Fault Diagnosis And Supervisory Control For Power Transmission Systems In Smart Grids

Electric power as an important impetus energy is widely used in the industrial and agricultural production. Its utilization is the major symbol in the second industrial rev-olution. Electric power is converted from other nature energy in remote power stations. Therefore, power transmission systems (PTSs) are constructed to transmit electric power from the remote power stations to cities. Traditional PTSs are constructed by using power transmission lines to connect a lot of distributed electric substations (ESs). These power transmission lines usually suffer from faults that can lead to large area blackouts because of the natural disasters and equipment failures. However, it is difficult to avoid fault occurrences in power transmission lines. The feasible methods are that the PTSs can automatically diagnose and restore faults if the faults occur in power transmission lines. Therefore, researchers propose the new generation smart grids. Smart grids are composed of traditional PTSs and distributed control systems (DCSs). Automatic fault diagnosis and restoration are two important characteristics for smart grids. In order to enhance the stability and reliability of smart grids, the reliable and safe DCSs are of extreme importance.Supervisory control theory is widely used in discrete event systems and DCSs are typ-ical discrete event systems. Many researchers propose a lot of DCSs and supervision control protocols based on the supervisory control theory to implement the automatic fault diagnosis and restoration for smart grids. Generally, the DCSs and supervision control protocols can be evaluated by considering their structural complexity and com-munication complexity. Most of DCSs are constructed by using communication lines to connect supervisors in smart grids. However, the structures of existing DCSs have high redundancy in communication lines and only a few of studies consider the redundancy optimization. Supervisory control protocols are an important component for DCSs, their properties are affected by the number of exchanged messages. In the existing studies, they rarely consider the optimization for the number of exchanged messages in the DCSs of smart grids, especially lack for the formal simulation of DCSs and the cor-rectness verification of supervisory control protocols. Petri nets have been widely used for the modeling, analysis, and verification of discrete event systems for their compact and graphical representation and rigorous mathematical verification methods. DCSs are typical discrete event systems. Therefore, Petri nets can be used to model, analyze, and verify the DCSs and supervisory control protocols of smart grids.This thesis addresses the strategies of automatic fault diagnosis and restoration in smart grids based on Petri nets. It also considers the optimization of system structures, com-munication lines, and the numbers of exchanged messages in DCSs and supervisory control protocols. The main results of this research are presented as follows.(1) The formalized definitions for the traditional PTSs and the faults that occur in power transmission lines are proposed by using Petri nets. First, the backbone of traditional PTSs is optimized by using hierarchical optimization techniques such that it is more simple and easy to be modeled. Then, traditional PTSs are formalized by defining them as a class of Petri nets. Furthermore, the faults that occur in power transmission lines are also defined based on Petri nets by analyzing the formalized structure of traditional PTSs. Moreover, a structure of semi-intelligent PTSs that have the characteristics of manual fault diagnosis and restoration are also formalized by using Petri nets. A semi-intelligent PTS is constructed by implementing emergency power transmission lines and electric power switches (EPSs) among ESs. The EPSs are controlled by corresponding control systems. Therefore, if a fault occurs in the input lines of an ES, the electric power of the ES can be supplied through the emergency power transmission lines.(2) A scheme is proposed to design hierarchical DCSs for the given semi-intelligent PTSs. By analyzing the structure of semi-intelligent. PTSs, a scheme is proposed to design the corresponding hierarchical DCSs for semi-intelligent PTSs based on the supervisory con-trol theory. Each ES of a semi-intelligent PTS is supervised by a supervisor to perform the automatic fault diagnosis and restoration. Furthermore, the communication lines of the proposed hierarchical DCSs are optimized based on the structure optimization theory. Therefore, the optimized DCSs not only can ensure the communication between any two supervisors but also reduce the number of communication lines. Then, the structural complexity is reduced and the costs for system implementation and mainte-nance are also saved. Moreover, power storage systems are combined with the ESs in semi-intelligent PTSs. They can be used to temporarily supply electric power for the faulty ESs to avoid large area blackouts during the fault restoration.(3) Two supervisory control protocols are proposed for fault restoration. By analyz-ing the semi-intelligent PTSs and hierarchical DCSs, a supervisory control protocol is proposed to optimize the fault restoration schemes. If there are several schemes that can be used to restore a fault, the optimal one can be computed and executed by this protocol. The purpose is to ensure that the available loads of semi-intelligent PTSs can be maximally used to avoid more serious large area blackouts. Another supervisory control protocol is proposed to optimize the number of exchanged messages based on the reconfigurable technology. This protocol can reduce the number of exchanged messages by dynamically analyzing and reconfiguring messages during the fault restoration.(4) Formal modeling, simulation, and correctness verification are very important analy-sis techniques for analyzing the stability and reliability of DCSs. The proposed structure of hierarchical DCSs and two supervisory control protocols are modeled, simulated, and verified in this thesis. First, a hierarchical DCS is modeled by using Petri nets based on the supervisory control protocols. Then, the processes of fault diagnosis and restoration are dynamically analyzed by injecting faults in the electric power transmission lines and the correctness of fault restoration is verified by using the mathematical methods of Petri nets. Moreover, to facilitate researchers correctly understanding the proposed hierarchical DCSs and protocols, a hierarchical DCS is modeled by using timed au-tomata based on the supervisory control protocols. The fault diagnosis and restoration processes are intuitively simulated and verified by using UPPAAL that is a simulation and verification tool based on timed automata. Therefore, researchers can intuitively observe and analyze the fault restoration processes.