Self-organization comprehensive real-time state evaluation model for oil pump unit on the basis of operating condition classification and recognition
| Autor: | Lu Wenqing, Xuchao Yu, Wei Liang, Laibin Zhang |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
0209 industrial biotechnology
Engineering Gaussian Aerospace Engineering Analytic hierarchy process 02 engineering and technology Fuzzy logic symbols.namesake Oil pump 020901 industrial engineering & automation Control theory 0202 electrical engineering electronic engineering information engineering Civil and Structural Engineering Basis (linear algebra) business.industry Mechanical Engineering Mixture model Backpropagation Computer Science Applications Control and Systems Engineering Signal Processing symbols 020201 artificial intelligence & image processing Artificial intelligence State (computer science) business |
| Zdroj: | Mechanical Systems and Signal Processing. 104:224-241 |
| ISSN: | 0888-3270 |
| DOI: | 10.1016/j.ymssp.2017.10.018 |
| Popis: | In oil transmission station, the operating condition (OC) of an oil pump unit sometimes switches accordingly, which will lead to changes in operating parameters. If not taking the switching of OCs into consideration while performing a state evaluation on the pump unit, the accuracy of evaluation would be largely influenced. Hence, in this paper, a self-organization Comprehensive Real-Time State Evaluation Model (self-organization CRTSEM) is proposed based on OC classification and recognition. However, the underlying model CRTSEM is built through incorporating the advantages of Gaussian Mixture Model (GMM) and Fuzzy Comprehensive Evaluation Model (FCEM) first. That is to say, independent state models are established for every state characteristic parameter according to their distribution types (i.e. the Gaussian distribution and logistic regression distribution). Meanwhile, Analytic Hierarchy Process (AHP) is utilized to calculate the weights of state characteristic parameters. Then, the OC classification is determined by the types of oil delivery tasks, and CRTSEMs of different standard OCs are built to constitute the CRTSEM matrix. On the other side, the OC recognition is realized by a self-organization model that is established on the basis of Back Propagation (BP) model. After the self-organization CRTSEM is derived through integration, real-time monitoring data can be inputted for OC recognition. At the end, the current state of the pump unit can be evaluated by using the right CRTSEM. The case study manifests that the proposed self-organization CRTSEM can provide reasonable and accurate state evaluation results for the pump unit. Besides, the assumption that the switching of OCs will influence the results of state evaluation is also verified. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect