Machine Learning for Long Cycle Maintenance Prediction of Wind Turbine.

Within Internet of Things (IoT) sensors, the challenge is how to dig out the potentially valuable information from the collected data to support decision making. This paper proposes a method based on machine learning to predict long cycle maintenance time of wind turbines for efficient management in the power company. Long cycle maintenance time prediction makes the power company operate wind turbines as cost-effectively as possible to maximize the profit. Sensor data including operation data, maintenance time data, and event codes are collected from 31 wind turbines in two wind farms. Data aggregation is performed to filter out some errors and get significant information from the data. Then, the hybrid network is built to train the predictive model based on the convolutional neural network (CNN) and support vector machine (SVM). The experimental results show that the prediction of the proposed method reaches high accuracy, which helps drive up the efficiency of wind turbine maintenance.

Existence of Multiple Solutions for a Quasilinear Biharmonic Equation.

Using three critical points theorems, we prove the existence of at least three solutions for a quasilinear biharmonic equation.

Alzheimer's disease phenotypes and genotypes associated with mutations in presenilin 2.

Mutations in presenilin 2 are rare causes of early onset familial Alzheimer's disease. Eighteen presenilin 2 mutations have been reported, although not all have been confirmed pathogenic. Much remains to be learned about the range of phenotypes associated with these mutations. We have analysed our unique collection of 146 affected cases in 11 Volga German families, 101 who are likely to have the same N141I mutation in presenilin 2 (54 genotyped confirmed). We have also assessed the detailed neuropathologic findings in 18 autopsies from these families and reviewed the world's literature on other presenilin 2 mutations; presenting a novel mutation that is predicted to lead to a premature truncation codon. Seven presenilin 2 mutations reported in the literature have strong evidence for pathogenicity whereas others may be benign polymorphisms. One hundred and one affected persons, with sufficient historical information from the Volga German pedigrees (N141I mutation), had a mean onset age of 53.7 years+/-7.8 (range 39-75) and mean age at death of 64.2 years+/-9.8 (range 43-88). These figures overlap with and generally fall between the results from the subjects in our centre who have late onset familial Alzheimer's disease or mutations in presenilin 1. Seizures were noted in 20 (30%) of 64 subjects with detailed medical records. Two mutation carriers lived beyond age 80 without developing dementia, representing uncommon examples of decreased penetrance. Two persons had severe amyloid angiopathy and haemorrhagic stroke. Eighteen cases had detailed histopathology available and analysed at our institution. Braak stage was five or six, amyloid angiopathy and neuritic plaques were common and more than 75% had Lewy bodies in the amygdala. TAR DNA-binding protein-43 inclusions were uncommon. In addition, a 58-year-old female with a 2 year course of cognitive decline and no family history of dementia has abnormal fludeoxyglucose-positron emission tomography imaging and a novel 2 base pair deletion in presenilin 2 at nucleotide 342/343, predicted to produce a frame-shift and premature termination. We conclude that mutations in presenilin 2 are rare with only seven being well documented in the literature. The best studied N141I mutation produces an Alzheimer's disease phenotype with a wide range of onset ages overlapping both early and late onset Alzheimer's disease, often associated with seizures, high penetrance and typical Alzheimer's disease neuropathology. A novel premature termination mutation supports loss of function or haploinsufficiency as pathogenic mechanisms in presenilin 2 associated Alzheimer's disease.