OPTIMIZATION OF THE ROBOT SPACE TRAJECTORY BY USING KINEMATICS, DYNAMICS, INTELLIGENT DAMPER AND PROPER NEURAL NETWORK
Journal Title: Proceedings in Manufacturing Systems - Year 2012, Vol 7, Issue 2
Abstract
In the robotics field one of the more important is the optimization of the space trajectory of the tool center point (TCP) without the undesired frequencies of the vibrations. The paper shows one new optimizing method by controlling the robots space trajectory by using the direct and inverse kinematics, direct and inverse dynamics, desired optimal Fourier spectrum and proper neural network. The proper neural network was established after the assisted research of the dynamic behavior, by trying to use some known networks type. The better solution of the neural network design, for solving the inverse kinematics and direct dynamics problems with the final goal of obtaining quickly the convergence process of the space trajectory and the Fourier target spectrum with the minimum errors without some of the frequencies of the spectrum was established by using the LabVIEW proper instrumentation, after the assisted research of all network parameters like the hidden layer data, amplifier gain, time delay and position in the network structure, recurrent links and number of neurons in each of all three used layers. The Sigmoid Bipolar Hyperbolic Tangent with Time Delay and Recurrent Links SBHTNN (TDRL) neural network type was proposed and used. The complex controlling of the space trajectory was made by using three neural networks of the same type: the first one to solve the inverse kinematics problem, the second for the direct dynamics problem by using some output data from the first one and the third one to establish the magnetorheological damper current intensity. All obtained results were verified by applying the simulation with LabVIEW instrumentation. Finally we obtained one optimal complex controller that can optimize the kinematics, dynamics and vibrations with trajectory errors smaller than 2%. The proper neural network, the controller design, the results and the virtual LabVIEW instrumentation could be used in many other mechanical applications.
Authors and Affiliations
A. Olaru
Keywords
Related Articles
- EP ID EP162823
- DOI -
- Views 80
- Downloads 0
Replacing grinding with hard turning ‒ a solution for green machining
he global competition forces the manufacturing industry players to permanently innovate, thus the pressure for eco-friendly and sustainable products arise. Once with the occurrence of novel geometries, materials and coat...
The development of a laboratory system to experiment methods to improve the production flows
Globalization of production and strong competition in the automotive industry requires manufacturers to offer their customers a wide range of products, of the best quality and at lower prices. This requires uniform produ...
Creep testing of epoxy resin-flax fiber composites
The present study proposes an experimental approach into determining the viscoelastic characteristics of a composite made out of an epoxy-resin matrix with long flax fiber reinforcement. In a first instance, the fabricat...
MONITORING AN ELECTRIC MOTOR INTEGRATED IN AN AUTOMATED SYSTEM
This paper presents the progress from the field of integrated automation and safety systems that assure a well functioning of manufacturing equipments. Nowadays, many devices have sensors that monitor operational paramet...
Profiling of the planing tool for cycloidal worms machining – CAD method versus analitycal method
Cycloidal worms are helical-cylindrical surfaces, having a constant pitch. They are used in helical pumps construction, usually as driving screws. This paper introduces a new method for profiling a helical-cylindrical su...