Scholarlink

<p>The reliable operation of power systems is heavily reliant on the optimal power flow (OPF), which becomes complex when integrating renewable energy sources (RESs) due to their output uncertainty. This paper presents an improved growth optimizer (IGO) to solve the OPF in a hybrid model including three types of RESs (solar photovoltaic, wind, and small hydro) and thermal power resources. Gaussian barebones (GB) and quasi-opposition-based learning (QOBL) methods are used to enhance how quickly the original GO finds solutions by adjusting the positions of individuals during both the learning and reflection stages. The original GO, like other meta-heuristic optimization methods, also suffers from system constraints being violated throughout the optimization process. To overcome this issue, the static penalty factor (SPF) is added to the objective functions. The available output powers of RESs are determined by three types of probability density functions (PDFs): the lognormal PDF for solar photovoltaic power, the Weibull PDF for wind power, and the Gumbel PDF for hydropower. The objective function in this study considers the overestimation and underestimation of RESs output power by including reserve cost and penalty cost, respectively. Additionally, a carbon tax on thermal power is added to the main objective function in order to decrease carbon emissions. The model is tested and evaluated on the modified IEEE-30 and IEEE-57 bus power networks using MATLAB software. The results of the IGO are compared with the results of the original GO and nine standard and recent optimization algorithms, in addition to the results of six other optimization algorithms used in published researches. The simulation results demonstrate that the GO and IGO outperformed other algorithms in all case studies. The results further demonstrate that in the IEEE 30 bus system, the IGO outperformed the original GO primarily in terms of convergence characteristics, and in the IEEE 57 bus system, the IGO exceeded the GO in terms of solution quality and convergence.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Optimal power flow analysis of power system considering renewable energy and load uncertainties based on an improved growth optimizer

  • 期刊论文

  • 发表日期：

    2026年4月21日

查看全文
  • Mohamed Farhat,
  • Salah Kamel,
  • Almoataz Y. Abdelaziz

摘要

The reliable operation of power systems is heavily reliant on the optimal power flow (OPF), which becomes complex when integrating renewable energy sources (RESs) due to their output uncertainty. This paper presents an improved growth optimizer (IGO) to solve the OPF in a hybrid model including three types of RESs (solar photovoltaic, wind, and small hydro) and thermal power resources. Gaussian barebones (GB) and quasi-opposition-based learning (QOBL) methods are used to enhance how quickly the original GO finds solutions by adjusting the positions of individuals during both the learning and reflection stages. The original GO, like other meta-heuristic optimization methods, also suffers from system constraints being violated throughout the optimization process. To overcome this issue, the static penalty factor (SPF) is added to the objective functions. The available output powers of RESs are determined by three types of probability density functions (PDFs): the lognormal PDF for solar photovoltaic power, the Weibull PDF for wind power, and the Gumbel PDF for hydropower. The objective function in this study considers the overestimation and underestimation of RESs output power by including reserve cost and penalty cost, respectively. Additionally, a carbon tax on thermal power is added to the main objective function in order to decrease carbon emissions. The model is tested and evaluated on the modified IEEE-30 and IEEE-57 bus power networks using MATLAB software. The results of the IGO are compared with the results of the original GO and nine standard and recent optimization algorithms, in addition to the results of six other optimization algorithms used in published researches. The simulation results demonstrate that the GO and IGO outperformed other algorithms in all case studies. The results further demonstrate that in the IEEE 30 bus system, the IGO outperformed the original GO primarily in terms of convergence characteristics, and in the IEEE 57 bus system, the IGO exceeded the GO in terms of solution quality and convergence.