Elsevier

Energy Conversion and Management

Volume 88, December 2014, Pages 573-581
Energy Conversion and Management

Distributed multi-agent scheme for reactive power management with renewable energy

https://doi.org/10.1016/j.enconman.2014.09.002Get rights and content

Highlights

  • A distributed multi-agent scheme is proposed to enhance the dynamic voltage stability.

  • A control agent is designed where control actions are performed through PI controller.

  • Proposed scheme is compared with the conventional approach with DSTATCOM.

  • Proposed scheme adapts the capability of estimation and control under various operating conditions.

Abstract

This paper presents a new distributed multi-agent scheme for reactive power management in smart coordinated distribution networks with renewable energy sources (RESs) to enhance the dynamic voltage stability, which is mainly based on controlling distributed static synchronous compensators (DSTATCOMs). The proposed control scheme is incorporated in a multi-agent framework where the intelligent agents simultaneously coordinate with each other and represent various physical models to provide information and energy flow among different physical processes. The reactive power is estimated from the topology of distribution networks and with this information, necessary control actions are performed through the proposed proportional integral (PI) controller. The performance of the proposed scheme is evaluated on a 8-bus distribution network under various operating conditions. The performance of the proposed scheme is validated through simulation results and these results are compared to that of conventional PI-based DSTATCOM control scheme. From simulation results, it is found that the distributed MAS provides excellence performance for improving voltage profiles by managing reactive power in a smarter way.

Introduction

Energy distribution networks are usually passive networks in which both real and reactive power always flow from medium to lower voltage levels. In recent years, the integration of renewable energy sources (RESs) has attracted a great deal of interest due to potential environmental and economic factors. The integration of RESs into the distribution networks adds new dynamic elements due to the variabilities and inherent uncertainties in the operation of RESs. The integration of wind energy generation poses a major challenge to the power industry due to the intermittent nature of wind [1]. Presently, 30 percent of the installed wind power is still being produced by induction generator (IG)-based wind turbines (WTs) which are directly connected to the grid as distributed generation (DG) and operate at an almost fixed-speed [2]. Therefore, it is extremely important to export some reactive power to the systems to maintain voltage profile since the IG requires a source of reactive power for its operation. The lack of reactive power support can reduce the energy transfer capability from the source to the load [3], [4]. Therefore, it is essential to manage the reactive power in an efficient way in order to maintain the dynamic voltage stability of distribution networks and this can be achieved by using an intelligent control scheme in a distributed multi-agent system (MAS) framework.

The reactive power control plays an important role in maintaining the voltage profile within specified limits and can be achieved through the dynamic compensating devices. Traditionally, shunt capacitors have been employed for reactive power compensation [5] but, in the case of variable loads and a high penetration of RESs, a fixed capacitor bank may often lead to either over or under compensation. Moreover, it cannot ensure dynamic voltage recovery during low voltage (LV) conditions due to the drop in VAr supports [6]. With the growing energy ratings obtained by solid-state devices, the static synchronous compensator (STATCOM) has emerged as one of the new generation of power electronic-based shunt flexible AC transmission system (FACTS) devices [7], [8], [9]. STATCOMs connected to distribution networks are commonly known as distribution or distributed STATCOMs (DSTATCOMs) and are widely used to regulate the line voltage at the point of common coupling (PCC) by providing appropriate reactive power support.

A variety of control approaches is proposed for reactive power control with most of those involves traditional control techniques. There are a few MAS-based approaches for energy management with RESs discussed in [10], [11], [12], [13], [14], [15] and recent research studies indicate that a few agent-based techniques have been used for reactive power management. A multi-agent approach for reactive power management with coordinated voltage control strategy is proposed in [16], [17], [18] using a set of multi-objective functions. A paradigm of distributed agent-based reactive power control scheme with shunt compensation using PI controller is discussed in [19], [20], [21] for voltage regulation. In [22], [23], a multi-agent based coordinated approach for secondary voltage control with reactive power management using STATCOM is provided. Also, a few more studies based on reactive power management using multi-agent frameworks can be found in [24].

Though all approaches [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23] so far discussed in this paper are surely functional, but there is no clear indication about the estimation of reactive power. However, the estimation of reactive power is a key issue which cannot be neglected as it is essential for initiating the control actions [25]. Moreover, the approaches [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23] do not exploit how the controller will respond for estimating reactive power by adopting to changes in the operating conditions of the system and tuning the PI controller gain based on this estimated information. This is the motivation for using a distributed MAS in which an agent-based control method ensures fast and desired responses through communication to maintain voltage stability with proper estimation of reactive power under various operating conditions.

The main focus of this paper is to design a distributed multi-agent scheme for reactive power management with wind power generation to maintain dynamic voltage stability. It is mainly based on designing a PI controller for a DSTATCOM to control both the DC-link voltage and reactive current. In this MAS, each intelligent agent represents a specific physical device model with a particular definition which performs a set of specific tasks to enhance the voltage stability. When a disturbance, e.g., fault, occurs in a distribution system, the agents estimate the required amount of reactive power for the system and use this information to determine the control objectives for the DSTATCOM controller. In this paper, the gains of the proposed agent-based PI controller are tuned using the well-known Ziegler–Nichols (Z–N) method where the gain parameters can be adjusted to track the reference values. The chosen control objectives are the DC-link capacitor voltage and reactive power. The performance of the designed scheme is demonstrated on an 8-bus test distribution network under various operating conditions, such as faults, different wind energy penetration levels and change in wind speeds, to ensure dynamic voltage stability within the energy distribution network.

The rest of the paper is organized as follows: Section 2 presents the distributed MAS architecture for distribution systems; Individual activity of each intelligent agent is presented in Section 3; Section 4 discusses the performance evaluation through simulation results; and finally, the paper is concluded with a brief remark in Section 5.

Section snippets

Distributed multi-agent framework for reactive power management

The MAS is a group of multiple interacting distributed intelligent agents within an environment. The autonomous agents can act in response to physical events in the surrounding environment by exerting independent control actions over the physical models. This MAS provides an effective platform for modeling autonomous decision-making entities for power grid operation [26], [27]. In power systems, physical devices, such as DG units, control equipments, loads and transformers, are installed in a

Activities of DN agent

In this paper, the DN agent is designed to capture the dynamics of the generators connected to the power systems. The generators are considered as wind farm-based IGs, the characteristics of which change depending on the intermittent nature of wind. Dynamic models of such wind farm-based generators are essential in order to adapt their changing characteristics using DN agent which provide its generated information to the RN agent for an actual wind energy generation count. The wind farm model

Performance evaluation of distributed MAS

In order to evaluate the effectiveness of the proposed distributed MAS approach, a 8-bus distribution network as shown in Fig. 6 is considered. It operates at 25 kV/50 Hz and includes IG-based wind farm units connected at bus-3 and bus-5 with nominal capacities of 6 MW. A DS with rating of 1000 MVA, 120 kV is connected at bus-1 and electrical loads at bus-7 and bus-8. The total real and reactive power loads on the system are PL = 5.5 MW and QL = 1.2 MVAr, respectively. The system data are also given in

Conclusion

The designed distributed multi-agent infrastructure for reactive power management maintains the voltage profiles of distribution networks with RESs under various operating conditions. To implement this scheme, the designed agent-based controller with a DSTATCOM model gathers information of the reactive power estimation to meet the control objectives. Different scenarios are considered to investigate the effectiveness of the designed scheme and from the simulation results, it is clear that the

References (39)

  • X. Li et al.

    Review of building energy modeling for control and operation

    Renew Sustain Energy Rev

    (2014)
  • J.B. Ekanayake et al.

    Comparison of 5th order and 3rd order machine models for doubly fed induction generator (DFIG) wind turbines

    J Electr Power Syst Res

    (2003)
  • Orchi TF, Pota HR, Hossain MJ. Stability improvement of wind farms using shunt and series compensation. In: 22nd...
  • J. Fan et al.

    The evolution of distribution

    IEEE Power Energy Mag

    (March 2009)
  • A. Moreno-Muoz et al.

    Grid interconnection of renewable energy sources: Spanish legislation

    Energy Sustain Dev

    (2010)
  • A. Elmitwally et al.

    An approach for placement and sizing of capacitor banks in distribution networks with distributed wind generation

    Int Trans Electr Energy Syst

    (2013)
  • P. Kundur

    Power system stability and control

    (1994)
  • Y. Xu et al.

    Adaptive pi control of STATCOM for voltage regulation

    IEEE Trans Power Delivery

    (2014)
  • N. Guerrero-Rodrguez et al.

    Modelling, simulation and experimental verification for renewable agents connected to a distorted utility grid using a real-time digital simulation platform

    Energy Convers Manage

    (2014)
  • Cited by (34)

    • Optimal scheduling of a wind energy dominated distribution network via a deep reinforcement learning approach

      2022, Renewable Energy
      Citation Excerpt :

      By improving the structure of wind turbines, intermittency can be reduced by dynamically utilizing the energy stored in the wind turbine system [5,6]. Some research activities are focused on fast climbing generators [7,8], energy storage devices [9], flexible loads [10–12]. Wind farm investors can be encouraged to take certain measures by giving fluctuation penalties.

    • Optimal energy management and control aspects of distributed microgrid using multi-agent systems

      2019, Sustainable Cities and Society
      Citation Excerpt :

      Plug-in rechargeable electric vehicles (PEV) lead to additional challenges to the grid operations because of the unexpected upsurge of electrical load in a form of vehicle charging. Management of electrical power by employing the MAS technique for executing a hybrid energy organization at broader altitude is conferred in (Mbodji, Ndiaye, & Ndiaye, 2016; Rahman, Mahmud, Pota, & Hossain, 2014). Additionally, through the droop and predictive control techniques in instant power dispatch, demand of electrical energy, production of electrical power, and power distribution arrangements along with costs at the MG level are focused in (Bruni, Cordiner, Mulone, Rocco, & Spagnolo, 2015; Ghalebani & Niasati, 2018).

    • An enhanced concept of Q-power management

      2018, Energy
      Citation Excerpt :

      Some grid codes strongly suggest operating DG units close to the unity power factor as the Q-power support from the generator will not be sufficient to cover the Q-power consumption by loads following a disturbance [21]. The advancement in power electronics made it feasible to control Q-power by using an intelligent control scheme in co-ordinated grids also with renewable energy sources (e.g. Ref. [22]). A comprehensive review of voltage and Q-management of a wind farm, including classification, techniques and methodologies was presented in Ref. [23].

    View all citing articles on Scopus
    View full text