The Design Of Hydroelectric Generating Sets Fault Diagnosis Systemand The Study Of Fault Diagnosis Technology

Fault is referred to as a case in which the system all or partly has departed from normal functions or has been invalidated while the real system output is incompatible to the system output expected or there is inconsistency between the system observation value and prognostication value in terms of the system model of behavior described. There are a few foundational features as followed: complexity, hierarchy, relativity, time-relayed, indefiniteness, regularity. Fault diagnosis is a kind of examination applied to system in terms of appropriate diagnosis strategies. You can get fault mode and premonition of diagnosis objects, distilling fault features. You can also estimate the subsystems or elements resulting in system and find out the reasons and characteristics resulting in these subsystems or elements departing normal functions or invalided and dope out the developing trends of the state according by predefined reasoning rules. Since the power manufacture is provided with considerable status in country economy, It is most important to ensure that the equipments of power plants run safely, stably and economically. Once the accident of Hydroelectric Generating Sets occurred, it will bring about great maintenance work and economic loss about power net, furthermore, it will affect the running of other departments of country economy and result in serious economic loss. If fault were detected in time and appropriate measures were adopted, there would be no more serious fault occurred. Fault diagnosis is a complex and high experienced technical work. There are many reasons for the faults of power equipments, so the faults are to be recognized and eliminated quickly, effectively and precisely, which is a work that deserves further study. Nowadays computer monitoring system is widely used in big power plants. The presently used monitoring system can nearly reflect the true value of the monitoring elements and giving an alarm for parameter out of limit, but when the fault happens to the engine, faced with enormous on-line monitoring data, the operators can hardly make a comprehensive analysis in a short time, it will be even harder for them to judge correctly the present state of Hydroelectric Generating Sets and to take correct measures. Hydroelectric Generating Sets Fault Diagnosis is developed based on the intentions as giving disposal measure, helping the operators to diagnose the faults and improving safety. Hydroelectric Generating Sets Fault Diagnosis System can make diagnose betimes and precisely according to the experts'suggestion, institute the measures that should be adopted, and dispose the faults that already happened and take necessary defense measures. This system chooses C++Builder as developing flat, and the whole system is based in Windows XP. The system adopts modularized designing scheme, and the independent part between different modules is controlled by the main module. There are two main modules: fault diagnosis module, and database administration module. The constitution model of the system takes database as the carrier, and the system is presented by database language. As to the other required knowledge or data, it remains the form of database. Setting up relation database concludes three parts: Firstly, setting up data constitution, that is, to indicate the data object with a series of two-dimension sheet. Secondly, restricting the two-dimension sheet, reducing redundancy, standardizing unreasonable design, hence making the database adopted efficiently. Thirdly, Operating the standardized data so as to diagnose the system easily. Standardizing the relation database is indispensable and important. A highly efficient database must have all the necessary data but withoutredundancy, which demands that we dispose every data sheet with standard. To standardize the database is, in fact, to standardize the data structure, so as to keep the integrality of the database and to minimize the data redundancy. In order to make use of the data reasonable and efficiently, we should reasonably connect the data sheet with one or more key words so as to establish the foundation for the fault diagnosis system and the data inquire and revise at usual time. The administration of the database includes the insertion of the knowledge, deletion and revision. When they want to insert, delete or revise the knowledge, the users can use the fault phenomenon window, fault type window, fault regulation window and so on. The maintenance of the database is to inspect redundancy, contradiction, and recycling, etc happened during the compilation of the database. Hydroelectric Generating Sets Fault Diagnosis System adopts downward menu under Windows standard. You can finish switching from window to window by tapping buttons. Button tapping is often used in realizing most of the function of subsystems. The Hydroelectric Generating Sets Fault Diagnosis System that the author developed is in fact a precise inference based on regulation inference. Based on this, the author analyzed the currently popular fault diagnosis technique, and introduced CBR( Case-Based Reasoning) mainly. CBR is a gradually arisen artificial intelligence methodology. It is easier to get knowledge, to understand, to enlighten thought. It has a better adaptability than traditional BBR and MBR. Based on case inference method: resolve the new problem by searching the previous case. This kind of diagnose method applies to the field in which field knowledge can hardly be regulated, but can easily beexpressed as case form, and a large amount of cases are accumulated. There are two important components for CBR: case and inference engine. The process of CBR inference has four steps: searches, reuse, revision and storage. The process of setting up CBR system includes the following techniques: CBR knowledge project, the searches of the cases, the revision of the cases and the study of the cases. Knowledge project refers to express, obtain and process knowledge. The searches of the cases contain the feature recognition, index, matches and reuse of the cases. The revision of the cases is to adjust and revise the problems so as to reflect the essential feature and to make it closer to reality. It is not only an important process of resolving problems, but also a crucial measure to study cases. In essence, it is the process of recognizing the cases and problem resolving, and new knowledge can be learned from the experience of resolving the problems. The author of the thesis developed and fulfilled regulation reference based Hydroelectric Generating Sets Fault Diagnosis System, and analyzed CBR technique, which formed a firm foundation for the future work.