R.A. Kosiński

Vision safety system based on cellular neural networks

This work is dedicated to develop a safety measurement for human–machine cooperative system, in which the machine region and the human region cannot be separated due to the overlap and the movement both from human and from machines. Modern production processes become more and more flexible. Therefore, there is a need that devices used in workplace also support flexibility as much as possible. Such characteristics have vision-based protective devices. We present a neural system for the advanced recognition of danger situation for safety control. The sequence of the images from two cameras located above the robot is presented to the system of cellular neural networks (CNNs) realized in the PC computer. They detect a new object appearing in a safety field, define its position with respect to the robot arm and perform the feature extraction of its image. Experiments conducted using artificial images (virtual environment) and low-quality images (internet cameras) indicate that our system can work in a real time and detect successively dangerous situations. We have found that the CNN is unable to detect a new object properly in the presence of high level of noise as a result of percolation type phase transition. An example of possible application of the system is presented.

Nonlinear dynamics and route to chaos for twisted domain walls

Abstract The nonlinear dynamics and the route to chaos are analyzed for twisted domain walls in a uniaxial bubble garnet films. Using the Hubert expression for the magnetic stray field coming from poles at the surfaces of the film, the route to chaos via intermittency is obtained.

Stereovision Safety System for Identifying Workers’ Presence: Results of Tests

This article presents the results of extensive tests of a stereovision safety system performed using real and artificial images. A vision based protective device (VBPD) analyses images from 2 cameras to calculate the position of a worker and moving parts of a machine (e.g., an industrial robot’s arm). Experiments show that the stereovision safety system works properly in real time even when subjected to rapid changes in illumination level. Experiments performed with a functional model of an industrial robot indicate that this safety system can be used to detect dangerous situations at workstations equipped with a robot, in human–robot co-operation. Computer-generated artificial images of a workplace simplify and accelerate testing procedures, and make it possible to compare the effectiveness of VBPDs and other protective devices at no additional cost.

Spatio-temporal diagrams for domain wall dynamics

Abstract Spatio-temporal diagrams are applied to the analysis of the motion of a domain wall in a thin magnetic film using numerical solutions of the equations of motion. The pattern entropy, calculated for these diagrams, allows one to discern periodic, quasiperiodic and chaotic wall motion.

Chaotic motion of a domain wall in the time dependent drive fields

Abstract Motion of a domain wall in a time dependent drive field was investigated using numerical solutions of equations of motion. Periodic, quasiperiodic and chaotic motions of the domain wall were found for various values of the amplitude of the drive field.

Chaotic motion of a domain wall with vertical Bloch lines

Abstract The motion of domain wall containing a pair of vertical Bloch lines under the influence of the periodical in time drive field may be chaotic provided the damping of the film is sufficiently small.

Intelligent Agents System for Evacuation Processes Modeling

In the paper an intelligent agents system is used for modeling evacuating processes of rooms and buildings. The motion of each agent is described using Langevin equation with the social force term, which make possible to take into account mental component in agents motion - connected with the level of hazard. In numerical simulations it is possible to observe the trajectories of agents and emergent phenomena accompanying evacuation. Calculation of times of evacuation for different rooms and buildings enable to discuss factors influencing the effectiveness of evacuation.

Stochastic resonance in spatiotemporal on–off intermittency

Abstract Stochastic resonance is investigated in a generic system with spatiotemporal on–off intermittency: a chain of coupled logistic maps with a time-dependent control parameter, driven by a spatiotemporal periodic signal. Spatiotemporal correlation function between the periodic signal and the output signal, reflecting the occurrence of laminar phases and chaotic bursts, has a maximum as a function of the mean value of the control parameter. For a given period and length of the periodic signal the height of this maximum can be increased by choosing an optimum coupling strength between maps. It is argued that the obtained result can be interpreted as an example of noise-free (dynamical) stochastic resonance in a system with spatiotemporal intermittency.

Dynamics of a Square Neural Network with Uniform and Nonuniform Excitations

Abstract Dynamics of a square neural network with some type of dilution of synaptic connections, as a response for periodic excitation coming from one excitatory neuron (uniform excitation) or two excitatory neurons (nonuniform excitation), is examined numerically. For both types of excitations three types of the time evolution of the network: periodic, quasiperiodic and chaotic, are found as a function of nonlinearity of a single neuron (g) and number of neighbours of each neuron (K). The route to chaos and the radius of influence of the excitatory neurons on a dynamics of neighbouring neurons is discussed.

On the effect of damping and in-plane magnetic field on the generation of vertical bloch lines in the domain wall

Abstract The relation between the critical value of the drive field, above which the generation of a vertical Bloch lines in a domain wall may occur, and the values of the external in-plane field as well as the damping parameter is found numerically for thin garnet film.