Hoist is an unusually complex system, and because of its special operating conditions and environment, the knowledge of hoist fault has the characteristics of large amount of data, time-varying, coupling and data redundancy, but the realization of the current hoist fault diagnosis method still depends mainly on the traditional fault diagnosis expert system and the experience of diagnostic experts,
The stability of the diagnosis results is only based on successful diagnostic examples and experience, resulting in the lack of scientific theoretical basis for diagnostic decisions, and the existence of low diagnostic efficiency, poor knowledge management and other drawbacks, can not fully adapt to the requirements of hoist fault diagnosis tasks, mainly reflected in:
1, Because the mine hoist system structure knowledge, technical Report, maintenance manual, data report and other information dispersion, did not achieve effective management and utilization, the backwardness of knowledge management led to the division of Mine hoist knowledge, resulting in easy loss of technology and knowledge sharing difficulties;
2, The use of traditional fault diagnosis methods, mainly based on the diagnosis experience of experts to diagnose, due to the differences of individual experts, resulting in field expert knowledge has a strong randomness, in addition, because it is subjective knowledge collation, inevitably with prejudice and error, coupled with the complexity and difficulty of communicating with domain experts,
The knowledge that leads to the acquisition often has incomplete characteristics, which further results in the incomplete diagnosis Knowledge base;
3. The establishment of traditional expert system lacks the guidance of systematic theoretical knowledge, has no uniform description of the concept of fault diagnosis, the description of diagnosis problem and the treatment of diagnostic knowledge, and does not form a unified understanding and norm in the representation, acquisition and renewal of knowledge, can not fully meet the needs of system diagnostic reasoning, and has weak expression ability. can not present the complete knowledge and other defects;
4, In reality, the failure of the hoist and the cause of the failure is not one by one corresponding, so it is necessary to carry out uncertainty reasoning, the existing fault diagnosis system reasoning method is simple, the reasoning strategy is not flexible, the most important thing is that the system does not have the ability to learn from oneself, can not improve itself.
