Narrow Results

In Internet of Things (IoT) environment, a number of sensors and devices exist and they have various and heterogeneous characteristics. Applications which provide a variety of services based on the sensors also have different service requirements. Therefore, a middleware that is located between sensors and application systems is needed for integrating two layers. This paper proposes a general purpose middleware for providing intelligent services based on heterogeneous sensors existing in IoT environment. The proposed middleware acquires and manages sensing data in real time. The middleware stores and manages heterogeneous sensors, node, and network metadata. Especially, the middleware in this paper includes a component for providing intelligent services based on inferencing the situation based on ontologies and rules.

Considering fuzziness, randomness, and the association between them, cloud model-based control is a new way to address uncertainty in the inference system. Similar to fuzzy control theory, this method includes an important step of dealing with the logic concept “and”, which is defined as the operation of soft-and between several antecedents and has not been scientifically solved in the current literatures. The traditional method of realizing soft-and is to use multi-dimensional cloud model theory, which lacks a theoretical basis. Based on the fuzzy and random theory, this paper proposes a novel approach using numeric simulation to calculate the soft-and in the cloud control system. In this method, the theory to determine the distribution of the minimum value between two random variables is applied. Compared with the traditional method, the considered approach is more reliable and reasonable, and its result is also in accordance with the standard fuzzy inference system.

A bottleneck of laboratory analysis in process industries including steelmaking plants is the low sampling rate. Inference models using only variables measured online have then been used to made such information available in advance. This study develops predictive models for key mechanical properties of seamless steel tubes, by strength, ultimate tensile strength and hardness. A plant in Brazil was used as the case study. The sample sizes of some steel tube families given namely, yield a particular property are discrepant and sometimes very small. To overcome this sample imbalance and lack of representativeness, committees of predictive neural network models based on bagging predictors, a type of ensemble method, were adopted. As a result, all steel families for all properties have been satisfactorily described showing the correlations between targets and model estimates close to 99%. These results were compared to multiple linear regression, support vector machine and a simpler neural network. Such information available in advance favors corrective actions before complete tube production mitigating rework costs in general.