A New Methodology for a Retrofitted Self-tuned Controller with Open-Source FPGA.

Servo systems are feedback control systems characterized by position, speed, and/or acceleration outputs. Nowadays, industrial advances make the electronic stages in these systems obsolete compared to the mechanical elements, which generates a recurring problem in technological, commercial and industrial applications. This article presents a methodology for the development of an open-architecture controller that is based on reconfigurable hardware under the open source concept for servo applications. The most outstanding contribution of this paper is the implementation of a Genetic Algorithm for online self tuning with a focus on both high-quality servo control and reduction of vibrations during the positioning of a linear motion system. The proposed techniques have been validated on a real platform and form a novel, effective approach as compared to the conventional tuning methods that employ empirical or analytical solutions and cannot improve their parameter set. The controller was elaborated from the Graphical User Interface to the logical implementation while using free tools. This approach also allows for modification and updates to be made easily, thereby reducing the susceptibility to obsolescence. A comparison of the logical implementation with the manufacturer software was also conducted in order to test the performance of free tools in FPGAs. The Graphical User Interface developed in Python presents features, such as speed profiling, controller auto-tuning, measurement of main parameters, and monitoring of servo system vibrations.

A PID-Type Fuzzy Logic Controller-Based Approach for Motion Control Applications.

Motion control is widely used in industrial applications since machinery, robots, conveyor bands use smooth movements in order to reach a desired position decreasing the steady error and energy consumption. In this paper, a new Proportional-Integral-Derivative (PID) -type fuzzy logic controller (FLC) tuning strategy that is based on direct fuzzy relations is proposed in order to compute the PID constants. The motion control algorithm is composed by PID-type FLC and S-curve velocity profile, which is developed in C/C++ programming language; therefore, a license is not required to reproduce the code among embedded systems. The self-tuning controller is carried out online, it depends on error and change in error to adapt according to the system variations. The experimental results were obtained in a linear platform integrated by a direct current (DC) motor connected to an encoder to measure the position. The shaft of the motor is connected to an endless screw; a cart is placed on the screw to control its position. The rise time, overshoot, and settling time values measured in the experimentation are 0.124 s, 8.985% and 0.248 s, respectively. These results presented in part 6 demonstrate the performance of the controller, since the rise time and settling time are improved according to the state of the art. Besides, these parameters are compared with different control architectures reported in the literature. This comparison is made after applying a step input signal to the DC motor.

Fuzzy Logic and Genetic-Based Algorithm for a Servo Control System.

Performing control is necessary for processes where a variable needs to be regulated. Even though conventional techniques are widely preferred for their implementation, they present limitations in systems in which the parameters vary over time, which is why methods that use artificial intelligence algorithms have been developed to improve the results given by the controller. This work focuses on implementing a position controller based on fuzzy logic in a real platform that consists of the base of a 3D printer, the direct current motor that modifies the position in this base, the power stage and the acquisition card. The contribution of this work is the use of genetic algorithms to optimize the values of the membership functions in the fuzzification of the input variables to the controller. Four scenarios were analyzed, in which the trajectory and the weight of the system were modified. The results obtained in the experimentation show that the rising and setting times of the proposed controller are better than those obtained by similar techniques that were previously developed in the literature. It was also verified that the proposed technique reached the desired values even when the initial conditions in the system changed.