Youbo Liu

Market evolution modeling for electric vehicles based on system dynamics and multi-agents

The deployment of electrical charging service network and power distribution network depend largely on the precisely electric vehicle prediction. System dynamics and agent based modeling are adopted in this paper to build the system model of the electric vehicle prediction through the systematic analysis of the various factors which affecting the development of electric vehicles. Case simulation is conducted to analysis the evolution of market share of electric vehicles. The key factors which affect the development of EVs in different stages, demonstration promotion, rapid growth and steady development, is evaluated through sensitivity analysis. It is found that the electric vehicles will present the fast growth in next decades, and it is mainly affected by government policies at initial period, then the development of technology and infrastructure construction will become the important factors in future.

A Chance-constrained Optimization Model for Determining Renewables Penetration Limit in Power Systems

Abstract In this article, an optimization model determining renewables penetration limit in power systems is presented. The penetration limit is defined as the enabling renewables output with quantified maximum capacity avoiding the violation of power system operation constraints. Thus, an optimal power flow (OPF)-based model with chance constraints is built and a framework including a Monte Carlo-based genetic algorithm is designed. Moreover, a transient stability verification and correction strategy based on trajectory sensitivity is proposed and modularized in the extended framework. The feasibility of the proposed methodology is verified using several test scenarios, and some related factors are investigated as well. The results indicate that renewables penetration limit can be increased by improving those studied factors.

A MapReduce Based High Performance Neural Network in Enabling Fast Stability Assessment of Power Systems

Transient stability assessment is playing a vital role in modern power systems. For this purpose, machine learning techniques have been widely employed to find critical conditions and recognize transient behaviors based on massive data analysis. However, an ever increasing volume of data generated from power systems poses a number of challenges to traditional machine learning techniques, which are computationally intensive running on standalone computers. This paper presents a MapReduce based high performance neural network to enable fast stability assessment of power systems. Hadoop, which is an open-source implementation of the MapReduce model, is first employed to parallelize the neural network. The parallel neural network is further enhanced with HaLoop to reduce the computation overhead incurred in the iteration process of the neural network. In addition, ensemble techniques are employed to accommodate the accuracy loss of the parallelized neural network in classification. The parallelized neural network is evaluated with both the IEEE 68-node system and a real power system from the aspects of computation speedup and stability assessment.

A sliding window‐based dynamic load balancing for heterogeneous Hadoop clusters

At present MapReduce computing model‐based Hadoop framework has gradually become the most famous distributed computing framework because of its remarkable features such as scalability, fault tolerance, data security, and powerful IO ability. However, Hadoop framework only supports limited load balancing policies, which may result in performance deterioration in heterogeneous clusters. Additionally Hadoop does not have advanced dynamic load balancing mechanism in enabling its optimal performance in dynamic environment. This paper presents a sliding window‐based dynamic load balancing algorithm, which specially aims at balancing the load among the heterogeneous nodes during the Hadoop job processing. The presented algorithm is evaluated in both simulated and physical environments. The experimental results show that the performances in terms of efficiency of Hadoop cluster can be significantly improved. Copyright

Researches on Electric Vehicles Access in Demonstration District Considering Network Losses

Introducing the EV into the electric distribution network .which has already become the hot spot in the smart grids. Its load has an effect on the involvement considering the uncertainties of load change. This paper presents the method of calculating network losses, considering the impact of EV and take into account the uncertainly of load. According to the view of slightly increased rate of network losses access points of EV is choose. Carry on sample and simulate into uncertainly of load by Monte Carlo. Compare to the effect from the loss of the electric distribution network whether introduce the EV into the distribution network or not. Through the typical example calculated and analyzed, we can draw the conclusion that under the linear incremental access, change rate of network losses is positive. In this article, EV access points and time provided.

Utilizing generator pair-wise potential energy functions for critical generator estimation

In this paper, a novel method is proposed for identifying the critical generator set following a system disturbance. The proposed method is specifically designed as an enhancement which works in conjunction with single machine equivalent (SIME) methods. The method utilizes an identical set of inputs, and the output yields an estimate of the set of critical generators, a requisite input for applying a SIME approach. The approach utilized proposes a set of generator pair-wise potential energy measures as a quantitative indicator of post-fault system separation. In this manuscript, a derivation is provided justifying the use of the set of pair-wise potential energy measures, since the set is shown to serve as an excellent analog to the total post-fault system potential energy. These pair-wise quantities are subsequently analyzed via the use of spectral graph theory, allowing an estimate of the critical generator set to be extracted from the energy measures. Due to the methods limited input data requirements, it can be implemented without the need for detailed model information, requiring only measured time-domain data from generator buses. This makes the method an excellent candidate for real-time stability estimation and corrective protection control schemes.

Analyzing the impacts of data quality and availability on system stability analysis using single machine equivalents

In this paper we propose to analyze the effects which errant or missing data may have on the stability analysis of a system during transient events, applying single machine equivalent (SIME) modelling for the analyses. As modern power systems are updated with the latest technologies, the deployment of PMUs is expected to increase the observability of the network both in real-time and for post-event analysis. However, there is a risk that while the availability of this new data will lead to better observability and controllability, there is an increasing risk of relying on the availability of this data for ensuring system security. Furthermore, advancements in the power system are expected to increase the number of active consumers, where the deployment of small-scale distributed non-synchronous generation or sources of unmodelled inertia will impact the swing of the system during transients, but may not be possible to account for in system models. While SIME models have been shown to be useful in many situations, in this paper we seek to quantify the errors introduced into the stability analysis due to the mentioned data quality issues. This analysis allows an assessment of the performance of SIME in situations where data quality is nonoptimal, allowing for a determination of the viability of SIME in more realistic scenarios.

Multi-agent based simulation architecture for distribution system regarding PEVs

The integrated system with EVs can be divided into three different stake-holders: Distribution Network Operators, Charge Station Operators and EV Owners, which are usually recognized as of different or even conflicting goals. In this paper, a multi-agent based solution is presented to model different entities by varying agents with individual parameters and goals. Firstly, a hierarchical and distributed EV management structure is proposed to deal with the large population of disperse EVs. Then based on research of interactions among different agents, an EV optimal charging scheduling algorithm is proposed using non-cooperative game theory, which considers the interests of each individual in a distributed way. Furthermore, the technical feasibility of the EV Management Multi Agent System developed in JADE is proved on a section of residential distribution network. The simulation results show that the proposed technology can integrate electric vehicles into the distribution systems without violating grid operational constraints.

A cloud computing framework for cascading failure simulation and analysis of large-scale transmission systems

In practical industry applications, computing complexity is frequently the primary concern of transmission system cascading failure simulation (CFS), because of high-order contingency combinations and probabilistic time-sequenced events caused by the impacts of uncertainty. In this paper, a cloud computing framework based on Hadoop/MapReduce integrated with BPA software is presented for performing high-efficiency parallel CFS and analysis. The most significant functions for CFS, including automatic action logic identification, pre-defined fault set scanning, fault chain searching, and system severity evaluation, were also carefully designed to contribute to the analysis procedure as precisely as possible for both quasi-steady and dynamic analyses. The complete architecture is implemented using Java. Two benchmark cases and one real transmission system numeric test verifies the feasibility of the proposed technology.

A critical evaluation of the application of HVDC supplementary control for system stability improvement

Bulk power transfer requirements over long distances within a country or between neighboring countries as well as rapid increase in the number of offshore wind farms, have increased the application of high voltage DC links. Apart from the controllable power flow function of HVDC systems, advanced controllers can be employed to extend their functionality for dynamic and transient stability improvement of the underlying AC systems. This paper compares and critically evaluates HVDC supplementary controller strategies for improving AC system stability. A Preliminary explanation of the controller design process for each controller type as well as a brief comparison on their application to damp the inter-area oscillations have been presented in this paper.