Hybrid Control over the State of a 3D Printer

The tools for soft computational technologies are based on the fuzzy systems, models of fuzzy neural networks, genetic algorithms, etc., which have advantages and drawbacks. In this paper, these tools are considered as applied to a 3D printer. From the point of view of control theory, a 3D printer is a complex nonlinear object, the mathematical description of which is a priori known, with one input and several outputs. During the operation of a 3D printer, it is required to provide continuous monitoring of such parameters as temperature of filament heating, rotation speed of a dc motor for feeding the filament to the extruder, and linear replacement of the carriage of a 3D printer. The operation of circuits is regulated by the controller of state, which, by analyzing the deviations of parameters, synchronously logically continuously controls the temperature of filament heating, feeding the filament to the extruder, and linear replacement of the carriage. The target function for all the circuits takes a given value of the adjustable parameters. The hybrid (fuzzy-neural) control over a 3D printer is based on the designing of a state controller using the special fuzzifier with the use of asymmetric sigmoid functions and the formation of layers to carry out fuzzy implication. The conversion of fuzzy information into determined information is carried out in a converter (decoder) that controls the printer voltage within the given range.