Research On Transient Stable Adaptive Control For Micro-grid Based On Multi Agent Technology

Micro grid technology can manage the distributed generation, load and storage device effectively which can propel the development of new energy, promote energy conservation and emissions reduction, improve the network architecture, and increase the operation reliability and economy of the system. With the development of the micro grid technology, a large number of distributed energy and power electronic devices connects to the power distribution network, the features of distributionand low inertia makes micro grid control complex as well as reducing the system's stability, So the related stability analysis and control for micro power grid is very necessary after failure.The stability research of the micro grid based on MAS to deal with the problem of distributed advantage gives full play to the theory of MAS. Based on the operating characteristics of micro-grid connected with distributed generations(DG), this paper proposes using multi-agent technology to solve the transient stability problem of the micro-grid running of grid-island when the short-circuit fault occurs and when the breaker closes and opens with multi-agent technology. Firstly, this paper constructs a multi-agent system framework with the layered control method at its core. Then this paper designs a model for solution based on multi-agent negotiation mechanism, which describes the architecture and functions of this model.This article introduces the work content of the micro grid from multiple temporal and spatial scale after the failure based on the multi-agent system framework. Which deals with the transient stability assessment after faults and the event processing after the assessment respectively as well as proposing a three-stage transient stability control method.The first stage is the transient stability assessment: Dynamic prediction algorithm based on support vector regression machine is using to prediction micro grid will lose stability or not which adopt the way of offline training and online prediction; At the same time, in order to prepare the following coordinated action of the underlying Agent in advance, this paper trains the action priority of the underlying Agent by combining the fuzzy mathematics with artificial neural network intelligent algorithm, which adapt to the different operating conditions cause of emergency. The second stage is the phase of distribution control in the middle Agent: The middle Agent coordinate with each other by using the negotiation mechanism of the contract net, where after using multi criteria evaluation method to determine the final task allocation.Focus on the action characteristic of the underlying Agent after transient stability assessment.In order to realize quantitative control in the third stage adapting the change of running environment and their own constraints, this paper creates the switching rule of the underlying Agent, who controlsthe supply of distributed generation and the change of the load Agent by combining middle Agent concentration control instruction with the neural network training.Finally, the simulation results of the case demonstrate that the proposed control method has good validity and feasibility. The Agents realize the transient stability of micro grid by coordinatingwith each other. This topic research has very important significance in how to ensure the stable operation of the micro grid and how to maintain the transient stability after fault.