Reliability assessment of power system generation adequacy with wind power using population-based intelligent search methods

Reliability of the generation system is an important aspect of planning for the expansion of future system capacity which ensures that the total installed capacity would be sufficient to provide adequate electricity, therefore, it is necessary to investigate the reliability of the power system. The reliability assessment of the adequacy of the generating system is normally calculated by using either analytical or simulation methods. The Monte Carlo simulation (MCS) method enables an accurate evaluation of reliability indices. The drawback of MCS is that it is not suitable for a system with large number of components in a system or high-reliability performance that require large computational effort which would take a long time to converge efficiently. This study sought to examine the performance of three newly proposed techniques, for reliability assessment of the power systems, namely Disparity Evolution Genetic Algorithm (DEGA), Binary Particle Swarm Optimisation (BPSO), and Differential Evolution Optimization Algorithm (DEOA). The proposed intelligent algorithms would rely on the population intelligent search (PIS) techniques considered as viable replacement for the MCS method in assessing the reliability indices of non-chronological system. The advantage of using these algorithms is obvious as they would speed up the computation to obtain higher accuracy with less computation effort. In recent years, the development of wind power to meet the demand for electricity has received considerable attention. However, this energy source differs considerably from the conventional generation sources because it is intermittent in nature and may lead to high-risk levels in the electrical system reliability. As such, the three novel PIS techniques (i.e., DEGA, DEOA, and BPSO) were proposed for reliability assessment of power generation systems with the integration of wind energy. These methods proved accurate in estimating the reliability indices, with less computation effort. In this study, analysis was made on the wind speed data characteristics and wind power potential assessment at three a given sites in Malaysia namely Mersing, Kudat, and Kuala Terengganu. Results have shown that Mersing and Kudat were suitable as wind sites. The findings of this research have provided evidence to support those of the previous studies were conducted separately for Mersing and Kudat indicating that these sites could potentially be utilised to construct a new wind power plant in Malaysia. Additionally, the present study has developed a prediction model of wind speed for these three sites in Malaysia. This model took into consideration the seasonal wind speed variation during the year, in the format of the combined method comprising the Weibull model with artificial neural network (ANN), so that the forecasting errors of wind values would be lower than those generated by using only the Weibull model. It was suggested that the wind power should be connected to the Roy Billinton Test System (RBTS), from two sites in Malaysia. The reliability indices obtained before and after the inclusion of the two farms to the system under consideration were compared. Based on this analysis, it was found that, with the inclusion of wind power from both sites, the reliability indices had slightly improved the reliability of RBTS.

Preview Text FK 2018 23 - IR.pdf Download (866kB) | Preview

Actions (login required)

View Item