processing (DSP) systems consisting of several computation resources communicating via some shared media. It reviews different methods of analyzing the performance of such communication-centric systems, including simulation, trace-based simulation, and symbolic methods. Finally, a component model that can be used to formally state the relationship between the usage and the processing capabilities of different computation and communication resources is established.     

Reduction of NO emissions in a turbojet combustor by direct water/steam injection: Numerical and experimental assessment

Numerical and experimental investigations are conducted to assess the benefits and drawbacks of both water (mist) and steam direct injection within the combustion chamber of a 200 N static thrust turbojet. For this purpose, a three-dimensional CFD model of the combustion process is implemented where pollutant emissions are calculated; in parallel, a test campaign on the turbojet at sea level static conditions is carried out. In both cases the refrigerant flow is injected directly into the combustor, outside the liner. The aim of the investigations is to evaluate the impact of increasing water and steam flows (ranging from 0% to 200% of the fuel mass flow) onto the emissions levels (NO and CO) of the engine.