An efficient hybrid approach of improved adaptive neural fuzzy inference system and teaching learning-based optimization for design optimization of a jet pump-based thermoacoustic-Stirling heat engine

ORIGINAL ARTICLE

An efficient hybrid approach of improved adaptive neural fuzzy

inference system and teaching learning-based optimization for design

optimization of a jet pump-based thermoacoustic-Stirling heat engine

Ngoc Le Chau1 • Thanh-Phong Dao2,3 • Van Anh Dang1

Received: 21 March 2018 / Accepted: 9 May 2019

Springer-Verlag London Ltd., part of Springer Nature 2019

Abstract

The acoustic streaming is a key drawback and eliminates the performance of a jet pump-based thermoacoustic-Stirling heat

engine. The present study deals with a new hybrid optimization approach to reduce the acoustic streaming energy. The

proposed work is an integration of Taguchi method (TM), adaptive neural fuzzy inference system (ANFIS), and teaching–

learning-based optimization (TLBO). The Taguchi method plays three important roles. The first role is to layout the

number of experiments. The second role is to identify the most appropriate parameters for ANFIS structure regarding the

number of input membership functions (MFs), types of input MFs, optimal learning method, and types of output MFs. In

order to determine the optimum parameters for the ANFIS structure, the root-mean-squared error, a performance criterion,

is minimized by using the TM. The final role of TM is to optimize the controllable parameters of the TLBO. Subsequently,

modeling between geometric parameters and acoustic streaming is established by the built ANFIS structure. Finally, the

TLBO is adopted by optimizing the design parameters. The outcomes of study revealed that the acoustic streaming is

relatively reduced. Based on Wilcoxon signed-rank test and Friedman test, it proves that the effectiveness of the proposed

hybrid approach is better to other evolutionary algorithms. The current approach is an efficient optimizer for complex

optimization problems.

Keywords Jet pump Acoustic streaming Optimization Taguchi method ANFIS Teaching–learning-based

optimization Statistical analysis

1 Introduction

Commonly, thermoacoustic devices have been employed

as the sound waves for a thermodynamic process instead of

utilizing the mechanical pistons [1]. This is very efficient

and useful technique to convert a waste heat into an elec￾trical energy or mechanical power where the Stirling cycle

is an important thermodynamic cycle. This process can cut

the manufacturing and maintenance costs to be handled.

Nowadays, there have been growing a lot of attention in

development of the Stirling engine-based device. In an

earlier attempt, the mechanical Stirling engine was

employed with two pistons and a regenerative heat

exchanger [2]. Later, a thermoacoustic-Stirling heat engine

was developed [3], but its performance efficiency was

restricted by the existing acoustic streaming [4]. The

acoustic streaming is defined as a steady flow resulted by

absorption of high amplitude acoustic oscillations. In

search of enhanced outcome, a jet pump was integrated and

used with engine to decrease the streaming energy [5]. It

was found that the structural parameters of jet pump played

an important role in suppressing the acoustic streaming.

Regarding the state of the art of the Stirling cycle in

recent era, owing to its capability to transfer the energy

without utilizing external mechanical actuators, the

& Thanh-Phong Dao

[email protected]

1 Faculty of Mechanical Engineering, Industrial University of

Ho Chi Minh City, Ho Chi Minh City, Vietnam

2 Division of Computational Mechatronics, Institute for

Computational Science, Ton Duc Thang University,

Ho Chi Minh City, Vietnam

3 Faculty of Electrical and Electronics Engineering, Ton Duc

Thang University, Ho Chi Minh City, Vietnam

123

Neural Computing and Applications

https://doi.org/10.1007/s00521-019-04249-y(0123456789().,-volV) (0123456789().,-volV)