Accurate network fault diagnosis in smart substations is key to strengthening grid security. To solve fault classification problems and enhance classification accuracy, we propose a hybrid optimization algorithm consisting of three parts: anti-noise processing (ANP), an improved separation interval method (ISIM), and a genetic algorithm-particle swarm optimization (GA-PSO) method. ANP cleans out the outliers and noise in the dataset. ISIM uses a support vector machine (SVM) architecture to optimize SVM kernel parameters. Finally, we propose the GA-PSO algorithm, which combines the advantages of both genetic and particle swarm optimization algorithms to optimize the penalty parameter. The experimental results show that our proposed hybrid optimization algorithm enhances the classification accuracy of smart substation network faults and shows stronger performance compared with existing methods.
Due to the difficulty of detecting traces of organic acid mixture in an aqueous sample and the complexity of resolving UV-Vis spectra effectively, a combinatory method based on a self-made radical electric focusing solid phase extraction (REFSPE) device, UV-Vis detection and partial least squares (PLS) calculation is proposed here. In this study, REFSPE was used to enhance the extraction process of analytes between the aqueous phase and the membrane phase to enrich the trace of mixed organic acid efficiently. Then, the analytes, which were eluted from the adsorption film by ethanol with the assistance of an ultrasonic cleaning machine, were detected with UV-Vis spectrophotometry. After that, the PLS method was introduced to solve the problem of overlapping peaks in UV-Vis spectra of mixed substances and to quantify each compound. The linearly dependent coefficients between the predicted value of the model and the actual concentration of the sample were all higher than 0.99. The limit values of detection for benzoic acid, phthalic acid and p-toluene sulfonic acid were found at 9.9 #22;g/L, 12.2 #22;g/L and 13.8 #22;g/L with the relative recovery values between 84.8% and 117.9%. The RSD (n = 20) values of each component are 1.17%, 1.11% and 0.86%, respectively. Therefore, the proposed combined method can determine traces of complex materials in an aqueous sample efficiently and has wonderful potential applications.