Fault tolerance in the framework of support vector machines based model predictive control

Rodil, Sergio Saludes; Fuente, José Luis de la

doi:10.1016/j.engappai.2010.06.008

Cited by 22 publications

(5 citation statements)

References 30 publications

(24 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The first step is to establish a high-fidelity predictive model. The datadriven methods such as neural network, [15][16][17] support vector machines, 18,19 random forests, 20 and subspace identification [21][22][23][24] are proposed in recent years. These methods learn the system information through input 1 and output data, which can refine the precision of predictive model.…”

Section: Introductionmentioning

confidence: 99%

Digital twin-based subspace model predictive control for thermal power plant

Zhao

Cai

Jiang

2023

Proceedings of the Institution of Mechanical Engineers, Part I:

View full text Add to dashboard Cite

In this article, a digital twin-based subspace model predictive control scheme is proposed for solving model mismatch caused by parameter perturbations of thermal power plants. First, the rational range of essential parameters in the nonlinear model is identified based on the nominal data of an actual 600 MW power plant, which provides necessary foundation for subsequent studies. Second, to improve the load-changing capacity under parameter disturbances, a novel digital twin-based subspace model predictive control algorithm is established by regarding the crucial parameter as extended input of predictive model. Finally, the small fluctuations during steady-state operation are effectively restrained through transforming weights and constraints slowly and smoothly at the end of load change. Simulation results illustrate that compared with the conventional subspace-based model predictive control, the proposed digital twin-based subspace model predictive control scheme can obtain a better performance in terms of both dynamic regulation rate and steady-state performance under parameter perturbations.

show abstract

Section: Introductionmentioning

confidence: 99%

Digital twin-based subspace model predictive control for thermal power plant

Zhao

Cai

Jiang

2023

Proceedings of the Institution of Mechanical Engineers, Part I:

View full text Add to dashboard Cite

show abstract

“…The data-driven method provides a promising solution for analytical redundancy which can extract the engine features and bridge the relationship between input and output from a large number of experimental data. Shallow learning methods, such as support vector machine (SVM) [18][19][20] and extreme learning machine (ELM) [21,22], have quick calculation speed; however, they cannot extract deep information. Jun Zhou et al [21] proposed an ELM-based method to provide analytical redundancy for sensor fault diagnostics.…”

Section: Introductionmentioning

confidence: 99%

Analytical Redundancy for Variable Cycle Engine Based on Variable-Weights-Biases Neural Network

et al. 2023

View full text Add to dashboard Cite

Due to the complex nature of a variable cycle engine (VCE), which has numerous control variables and working modes across a broad flight envelope, coupled with the whole engine’s degradation, the analytical redundancy method based on component-level models may not provide an accurate estimation of the sensors. Variable-weights-biases neural network (VWB Net) is proposed to construct VCE’s analytical redundancy. Unlike conventional networks whose weights and biases are fixed, VWB Net’s variable-weights and variable-biases are functions of input which greatly increase its nonlinear mapping capability by integrating input information. Variable-biases can also be used to eliminate the error between actual sensor output and estimated value quickly at the terminal node. Compared with the BP network and Dense net, VWB Net has fewer parameters, faster calculation speed, and higher accuracy. Digital simulation results of VCE parameter estimation demonstrate that VWB Net’s average relative errors are under 0.27% with calculation and parameter efficiency at least 166 times higher than that of Dense net. Hardware in the loop simulation further verifies VWB Net’s estimation accuracy and real-time calculation.

show abstract

“…In the on-line controller selection approach, the controllers associated with cer-tain/predetermined faulty conditions are computed in an off-line manner in the design stage and they are selected in an on-line manner based on real-time data from the FDD algorithm [15]. In the on-line controller calculation approach, the controller parameters are calculated in an on-line manner right after the occurrence of the fault [14,16].…”

Section: Introductionmentioning

confidence: 99%

A fault detection, diagnosis, and reconfiguration method via support vector machines

Kabaoğlu¹

2015

Turk J Elec Eng & Comp Sci

View full text Add to dashboard Cite

This paper presents a fault detection, diagnosis, and reconfiguration method based on support vector machines. This method is appropriate for certain or predetermined faults and involves a fault detection and diagnosis unit and an online controller selection type reconfiguration mechanism. In this method, when a fault is detected and diagnosed by the fault detection and diagnosis unit, a suitable controller, which has been determined via an optimization algorithm in an off-line fashion, is activated to maintain proper closed-loop performance of the system in an on-line manner. In the detection, diagnosis, and reconfiguration stages of the method, support vector classification and regression machines are used and the performance is tested on a simulation model of a two-tank level control system for various fault scenarios.

show abstract

Fault tolerance in the framework of support vector machines based model predictive control

Cited by 22 publications

References 30 publications

Digital twin-based subspace model predictive control for thermal power plant

Digital twin-based subspace model predictive control for thermal power plant

Analytical Redundancy for Variable Cycle Engine Based on Variable-Weights-Biases Neural Network

A fault detection, diagnosis, and reconfiguration method via support vector machines

Contact Info

Product

Resources

About