Application Of Membrame Computing In Fault Diagnosis And Economic Load Dispatch Of Power Systems

Membrane computing, also known as P systems, is a novel class of distributed parallelcomputing models, which are inspired by the structure and the functioning of the living cell aswell as from the cooperation of cells in tissues, organs. A large number of P systems andvariants have been proposed in recent years. According to the structure of membrane, Psystems can be classified into three categories: cell-like P systems, tissue-like P systems,neural-like P systems. Some scholars use P systems to design approximate optimizationcalculation in order to solve problems in practical applications. Thence, in this paper,membrane computing is attempted to be applied in fault diagnosis and economic loaddispatch (in short, ELD) of power systems. This strategy provides a new idea for faultdiagnosis and ELD, and extends the research areas of membrane computing. The mainresearch contents are shown as follows:(1) The inference algorithms and learning algorithms of adaptive fuzzy spiking neural Psystems (in short, AFSN P systems) are introduced effectively. Moreover, the matrixrepresentation of AFSN P systems is presented. It introduces how the reasoning algorithm ofAFSN P systems can be represented by the computing of matrix. Thence, the process ofdiagnosis based on AFSN P systems can be expressed by matrix successfully to improve therate of diagnosis eminently.(2) The novel AFSN P systems are applied to deal with the fault diagnosis problems ofpower systems and the uncertainty of action messages about protective relays and breakers.Firstly, search the passive regions by using the method of wiring analysis, and the faultelements will be in the passive regions. Secondly, build the corresponding fault diagnosismodels and determine the certain values of operated protective relays and tripped circuitbreakers. Thirdly, acquire the fault probability of element through fuzzy reasoning algorithmof AFSN P systems. Therefore, the fault elements can be judged via the fault probabilityof element.(3) Aiming at the problems of above fault diagnosis models, a transfer function is ledinto the rule neuron of AFSN P systems so that the diagnosis is more accurate. Furthermore,the previous fault diagnosis models based on AFSN P systems are simplified by us so that itdoes not need to re-modeling for different elements and different grids. Therefore, it has goodcapacity of the grid topology. The learning algorithm of AFSN P systems is adopted toadjust weights of AFSN P systems. Thus the diagnosis results are more close to actualsituation. Moreover, particle swarm optimization algorithm (in short, PSO) is introduced intothe learning algorithm of AFSN P systems, thus the convergence speed of diagnosis has a bigprogress. An example of4-node system is given to verify the effectiveness of this method. Compared with the existing methods, this method has faster diagnosis speed, higher accuracyand strong ability to adapt to the grid topology changes.(4) A novel artificial fish swarm algorithm based on P (in short, AFSAPS) systems isproposed to overcome falling into the local optimum solution and low optimization precisionin this paper. The algorithm combines the evolutionary rules of artificial fish swarm algorithm(in short, AFSA) with the hierarchical membrane structure and communication rules of Psystems based on the framework of AFSA and P systems. Extensive experiments withfunction optimization are investigated by the proposed algorithm. These experimental resultsshow the feasibility and effectiveness of the proposed algorithm. To reflect the practical andacademic significance of this algorithm, AFSAPS is proposed to deal with the ELD problem,and the simulation experiments are carried out on13-unit system and40-unit system.