Control Engineering Practice 98 (2020) 104356 Contents lists available at ScienceDirect Control Engineering Practice journal homepage: www.elsevier.com/locate/conengprac An adaptive nonlinear observer for sensorless wind energy conversion system with PMSG Aziz Watil a,∗ , Abdelmounime El Magri a , Abdelhadi Raihani a , Rachid Lajouad a , Fouad Giri b a SSDIA Laboratory, ENSET Mohammedia, Hassan II University of Casablanca, BP 159 Bd Hassan II, Mohammedia, Morocco b Normandie Univ, UNICAEN, ENSICAEN, LAC, 14000 Caen, France ARTICLE INFO Keywords: Adaptive nonlinear observer Wind energy conversion systems Synchronous aero-generators AC/DC/AC converter ABSTRACT This paper addresses the problem of designing an adaptive nonlinear observer for wind energy conversion systems (WECSs). The study targets specifically a synchronous aero-generator connected to the grid using AC/DC/AC converter. The main objective of the present work is to establish a different approach for estimating the rotor speed, position and the turbine torque. Furthermore, the proposed technique aims to provide measures of the grid voltage, considered generally as accessible. Indeed, the considered voltage values in general are supposed to be the same as given by the transformer. However, transformers are generally positioned at long distances from measuring points. Unlike previous studies, the grid frequency is considered as an unknown parameter and time varying. Thus, electrical currents are the only needed measures. This proposed strategy, considered as sensorless, is validated by a formal analysis proving that all objectives are achieved. Finally, several simulations have conclusively proven that the proposed observer enjoys additional high robustness qualities. 1. Introduction No one can deny the importance of wind energy for protecting the planet against greenhouse gases. Wind power is one of the fastest- growing renewable energy technologies. Usage is on the rise world- wide, in part because costs are falling. Many parts of the world have strong wind speeds, but the best locations for generating wind power are sometimes remote ones. Two solutions are considered: moving towards the offshore wind, which offers tremendous potential or uses nearby stations with a suitable optimization system (Council et al., 2007; van Grieken & Dower, 2017; Martin, 2010). Compared to other aero-generators, permanent magnet synchronous generator (PMSG) has many advantages such as; its simple struc- ture, high power density, high efficiency and absence of rotor losses. Moreover, PMSGs are used in variable speed wind turbines (VSWTs), allowing the generator to be directly coupled to the wind turbine, without using a gearbox which would decrease the availability of the system, increase its weight and its need for maintenance (Al Tahir et al., 2016; El Magri et al., 2013b; Elmagri, Giri, Abouloifa, & Chaoui, 2010; Giri, 2013). The problem of power efficiency maximization has been approached in recent years by different control strategies (Bianchi, De Battista, & Mantz, 2006). Different non-linear and robust control techniques have been used in the literature, for instance (Al Tahir et al., 2016; Arias et al., 2019; ∗ Corresponding author. E-mail addresses: azizwatil@gmail.com (A. Watil), magri_mounaim@yahoo.fr (A. El Magri), abraihani@yahoo.fr (A. Raihani), dsa.lajouad@gmail.com (R. Lajouad), fouadgiri@yahoo.fr (F. Giri). El Magri et al., 2013b; Giri, 2013; Lajouad, Giri, Chaoui, El Fadili, & El Magri). All cited control algorithms require feedback information about wind velocity, rotor position and speed, grid frequency and also the grid electromotive forces (EMFs). The easiest way to implement these control strategies is to install sensors in the rotor side and grid side, but this would increase the system cost and reduce the reliability of the control system and usually induce significant errors such as stochastic noise, cyclical errors, limited responsiveness and sometimes it is not possible to put sensors for every state variable. This can be due to the unavailability of sensors to measure a particular state variable. Accordingly, the state observers have been used to get online estimates of needed states using only measurements of the electrical variables (voltages and currents). For these reasons, the so-called sensor-less control has attracted a great deal of attention recently and induced an intensive research activity in control community (Giri, 2013). Indeed, many contributions on this topic have recently appeared for different types of electrical machines. Putting aside Luenberger observers that can be naturally applicable only for small variations around the operating point, Many methods have been proposed to design observers for nonlinear systems. The first one, which has met a great success, is based on the Kalman filter used as a nonlinear observer. Nevertheless, the major issue with this approach is to guarantee the boundedness of the observability https://doi.org/10.1016/j.conengprac.2020.104356 Received 30 March 2019; Received in revised form 24 February 2020; Accepted 25 February 2020 Available online xxxx 0967-0661/© 2020 Published by Elsevier Ltd.