Study On A Power Grid Fault Diagnosis Method Based On Rough Set Theory And Quantum Neural Networks

Substation automation and unattended scheduling is a hot topic area in power automation today. With the continuous expansion of the scale of power grids, especially when the power grid failure is a complex failure or abnormal movements automatic devices, switches, protection exists misoperation, and refused to move because of channel interference and the loss of information, and many other factors of uncertainty, the power system response will be complex, fault diagnosis to the power grid caused a lot of problems. Therefore, it is necessary to develop a small interference by the wrong message, strong fault tolerance of fault diagnosis, scheduling staff to assist rapid fault identification, and ensure that the security and stability of the power system operation. This system guarantees safe operation and enhance power supply reliability is a very interesting study.Previous studies showed that, even fault-tolerant strong artificial neural networks, pattern recognition deal with the issue of cross-effect data are not ideal, errors in the data sample multiple faults in particular to the situation in the diagnosis credibility greatly reduced. In this paper, learn from quantum information theory, as well as in-depth study of a typical artificial neural network based on the proposed use of incentive-based multi-function quantum pattern recognition and neural network fault diagnosis. Quantum neural networks learned from quantum mechanics is related to the concept of a new type of neural network, continuously updated through different layer neuron connections, as well as the hidden layer neurons quantum intervals with a view to increasing the fault tolerance of purpose. The experiments show that the method misoperation information on the presence of incomplete data and that strong in traditional neural network features. There is a certain method of this error message fault decision table with good recognition capability to significantly improve the accuracy of fault diagnosis.On this basis, in order to enhance the speed and fault diagnosis quantum neural network to verify the diagnosis, taking into account the fault of the complexity of decision-making table, this paper can be rough set theory identification matrix reduction, as a quantum neural network diagnostic system pretreatment method. Experiments show that the method reduces the time for training and network training complexity of the diagnostic accuracy and direct the use of quantum neural network method basically the same, retained the quantum neural network fault-tolerant better performance.Finally, the typical in-depth study of existing methods of artificial intelligence, on the basis of a large-scale urban power grids for example, using VC + + and MATLAB programming based on rough sets and quantum neural network algorithm program, and its application to The grid model, computing results with other methods of artificial intelligence data for a comprehensive comparative analysis. The experimental results show that the method has higher accuracy and fault-tolerant, can be a model for effective power network analysis and diagnosis of high theory and practical value.