Font Size: a A A

Research On State Awareness And Fault Restoration Methods For Active Distribution Networks

Posted on:2020-08-08Degree:DoctorType:Dissertation
Country:ChinaCandidate:J J XuFull Text:PDF
GTID:1362330611455334Subject:Electrical engineering
Abstract/Summary:PDF Full Text Request
The distribution network is linked with consumers directly,with the objective to ensure the quality of power supply and customer service,as well as improve the efficiency of the power system operation.Recently,the increasing penetration of distributed generations are all instrumental in promoting low-carbon operation of the distribution network.Meanwhile,they also bring new challenges and issues to the existing distribution network state estimation,fault location and power service restoration,etc.Under the framework and the objective of the active distribution network self-healing control,several research works such as the state awareness and the fault restoration,which are considered as the keys to ensure the security,reliability and economy of the network operation,as well as to improve the ability of renewable energy accommodation for the active distribution network.Driven by this motivation,on the basis of summarizing the research status and shortcomings of relevent technologies in the distribution network,such as the optimal meter placement,the state estimation,the fault location and restoration,novel state awareness and fault restoration methods for the security,reliability and economy operation of active distribution network have been established in this thesis.The main contributions are stated as follows:(1)The network observability is defined and analyzed separately,and then the relationship between the network observability and the real-time meter placement is deduced.After that,a mathematical model of optimal meter placement for the active distribution network considering the network observability is established,aiming at searching the minimum number of the feeder terminial unit(FTU)and the micro-phasor measurement unit(?PMU).The resultant optimization problem is addressed by using the covariance matrix adaptation evolution strategy(CMA-ES),and several alternative schemes of real-time meter placement is obtained preliminary.After that,the relationship between system power loss and locations of real-time meter placement is analyzed and quantized,and several alternative schemes of real-time meter placement are screened based on the requirement of the system power loss.The final scheme of the real-time meter placement for the active distribution network operation which can simultaneously meet the minimum number and the best location of meters is determined.(2)The uncertainties of photovoltaic(PV)generation outputs and wind turbine(WT)generation outputs are analyzed and interval modeled,and then an interval state estimation under consideration of uncertainties of system measurements for active distribution network is proposed on the basis of the well known unknown-but-bounded(UBB)theory.To achieve the estimated results precisely,a linear programming algorithm based on the iterative operation is proposed to solve this interval state estimation model.Moreover,a sparse matrix technology is mixed in order to improve the speed of this programming.(3)A new linear state estimation model for the active distribution network is established,which takes into account the multi-source real-time measurements such as the ?PMU data and the FTU data,as well as the new pseudo-measurements such as the DG outputs.In this context,the computational efficiency of the state estimation can be improved.Meanwhile,the principle and implementation process of the linear state estimation based single-phase ground faulted line identification method are presented.Demonstrating the fault point on the power line with the virtual bus,forming the extended system measurement vector,system state variable vector and bus admittance matrix based on the specific network topology,and then the trail of fault point is traversed by evaluating the augmented state estimation results.(4)The overall idea and framework of the dynamic robust restoration for the active distribution network are analyzed and introduced,and the uncertainties of DG outputs and load demands are modeled and analyzed based on the interval-affine number.After that,a two-stage dynamic robust restoration model is established,which takes maximizing the loads' power recovery as the objective function and taking the safe operation of the network as the constraints.In order to solve the mathematical model accurately,a piecewise linear approximation method based on the optimum equidistances is introduced to relax the quadratic non-linear constraints into linear ones,and then the original problem is further equivalent into a bi-level mixed integer linear programming problem according to the duality theorem.On this basis,a decomposition algorithm named column-and-constraint generation(C&CG)in a master-sub structure is utilized to solve the problem.
Keywords/Search Tags:active distributed network, self-healing control, optimal meter placement, interval state estimation, single-phase-to-ground faulted line identification, dynamic service restoration, robust optimization
PDF Full Text Request
Related items