Algorithm GATS Used In Grid Fault Diagnosis

With the development of economy, customers are stricter with the power quality. In addition, with the advent of the times of large grid having the characteristics of high voltage, AC/DC inter-connection and national connecting, the requirement on fault decision and fault restore is higher. A good fault diagnosis system is very important in fault decision for operators.At present, SCADA system (Supervisory Control And Data Acquisition System) is available in dispatching center of every city of China. The RTU telemetry, tele-signalling, watt and digital data are translated to systems. Substation integrated atomization techniques combine relay protection system and power grid atomization system, and microcomputer based protection device are all digitalize. Computer network were built among dispatching center, immediate power station and substation. Oscillate data is translated to dispatching center through public telephone network or other channels. The time synchronization of the grid is implemented by means of GPS system and hence the time difference between Oscillate devices and protection devices is minimized, which insures the correctness of the power system fault diagnosis. All these above form the foundation of the fault diagnosis information sources.Hierarchical diagnosis is adapted by virtue of the characteristics of power system information source. Breaker information received first is used to form the network-topology. The network is split into active and passive networks. The new added passive components are considered to be the suspected faulty components. The faulty components can be identified if there is only one group solution. Otherwise, the groups of the suspected faulty components are considered to be the initial solutions of the optimization algorithm for the next step.As mutli-solution exists in some cases of fault diagnosis, the causes of mutli-solution are analyzed. On this base, a more reasonable objective function is proposed. The effect of each faulty component on the improved objective function is analyzed. The cases are diagnosed by means of the improved objective function and the results shows that the mutli-solution problem can be overcome.Finally, based on the improved optimization objective function of fault diagnosis, the Genetic Algorithm -Tabu Search (GATS) method is introduced for the purpose of fault diagnosis of power systems. Genetic Algorithm(GA) has good global search ability, while Tabu Search(TS) Algorithm has good local search ability. Integrating the advantages of the two algorithms, a new hybrid algorithm, named as GATS, can be obtained. Using the function of Shcaffer, the advantages of GATS are verified in this paper. The results of the tests show that the GATS algorithm is better than the conventional methods in terms of the sensitivity of the initial solution and the dependency of initial parameters. The satisfactory results can be obtained when this method is used in the event of abnormal operation of protective relay and circuit breaker, or the multiple-fault scenarios. Furthermore, the problems of the fault zone identification and algorithms used in the fault zone identification are introduced.