Special issue on patterns and autonomous control

Special issue on patterns and autonomous control

Robotics and Autonomous Systems 49 (2004) 135–136 Foreword Special issue on patterns and autonomous control In this special issue, within the first ...

31KB Sizes 0 Downloads 88 Views

Robotics and Autonomous Systems 49 (2004) 135–136

Foreword

Special issue on patterns and autonomous control In this special issue, within the first four articles, the authors investigate certain aspects of knowledge representation and reasoning with conceptual and technical topics which includes discovery of patterns in data, with learning schemas, and some applications of fuzzy sets from patterns, matching with strings, etc. Within the remaining seven articles, the authors investigate various applied topics in production and process industries and manufacturing, including robotics and control.

1. Within part 1 In “Discovering useful and understandable patterns in manufacturing data”, the authors present a novel, perception-based method, called “Automated Perception Network”, for automated construction of compact and interpretable models from highly noisy data. It is applied to yield data of two semi-conductor products. In “Learning associations between natural groupings of input and output with neuro-fuzzy structure”, the authors propose an approach that discovers associations between natural groups of input and output. First a Fuzzy C-Means algorithm identifies the fuzzy clusters then a learning algorithm determines fuzzy rules by a neuro-fuzzy architecture. The proposed approach is applied to a highly non-linear function and the results are compared to ANFIS and NN approaches. In “3-Dimensional curve similarity using string matching”, the authors present an approach to measuring similarity between 3D-curves. They validate the proposed approach with experiments that use synthetic and digitized data. 0921-8890/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.robot.2004.09.001

In “A fuzzy system modeling algorithm for data analysis and approximate reasoning”, the authors introduce an algorithm for data analysis and approximate reasoning. The performance of the proposed algorithm compared with two other algorithms that are applied to two benchmark data set from current literature.

2. Within part 2 In “Determination and exchange of supply information for co-operation in a complex production network”, the authors propose a system that supports co-operation in complex production networks by enabling companies to determine and exchange supply information with their customers. In its applications, a business process is implemented as a pilot system and evaluated by the user companies. In “BOFY—fuzzy logic control for the basic oxygen furnace (BOF)”, the authors propose a fuzzy system modeling for the control of BOF processes. It is applied to real-life integrated steel plant data. The application results, in terms of acceptable level of compatibility, are compared to the empirical BOF data and targeted steel composition. In “Development of a new dioid algebraic model for manufacturing systems with the scheduling decision making capability”, the authors introduce a decision making capability to a dioid algebraic model which captures the discontinuous nature of discrete event dynamic systems. The model has a sequencing capability for some job descriptions with resource unavailabilities, technological constraints and alternative process plans.

136

Foreword / Robotics and Autonomous Systems 49 (2004) 135–136

In “Investigation of center of mass by using magic squares and its possible engineering applications”, the authors first review a general method for the generation of magic squares. Next they discuss its applications to genetic engineering, operations research, intelligent computation, physics, etc. In particular, they demonstrate how the center of mass can be analysed by these magic squares. In “A study of neural network based inverse kinematics solution for a 3-joint robot”, the authors first generate many initial and final points in the work volume of a robotic manipulator. Next all the angles with coordinates (x, y, z) are recorded as a training set for a neural network. An online working feature of this neural network provides a successful solution. In “A simulation comparison of intelligent control algorithms on a direct drive manipulator”, the authors investigate the control of a non-linear system with neural networks and fuzzy logic. The performance of the proposed method is compared with respect to trajectory

tracking, computational complexity, design complexity, RMSE, training time and payload variation. In “Optimization of resistive loading of EMI/EMC near field probe”, the authors investigate “near field electromagnetic measurements” with an application of fuzzy logic algorithm. Fuzzy inference system results are compared with the experimental results. At the end two new equations are generated with a fuzzy inference system: one for circular loop probes and the other for square loop probes. ¨ I.B. Turksen∗ , H. Tas¸kin, C. Kubat, E. Oztemel Department of Mechanical and Industrial Engineering, University of Toronto 5 King’s College Road Toronto, Ont., Canada M5S 3G8 ∗ Corresponding author. Tel.: +1 416 978 6420 fax: +1 416 978 3453 E-mail address: [email protected] (I.B. Turksen)