RESUMO
Insulators installed outdoors are vulnerable to the accumulation of contaminants on their surface, which raise their conductivity and increase leakage current until a flashover occurs. To improve the reliability of the electrical power system, it is possible to evaluate the development of the fault in relation to the increase in leakage current and thus predict whether a shutdown may occur. This paper proposes the use of empirical wavelet transform (EWT) to reduce the influence of non-representative variations and combines the attention mechanism with a long short-term memory (LSTM) recurrent network for prediction. The Optuna framework has been applied for hyperparameter optimization, resulting in a method called optimized EWT-Seq2Seq-LSTM with attention. The proposed model had a 10.17% lower mean square error (MSE) than the standard LSTM and a 5.36% lower MSE than the model without optimization, showing that the attention mechanism and hyperparameter optimization is a promising strategy.
RESUMO
BACKGROUND: Bat rabies surveillance data and risk factors for rabies spillover without human cases have been evaluated in Curitiba, the ninth biggest city in Brazil, during a 6-year period (2010-2015). A retrospective analysis of bat complaints, bat species identification and rabies testing of bats, dogs and cats has been performed using methodologies of seasonal decomposition, spatial distribution and kernel density analysis. RESULTS: Overall, a total of 1003 requests for bat removal have been attended to, and 806 bats were collected in 606 city locations. Bat species were identified among 13 genera of three families, with a higher frequency of Nyctinomops in the central-northern region and Molossidae scattered throughout city limits. Out of the bats captured alive, 419/806 (52.0%) healthy bats were released due to absence of human or animal contacts. The remaining 387/806 (48.0%) bats were sent for euthanasia and rabies testing, which resulted in 9/387 (2.32%) positives. Linear regression has shown an increase on sample numbers tested over time (regression: y = 2.02 + 0.17×; p < 0.001 and r2 = 0.29), as well as significant seasonal variation, which increases in January and decreases in May, June and July. The Kernel density analysis showed the center-northern city area to be statistically important, and the southern region had no tested samples within the period. In addition, a total of 4769 random and suspicious samples were sent for rabies diagnosis including those from dogs, cats, bats and others from 2007 to 2015. While all 2676 dog brains tested negative, only 1/1136 (0.088%) cat brains tested positive for rabies. CONCLUSION: Only non-hematophagous bats were collected during the study, and the highest frequency of collections occurred in the center-northern region of the city. Rabies spillover from bats to cats may be more likely due to the registered exposure associated with cats' innate hunting habits, predisposing them to even closer contact with potentially infected bats. Although associated with a very low frequency of rabies, cats should always be included in rabies surveillance and vaccination programs.