Cooperative-Predictive Control Of Hybrid Microgrids

Driven by energy,environment and economy in modern society,renewable energy has developed vigorously all over the world.Further,it has promoted the profound change of power system operation mode.Microgrid(MG)is considered to be one of the most potential key technologies to deal with this change.The AC-DC hybrid MG has flexible interface,containing respective power supply advantages of AC and DC systems.Therefore,AC-DC hybrid system will become a long-term form of MG.However,the AC-DC hybrid MG is interactively coupled with heterogeneous subgrids,integrating different characteristics of various distributed resources which has different time constants.In addition,all the "source" and "load" are connected through power electronic converters.The above features bring the flexibility of energy conversion,meanwhile,they also lead to significant changes in the dynamic characteristics of the system,resulting in problems such as stability issues and power sharing problem.Control technology is the key to realize the optimal allocation of distributed resources on the premise of ensuring the stable operation of the system.In view of this,this thesis focuses on the control technology of AC-DC hybrid MG.The following summarizes the research content and main contributions of this thesis.Firstly,a hybrid MG control architecture based on predictive control is proposed.The classical linear control strategy obtains poor dynamic performance due to the cascaded control loop structure.Alternatively,predictive control directly selects the optimal voltage vector through the evaluation of the cost function,achieving ultra fast dynamic performance.Therefore,combined with virtual impedance technology,droop control and secondary control,predictive provides a new solution.Furthermore,this thesis proposes an improved predictive control applied to microgrid,and verifies its feasibility.(ⅰ)Predictive control considering dead-time effect.(ⅱ)A full state variable predictive control.Finally,the effectiveness of proposed methods are verified by experimental data.Secondly,we consider a complex connection in AC and DC networks,namely meshed network.The mathematical models of internal electrical network and communication network are established respectively using graph theory.Then,the control objectives in heterogeneous subgrid are described.We revisit the classical secondary control and its working principle.On this basis,a distributed unified cooperative control is proposed.The improved consensus algorithm is used to observe the global average voltage,which eliminates the constraints of the initial value,so that the observed value has strong robustness to agents’ arrival and departure.The proposed method realizes the objectives of voltage regulation and power sharing at the same time,significantly simplifies the control architecture.It only needs the current information of its neighbours,and the parameter design is intuitive,simple and easy to implement.The mathematical model of DC microgrid is established,and the input-state stability is analyzed.Finally,we verify our method through hardware in the loop test.It is shown that the proposed method can achieve accurate voltage regulation and power sharing.It is robust to communication failure,meanwhile it is flexible to realize plug and play feature.Thirdly,the distributed cooperative control and direct predictive circulating current control for multiple parallel interlinking converters are proposed.The electric network of the whole hybrid MG is regarded as a multi-agent system,and the distributed generators and interlinking converters are defined as agents.This innovation in system structure significantly simplifies the power interaction of heterogeneous subgrids.Based on this,a distributed cooperative control of interconnected converters is proposed.Specifically,the key parameters reflecting the load condition of the distributed generation of the subgrid are transmitted to the interlinking converter through the communication network.Then,the reference value of the active power of the interconnected converter is obtained by establishing a closed-loop negative feedback mechanism.Finally,a direct predictive circulating current control is proposed,which not only realizes the fast and accurate tracking of active and reactive power,but also effectively suppresses the zero sequence circulating current.The proposed method is a completely distributed control,which has significant engineering practical value.Lastly,a distributed cooperative control based on event self trigger technique is proposed.In order to reduce the dependence of distributed control system on communication,this thesis studies event trigger mechanism.Firstly,the voltage error is defined,and the weighted quadratic form of the node voltage vector is constructed as a Lyapunov function candidate.The dynamic event trigger conditions are set by the Lyapunov stability criterion.In addition,by predicting the evolution trajectory error of node voltage,the continuous monitoring of DC bus is avoided.The hardware in the loop test structure shows that the proposed scheme significantly reduces the communication times of the distributed control system on the premise of ensuring the system stability and good control performance.