UMP Institutional Repository

Optimization and control of hydro generation scheduling using hybrid firefly algorithm and particle swarm optimization techniques

Hammid, Ali Thaeer (2018) Optimization and control of hydro generation scheduling using hybrid firefly algorithm and particle swarm optimization techniques. PhD thesis, Universiti Malaysia Pahang.

[img]
Preview
Pdf
Optimization and control of hydro generation scheduling using hybrid firefly algorithm and particle swarm optimization techniques - Table of contents.pdf - Accepted Version

Download (317kB) | Preview
[img]
Preview
Pdf
Optimization and control of hydro generation scheduling using hybrid firefly algorithm and particle swarm optimization techniques - Abstract.pdf - Accepted Version

Download (68kB) | Preview
[img]
Preview
Pdf
Optimization and control of hydro generation scheduling using hybrid firefly algorithm and particle swarm optimization techniques - References.pdf - Accepted Version

Download (191kB) | Preview

Abstract

The fundamental requirement of hydropower system scheduling is to determine the optimal amount of generated powers for the hydro unit of the system in the scheduling horizon of 1 year or few years while satisfying the constraints of the hydroelectric system. Annual hydro generation scheduling (AHGS) is a complicated non-linear, non-convex and non-smooth optimization problem with discontinuous solution space. The model considers daily water inflows, limits on reservoir level, power generation depends on the available head of hydro units caused by power variations, start-up, and shut-down of hydro units. Moreover, hydro generation prediction typically has composite structures such as non-linearity, non-stationarity, and fluctuation due to unexpected variable of input parameters, which converts its prediction to be very tough. Artificial intelligence (AI) methods are normally selected to deal with this problem. However, they are suffering from partial optimization, falling in solutions of local minima, and low speed of convergence. To deal with these problems, this thesis introduces three approved intelligent controllers for hydropower generation. Firstly, a hybrid algorithm namely firefly particle swarm optimization (FPSO) and series division method (SDM) based on the practical swarm optimization and the firefly algorithm is proposed. In the FPSO method, the local search is performed through the modified light intensity attraction step with PSO operator. Secondly, this approach hybridizing the FA with the rough algorithm (RA), where RA is used to control the steps of randomness for the FA while optimizing the weights of the standard BPNN model. After that, the stationary simulation prediction model is obtained. Thirdly, a novel normalized firefly fuzzy control method (NFANN) is designed for stability control of a hydro-turbine system. Moreover, the more relaxed and simplified sufficient stability conditions are given as a new set of right-angle triangle membership function (RFANN), which has been guaranteed by strict mathematical derivation. The proposed methods tested on raw data of hydropower plant of Himreen Lake Dam. The optimal hydropower generation that observed is increased to the maximum over the actual value by PSO, SD-PSO, SD-FA, and FPSO increased by 1.5%, 2.3%, 3.1%, 2.5% respectively. The proposed SD-FA controller showed better and robustness compared to the other algorithm, which could resist the random disturbances.

Item Type: Thesis (PhD)
Additional Information: Thesis (Doctor of Philosophy of Engineering in Electrical and Electronics) -- Universiti Malaysia Pahang – 2018, SV: ASSOC. PROF. DR. MOHD HERWAN BIN SULAIMAN, NO. CD: 11375
Uncontrolled Keywords: Hydro generation scheduling; hybrid algorithm
Subjects: T Technology > TK Electrical engineering. Electronics Nuclear engineering
Faculty/Division: Faculty of Electrical & Electronic Engineering
Depositing User: Mrs. Sufarini Mohd Sudin
Date Deposited: 02 Jan 2019 02:19
Last Modified: 02 Jan 2019 02:19
URI: http://umpir.ump.edu.my/id/eprint/23423
Download Statistic: View Download Statistics

Actions (login required)

View Item View Item