APPLICATION OF GENETIC EXPRESSION PROGRAMMING IN URBAN DRINKING WATER

Purpose The purpose of this paper is to understand the factors that affect the performance of projects being implemented on the public-private partnership (PPP) framework, with specific reference to urban drinking water sector in India. Design/methodology/approach A listing of factors that have a bearing on project performance have been developed based on a review of the literature. Through a survey, seven factors that are relevant to the Indian context were determined. Interviews were then conducted across a cross-section of government agencies, financial institutions, development agencies, private sector entities and consultants to understand the relative importance of these attributes. The analytical hierarchy process was used to develop relative weights of these factors. Findings Ranking and relative weights of the factors in descending order are stakeholder consent and support for water PPP projects (22.1 percent), appropriate project structure (17.4 percent), availability of realistic baseline information (16.2 percent), reasonable water tariffs (13.9 percent), public sector capacity (13.0 percent), well-developed market (9.5 percent) and water sector regulator (7.9 percent). Differences in perceptions amongst various stakeholder groups were also found. Research limitations/implications Water sector has not matured, and with the advent of newer formats of implementation models, there could be significant changes in the sector. As the number of projects available for study is limited, this exercise can be undertaken periodically and updated in relation to experiences in other infrastructure sectors. Practical implications This analysis provides inputs to policymakers and project proponents for structuring more sustainable urban drinking water PPP projects. Originality/value Indian infrastructure PPP market is attracting increased attention from researchers, though not much emphasis is being given to urban drinking water sector. This paper aims to contribute toward filling this research gap.

Download Full-text

A Hybrid Genetic Expression Programming and Genetic Algorithm (GEP-GA) of Auto-modeling Electrical Equivalent Circuit for Particle Structure Measurement with Electrochemical Impedance Spectroscopy (EIS)

IEEE Sensors Journal ◽

10.1109/jsen.2021.3106160 ◽

2021 ◽

pp. 1-1

Author(s):

Jianfen Wan ◽

Hongrun Yin ◽

Kai Liu ◽

Chengjun Zhu ◽

Xiaoxiang Guan ◽

...

Keyword(s):

Genetic Algorithm ◽

Electrochemical Impedance Spectroscopy ◽

Impedance Spectroscopy ◽

Equivalent Circuit ◽

Electrochemical Impedance ◽

Particle Structure ◽

Electrical Equivalent Circuit ◽

Genetic Expression ◽

Genetic Expression Programming ◽

Structure Measurement

Download Full-text

On the Application of Machine Learning in Savonius Wind Turbine Technology: An Estimation of Turbine Performance Using Artificial Neural Network and Genetic Expression Programming

Journal of Energy Resources Technology ◽

10.1115/1.4051736 ◽

2021 ◽

pp. 1-47

Author(s):

Umang H. Rathod ◽

Vinayak Kulkarni ◽

Ujjwal K. Saha

Keyword(s):

Neural Network ◽

Machine Learning ◽

Artificial Neural Network ◽

Speed Ratio ◽

Stream Velocity ◽

Cross Sectional ◽

Genetic Expression ◽

Genetic Expression Programming ◽

Artificial Neural ◽

Input Variables

Abstract This paper addresses the application of artificial neural network (ANN) and genetic expression programming (GEP), the popular artificial intelligence and machine learning methods, in order to estimate the Savonius wind rotor's performance based on different independent design variables. Savonius wind rotor is one of the competent members of the vertical axis wind turbines (VAWTs) due to its advantageous qualities such as direction independency, design simplicity, ability to perform at low wind speeds, potent standalone system. The available experimental data on Savonius wind rotor have been used to train the ANN and GEP using MATLAB R2020b and GeneXProTools 5.0 software, respectively. The input variables used in ANN and GEP architecture include newly proposed design shape factors, number of blades and stages, gap and overlap lengths, height and diameter of the rotor, free stream velocity, end plate diameter and tip speed ratio, besides cross-sectional area of wind tunnel test section. Based on this, the unknown governing function constituted by the aforementioned input variables is established using ANN and GEP to approximate/forecast the rotor performance as an output. The governing equation formulated by ANN is in the form of weights and biases, while GEP provides it in the form of traditional mathematical functions. The trained ANN and GEP are capable to estimate the rotor performance with R2 ≈ 0.97 and R2 ≈ 0.65, respectively, in correlation with the reported experimental rotor performance.

Download Full-text