Shouzhen Zhu

Aggregation Model-Based Optimization for Electric Vehicle Charging Strategy

This paper presents an aggregation charging model for large numbers of electric vehicles (EVs). A genetic algorithm (GA) is employed to obtain the stochastic feature parameters of the aggregation model, and a charging strategy based on the aggregation model is developed to reduce the power fluctuation level caused by EV charging. In addition, an updatable optimization method is proposed to track the variation of the EV charging characteristics. The proposed charging strategy and optimization method are validated by the simulation results.

Expansion Planning of Active Distribution Networks With Centralized and Distributed Energy Storage Systems

A multistage active distribution network (ADN) planning model that is integrated with the application of energy storage system (ESS) is presented in this paper. Both the long-term investment cost and short-term operation conditions of ADN are considered in the proposed model. The power supply reliability improvement brought by ESS is also analyzed. At each planning stage, the operation conditions are divided into several typical day scenarios and an extreme condition scenario which are based on the load forecast data. The long-term expansion planning decisions including those for replacing and adding circuits, introducing ESS, and considering ADN short-term operation strategies for the ESS charging and discharging are optimized together in the proposed model. The ESS benefits pertaining to peak load shaving and power reliability enhancement are demonstrated using numerical cases, in which the centralized and the distributed ESS are considered as options for the ADN implementation. The effectiveness of the proposed model is demonstrated through various discussions in the paper.

Power system dynamic load identification and stability

Proper power system load models are playing more important roles in power system stability analysis in todays stressed power systems. Different load models may cause a large difference in stability analytical results. Measurement based load modeling gives a closer look at the real power system loads and their dynamic characteristics. In this paper, genetic algorithms and evolutionary programming based system identification is used to locate the best available parameters for proper load modeling based on real field test data taken in North China.

Multi-Stage Planning of Active Distribution Networks Considering the Co-Optimization of Operation Strategies

This paper presents a multi-stage and multi-load-scenario Active Distribution Network (ADN) expansion planning model. The proposed model considers the applications of new distributed generation (DG) and construction of feeders at the planning level, and the utilization of DG supply and topology reconfiguration of distribution network (DN) including microgrid at the operation level. The proposed co-optimization model incorporates both investment and operation costs of ADN in the objective function. The ADN operation at each time stage is divided into several scenarios based on the load forecast data, in which the optimal reconfiguration of the ADN topology and the power output of DG units are calculated. The co-optimization considers investment decisions at each planning stage and operation strategies at each ADN loading scenario. The benefits of introducing DG and network construction as planning options, and DG supply and network reconfiguration as operation strategies, are discussed. The effectiveness of the proposed co-optimization of investment-operation is demonstrated using the numerical results.

Study on microgrid operation modes switching based on eigenvalue analysis

The purpose of this paper is to investigate the characteristics of the switching caused by the microgrid transitions based on the eigenvalue analysis. The factors, including control parameters, line parameters and load parameters, which have impact on the eigenvalues of the two operation modes of the microgrid are studied. In addition, the parameter sensitivity analysis for the dominant eigenvalues is carried out under different load levels in two operation modes of microgrid. The results show: the line parameters and the load parameters have less impact on the stability of the microgrid system than control parameters in either the grid-connected mode or islanded mode; the parameter sensitivity of kpi, the proportional parameter of inner-loop current controller, for the dominant eigenvalue is increased after the microgrid switches from grid-connected mode to autonomous mode. This may cause the microgrid instable in the autonomous mode even if transition from a stable grid-connected mode. Therefore, the parameter kpi should be adjusted to a suitable value to maintain the microgrid stable during the switching, once the island state of microgrid is detected. By comparing the change of the eigenvalues distribution in the modes switching of micro-grid, some smooth-switching control strategies are proposed. Finally several time-domain simulations are present to verify the above result.

Online clustering modeling of large-scale photovoltaic power plants

This paper presents an online clustering modeling method for large-scale photovoltaic (PV) power plants. The proposed method utilizes the defined feature distance of inverter control parameters as the clustering index to derive the equivalent PV plant model. Based on the offline parameter database and the online matching method, the feature distance weighted by online parameter sensitivity is obtained to cluster PV generation units by using the K-means clustering algorithm. The method to acquire equivalent parameters of each clustered PV model is also presented. Simulation results show that the proposed online modeling method is effective and can track the dynamic characteristics of PV power plants accurately.

The Impact of Wake Effect on the Aggregated Modeling of Wind Farm

With the increasing scale of wind generations, the wind farms begin to influence power system. To facilitate the investigation of impact of a wind farm on the power system, an integrated model for large wind farm is necessary. The Jenson model and Lissaman model and the wind farm aggregated model based on the eigenvalues analysis are applied in the simulation system to analyze the impact of wind farm on the active power output of whole wind farm and terminal voltage of wind generators and the rotor dynamic behavior of wind generators.

A stability enhancement method for inverter-based distributed generation systems

Large-scale integration of distributed generations (DGs) may degrade the stability of the power distribution system. This paper proposes a control method to improve the system stability of grid-connected inverter-based distributed generation (DG) systems. The proposed method is to add a stabilizer inside the inverter interface controller of a DG unit to supply auxiliary current reference signal for the inverter current controller and thus improve the stability performance. The input signal of the stabilizer is the grid frequency. A series of small-signal analysis and time-domain simulations have been conducted to verify the effectiveness of the proposed stabilizer. The oscillations of the DG system are greatly damped with the stabilizer.

Online clustering modeling of photovoltaic power plant with LVRT control function

This paper proposes an online clustering modeling method for photovoltaic (PV) power plant considering low voltage ride-through (LVRT) control. The proposed method takes the distance of the unit impulse response curves of the inverter controllers as the clustering index to build the clustering models of the inverter controllers. After all LVRT controllers are simplified and described by an equivalent LVRT model, the clustering equivalent model of the PV power plant can be established which is composed of the clustering models of inverter controllers combined with their own equivalent LVRT models. The simulation results show that the proposed method has a strong capability to track the dynamic characteristics of the PV power plant while the output distribution of PV units changes.

Comprehensive coordinated model of active distribution network planning

Active distribution network (ADN) has gradually become a trend of future distribution network. The increasing distributed generations (DGs), demand-side controllable intelligent devices, energy storage devices and advanced secondary equipment are not fully considered in traditional planning methods. Key elements of ADN planning are introduced. A modified planning process is put forward. Load forecasting with consideration of higher DGs penetration and more controllable loads is discussed. Substation site selection and grid topology optimization considering flexibility of network are studied. A comprehensive coordinated planning model of ADN primary and secondary system is proposed. A pilot project is taken as an example to validate the whole model.