Engineering Applications of Artificial Intelligence 91 (2020) 103593 Contents lists available at ScienceDirect Engineering Applications of Artificial Intelligence journal homepage: www.elsevier.com/locate/engappai Development and implementation of induction motor drive using sliding-mode based simplified neuro-fuzzy control ✩ Rabi Narayan Mishra a,∗ , Kanungo Barada Mohanty b a Department of Electrical and Electronics Engineering, Silicon Institute of Technology, Bhunaneswar, India b Department of Electrical Engineering, National Institute of Technology, Rourkela, India ARTICLE INFO Keywords: Induction motor Neuro-fuzzy sliding-mode control (NFSMC) Feedback linearization ABSTRACT This paper develops a sliding-mode based simplified structure of neuro-fuzzy speed and torque compensator incorporated with an induction motor (IM) drive deploying feedback linearization (FBL). The intuitive linearization technique with the proposed simplified structure neuro-fuzzy sliding-mode control (NFSMC) considerably improves the torque and speed responses under system uncertainty and outer load disturbance, giving optimal system performance. This proposed technique also has high computational efficiency due to single error input over conventional one and thus can easily be applied for industrial uses. The parameter tuning of the simplified neuro-fuzzy control (NFC) is done by sliding-mode control (SMC) based adaptive mechanism. The proposed simplified method based linearized drive is simulated as well as experimentally investigated using low-cost DSP2812. The responses prove that the drive system performance characteristics using proposed simplified NFSMC is well-preserved compared to that of conventional one. Additionally, it provides optimal dynamic performance and is robust in terms of parameter variations and peripheral load disturbance. 1. Introduction Industries requirement for the efficient adjustable speed drive needs to have properties like quick dynamic response, extensive torque chat- tering reduction, fast recovery of speed under load disturbance, differ- ent modes of operation and speed range, and insensitivity to system and parameter uncertainties. Though IMs have been mostly used for industrial applications, its control mechanism is difficult due to its non- linear characteristics, parameter variations and difficulty in measuring the rotor quantities (Bose, 2008). Field oriented control techniques of IM have been a great revolution for industries’ purposes. Bose (2008) and Blaschke (1972). However, many transformations and nonlinear dynamics make the controller implementation complicated (Kim et al., 1992). The robust linearization control, variable structure control were also applied successfully in power electronics and drives field (Isidori et al., 1981; Luckjiff et al., 2001; Krezminski, 1987; Chiasson, 1998; Zhang et al., 2010; Alonge et al., 2016; Lascu et al., 2004). The FBL controller simplifies the controller design as it transforms the nonlinear system to its equivalent linear system by choosing an- other proper coordinate system. Though IM with FBL shows fast dy- namic response by resolving the coupling issue, they are very prone to parameter variations, load perturbations, and system modeling error. ✩ No author associated with this paper has disclosed any potential or pertinent conflicts which may be perceived to have impending conflict with this work. For full disclosure statements refer to https://doi.org/10.1016/j.engappai.2020.103593. ∗ Corresponding author. E-mail address: rabimishra2014@gmail.com (R.N. Mishra). Though IM drive successfully found out the solution for coupling issue, leading to fast dynamic response, they are mostly affected by error in modeling, system parameter uncertainties and outer load disturbances. This gives rise to the various controllers with feedback linearized IM drive. Further, the implementation of PI controller with feedback linearization shows imperfect decoupling of flux and torque during transient condition, and weak dynamic response as the controller is sensitive to parametric uncertainties and external load disturbance. In Luckjiff et al. (2001), two linearization techniques are derived from a single control quantity. Further, in Boukas and Habetler (2004), the sensitivity analysis shows that load disturbances, measurement error and detuned parameters may lead to low controller performance. To overcome this, a robust control technique based on sliding-mode con- cept is implemented with the power electronics field and motor drives leading to optimized performance even with uncertainties of plant. But, it produces chattering effect in the system which is the major drawback of sliding mode control (SMC) (Wai et al., 2001; Liutanakul et al., 2008; Lascu et al., 2017; Swain et al., 2017). So, gradually intelligent control approaches have been evolved to fix these issues as they are based on human knowledge in operating IM and thus results in robust, error-free and independent system model (Mir et al., 1994; Orlowska-Kowalska et al., 2010; Uddin and Chy, 2010). https://doi.org/10.1016/j.engappai.2020.103593 Received 19 June 2017; Received in revised form 9 January 2020; Accepted 2 March 2020 Available online xxxx 0952-1976/© 2020 Elsevier Ltd. All rights reserved.