Hongshan Zhao

Fault diagnosis for substation automation based on Petri nets and coding theory

In this paper, coding-based methodology for faults monitoring of discrete event systems is applied to electric power system. Using Petri nets and coding theory to perform the fault diagnosis is further studied. One feeder of the substation is modeled by Petri net using real time discrete events, and redundant places are introduced to form structure redundancy and facilitate fault diagnosis. Based on the previous work, the feeders all possible failures are analyzed and a new method to model these failures is given. Then parity check from coding theory is used to form an encoded Petri net, therefore faults can be detected and identified through error syndrome. Method to construct important matrix, generator matrix, is presented. The simulation is simple, fast and shows very high accuracy, as combining with error correction theory. The result of simulation shows that the scheme has good performance in real-time substation fault diagnosis.

Optimization maintenance of wind turbines using Markov decision processes

As the wind power industry is getting more mature, and wind farms are taking their place as one of the mainstream options for new power generation, the demands for ensuring the wind turbines of reliable and economical operation are getting higher. Maintenance is indispensable to the core business objectives of the wind industry. Maintenance optimization is a means to determine the most cost-effective maintenance strategy. After studied the exiting maintenance method of wind turbines, this brief paper presents a method which incorporates modeling of equipment deterioration, inspection and minor & preventive maintenance for optimal maintenance decision. Using Markov decision processes, we provide an optimal cost-effective maintenance policy including maintenance actions based on the condition revealed at the time of inspection and inspection interval. Furthermore, an example is presented to illustrate the implementation of proposed method.

Modeling and analysis of power system events

The events in power systems are an important dynamic behavior, and there is no good approach of modeling and analysis at all times due to the complexity of its behaviors. In this paper, a mathematic definition of the event is given after analyzing various power system events, and also gives a model framework of power system events, which is based on formal language theory. The hybrid dynamic behavior between power system discrete control actions and physical system is analyzed by using this framework, and a cascading event sequence occurred in the 8.14 blackout is used as an example to discuss the controllable of events. Then, the synthesis of supervisor of power system discrete events is also discussed

Reliability evaluation for distribution system connected with distributed generations

This paper presents a reliability index calculation method for distribution system connected with Distributed Generations (DGs). First, the sequence model of DGs and load are established on the assumption that the time unit of power changes is an hour. Then, on the basis of the concept of segmentation, four types of data structures are presented to represent the topological structure and parameters of the distribution system. According to the differences of each segment affected by the fault, the distribution system will be divided into six types of areas, and the outage time of each load point are determined by the result of area division and load optimization in the island. At last, the synchronous sampling of DG output and load size is realized by time sequential Monte Carlo simulation and the reliability indices are calculated.

A fast multi-step prediction and rolling optimization excitation control method for multi-machine power system

Original scientific paper A fast excitation predictive control method for multi-machine power system is presented. The multi-step prediction technique is realized via system dynamic model. Some inequality constraints on states, inputs and outputs are considered in rolling optimization. The Gramian balanced reduction technique and the improved optimization algorithm are used in order to save the time of open-loop optimization in model predictive control. A 50machine power system is used to verify the effectiveness of this approach. Compared with simulated results under different controllers, this method can greatly reduce the calculating-time. The voltages of generator terminals are maintained within the set points. The stability of power system is improved.

Nonlinear power system model reduction based on empirical Gramians

Abstract An effective nonlinear model reduction approach, empirical Gramians balanced reduction approach, is studied, to reduce the computation complexity in nonlinear power system model application. The realization procedure is: firstly, computing the empirical controllable and observable Gramians matrices of nonlinear power system model, secondly, by these two matrices, computing the balance transformation matrix to obtain the balanced system model of the original model, then, computing the controllable and observable matrices of the balanced system to obtain the diagonal Hankel singular matrix. Finally, deciding the lower-order subspace to obtain the reduced power system model. A 15-machine power system model is taken as an example to perform the reduction simulation analysis.

Reliability assessment of distribution network considering preventive maintenance

Maintenance strategy impacts the reliability of distribution network directly. Common distribution network maintenance strategies contain corrective maintenance and preventive maintenance. However, the existing reliability assessment methods ignore the effects of preventive maintenance. This study showed the influence of preventive maintenance on the system failure rate by describing the preventive maintenance strategy with a reliability function and proposed a reliability evaluation method considering preventive maintenance of the distribution network. The method adopted the Sequential Monte Carlo method to simulate changes of elements lifetime and the operation condition of distribution network elements and used two kinds of clocks to simulate the state change and the reliability of elements respectively. The proposed method was used to evaluate the reliability of IEEE RBTS BUS6 system with preventive maintenance strategy. Finally, the effects of reliability threshold on the frequency and cost of maintenance and distribution network reliability index were analyzed. A reasonable range of reliability threshold is given.

Feature extraction method of transformer vibration based on ensemble empirical mode decomposition subband

To extract the abnormal features of transformer in the early stage, the envelope demodulation method based on ensemble empirical mode decomposition (EEMD) and mutual information criterion was proposed. Firstly, the transformer vibration signal is decomposed into several intrinsic mode functions (IMFs) by the EEMD method. Then the singular value decomposition (SVD) is used to estimate the number of vibration source and mutual information criterion is performed to select the IMFs. Finally, extract the vibration features of the transformer through envelope demodulation of the IMFs. The method has good adaptability and can extract the weak vibration signal. And the validity of the method is verified by analyzing the measured transformer vibration signal.

The condition monitoring of wind turbine gearbox based on cointegration

Condition monitoring of wind turbine gearbox based on cointegration by vibration signals is investigated in this research. Firstly, we establish cointegration model of gearbox in normal condition and analyze the stability of the residuals calculated by cointegration model. According to the principle of cointegration, once gearbox failure occurs, the cointegration relationship is broken and the stability of residuals changes. Therefore, we set thresholds of residuals as the early failure warning condition. Finally, through the simulation analysis of gearbox fault data, the computational results verify the effectiveness of proposed approach in monitoring condition of wind turbine gearbox.

Nonlinear power system model reduction based on empirical gramians

An effective nonlinear model reduction approach, empirical Gramians balanced reduction approach, is studied, in order to reduce the computation complexity in nonlinear power system model application. The realization procedure is: firstly, computing the empirical controllable and observable Gramians matrices, secondly, computing the balance transformation matrix to obtain the balanced system model of the original model, then, and obtaining the diagonal Hankel singular values of balance transformation matrix. Finally, deciding the lower-order subspace to obtain the reduced model. A 15-machine nonlinear power system model is taken as an example to perform the reduction simulation analysis. Simulation result shows that, under the condition to maintain the stability and input-output dynamic behavior of the original system, the order of the reduced system model could be as low as 1/5 of that of the original system model, verifying the effectiveness of the proposed approach.