Iterative Approach to Basic Reluctance Motor Control J.C. QUADRADO Instituto Superior de Engenharia de Lisboa Rua Conselheiro Emídio Navarro PORTUGAL Abstract: - A feedback gain tuning method is proposed to control electromechanical actuators that have the characteristic applicable to iterative trial movements. The tuning algorithm is composed of an optimisation procedure based on the Complex Method and does not require any dynamic model of controlled systems. It tunes the feedback gain under an arbitrary performance index and is applicable to both linear and non-linear control laws. Furthermore, the algorithm is concise and has a few adjusting parameters. The tuning process is strait forward and allows on-site tuning for actual industrial systems. The algorithm is applied to a basic reluctance motor control, using simulations, which attest that the proposed tuning method is effective for this system. Key-Words: - Reluctance Motor, Control, Iterative Trial, Electromechanical Actuators, Complex Method. 1 Introduction It is generally difficult to exactly represent an electromechanical actuator by a mathematical model. It depends on many factors: estimated parameter errors of the model, sensing errors based on, for example, an encoder resolution, and non- linear characteristics such as magnetic saturation, backlashes between mechanically connected parts and reduction gear systems, friction and so on. On the other hand, there are the cases where, based on the knowledge of the dynamic structure of a controlled system, a specific type of controller can be used e.g. PID controller. The proposed approach in this paper decides the feedback gains of a specified controller without mathematical models. If the controller’s structure is specified, the design question is how to tune the gains to improve the response of the controlled system. For this purpose a given performance index evaluating is necessary. The desired gain tuning method should display the following features: 1 - The dynamics of controlled system can be unknown. However, if it is known, the tuning process can be improved and achieved a faster convergence; 2 – Should be applicable to any type of controllers and, therefore, has no requirement as to whether they are linear or not. But the performance index has to be known; 3 - The algorithm should be concise and have few adjusting parameters; 4 - The tuning convergence should allow on-site gain tuning. Since movements of electromechanical actuators are not exactly reproducible and measured data has some errors, the gradient calculation of a specified performance index is a good option for the gain optimisation procedure. Some authors have shown that the Complex Method [1] can be used in electromechanical devices for different purposes [2]. These approaches show that the Complex Method has a simple optimisation procedure and requires no gradient calculation in the optimisation process. In this paper a gain-tuning algorithm is presented that enables a feedback gain tuning method of electromechanical actuator control systems. The effectiveness of the proposed algorithm is also confirmed by simulation results of a basic reluctance motor control. 2 Gain Tuning Algorithm By considering the gain tuning process as an optimisation problem of a specified performance index, the performance index definition is mandatory. Using this approach an algorithm can be derived by the Complex Method. Furthermore, due to the algorithm characteristics, a scaling process should be referred. 2.1 Performance Index Considering the tune gains of a controller based on the performance index that evaluates response movements obtained by specified controller gains. If its possible to decide a control rule from the known