Research On Distributed And Intelligent Three-Phase Parallel Inverters System Based On Multi-Agent Theory

Despite the completely rapid development of multi-agent theory and its application in the field of power system, multi-agent technique has not been investigated in depth in parallel inverters system. Based on three-phase parallel inverters system without mutual communication, multi-agent theory is applied in this paper, aiming to actualize flexible power weighting distribution with different power ratings in multi-inverter system.Modeling of three-phase single inverter based on generalized state space averaging method is analyzed in detail in the first instance. In succession, PI dual-loop controller is further designed based on generalized state space averaging model. Moreover, inductor current moving averaging method based on generalized state space averaging model is presented, adjusting the PI controller applicable to three-phase unbalance and nonlinear loads. Finally, the contrastive experiments demonstrate the accuracy of the proposed model, and the proposed inductor current moving averaging method could track the change of current quickly so as to preferably improve the dynamic characteristics under different load switching.In order to combine three-phase parallel inverters system with multi-agent technique, on the basis of the proposed feedback linearization control strategy based on generalized state space averaging model, control strategy of power weighting distribution is further investigated. Additionally, the power weighting distribution control strategy for three-phase parallel inverters system based on multi-agent theory is illustrated minutely, and then working principle of power weighting distribution for three-phase parallel inverters system based on multi-agent technique is further depicted. Afterwards, the modeling method and parameters optimization strategy are ulteriorly presented. Ultimately, experimental results validate the effectiveness of the proposed feedback linearization control strategy, power weighting distribution control strategy, modeling method and parameters optimization strategy.Stability and dynamic performance of parallel inverters system principally depend on the choice of PQ droop coefficients. The static and dynamic analysis method for three-phase parallel inverters system based on multi-agent theory is put forward in this paper. Small signal model of three-phase parallel inverters system based on multi-agent technique is presented in the first instance. Moreover, the droop coefficients are selected reasonably by means of root locus analysis. In the end, experimental results perfectly indicate that the designed droop coefficients could preferably improve the dynamic performance of parallel inverters system.Aiming to actualize flexibility and autonomy of parallel inverters system, the flexible power distribution strategy for three-phase parallel inverters system based on multi-agent theory is investigated in this paper. According to the different sources of power weighting coefficient distribution commands, master-slave flexible power weighting coefficients distribution and autonomous flexible power weighting coefficients distribution are illustrated respectively in this paper. Experimental results depict that these two distribution ways based on multi-agent technique are able to perfectly guarantee the flexibility and autonomy of power distribution for parallel inverters system. In addition, the dynamic performance of autonomous flexible power weighting coefficients distribution is better than that of master-slave flexible power weighting coefficients distribution...