Data-driven diagnosis of high temperature PEM fuel cells based on the electrochemical impedance spectroscopy: Robustness improvement and evaluation

Utilizing machine learning techniques for data-driven diagnosis of high temperature PEM fuel cells is beneficial and meaningful to the system durability. Nevertheless, ensuring the robustness of diagnosis remains a critical and challenging task in real application. To enhance the robustness of diagnosis and achieve a more thorough evaluation of diagnostic performance, a robust diagnostic procedure based on electrochemical impedance spectroscopy (EIS) and a new method for evaluation of the diagnosis robustness was proposed and investigated in this work. To improve the diagnosis robustness: (1) the degradation mechanism of different faults in the high temperature PEM fuel cell was first analyzed via the distribution of relaxation time of EIS to determine the equivalent circuit model (ECM) with better interpretability, simplicity and accuracy;(2) the feature extraction was implemented on the identified parameters of the ECM and extra attention was paid to distinguishing between the long-term normal degradation and other faults;(3) a Siamese Network was adopted to get features with higher robustness in a new embedding. The diagnosis was conducted using 6 classic classification algorithms—support vector machine (SVM), K-nearest neighbor (KNN), logistic regression (LR), decision tree (DT), random forest (RF), and Naive Bayes employing a dataset comprising a total of 1935 collected EIS. To evaluate the robustness of trained models: (1) different levels of errors were added to the features for performance evaluation;(2) a robustness coefficient (Roubust_C) was defined for a quantified and explicit evaluation of the diagnosis robustness. The diagnostic models employing the proposed feature extraction method can not only achieve the higher performance of around 100% but also higher robustness for diagnosis models. Despite the initial performance being similar, the KNN demonstrated a superior robustness after feature selection and re-embedding by triplet-loss method, which suggests the necessity of robustness evaluation for the machine learning models and the effectiveness of the defined robustness coefficient. This work hopes to give new insights to the robust diagnosis of high temperature PEM fuel cells and more comprehensive performance evaluation of the data-driven method for diagnostic application.

Water quality evaluation based on improved fuzzy matter-element method

For natural water, method of water quality evaluation based on improved fuzzy matter-element evaluation method is presented. Two important parts are improved, the weights determining and fuzzy membership functions. The coefficient of variation of each indicator is used to determine the weight instead of traditional calculating superscales method. On the other hand, fuzzy matter-elements are constructed, and normal membership degrees are used instead of traditional trapezoidal ones. The composite fuzzy matter-elements with associated coefficient are constructed through associated transformation. The levels of natural water quality are determined according to the principle of maximum correlation. The improved fuzzy matter-element evaluation method is applied to evaluate water quality of the Luokou mainstream estuary at the first ten weeks in 2011 with the coefficient of variatiola method determining the weights. Water quality of Luokou mainstream estuary is dropping from level I to level II. The results of the improved evaluation method are basically the same as the official water quality. The variation coefficient method can reduce the workload, and overcome the adverse effects from abnormal values, compared with the traditional calculating superscales method. The results of improved fuzzy matter- element evaluation method are more credible than the ones of the traditional evaluation method. The improved evaluation method can use information of monitoring data more scientifically and comprehensively, and broaden a new evaluation method for water quality assessment.

Improved AHP Method and Its Application in the Evaluation of Agricultural Mechanization

On the basis of analyzing the shortage of scale measure of the current AHP, and following base principIe of two comparison: principle of reward-punishment and principle of conversion, translation function and conciliation coefficient have been structured. This article advances a new measure which structuring translation function and conciliation coefficient. The result of application of improved AHP on the synthesis evaluation of agricultural mechanization is satisfactory.