Abstract
This paper presents first a newly developed clustered neural network, which incorporates self-organization capacity into the well-known common multilayer perceptron (MLP) architecture. With this addition, it is possible to reduce significantly overall memory degradation of the neuro-controller during on-line training. In the second part of the paper, this clustered multilayer perceptron (CMLP) network is applied and compared to the MLP through modeling and simulations of machining processes. Simulation results presented using machining data demonstrate that the CMLP possesses more powerful modeling capacity than the standard MLP, offers better adaptability to new operating conditions, and finally performs more reliably. During on-line training with machining data about 65% degradation of previously learned information can be observed in the MLP as opposed to only 11% for the CMLP. Finally, an adaptive control scheme intended for on-line optimization of the machining processes is presented. This scheme uses a feed forward CMLP inverse neuro-controller which learns off-line and on-line the relationships between process inputs and output under simulated perturbations (i.e., tool wear and non-homogeneous workpiece material properties). The first results using the CMLP inverse neuro-controller are promising
Highlights
The need for intelligent and performing AC systems has led many investigators toward the development of more advanced schemes aiming at objectives such as: optimization of machining cost or improvement of part quality and surface finish and part dimensions
Considerable work has been realized on the development of AC systems for optimization using neural networks to achieve better performance and more robust controllers [3, 4 and 5]
Over the last few years, multilayer perceptron (MLP) neural nets using online adaptation have been successfully applied in prototype AC systems [6] and have been implemented and sold as an option in a commercial controller
Summary
The need for intelligent and performing AC systems has led many investigators toward the development of more advanced schemes aiming at objectives such as: optimization of machining cost or improvement of part quality and surface finish and part dimensions. Considerable research effort and several schemes have been realized over the last 30 years [1, 2], their applicability and robustness for industrial machining control is still extremely limited for several reasons. Modeling performance of the clustered neural network Non-Linear Modeling and Simulation At this stage, the capacity for non-linear modeling and adaptability of the CMLP network is tested and compared to that of MLP network through off-line and on-line training respectively. The CMLP network is expected to show a greater resistance with respect to the memory degradation problem of the stored information This resistance will be qualified during on-line training. These latter gradually improve their performances with the sequential training of the new data. About 11% and 65% memory degradation can be observed in CMLP and MLP respectively
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
