Naoto Yorino

A solution method of unit commitment by artificial neural networks

The authors explore the possibility of applying the Hopfield neural network to combinatorial optimization problems in power systems, in particular to unit commitment. A large number of inequality constraints included in unit commitment can be handled by dedicated neural networks. As an exact mapping of the problem onto the neural network is impossible with the state of the art, a two-step solution method was developed. First, generators to be stored up at each period are determined by the network, and then their outputs are adjusted by a conventional algorithm. The proposed neural network could solve a large-scale unit commitment problem with 30 generators over 24 periods, and results obtained were very encouraging. >

A new formulation for FACTS allocation for security enhancement against voltage collapse

This paper proposes a new formulation for var planning problems, including the allocation of FACTS devices. A new feature of the formulation lies in the treatment of security issues. Different from existing formulations, we directly take account of the expected cost for voltage collapse and corrective controls, in which the control effects by the devices to be installed are evaluated together with the other controls such as load shedding in contingencies to compute an optimal var planning. The inclusion of load shedding into the formulation guarantees the feasibility of the problem. The optimal allocation by the proposed method implies that the investment is optimized, taking into account its effects on security in terms of the cost for power system operation under possible events occurring probabilistically. The problem is formulated as a mixed integer nonlinear programming problem of a large dimension. The Benders decomposition technique is tested where the original problem is decomposed into multiple subproblems. The numerical examinations are carried out using AEP-14 bus system to demonstrate the effectiveness of the proposed method.

FACTS Devices Allocation With Control Coordination Considering Congestion Relief and Voltage Stability

This paper presents an optimal allocation method for flexible ac transmission system (FACTS) devices for market-based power systems considering congestion relief and voltage stability. The purpose of the FACTS devices installation is to provide benefit for all entities accomplished by both minimizing annual device investment cost and maximizing annual benefit defined as difference between expected security cost (ESC) with and without FACTS devices installation. Different from previous approaches, the proposed method accurately evaluates the annual cost and benefits obtainable by FACTS devices installation by formulating a large-scale optimization problem that contains power flow analyses for a large number of system states representing annual power system operations. In addition, dynamic state transitions caused by specified contingencies are also simulated in the optimization problem to evaluate the effect of FACTS control actions as well as the other coordinated controls. The expected cost consists of operating cost under normal and contingency states along with their related probabilities to occur. Maximizing social welfare is the objective for normal state while minimizing compensations for generations re-scheduling and load shedding as well as maximizing social welfare are the objectives in case of contingency. Although installation cost of FACTS devices is required, they are useful as cost free means, which can reduce effectively the annual costs for generations re-scheduling and load shedding.

An interaction problem of distributed generators installed in a MicroGrid

This paper investigates an interaction problem that might be induced from various kinds of distributed generators installed in a MicroGrid. One of the most important features of MicroGrid is islanding operation. MicroGrid does not include large central generators and all distributed small generators have to share all loads existing in the MicroGrid. In this case, undesirable mutual oscillation problem might appear when MicroGrid goes into islanding mode. In this paper, MicroGrid itself and several types of distributed generator are modeled by using Matlab/Simulink. A typical case is simulated and examined by using the developed model.

An investigation of voltage instability problems

The authors investigate the voltage stability problem based on singular perturbation theory. Possible voltage instability patterns are classified into four types (I, II-1, II-2S, and II-2D) according to the mechanism causing voltage instability. Several features of each type of instability are studied as well as their analysis methods. Voltage instability tends to begin with type I and then leads to one of the remaining types. The load flow Jacobian can be an effective index to approximately assess type I and II-2S instabilities, while types II-1 and II-2D require direct nonlinear analyses and eigenvalue analyses, respectively. The validity of the classification proposed has been verified through numerical simulations and theoretical analyses which take into account dynamic characteristics of generating units, loads, and tap-changing transformers. >

A method to approximate a closest loadability limit using multiple load flow solutions

A new method is proposed to approximate a closest loadability limit (CLL), or closest saddle node bifurcation point for power systems, using a pair of multiple load flow solutions. More strictly, the obtainable points by the method are the stationary points including not only CLL but also farthest and saddle points. An operating solution and a low voltage load flow solution are used to efficiently estimate the node injections at a CLL as well as the left and right eigenvectors corresponding to the zero eigenvalue of the load flow Jacobian. They can be used in monitoring loadability margin, in identification of weak spots in a power system and in the examination of an optimal control against voltage collapse. Most of the computation time of the proposed method is taken in calculating the load flow solution pair. The remaining computation time is less than that of an ordinary load flow.

A predictor/corrector scheme for obtaining Q-limit points for power flow studies

This work proposes a new continuation power flow method tracing QV constraint exchange points (CEP), at which generators regulating voltages hit the reactive power limits. The proposed method is based on a predictor/corrector scheme to obtain CEPs in succession. The condition for Q-limit immediate instability is derived and used in the algorithm, where the stability of the obtained CEP is checked in each iteration. The point of collapse method is also combined in the algorithm to detect a saddle node bifurcation. The effectiveness of the proposed method is demonstrated through numerical examinations in IEEE 118 bus systems.

High-Speed Real-Time Dynamic Economic Load Dispatch

A large amount of renewable energy penetration may cause a serious problem in load dispatch in the future power system, where the amount of controllable generators will decrease while disturbances increase. Therefore, a new economic load dispatch (ELD) method is required in order to make the best use of the ramp-rate capability of existing generators to cope with the disturbances caused by loads as well as by renewable energy generations. This paper proposes a new dynamic ELD method to meet the general requirements for real-time use in a future power system, where load following capability is critically limited. The method is also satisfactory from an economical point of view, and is suitable for high-speed online application due to fast and steady computation time. The proposed method has been successfully tested on several systems supplying a typical morning to noon demand profile.

A generalized analysis method of auto-parametric resonances in power systems

The envelope equation derived earlier is generalized to analyze a resonance among multiple modes by using the method of multiple scales, an excellent method for analyzing large-scale nonlinear systems. It is demonstrated through numerical simulations that the generalized envelope equation is valid for predicting the resonance phenomena themselves and also for examining the critical factors responsible for severe resonances. The factors chosen for study are the heaviness of the load, small dampings, and disturbances. The method is promising for the analysis of general multimode resonances. >

Optimal Distributed Generation Allocation in Distribution Systems for Loss Minimization

An efficient analytical (EA) method is proposed for optimally installing multiple distributed generation (DG) technologies to minimize power loss in distribution systems. Different DG types are considered, and their power factors are optimally calculated. The proposed EA method is also applied to the problem of allocating an optimal mix of different DG types with various generation capabilities. Furthermore, the EA method is integrated with the optimal power flow (OPF) algorithm to develop a new method, EA-OPF which effectively addresses overall system constraints. The proposed methods are tested using 33-bus and 69-bus distribution test systems. The calculated results are validated using the simulation results of the exact optimal solution obtained by an exhaustive OPF algorithm for both distribution test systems. The results show that the performances of the proposed methods are superior to existing methods in terms of computational speed and accuracy.