Neural Network-Based Control Framework for SISO Uncertain System: Passive Fault Tolerant Approach

Raval, Sejal; Patel, Himanshukumar R.; Shah, Viral B.

doi:10.1007/978-3-030-51156-2_121

Cited by 5 publications

(4 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Additionally, statistical analysis and a Friedman rank test confirm the superiority of the JOA algorithm-based PID controller over other controllers. In future, fractional order controllers 26,27 , fuzzy controllers [28][29][30][31][32][33] and learning based controllers 34,35 , along with learning based optimization techniques 36 , can be implemented for optimal generation control. Z 0102 0.00447 0.00426 0.00364 0.00336 0.00425 0.00346…”

Section: Discussionmentioning

confidence: 99%

Investigation of rank order centroid method for optimal generation control

Varshney,

Waghmare,

Singh

et al. 2024

Sci Rep

View full text Add to dashboard Cite

Multi-criteria decision-making (MCDM) presents a significant challenge in decision-making processes, aiming to ascertain optimal choice by considering multiple criteria. This paper proposes rank order centroid (ROC) method, MCDM technique, to determine weights for sub-objective functions, specifically, addressing issue of automatic generation control (AGC) within two area interconnected power system (TAIPS). The sub-objective functions include integral time absolute errors (ITAE) for frequency deviations and control errors in both areas, along with ITAE of fluctuation in tie-line power. These are integrated into an overall objective function, with ROC method systematically assigning weights to each sub-objective. Subsequently, a PID controller is designed based on this objective function. To further optimize objective function, Jaya optimization algorithm (JOA) is implemented, alongside other optimization algorithms such as teacher–learner based optimization algorithm (TLBOA), Luus–Jaakola algorithm (LJA), Nelder–Mead simplex algorithm (NMSA), elephant herding optimization algorithm (EHOA), and differential evolution algorithm (DEA). Six distinct case analyses are conducted to evaluate controller’s performance under various load conditions, plotting data to illustrate responses to frequency and tie-line exchange fluctuations. Additionally, statistical analysis is performed to provide further insights into efficacy of JOA-based PID controller. Furthermore, to prove the efficacy of JOA-based proposed controller through non-parametric test, Friedman rank test is utilized.

show abstract

Section: Discussionmentioning

confidence: 99%

Investigation of rank order centroid method for optimal generation control

Varshney,

Waghmare,

Singh

et al. 2024

Sci Rep

View full text Add to dashboard Cite

show abstract

“…For the best (10) Fig. 4 Structure of NN model for five inputs [23] fitting of data on this network, the optimal value of the coefficient regression (R) is equal to 1 and its value close to 0 is absolutely unacceptable. The Regression plot [19] reflects the relationship between input and output parameters, which is given in the equation given: From the regression plot, it is evident that perhaps the regression plot for the ANN getting a configuration of five neurons in the hidden layer has a R value of 1.…”

Section: Implementation Setupmentioning

confidence: 99%

“…Proposed methodology for PFTC[23] Data generation system for SISO uncertain level control system model[23] …”

mentioning

confidence: 99%

See 1 more Smart Citation

Fault-Tolerant Controller Comparative Study and Analysis for Benchmark Two-Tank Interacting Level Control System

2021

Self Cite

View full text Add to dashboard Cite

In recent times fault-tolerant controller is a prime choice because of the handling capability of the multifunction into the closed-loop system irrespective of the faults and the external process disturbances in various engineering applications. This paper presents innovative fault-tolerant control scheme using a neural network, and the main contribution of this work is to design a feedforwarded back propagation neural network for controlling the actuator and system component (leak) faults into the level control process. By adopting the neural network as a fault-tolerant controller the efficacy of the fault-tolerant control scheme is dominating to the conventional scheme proposed by Dutta et al. (Real-time linear quadratic versus sliding mode liquid level control of a coupled tank system"

show abstract