S.I.: EMERGING INTELLIGENT ALGORITHMS FOR EDGE-OF-THINGS COMPUTING Fault classification and detection in wind turbine using Cuckoo- optimized support vector machine A. Agasthian 1 • Rajendra Pamula 1 • L. A. Kumaraswamidhas 2 Received: 22 May 2018 / Accepted: 9 August 2018 Ó The Natural Computing Applications Forum 2018 Abstract Fault detection in wind turbine which is identified with complete system monitoring under multi-fault scenario is proposed. When a fault is detected, its types and location are recognized for easy maintenance. Fault in wind turbines is caused due to the high speed of gearbox, generator bearing and the failures occurred in various parts. In wind farm, wind turbine condition monitoring is used to reduce the maintenance cost and also improves the accuracy. Generally, in wind turbine gearbox condition monitoring using sensor is a gainful method to monitor wind turbine performance and fault. This paper nominates a method to decide the parameters for support vector machine (SVM) in wind turbine called Cuckoo search optimization (CSO). The combination of optimization technique with classification technique is evaluated. MATLAB platform was used to evaluate the various faults under fixed value and gain factor conditions. Comparing the accuracy with SVM, particle swarm optimized SVM and k-nearest neighbor, the proposed fault detection and fault isolation technique (CSO-SVM) is improved by 2.5%, 3.5% and 6.5%, respectively. The result shows the CSO model based on SVM algorithm accomplishes the most accurate fault detection than the past models. Keywords Wind turbine  Cuckoo search optimization (CSO)  Support vector machine (SVM)  Sensor fault  Particle swarm optimization (PSO)  Data mining  Neural network (NN) 1 Introduction In the previous decade, the wind power generation has grown rapidly and enormous numbers of wind turbines have been conveyed. With the development of wind power, the reliability and operation condition are the important issues which have attracted lot of attention in research. In wind turbine, a gearbox is a fundamental part, and the failure rate is low, yet the maintenance procedure is complex [1]. Gear box of wind turbine in the working environment can be extremely complex with its performance affected by dif- ferent factors such as wind speed fluctuations and load [2]. Also, a sensor is used to monitor the condition of the wind turbine. Fault detection strategy depends on moder- ately extensive data volume for model preparing, and these techniques will not be able to perceive different failure modes. Monitoring the condition and diagnosing the faults of rotating machine is essential to run the machine at high reliability without any trouble. In order to evaluate the machine, the condition monitoring approach continuously monitors the functions of rotating machines via prior data. The parts which are under the monitoring are wind farm turbines, compressors, induction motor, chemical engi- neering power plant and oil refineries. Blade fault occurs due to lightning, burning and vibration cracking [3]. A fault tolerance system consists of three essential components, namely fault detection, isolation and analysis, to ensure the reliability of the system. Other fault in electricity is not important but still is to be considered due to power device fault. Various fault detection methods such as signal analysis and artificial intelligence also exist. & A. Agasthian agasthis2003@gmail.com Rajendra Pamula rajendra@iitism.ac.in L. A. Kumaraswamidhas lakdhas1978@iitism.ac.in 1 Department of Computer Science and Engineering, Indian Institute Of Technology (ISM) Dhanbad, Dhanbad 826004, Jharkhand, India 2 Department of Mining Machinery Engineering, Indian Institute Of Technology (ISM) Dhanbad, Dhanbad 826004, Jharkhand, India 123 Neural Computing and Applications https://doi.org/10.1007/s00521-018-3690-z