The structural diagram knowledge base of the expert system is the knowledge storage component, which is the core of the whole system, and all operations are carried out around it. This system uses SQLServer2000 to build a knowledge base. The knowledge base contains a set of fault cases for all equipment in the air compressor system, which gathers the knowledge and experience of teaching materials, experts and maintenance personnel, and basically covers all possible failure cases of various subsystems and equipment of the air compressor. When dealing with practical problems, the expert system starts from the database, calls the corresponding knowledge in the knowledge base, and infers through the reasoning mechanism to obtain the required conclusions. The database is used to store the initial state of the problem being solved and its known facts, intermediate results of inference, and conclusions. This system uses SQLServer2000 as the underlying database server, respectively, part of the knowledge base table serial number fault specific fault name fault reason analysis fault handling method 123456 air compressor unit vibration large intercooler effect is not good air compressor exhaust temperature too high rotor dynamic balance In the bad manufacturing production, the dynamic balance impeller scale is not corrected, the dynamic balance is lost, and the dynamic balance is recalibrated; the impeller is cleaned and the rotor balance is corrected. The gear teeth mesh with the bad tooth surface and cause the vibration to overhaul. The temperature of the cooling water in the gearbox is too high. The temperature difference between the cooling water temperature and the air in the cooler is reduced, and the heat transfer amount is reduced, causing the air temperature of the cooler to rise. Reduce the cooling water temperature, use deep well water as much as possible, and strengthen the cooling measures for circulating water. Cooling area reduces rib peeling off, heat transfer area is reduced. Repair or replacement of ribbed tube air compressor. % inlet temperature is too high. Air compressor level vent temperature is too high. Air compressor level vent cooling and cooling is improved. The small amount of intercooler is fouled, resulting in too little cooling water, and the heat exchanger does not adequately clean the intercooler to remove scale and grease. Establish three databases of real-time data, historical data, and fault data. The inference engine is actually a set of programs. It starts the relevant rules in the knowledge base according to the current running state of the air compressor system, refreshes the dynamic database and saves the inference trajectory to explain the diagnosis results. In fact, it uses the knowledge of the diagnostic knowledge base. According to the signs of the running state of the equipment, the historical data of the equipment is compared, reasoned and diagnosed to solve the strategy. The interpreter can explain the behavior of the expert system to the user, including explaining the correctness of the inference conclusions and the reasons for the system to output other candidate solutions. The knowledge acquisition part is mainly used to acquire knowledge from experts and provide means for modifying and expanding the knowledge base. It has the ability to maintain knowledge consistency and extract new knowledge from a large amount of knowledge in advanced systems (machine self-learning). . The human-machine interface is used for interaction between the system and the user, allowing the user to select a fault phenomenon through a friendly graphical interface, and evaluate the diagnosis result, and confirm whether the diagnosis is saved as a valid history record <4>. Construction of the Expert System Construction of the Knowledge Base The construction of the knowledge base is a core part of the fault diagnosis expert system. Due to the complexity of knowledge and the various fault mechanisms intertwined, the construction of the knowledge base cannot be accomplished overnight, and repeated modifications are needed to gradually achieve a better diagnostic effect. Inference engine and inference process design Inference engine is divided according to control strategy. Commonly used inference methods mainly include forward reasoning, reverse reasoning and forward and reverse mixed reasoning. This system uses forward and backward hybrid reasoning. Forward Reasoning (Data-Driven Control Strategy) The basic idea of ​​forward reasoning is to use the original data or the original symptom, to use the rule forward, and to reason in the direction of the conclusion. Based on the original symptom, the inference engine searches for a rule that can match it in the knowledge base. If the match is successful, the rule is activated, and the conclusion part of the rule is added to the knowledge base as a new rule. Repeat the above process until there are no matching diagnostic rules. In the diagnostic system of forward reasoning, the user needs to organize the symptoms of the symptoms related to the diagnosis object into rules, input them into the knowledge base, and then make inferences. Backward reasoning (target-driven control strategy) The basic idea of ​​reverse reasoning is to first make assumptions and then proceed to the conclusions to find matching rules. If the match is successful, the condition of the rule is taken as an intermediate result, and then the matching rule is found until a matchable original symptom is found, which in turn is considered to be true. If it cannot be verified, it is assumed that the fault does not exist, the reasoning fails, and the hypothetical fault needs to be re-proposed. Forward and Reverse Mixed Inference Both of the above inferences are one-way inference and are two extreme methods in the control strategy. Simple forward reasoning, the purpose is not strong, the search efficiency is low; the simple backward reasoning, the initial hypothesis is blind. An effective way to solve this problem is to combine forward reasoning and backward reasoning, that is, forward and reverse mixed reasoning. Fault Diagnosis Example The fault diagnosis module first determines which type of parameter is the problem based on the alarm information. After selecting the parameter, the diagnosis can be made step by step according to the prompt of the system until the diagnosis conclusion is obtained. According to the above, the alarm message appears that the air compressor exhaust temperature exceeds the warning value, that is, the air compressor exhaust temperature is too high. After clicking the Next button, the interface as shown appears. After that, follow the prompts to make a diagnosis step by step. Light Mounts,Offset Ring Mount,Tactical Light Ring Mount,Round Disk Light Guangzhou Miaozhun Jie Trade Co.,Ltd. , https://www.focuhunter.store