Fault tolerant multi-layer neural networks with GA training

Sugawara, Eiko; Fukushi, Masaru; Horiguchi, S.

doi:10.1109/dftvs.2003.1250128

Cited by 9 publications

(4 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, with Equation (10) defined, the error partial derivatives are transformed and, by employing the MIT rule, the update rules for the adaptive feedforward and adaptive feedback gain are written. This can be observed in Equations (11) and (12…”

Section: Model Reference Adaptive Controlmentioning

confidence: 75%

“…Recently genetic algorithms have been applied in fault-tolerant control as a strategy to optimize the controlled system in order to accommodate system failures. For instance, in [11] a FTC approach using multi-layer ANNs with a GA was presented. The proposed FTC scheme uses hardware redundancy and weight retraining based on a GA in order to reconfigure the ANN to accommodate the fault.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Combining Artificial Intelligence and Advanced Techniques in Fault-Tolerant Control

Martínez¹,

Garza-Castañón²

2011

JART

View full text Add to dashboard Cite

We present the integration of artificial intelligence, robust, nonlinear and model reference adaptive control (MRAC)methods for fault-tolerant control (FTC). We combine MRAC schemes with classical PID controllers, artificial neuralnetworks (ANNs), genetic algorithms (GAs), H∞ ∞ ∞ ). The third scheme is an MRAC controller with a sliding mode controller (SMC). The fourth scheme is an MRAC controller with an ANN. The fifth scheme is an MRAC controller with a PID controller optimized by a GA. Finally, thelast scheme is an MRAC classical control system. The objective of this research is to generate more powerful FTCmethods and compare the performance of above schemes under different fault conditions in sensors and actuators.An industrial heat exchanger process was the test bed for these approaches. Simulation results showed that the useof Pattern Search Optimization and ANNs improved the performance of the FTC scheme because it makes the controlsystem more robust against sensor and actuator faults.control(H controls and sliding mode controls. Six different schemas are proposed: the first one is an MRAC with an artificial neural network and a PID controller whose parameters weretuned by a GA using Pattern Search Optimization. The second scheme is an MRAC controller with an H

show abstract

Section: Model Reference Adaptive Controlmentioning

confidence: 75%

Section: Introductionmentioning

confidence: 99%

Combining Artificial Intelligence and Advanced Techniques in Fault-Tolerant Control

Martínez¹,

Garza-Castañón²

2011

JART

View full text Add to dashboard Cite

show abstract

“…Approaches in this category include constraining the weights to lie within a limited range toward an even weight distribution [204]- [206]. A fourth method is to combine the training process and fault tolerance objective into an optimization problem solved by nonlinear optimization algorithms with the aim to learn a network model that performs the desired task and at the same time fulfills fault tolerance constraints [207]- [210]. A fifth method proposed in [211] considers a constructive training in the presence of faults, where neurons are incrementally added whenever the network fails to learn until a satisfactory learning or a userdefined maximum network size is reached.…”

Section: ) Model Training Modificationmentioning

confidence: 99%

Testability and Dependability of AI Hardware: Survey, Trends, Challenges, and Perspectives

Liu

Stratigopoulos³

2023

IEEE Des. Test

View full text Add to dashboard Cite

In recent years, there has been an expedited trend in embracing bold and radical innovation of computer architectures, aiming at the continuation of computing performance improvement despite the slowed-down physical device scaling. One new frontier in this field focuses on Artificial Intelligence (AI) hardware. While functionality of AI hardware still remains the main focus, testability and dependability of these new architectures need to be addressed before the mainstream adoption. This survey paper covers the state-of-the-art in research and development of dependability and testability solutions for AI hardware including digital or analog implementations of Artificial Neural Networks (ANNs) and Spiking Neural Networks (SNNs), used in accelerators and neuromorphic designs. Trends, challenges and perspectives are also discussed in this paper.

show abstract

“…Inconsistent mutation is adopted for maintaining diversity in the population so as to generate a new solution randomly which allows engineers to obtain global optimal solution. Based on the analysis of the weights associated with wafer fabrication process suggested by Sugawara (2003), the corresponding weight can be obtained from the weight table.…”

Section: <Insert Figure 6>mentioning

confidence: 99%

Development of a co-operative distributed process mining system for quality assurance

Lau

Kwok

et al. 2008

International Journal of Production Research

View full text Add to dashboard Cite

In this paper, a cooperative distributed process mining system (CDPMS) is developed to streamline the workflow along the supply chain in order to offer shorter delivery times, more flexibility and higher customer satisfaction with learning ability. The proposed system is equipped with the "distributed process mining" feature which is used to discover the hidden relationships among each working decision in distributed manner. This method incorporates the concept of data mining and knowledge refinement into decision making process for ensuring "doing the right things" within the workflow. An example of implementation is given, based on the case of slider manufacturer.

show abstract

Fault tolerant multi-layer neural networks with GA training

Cited by 9 publications

References 8 publications

Combining Artificial Intelligence and Advanced Techniques in Fault-Tolerant Control

Combining Artificial Intelligence and Advanced Techniques in Fault-Tolerant Control

Testability and Dependability of AI Hardware: Survey, Trends, Challenges, and Perspectives

Development of a co-operative distributed process mining system for quality assurance

Contact Info

Product

Resources

About