Search

This paper proposes a model‐free decision algorithm for battery energy storage system (BESS) charging/discharging using deep reinforcement learning (DRL) to regulate off‐grid frequency fluctuation. This method is novel since the frequency regulation problem is cast in an off‐grid system to a deep Q‐network framework, which avoids directly solving a specific optimization model or model‐based control. The advantage of the proposed method is that the agents learn to comprehensively make short‐term forecasts of demand and PV output, analyse trends in forecast errors, and create control signals within the framework of DRL. A reward function with two criteria, frequency target value and frequency constraint value, guides the agent to suppress the frequency fluctuation by adjusting the weight coefficient. The effectiveness of the proposed method is verified using computational simulations with the off‐grid system model and actual power consumption data. The proposed method enables controlling the BESS to satisfy the balance of supply and demand on the off‐grid, thus realizing a control method that does not require predictive information. Moreover, the frequency control can be achieved more accurately in a short computation time than the prediction‐based method using the seasonal auto‐regressive integrated moving average. [ABSTRACT FROM AUTHOR]

With increasing amounts of wind power generation installed, the steep fluctuation of wind power generation output, called ramp events, causes serious problems for power system operation. Controlling fluctuations is an important issue for increasing the amount of wind power generation as a wind farm (WF) in the future. The authors reported the scheduled operation method of WF using a battery energy storage system (BESS) and forecast data of wind power generation output. In this paper, the authors propose a new scheduled operation method of WF. In particular, we propose the application of deep reinforcement learning to decide the output schedule of WF. Moreover, we compare the conventional method, the reinforcement learning method, and the deep reinforcement learning method in terms of the number of ramp events. In addition, we calculate the reducing effect of the storage capacity of BESS. © 2021 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC. [ABSTRACT FROM AUTHOR]

Recently, the penetration of wind farm (WF) has been increasing because of the global warming and the lack of the fossil energy resource. However, WF output fluctuate due to the weather and the seasonality. Specially, the long‐term fluctuation is called "Ramp event." It's difficult for WF output to dispatch to the power grid without controlling the ramp events. This article proposes the method of scheduled operation to mitigate the ramp events and reduce the imbalance energy with the stochastic programming. In this method, it's necessary to use the WF output scenarios. However, there has never been studied the method of generating scenarios with the temporal correlation so far. Therefore, we propose the production method of generating scenarios considering the temporal correlation with copula. Finally, this article shows the validity and the effectiveness of these methods through computational simulations using the integrated WF data of Tohoku area. [ABSTRACT FROM AUTHOR]

Abstract: Recently, wind power generation are installed as a countermeasure for the environmental issue in Japan. However, wind power generation output is fluctuated due to wind condition. Moreover, it is some negative impact for system operation such as the economic load dispatching control and the load frequency control. In Japan, NEDO has started the project as a countermeasure for massive penetration of wind power generation in 2014, called “R&D Project on Grid Integration of Variable Renewable Energy.” In this paper, we proposed the output fluctuation suppression control method of wind power generation using the battery and forecast data of wind generation output. Specifically, the target value of wind power generation output is calculated to decrease the deviation output from fluctuation suppression code for ramp event using deterministic optimization method. The effectiveness of proposed method is evaluated through the computational simulation using forecast data and actual wind power generation output in Tohoku area. [ABSTRACT FROM AUTHOR]

SUMMARY In this paper, a new technique of daily peak load forecasting is proposed. At first, we propose nonlinear correction T method as improvement technique of T method that is one of the multivariate analysis techniques, called Mahalanobis-Taguchi (MT) system, suggested in Quality Control. Moreover, we examine the daily peak load forecasting technique, which is based on data in the last year on a forecasting day, by T method and nonlinear correction T method. Furthermore, we examine the technique, which is based on the data of most recent several weeks of a forecasting day, by the T method and Nonlinear correction T method. In addition, the effectiveness of the proposed method by comparing the multiple regression analysis is discussed. [ABSTRACT FROM AUTHOR]

SUMMARY Recent years, installation of photovoltaic (PV) system is active by the environmental problem and feed in tariff system. Due to PV's output fluctuation, the static var compensator (SVC) made by power electronics technology is planning to install in distribution system. The SVC is expected for maintain the voltage. But SVC differs in a theory of operation of the existing voltage controller load ratio control transformer (LRT) and step voltage regulator (SVR). Because the SVC operation makes the system impedance changes, the existing voltage controller may not operate effectively. In this study, the intelligent-control method which used the radial basis function (RBF) network is proposed. This method can reduce the effects of the SVC operation by including the performance characteristics of SVC in the conventional method. An advantage of unnecessary arrangement attention of the SVC setup node and easy cooperation with the existing equipment is obtained by applying this method. [ABSTRACT FROM AUTHOR]

A large number of dispersed generators (DGs) are expected to be installed in distribution systems. Therefore, state estimation is an important problem for stable and reliable system operation. However, it is difficult to estimate the total power of DGs connected to a load-side system from a metering spot on the distribution line because at the metering spot only the sum of the active power from various loads and DGs can be measured. In this paper, we propose an estimation method for unknown DG outputs connected to a distribution system. This method enables to estimate DG outputs by analyzing power flow data measured at one spot, using independent component analysis (ICA). The estimation by ICA needs the same number of observations as estimations. However, the observation spots are extremely limited in existing distribution systems. Thus, we propose an estimation method which makes it possible to estimate DG outputs and load changes from single observations by using known information on the load power and a priori knowledge of insolation. © 2012 Wiley Periodicals, Inc. Electr Eng Jpn, 180(1): 36-45, 2012; Published online in Wiley Online Library (). DOI 10.1002/eej.21275 [ABSTRACT FROM AUTHOR]

Particle swarm optimization (PSO) is one of the most powerful methods for solving unconstrained and constrained global optimization problems. A cluster-structured PSO with interaction and adaptation is proposed in this paper, and the cluster structure, interaction, and adaptation of the proposed PSO are analyzed by numerical simulations. The feasibility and advantages of the proposed cluster-structured PSO are demonstrated by numerical simulations using some typical global optimization test problems. © 2011 Wiley Periodicals, Inc. Electron Comm Jpn, 94(11): 9-17, 2011; Published online in Wiley Online Library (). DOI 10.1002/ecj.10379 [ABSTRACT FROM AUTHOR]

This paper presents a new form of Particle Swarm Optimization (PSO) based on the concept of tabu search (TS). In PSO, when a particle finds a local optimal solution, all of the particles gather around that one, and cannot escape from it. On the other hand, TS can escape from the local optimal solution by moving away from the best present solution. The proposed Tabu List PSO (TL-PSO) is a method for combining the strong points of PSO and TS. This method stores the history of pbest in a tabu list. When a particle has a reduced searching ability, it selects a pbest of the past from the historical values, which is used for the update. This makes each particle active, and the searching ability of the swarm makes progress. The proposed method was validated by numerical simulations with several functions that are well known as optimization benchmark problems for comparison to the conventional PSO method. © 2010 Wiley Periodicals, Inc. Electr Eng Jpn, 172(4): 31–37, 2010; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/eej.20966 [ABSTRACT FROM AUTHOR]

This paper presents construction of Lyapunov functions for power systems based on solving the Linear Matrix Inequality (LMI) derived from the Lyapunov stability theorem considering the dynamics of load characteristic and AVR control system. The proposed Lyapunov function is constructed as a quadratic form of state variables and an integral term which satisfies the curl equation and the sector condition. An induction machine and a synchronous machine are considered as load characteristics. One-machine one-load infinite bus system is considered taking into account the flux decay effects and AVR with one time constant of the generator. To verify the proposed Lyapunov function, the transient stability assessment is shown. The critical clearing times given by the proposed Lyapunov function are compared with those obtained by the numerical integration method, and they are shown to be practical. © 2007 Wiley Periodicals, Inc. Electr Eng Jpn, 158(4): 42– 50, 2007; Published online in Wiley InterScience (www.interscience. wiley.com). DOI 10.1002/eej.20328 [ABSTRACT FROM AUTHOR]

This paper presents a systematic procedure of constructing a Lyapunov function for power systems with induction machine dynamics characteristics by the state function method. Constructing a Lyapunov function for power systems with induction machines is a very difficult problem. In the composition process of the proposed Lyapunov function, the slipping dynamic characteristic of the induction machine is implicitly expressed in the swing equation of power systems. As a result, the dynamic characteristic of an induction machine becomes equivalent to a synchronous machine with a voltage adjustment mechanism to which a dissipation function is added. Using this proposed system description, the Lyapunov function is composed by the state function method, which is based on an expanded energy function, under the condition of some approximations. The effectiveness of the proposed Lyapunov function is verified numerically for multimachine power systems with induction machine loads. © 2005 Wiley Periodicals, Inc. Electr Eng Jpn, 151(1): 34–42, 2005; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/eej.10372 [ABSTRACT FROM AUTHOR]

Discusses the combination of fuzzy control and optimal control to control the cement kiln. Application of modern control theory to power system stabilization; Process of determining the optimal control; Difficulty of determining the switching-over point.