Study On Semi-qualitative Method Based Fault Diagnosis Approachesof Hybrid Systems

Due to the increasing complexity and riskiness, modern control systems are often subjected to unexpected changes. In recent years, in order to maintain the reliability and safety of control systems, fault diagnosis techniques went through a dynamic and rapid development and are currently becoming an important research area in automatic control field.Modern systems become more and more complex. Most of the real dynamic systems include some of the variables taking continuous values (e.g. the real-value set) and others taking discrete values (e.g. the symbols set), and these continuous variables and discrete variables may be interactive with each other, such as airplanes, automobiles, chemical processes, manufacturing plants, elevators, etc. Such complex dynamic systems can be described by hybrid systems. A hybrid system consists of continuous dynamical subsystems, discrete dynamical subsystems and interaction subsystems. With the increasement of control systems' complexity and the requirement on high safety, the fault diagnosis technique of hybrid systems has became a hot topic in recent years. In most cases, traditional quantitative model based fault diagnosis approaches suffer from the disadvantages suac as, the a priori knowledge is hardly complete, or, even if this is the case, it might be too complex to directly deal with. In order to solve these problems, qualitative model based approach can be applied. The approach is based upon the concept of qualitative model, which may be inaccurate but can be obtained easily. In the past 2 decades, qualitative model based fault diagnosis has been paid much attention. This thesis studies the semi-qualitative method based fault diagnosis approaches of hybrid systems. The main contributions of this thesis are:1. A survey on semi-qualitative method based fault detection and fault diagnosis techniques of hybrid systems is presented, which introduces the background and the objective of these techniques and discusses their advantages and disadvantages. Finally, the developing trend of the techniques is pointed out.A semi-qualitative approach to diagnosis of gradually changing faults of a class of hybrid systems is proposed. Under the framework of hybrid bond graph, some improvements are further carried out. The strong tracking filter as a part of hybrid observer is introduced to insure the hybrid behaviors being tracked more exactly between different modes. In the qualitative reasoning, an improved temporal causal graph is proposed by integrating the fuzzy set knowledge. This makes the original method be extended to diagnosis of gradually changing faults of hybrid systems, and the reasoning time is greatly reduced at the same time.On the basis of a three-tank-system, a semi-qualitative approach to diagnosis of sensor and actuator faults of a class of hybrid systems is proposed. In the qualitative reasoning, an improved temporal causal graph with additional nodes is introduced. This makes the original method be extended to diagnosis of sensor and actuator faults of hybrid systems.A semi-qualitative approach to diagnosis of multiple faults of a class of hybrid systems is proposed. Computer simulations are done for multiple faults of components, which demonstrate the effectiveness of the proposed approach.In order to solve the fault identification problem of hybrid systems, a fuzzy sets and fuzzy entropy integrated approach to identifing the faults in hybrid systems is proposed. The basic concepts of the fuzzy entropy theory are first briefly introduced, and then we describe how the fuzzy entropy theory is combined with fuzzy logic to form a new fault identification scheme. An application example, fault detection and diagnosis system for a three-tank-system is presented, and the simulation results demonstrate the effectiveness of the proposed approach, which also show that the fuzzy theory has wide application prospect in the semi-qualitative knowledge based fault diagnosis.