We apply the recently improved version of the 0-1 test for chaos to real experimental time series of laser droplet generation process. In particular two marginal regimes of dripping are considered: spontaneous and forced dripping. The outcomes of the test reveal that both spontaneous and forced dripping time series can be characterized as chaotic, which coincides with the previous analysis based on nonlinear time series analysis.
COBISS.SI-ID: 11930139
Organizing and optimizing production in small and medium enterprises (SMEs) with small batch production and many different products can be very difficult. This paper presents an approach to organize the production cells by means of clustering manufactured products into groups with similar product properties. Several clustering methods are compared, including the hierarchical clustering, k-means and self-organizing map (SOM) clustering. Clustering methods are applied to production data describing 252 products from a Slovenian company KGL. The best clustering result, evaluated by an average silhouette width for a total data set, is obtained by SOM clustering. In order to make clustering results applicable to the industrial production cell planning, an interpretation method is proposed. The method is based on percentile margins that reflect the requirements of each production cell and is further improved by incorporating the economic values of each product and consequently the economic impact of each production cell.
COBISS.SI-ID: 12359195
In laser generation of droplets from a metal wire, it has been observed experimentally that droplet detachment can be influenced by interaction of the laser pulse frequency and eigen dynamics of a pendent droplet. In laser pulse frequency range between 70 Hz and 150 Hz, the dynamics of the pendent droplet resembles that of a spring-mass system. Based on experimental observations a nonlinear model of droplet generation in the form of spring-mass oscillator is developed. Model based simulation of pendent droplet dynamics, and especially predicted time of droplet detachment and detached droplet diameter are in high agreement with experimental results.
COBISS.SI-ID: 12574491
Clustering-ensemble methods have emerged recently as an effective approach to the problem of clustering, which is one of the fundamental data-analysis tools. Data clustering with an ensemble involves two steps: generation of the ensemble with single-clustering methods and the combination of the obtained solutions to produce a final consensus partition of the data. In this paper we first propose a novel clustering method, based on Kohonen's self-organising map and gravitational algorithm, and, second, investigate its performance in the generation of a clustering ensemble. The proposed method is able to discover clusters of complex shapes and determines the number of clusters automatically. Furthermore, its stochastic nature is beneficial in the construction of a diverse ensemble of partitions. Promising results of the presented method were obtained in comparison with three, relevant, single-clustering algorithms over artificial and real data sets.
COBISS.SI-ID: 9315924
In recent years there has been a growing interest in hardware neural networks, which express many benefits over conventional software models, mainly in applications where speed, cost, reliability, or energy efficiency are of great importance. These hardware neural networks require many resource-, power- and time-consuming multiplication operations, thus special care must be taken during their design. Since the neural network processing can be performed in parallel, there is usually a requirement for designs with as many concurrent multiplication circuits as possible. One option to achieve this goal is to replace the complex exact multiplying circuits with simpler, approximate ones. The present work demonstrates the application of approximate multiplying circuits in the design of a feed-forward neural network model with on-chip learning ability. The experiments performed on a heterogeneous Proben1 benchmark dataset show that the adaptive nature of the neural network model successfully compensates for the calculation errors of the approximate multiplying circuits. At the same time, the proposed designs also profit from more computing power and increased energy efficiency.
COBISS.SI-ID: 9160276