Difficult to achieve, is still in the stage of theoretical research and argumentation.
According to the existing monitoring methods and technical conditions of the external bridge power plant, the developed fault diagnosis expert system is hereinafter referred to as the system can diagnose 106 faults that may occur in the generator and its hydrogen-oil water system, and develop the fault diagnosis method. Trend management methods and preventive measures give relevant expert advice. The expert system diagnosis is based on a total of 273 monitoring quantities including online real-time online non-real-time and offline data. The inference mechanism of forward-reverse mixed reasoning is used, and the knowledge method knowledge base based on production knowledge rules includes 1.655 million. The remaining rule sets the online real-time online non-real-time and offline data collection and analysis status monitoring fault diagnosis fault handling in the multi-task multi-client multi-information multi-information fault diagnosis expert system.
The main purpose of generator fault diagnosis is to identify the state of the generator based on various monitoring data of the generator and other test information, and to detect the fault early and take effective measures in time to prevent the problem before it is ensured. Reasonable maintenance on the basis of reliable operation of the generator.
The physical model of the system is in the supervision of the statistical analysis of the generator accidents of nearly 2, 3 to 100, Jia and above, and the preventive measures and the quality of the generator manufacturing.
On the basis of analysis and research, with reference to relevant domestic and foreign materials, the use of generator operation and maintenance test design structure and manufacturing process and many other subject areas knowledge, with reference to national standard industry standards and generator related regulations, the development of the failure mechanism causes failures The consequences of threats and preventive measures have been thoroughly analyzed. At the same time, the coupling of multi-factor and multi-type faults, the master-slave connection between faults and the possibility of concurrent development have been made, and 9 major generators and hydrogen oil water systems have been realized. A diagnosis of some of the six faults that may occur. In the process of realization, the following effective measures are taken for the characteristics of the generator fault and the difficulties in the diagnosis process, so that the contents of the physical model can be fully realized.
1.1 Comprehensive analysis of various fault precursor production practices, generator faults have a gradual character.
On the back, there is a fault behind the sudden accident, and there is even a relatively long development process. Therefore, the early diagnosis of generator faults must be integrated with the fault precursors of online real-time monitoring, and the precursors of faults obtained by online non-real-time monitoring offline test and other diagnostic information. In the inference mechanism of the system, in addition to the online real-time precursor and the online non-real-time aura as the reasoning process of the primary harbinger, the offline precursor is also considered as the primary harbinger, and these precursors are often early failures. A diagnostic method for precursors.
1.2 Combination of online and offline diagnostic methods Due to the complex structure of the generator and the limitations of existing monitoring methods, for some faults, only relying on online real-time and online non-real-time aura, only phased diagnosis results can be obtained. The final conclusion of the diagnosis can only be obtained by combining the offline precursors obtained through the test inspection after the shutdown. It can also be said that the generator failure has a certain concealment in some cases. According to this, the system adopts two methods of online and offline diagnosis, which inherit each other and inspire each other to realize comprehensive diagnosis of faults. Aspects based on online precursors including online real-time and online non-real-time precursors to derive the current diagnostic status of the generator state, and give expert advice on operational measures to enhance monitoring and prevent the progress of failures; Test the protocol and make the final diagnosis based on the resulting harbinger, and vice versa.
1.3 Multi-input multi-dimensional network structure is used to realize the reasoning process In the actual production, the occurrence and development process of generator faults are uncertain. For the same kind of fault, the first occurrence of the primary aura may be different, and the procedural precursors appear. The order is also different. In addition, there is no corresponding simple relationship between failure and aura, a combination of precursors or precursors, which may presuppose the possibility of multiple faults, the existence of faults, and the possibility of multiple precursors or multiple precursor combinations. This is the generator. The complexity of the failure. These two characteristics bring great difficulties to the inference diagnosis of the expert system, and it is easy to cause missed or misjudged. In response to this question, the system uses multi-entry multi-dimensional network structure in the fault reasoning analysis and judgment to realize the adaptive network function in the reasoning process, so that the system has the human expert to solve the practical thinking when facing different situations. ability.
1.4 Propose the concept of multi-level faults and fault-like precursors to solve faults. Some generator faults and faults are induced and compliant. It is not enough to rely on the relationship between the precursory precursor combination or the combination of the failure and the precursory precursor combination, and the relationship between the faults must be studied. In order to solve this problem, it is integrated into the reasoning and diagnosis process, which makes this difficult to solve.
1.5 Resolving the concurrency of the fault The generator sometimes has two or more faults concurrently, that is, the generator fault has concurrency, which easily leads to missed judgment.
To this end, the system uses knowledge rule algorithms or expert advice to achieve a correct diagnosis of concurrent failures.
1.6 The concept of the aura of accidents is proposed. The external causes cause the accidents caused by external causes, such as short-circuiting of the external two-phase short circuit of the generator or by the switch failure and causing other faults. To this end, the system proposes the concept of accident-like aura, and uses it as a primary aura trigger trigger inference engine to detect hidden dangers early to avoid the occurrence of induced faults, and to diagnose and propose the state of the generator after the accident. And preventive measures.
1.7 In combination with the actual, scientifically determine the threshold of the threshold of failure is the diagnostic basis of the expert system, so the threshold value of the precursory threshold is the various specifications of the manufacturer according to the national standard industry standards and other relevant regulations and domestic In recent years, the relevant experience summary, etc., after careful consideration and final determination, meets the actual needs of the site.
2.1 Generator fault diagnosis expert system software function can diagnose 106 kinds of faults that may occur in the generator of hydrogen-hydrogen-hydrogen cooling and its hydrogen-oil water system, and give relevant methods for the treatment of fault diagnosis methods and preventive measures. Expert advice. Facing the difficult practical application field of generator fault diagnosis, through the inductive rules of more than 1.65 million pieces, the domain knowledge of the design and structure of the generator operation test and the manufacturing process, especially the field experts in each The decades of practical experience and theoretical foundation and advanced experience at home and abroad are applied to the solution of practical questions.
Realize online and offline comprehensive diagnosis of multiple data sources.
According to the actual situation of Waigaoqiao Power Plant, the data source used by the system is online real-time data online non-real-time data offline data. The system can collect and process data in real time online to obtain precursors; obtain online non-real-time data that has not entered the data collection system by human-machine session, and obtain precursors through manual monitoring and judgment; and can also pass test data and test results. Offline data is a precursor. Through a variety of data sources and online diagnostic offline diagnostics to inherit the purpose of mutual inspiration diagnosis... 3 fault classification processing function. In order to enable users to prioritize when dealing with faults, the system classifies and ranks faults.
According to the relevant regulations and operating experience, the generator fault is divided into 4 levels according to the severity. It must be immediately stopped for processing. The first level should immediately reduce the load. The second level should report to the leader that the load is reduced to the third level. For level 4. When a fault of level 123 occurs, different alarm sounds and flashing of the picture image are issued according to the low level of the fault, respectively, to alert the user. If there are 2 or more trend features. In order to help users monitor generator status and analyze fault development trend at any time, the system provides generator stator winding temperature generator wire rod stator cold water outlet temperature generator lead-out line casing temperature generator stator cold water inlet pipe inlet and outlet temperature hydrogen pressure hydrogen purity There are 9 kinds of monitoring trends that need to observe the changing trend of the fault bar temperature and the corresponding bar cooling water outlet temperature fault casing temperature. At the same time, it also provides the fault bar position display and the fault bar position fault casing position, so that the user can see the fault location.
Early warning function. This system is! Multi-client expert system, in order to avoid the user missing information about the generator failure when not logging in to the system, when the expert system detects that the generator is faulty, the early warning system issues an audible alarm to all the clients, and the user logs in the fault. Diagnose expert systems for viewing and processing.
Fault recovery function. The system has a fault-recovery function according to the classification of time operator faults, which can record the whole process of inference diagnosis and user processing faults in detail, and provides a report-like printing function with the style of the record, so that the user can summarize the fault diagnosis processing regularly. Method, learn successful experience and learn the lessons of failure in troubleshooting... 7 query function. The system classifies the process data historical data in detail, provides seven kinds of classification query functions, and provides a double-like 5-style newspaper inference backtracking function for the user's query results. In order to avoid the user's negligence, the error in the answer to the human-machine session leads to the wrong diagnosis result, the inference engine of the system has a backtracking function, providing the user with the opportunity of minor modification errors.
The identification function of the working state of the sensor. Among them, the species is the automatic identification function of the expert system. It can automatically identify whether the sensor is working properly by checking and judging the sensor data. The other is that the expert system asks the user to determine whether the working condition is correct or not, and check the related meter. In case of failure, the user can submit this information to the expert system. When the expert system recognizes and confirms that the working state of a sensor is abnormal, the monitoring amount will be blocked and will not be involved in the diagnostic reasoning process until the normal state is restored.
Explain the function. In the diagnostic reasoning process, in addition to the current aura and diagnostic results, the diagnostics column of the form explains the knowledge of how to obtain this diagnosis. In addition, in the expert suggestion column, the current state of the generator is diagnosed. The measures to be taken during the operation when the final diagnosis result has not been obtained are made. The measures for the failure have been made when the final diagnosis result has been obtained. A detailed explanation.
Online help function. The system has detailed online help with 3 styles.
2.2 System network structure design Waigaoqiao Power Plant has built an information management system touch 15 and data acquisition system 0 people 3, the system directly uses the existing network structure of the power plant for diagnosis. The pre-processing and inference engine of the system runs on the database server, and the post-processing, that is, the client program runs on each workstation terminal in the power plant, 15 .
2.3 System Function Module The system adopts a hierarchical hierarchical structure, and its overall structure 2, in general, the large module is divided into the pre-processing inference engine post-processing and management module. The interface between the modules takes the form of a data interface.
Each expert system is faced with a knowledge update. Therefore, it is necessary to consider the convenience of supplemental updating of knowledge in programming.
The operating conditions of the generator are relatively complicated. The correct understanding of the development of the cause of the fault and the correct understanding of the treatment method is a long-term accumulation and continuous updating process from theoretical practice and practical theory. Especially at present, with the development of high technology, The capacity of the generator is rapidly increasing, the manufacturing process of the design material is improving day by day, and the monitoring method is gradually improved, and the fault diagnosis method should be continuously updated.
The number of data interfaces is stable, and some early and early faults that are not easy to find are found in the operation, and they are processed in time, which is highly praised by relevant personnel at all levels of the power plant.
The generator fault diagnosis expert system is a software project applied to actual production. Based on the power plant demand, the software function is designed, combined with the existing technical monitoring conditions of the power plant, and realized under the premise of not adding any hardware facilities. The diagnostic function of 106 faults that may occur in the generator is comprehensive and has great practical value.
For the purpose of this system, the system has set up fault recovery and query functions, providing first-hand information for accumulating practical experience, editor of expert system knowledge base and adaptive network function for knowledge rules and reasoning. The expansion of the machine creates convenient conditions.
The system was put into trial operation at No. 2 generator of Waigaoqiao Power Plant in December 2000. During operation, the server program runs on the database server connected to the elbow 13 network of the power plant, and collects online real-time monitoring data directly from the elbow 15 . The client is installed in the main control room, several in the power generation department, several in the maintenance department, and various levels in the power plant. Several managers. The operation results show that the generator fault diagnosis system is running well. The 2001 annual meeting of the Power System Committee jointly organized by the China Electric Engineering Society Power System Committee and the Fujian Electric Power Bureau was held in Xiamen on September 2528. boil. More than 80 experts and scholars participated in the technical exchanges and conducted technical consultations on the development of power systems in Fujian Province. More than 40 papers were presented at the conference.
The meeting was supported by many old leading experts in the electric power industry; the activities of the member units were positive and the discussion was enthusiastic; in addition to the pure academic exchanges, the specific questions on the development of the Fujian power grid were also consulted, and many valuable suggestions were put forward. Out of the road of innovation that combines development and practice.
The meeting concluded that 1 according to the development of China's power industry, the power grid should still maintain a certain development speed, grid size and Jiang Jianguo, Shi Jiayan. The state of the motor is monitored by the elbow. Beijing Water and Electric Power Press, Liu Yaxiong. Research on Generator Operation State Assessment Expert System Beijing China Electric Power Research Institute, 1997.
Cheng Daolai, Wu Hao, Lv Tingyan, et al. The status quo and development direction of domestic power station fault diagnosis system 1 Power Engineering, 1999, 1.
Zhou Keyi, Shao Aiju, Ge Cong. Overview of performance monitoring and diagnosis of foreign power stations
Yellow tree red. State-of-the-art maintenance system and development status of equipment in thermal power plants
Responsible editor Li Bo's technical level should maintain rapid development, and the development of the power market and the promotion of grid interconnection should be based on improving economic efficiency. 2 In order to do a good job in national grid interconnection, we should pay attention to the medium and long-term planning of the power system. The planning should have a sense of advancement. From the overall situation of the whole system, sensitivity analysis should be made to recognize that a reasonable grid structure is the basis for the safety and stability of the power system. 3 Select the principle of unified planning and dispatching on the basis of the scale and model of China's power market development. 4 For some networked systems, the problem should be studied and improved on the basis of sufficient analysis. 5 The current network analysis and simulation methods still have some problems, such as the load model, etc. should strive to improve.
Strategy, man-machine, squatting, man-machine, arbitrage, bamboo ç ° 处 啼 åª ç ç ç ç ç 推 推 推 推 推 å…ˆ å…ˆ å…ˆ å…ˆ å…ˆ å…ˆ å…ˆ å…ˆ å…ˆ å…ˆ å…ˆ å…ˆ
The Co2 Laser Engraving Machine is suitable for marking large non-metallic materials, such as paper packaging, plastic products, label paper, leather fabrics, glass ceramics, resin plastics, bamboo products, PCB boards, etc.
Advantages of co2 laser engraving machine processing:
1. Wide range: Carbon dioxide laser can engrave and cut almost any non-metallic material. And the price is low!
2, safe and reliable: the use of non-contact processing, will not cause mechanical extrusion or mechanical stress on the material. No [knife marks", no damage to the surface of the workpiece; no deformation of the material;
3, accurate and meticulous: processing accuracy can reach 0.02mm;
4, saving environmental protection: beam and spot diameter is small, generally less than 0.5mm; cutting processing saves materials, safety and health;
5, the effect is consistent: to ensure that the same batch of processing results are exactly the same.
6, high-speed and fast: can immediately engrave and cut according to the computer output pattern.
7, low cost: not limited by the amount of processing, for small batch processing services, laser processing is cheaper.
CO2 Laser Engraving And Cutting Machine
Co2 Laser Engraving And Cutting Machine,Co2 Laser Engraving Machine,MDF Co2 Laser Cutting Machine,Plywood Co2 Laser Cutting Machine
Jinan Vmade CNC Machine Co.,Ltd , https://www.vmadecnc.com