Maritime Autonomous Surface Ships (MASS) perfectly fit into the future vision of merchant fleet. MASS autonomous navigation system combines automatic trajectory tracking and supervisor safe trajectory generation subsystems. Automatic trajectory tracking method, using line-of-sight (LOS) reference course generation algorithm, is combined with model predictive control (MPC). Algorithm for MASS trajectory tracking, including cooperation with the dynamic system of safe trajectory generation is described. It allows for better ship control with steady state cross-track error limitation to the ship hull breadth and limited overshoot after turns. In real MASS ships path is defined as set of straight line segments, so transition between trajectory sections when passing waypoint is unavoidable. In the proposed control algorithm LOS trajectory reference course is mapped to the rotational speed reference value, which is dynamically constrained in MPC controller due to dynamically changing reference trajectory in real MASS system. Also maneuver path advance dependent on the path tangential angle difference, to ensure trajectory tracking for turns from 0 to 90 degrees, without overshoot is used. All results were obtained with the use of training ship in real–time conditions.

This study investigated the relationship between the parameters of the DLP manufacturing process and the structure of photopolymerizable acrylic resins. Four different process parameters were established to produce different thin-walled acrylic sample series: exposure time, layer thickness, area offset, and number of transition layers. The structure and the surface of the obtained samples were examined with the use of the FTIR–ATR method and an optical microscope, respectively. It was proved that extension of the exposure time increases the density of crosslinking and sample thickness. A decreasing crosslinking density due to rising layer thickness is observed. The area offset affects only the dimensions of the sample, predictably reducing the dimensions of the sample as the compensation increases. The absence of transition layers proved unfavorable in many respects, both structurally and geometrically.

The paper presents analysis of the positivity for a two-dimensional temperature field. The process under consideration is described by the linear, infinite-dimensional, noninteger order state equation. It is derived from a two-dimensional parabolic equation with homogenous Neumann boundary conditions along all borders and homogenous initial condition. The form of control and observation operators is determined by the construction of a real system. The internal and external positivity of the model are associated to the localization of heater and measurement. It has been proven that the internal positivity of the considered system can be achieved by the proper selection of attachment of a heater and place of a measurement as well as the dimension of the finite-dimensional approximation of the considered model. Conditions of the internal positivity associated with construction of real experimental system are proposed. The postivity is analysed separately for control and output of the system. This allows one to analyse the positivity of thermal systems without explicit control. Theoretical considerations are numerically verified with the use of experimental data. The proposed results can be applied i.e. to point suitable places for measuring of a temperature using a thermal imaging camera.

Illegal elements use the characteristics of an anonymous network hidden service mechanism to build a dark network and conduct various illegal activities, which brings a serious challenge to network security. The existing anonymous traffic classification methods suffer from cumbersome feature selection and difficult feature information extraction, resulting in low accuracy of classification. To solve this problem, a classification method based on three-dimensional Markov images and output self-attention convolutional neural network is proposed. This method first divides and cleans anonymous traffic data packets according to sessions, then converts the cleaned traffic data into three-dimensional Markov images according to the transition probability matrix of bytes, and finally inputs the images to the output self-attention convolution neural network to train the model and perform classification. The experimental results show that the classification accuracy and F1-score of the proposed method for Tor, I2P, Freenet, and ZeroNet can exceed 98.5%, and the average classification accuracy and F1-score for 8 kinds of user behaviors of each type of anonymous traffic can reach 93.7%. The proposed method significantly improves the classification effect of anonymous traffic compared with the existing methods.