Research On Fault Diagnosis System For The Anode Of Aluminum Reduction Cell

The anode of aluminum reduction cell is called "the heart of the cell", which can directly or indirectly influence the production process. It also influences the economic and technical index of the production such as electric current efficiency, the total electricity energy consumption, carbon consume per ton aluminum. Thus, the study of supervising the working state of the anode and its fault diagnosis is very important, which attracts many researchers to study. It has been a focus to the relative researchers. These studies can guarantee the safety and the high work efficiency of production by predicting the fault even the prevent of the fault. The paper mainly investigates the phenomenon of, the signal character of, the diagnosis model of the reduction cell anode fault, it can establish an solid base for further development and study.To improve the accuracy of the fault diagnosis for the working state of the anode, we make use of the signal of the equidistant voltage drop for the anode leader beside traditional signal (the voltage of the cell) . The equidistant voltage drop for the anode leader is more sensitive to the anode fault. The mechanism for the fault of the working state of anode is studied in this paper. The spectrum of the sampled signals is got using the technique of amplitude-frequency analysis, then the corresponding IIR Butterworth digital filter is designed to separate the low-frequency part and the high-frequency part using digital filter design technique. Based on the results of the filtering, characteristic parameters relative to the fault of the anode working are gained.Fuzzy comprehensive evaluation method is used to diagnose the fault of the working state of the aluminium reduction cell anode. This method synthesizes symptom factors to diagnose the fault thus improves the accuracy of the diagnosis result. The determine of the character parameters, symptom factors, membership function, the fuzzy relationship between fault and symptom factors, weights and the diagnosis rules are introduced in detail.After analyzing the traditional way of anode effect prediction, this paper develops an new way to predict the anode effect: the integrative way based on fuzzy mathematic by extracting the predicting character parameters from two types signals sampled and integratingfuzzy. The principle and the process of the new way are discussed in this paper.The author designed and realized the software system of diagnosing the fault of the working state of the aluminium reduction cell anode. The software is mainly composed of three modules: monitor module, diagnosis module and anode effect predicting module. The soft platform and the development environment used are discussed briefly, then the database design and realization, multi-thread data processing and serial port communication are described in detail. The software realizes the following functions: real-time monitoring, data processing with digital filter, the history curve query and the report forms query, diagnosis of the anode state, Anode Effect prediction, and etc. The software interface is also given in this paper.