Fuzzy Digital Filter Algorithm

Abstract

In this paper, an experimental model of a robotic manipulator device is presented. The robot's movement is carried out by feeding power to the servo motors, which are attached to the robot's working mechanisms: the base, the left and right joints and the robotic manipulator's gripper. During the operation of the device, due to uneven surfaces in the contacts connection sockets, a contact bounce occurs, which decreases the positioning accuracy of the manipulator links. To solve this problem, a fuzzy digital filter was developed. The paper proposes a fuzzy digital filter algorithm, consisting of six stages. The first stage consists of calculating the degrees of truth of the input and output variables. A fragment of the program code is presented, which carries out the determination of the degrees of truth for the terms of the second input variable. The graphics of membership functions for input and output variables are presented. The second stage calculates the degrees of truth for the fuzzy rules preconditions. A fuzzy rules base is put together so that it carries out the fuzzy digital filter operation. The third stage consists of finding the degree of truth of the fuzzy rules conclusions. The fourth step is to calculate the exact precise value of the digital filter coefficient. In the fifth step the calculation of the output voltage transmitted to the servo motors is carried out. At the sixth step, the resulting output voltage is converted to a rotation angle value for the servo motors. The paper presents the experimental studies showing the efficacy of the fuzzy digital filter. The graphics of the robotic manipulator base rotation transition process and the graphics of the input variables dependence on the output variables are presented, based on soft and hard calculations. The main conclusions on the results obtained in the experimental studies were made.