With the increasing number of electric vehicles (EVs), the disordered charging of a large number of EVs will have a large influence on the power grid. The problems of charging and discharging optimization management for EVs are studied in this paper. The distribution of characteristic quantities of charging behaviour such as the starting time and charging duration are analysed. The results show that charging distribution is in line with a logarithmic normal distribution. An EV charging behaviour model is established, and error calibration is carried out. The result shows that the error is within its permitted scope. The daily EV charge load is obtained by using the Latin hypercube Monte Carlo statistical method. Genetic particle swarm optimization (PSO) is proposed to optimize the proportion of AC 1, AC 2 and DC charging equipment, and the optimal solution can not only meet the needs of users but also reduce equipment investment and the EV peak valley difference, so the effectiveness of the method is verified.
This paper discusses three variants of how e-mobility development will affect the Polish Power System. Multivariate forecasts of annual new registrations of electric vehicles for up to seven years are developed. The forecasts use the direct trend extrapolation methods, methods based on the deterministic chaos theory, multiple regression models, and the Grey model. The number of electric vehicles in use was determined for 2019‒2025 based on the forecast new registrations. The forecasts were conducted in three variants for the annual electric energy demand in 2019‒2025, using the forecast number of electric vehicles and the forecast annual demand for electric energy excluding e-mobility. Forecasts were conducted in three variants for the daily load profile of power system for winter and summer seasons in the Polish Power system in 2019‒2025 based on three variants of the forecast number of electric vehicles and forecast relative daily load profiles.
From the perspective of a virtual power plant (VPP) with electric vehicles (EVs), a self-scheduling strategy considering the response time margin (RTM) and state of charge margin (SOCM) is proposed. Firstly, considering the response state of the state of charge (SOC) and charge-discharge state of EVs, a VPP based response capacity determination model of EVs is established. Then, RTM and SOCM indexes are introduced on the basis of the power system scheduling target and the EV users’ traveling demands. The RTM and SOCM indices are calculated and then are used to generate a priority sequence of responsive EVs for the VPP. In the process of the scheduling period and rolling iteration, the scheduling schemes of the EVs in the VPP for multiple time periods are determined. Finally, the VPP self-scheduling strategy is validated by taking an VPP containing three kinds of EV users as an example. Simulation results show that with the proposed strategy, the VPP is able to respond to the scheduling power from the power system, while ensuring the traveling demands of the EV users at the same time.
Inaccurate estimation in highway projects represents a major problem facing planners and estimators, especially when data and information about the projects are not available, and therefore the need to use modern technologies that addresses the problem of inaccuracy of estimation arises. The current methods and techniques used to estimate earned value indexes in Iraq are weak and inefficient. In addition, there is a need to adopt new and advanced technologies to estimate earned value indexes that are fast, accurate and flexible to use. The main objective of this research is to use an advanced method known as artificial neural networks to estimate the TSPI of highway buildings. The application of artificial neural networks as a new digital technology in the construction industrial in Republic of Iraq is absolutely necessary to ensure successful project management. One model built to predict the TCSPI of highway projects. In this current study, artificial neural network model were used to model the process of estimating earned value indexes, and several cases related to the construction of artificial neural networks have been studied, including network architecture and internal factors and the extent of their impact on the performance of artificial neural network models. Easy equation was developed to calculate that TSPI. It was found that these networks have the ability to predict the TSPI of highway projects with a very outstanding saucepan of reliability (97.00%), and the accounting coefficients (R) (95.43%).