Efficiency optimization of induction motor drive using Artificial Neural Network

Induction motors are the workhorse of industry, have good efficiency at rated load, but long duration usage of IM at partial load shows poor efficiency which leads to waste in energy and revenue as well. These motors are reliable, robust, high power/mass ratio and economic, hence replaced all other motors in the industry, so even minute increment in induction motor efficiency can have a major impact on consumption of electricity and saving of revenue, globally. This paper utilizes, a combination of two key concepts of efficiency optimization-loss model control (LMC) and search control (SC) for efficient operation of induction motors used in various industrial applications, in aforesaid load condition. At first, to estimate optimal Ids values for various load conditions, an optimal Ids expression in terms of machine parameters and load parameters, based on machine loss model in d-q frame along with classical optimization technique, is utilized. Secondly, an offline trained artificial neural network (ANN) controller is used to reproduce the optimal Ids values, in run-time load condition. This eliminates run-time computations and perturbation for optimal flux, as in conventional SC method. The (ANN) optimal controller is designed for optimal Ids as output, while providing load torque and speed information as inputs. The training is performed in MATLAB and good accuracy of the training model is seen. Dynamic and steady-state performances are compared for proposed optimal (optimal Ids) operations and conventional vector operations (constant Ids), with the help of a simulation model, developed in MATLAB. Excellent dynamic response in load transients as well as superior efficiency performance (1- 18%) at steady-state, for a wide range of speed and torque in simulation is attained. Assimilated with similar earlier work, the proposed methodology offers effortless implementation in real-time industrial facilities, ripple free operations, fast response and higher energy savings.