Yann-Chang Huang

Optimal active-reactive power dispatch using an enhanced differential evolution algorithm

This paper presents an enhanced differential evolution (EDE) algorithm to alternatively solve the optimal active-reactive power dispatch (ARPD) problem. Theoretically, there has a coupling relation between active and reactive power dispatches. However, due to a high X/R ratio existing in the transmission line, the problem of ARPD can then be decomposed into two individual sub-problems by the decoupling concept, i.e., active and reactive power dispatch problems. In this paper, a differential evolution algorithm enhanced by combining variable scaling mutation and ant algorithm was proposed to alternatively solve the ARPD problem in a single simulation run. Due to the superiority in optimization, the EDE method is very adequate to deal with the ARPD problem. The proposed approach has been verified on the IEEE 30-bus 6-generator system. Testing results indicate the proposed EDE was superior to the existing methods in terms of active power loss, operating cost, and convergence time.

Evolving radial basic function neural network for fast restoration of distribution systems

This paper presents an optimal radial basic function (RBF) neural network for fast restoration of distribution systems under different load levels. Basically, service restoration of distribution systems is a stressful and urgent task that must be performed by system operators. In this paper, a RBF network evolved by an enhanced differential evolution (EDE) algorithm is developed to achieve the fast restoration of distribution systems. The proposed scheme comprises training data creation phase and network construction phase. In the training data creation phase, a heuristic-based fuzzy inference (HBFI) method is employed to build the restoration plans under various load levels. Then an optimal RBF network is constructed by the EDE algorithm in the network construction phase. Once the RBF network is constructed properly, the desired restoration plan can be produced as soon as the inputs are given. The proposed method has been verifiedd on a typical distribution system of the Taiwan Power Company (TPC). Results show the proposed method can provide better convergence performance and forecasting accuracy than the existing methods.

Fault Diagnosis of Power Transformers Using Rough Set Theory

This paper has presented an effective and efficient approach to extract diagnosis rules from inconsistent and redundant data set of power transformers using rough set theory. The extracted diagnosis rules can effectively reduce space of input attributes and simplify knowledge representation for fault diagnosis. The fault diagnosis decision table is first built through discretized attributes. Next, the genetic algorithm based optimization process is used to obtain the minimal reduct of symptom attributes. Finally, the rule simplification process is adapted to achieve the maximal generalized decision rules derived from inconsistent and redundant information. Experimental results demonstrate that the proposed approach has remarkable diagnosis accuracy than the existing method.

Data mining for oil-insulated power transformers: an advanced literature survey

Knowledge discovery in database and data mining (DM) have emerged as high profile, rapidly evolving, urgently needed, and highly practical approaches to use dissolved gas analysis (DGA) data to monitor conditions and faults in oil‐immersed power transformers. This study reviews different DM approaches to oil‐immersed power transformer maintenance by discussing historical developments and presenting state‐of‐the‐art DM methods. Relevant publications covering a broad range of artificial intelligence methods are reviewed. Current approaches to the latter method are discussed in the field of DM for oil‐immersed power transformers. In this paper, various DM approaches are discussed, including expert systems, fuzzy logic, neural networks, classification and decision, and hybrid intelligent‐based diagnostic systems that apply the DGA database.

A New Approach for Generator Maintenance Scheduling in Deregulated Power Systems

This paper presents a parallel-refined simulated annealing (PRSA) approach for the generator maintenance scheduling (GMS) problem to maximize profit objective function. The merit of simulated annealing (SA) is preserved and the deficiency is pruned away in the proposed PRSA. The proposed PRSA is an effective method because the multiple search trajectories parallel and refined in each stage. The near optimal solution for each trajectory in current state can be derived from the local optimal solutions by using the parallel optimization process. The proposed parallel searching strategy can obtain optimal or near-optimal solution for the GMS problem. This paper has demonstrated the effectiveness and feasibility of applying the proposed approach for the 50-unit GMS problem.

A New Algorithm for Power System Scheduling Problems

This paper presents a parallel-refined simulated annealing (PRSA) approach for the generator maintenance scheduling (GMS) problem to maximize profit objective function. The merit of simulated annealing (SA) is preserved and the deficiency is pruned away in the proposed PRSA. The proposed PRSA is an effective method because the multiple search trajectories parallel and refined in each stage. The near optimal solution for each trajectory in current state can be derived from the local optimal solutions by using the parallel optimization process. The proposed parallel searching strategy can obtain optimal or near-optimal solution for the GMS problem. This paper has demonstrated the effectiveness and feasibility of applying the proposed approach for the 50-unit GMS problem.

Vibration Fault Diagnosis of Rotating Machinery in Power Plants

This paper presents a novel data mining approach for fault diagnosis of turbine-generator units. The proposed rough set theory based approach generates the diagnosis rules from inconsistent and redundant information using genetic algorithm and process of rule generalization. In this paper, a fault diagnosis decision table is obtained from discretization of continuous symptom attributes in the data set. Then, the proposed genetic algorithm is used to achieve the minimal reduct from the discretized symptom attributes. In addition, a set of maximal generalized decision rules is obtained from the proposed rule generalization process.

Deterministic and probabilistic wind power forecasting using a hybrid method

This paper proposes a hybrid method for probabilistic wind power forecasting. The proposed approach consists of data classification, deterministic forecasting and probabilistic forecasting stages. In the data classification stage, a fuzzy k-means clustering algorithm is used to classify the historical time series of wind power into various wind classes. Several support vector regression (SVR) models that correspond to diverse wind speeds are then established to train the collected data in the deterministic forecasting stage. An enhanced harmony search (EHS) algorithm is presented to estimate the parameters for each SVR model. Using the wind speed forecasts given by Taiwan Central Weather Bureau (TCWB) for every three hours, a corresponding forecasting model is then used to produce wind power forecasts for future 3 hours in steps of 15 minutes. To assess the risk that is associated with forecasting errors, an EHS-based quantile regression (QR) method is used to provide the confidence intervals for forecasted values in the probabilistic forecasting stage. During testing on a practical wind power generation systems, the proposed method gives better forecasting accuracy and produces more reasonable confidence intervals than existing methods.

A hybrid method for one-day ahead hourly forecasting of PV power output

This paper proposes a hybrid method combining support vector regression (SVR) and fuzzy inference method for one-day ahead hourly forecasting of photovoltaic (PV) power output. The proposed method comprises training stage and forecasting stage. In the training stage, a number of SVR models are used to learn the collected input/output data sets. To achieve accurate forecast, the fuzzy inference method is used to select an adequate trained model in the forecasting stage, according to the weather information collected from Taiwan Central Weather Bureau (TCWB). The proposed approach is verified on a practical PV power generation system. Numerical results show that the proposed approach achieves better forecasting accuracy than the simple SVR and traditional artificial neural network (ANN) methods.

Optimal Power Flow Control Using a Group Search Optimizer

This paper proposes a group search optimizer (GSO) for optimal power flow (OPF) control based on a flexible AC transmission system (FACTS). FACTS has been successfully applied to steady-state control of power system, which determines the optimal location of FACTS devices and their associated values in the transmission lines. To solve the optimal solution of FACTS devices, a GSO inspired by animal searching behavior is used in this paper. GSO is a population-based optimization algorithm which has been successfully applied to deal with optimization problem. The proposed method is verified using the IEEE 30-bus 41-transmission system. The results demonstrate that the proposed method improves the total transfer capability and provides better steady-state control of power systems than existing methods.