Basit ve sağlam yapıları nedeniyle asenkron motorlar, endüstriyel uygulamalarda en yaygın elektrik motorlarıdır. Artan malzeme maliyetleri, elektrikli otomobiller ve beyaz eşya sektörü gibi alanlarda kompakt yapıda elektrik motorlarına duyulan ihtiyaç ve değişken hızlı tahriklerin yaygın kullanımı nedeniyle makinanın çeşitli kısımlarındaki sıcaklık artışının belirlenmesi büyük önem taşımaktadır. Bu çalışmada, asenkron motorlarda sürekli halde sıcaklık dağılımını ve ısıl geçici rejimi belirlemek üzere akım kaynaklı elektriksel devre benzeşimine dayanan bir ısıl model oluşturulmuştur. Isıl modeli oluşturmakta kullanılan kritik parametreler belirlenmiş ve bu parametrelerin hesabı incelenmiştir. Farklı yükleme durumları için elde edilen sonuçlar deneysel yolla elde edilen verilerle kıyaslanarak modelin doğruluğu gösterilmiştir. Asenkron motorun sürekli rejimdeki işletme büyüklüklerini hesaplayan bir tasarım programı yazılmıştır. Gerek imalatçı, gerek kullanıcı açısından motor performansının iyileştirilmesi ve maliyetin düşürülmesi büyük önem taşımaktadır. İşletme maliyetinin toplam maliyetin büyük bir kısmını oluşturması nedeniyle, işletme maliyetinin, motorda kullanılan aktif malzemelerin iyileştirilmesi ve motor tasarım değişkenlerinin optimizasyonu yoluyla azaltılması büyük önem taşır. Isıl model, asenkron motor tasarım programıyla birleştirilerek motor tasarım optimizasyonu için kullanılmıştır. Sabit güçte, mevcut tasarıma göre daha yüksek verimli, kompakt ve sıcaklık artışı minimumlaştırılmış bir motor tasarımı gerçekleştirilmiştir. Optimizasyon yöntemi olarak ısıl işlem benzeşimi kullanılmış ve asenkron motorun çok amaçlı tasarım optimizasyonu için basit ve verimli bir yapay ısıl işlem algoritması geliştirilmiştir.

Induction motors are most common electric motors for industrial application due to their simple construction and robustness. Because of increased material costs, need of compactnes for many applications and extensive use of variable speed drives, the temperature rise should be investigated in various part of an induction motor. An useful method of thermal analysis is to use an equivalent electrical circuit. By electro-thermal analogy based on wellknown Poison and Laplace equations, thermal behaviour of induction motor could represented through electrical elements, such as resistors, capacitors and independenet current sources. Induction motor can be geometrically divided to eight components which are connected to each other via thermal resistances. Each part has a bulk thermal storage. In the network, a node connotes the mean temperature of the component. All network parameters are derived from dimensional information of motor components, thermal properties of materials used in motor construction and several thermal constants such as thermal conduction and convection coefficents. Different operating power losses of motor form the current sources which represent the heat sources. The overall network can be represented as a linear set of 8 equations. A thermal model for estimating the steady-state temperature distribution and transient thermal behaviour of a 3-phase induction motor was developed in this study. Critical parameters of the model such as convection film coefficents of air gap, end cap air and frame, frame-stator yoke thermal contact coefficent were determined and their computations and experimental methods were discussed. The accuracy of the model for both steady state and transient cases was proven by comparing the calculated results that obtained from various load conditions with those of the experimental results. The temperatures at different points of motor stator and body were obtained using thermocuple sensors and infrared thermometer. The results were also verified by comparing a 2D finite element model. To show the sensivity of the model to its component, a sensivity analaysis was conducted. It is important to improve the performance and to reduce the cost of the machine from the both manufacturers and custormers point of view. Operating cost is the main part of the total cost. It is achieved by increasing motor efficiency through improving the materials and optimizing the motor design variable. The thermal model integrated with design program was used for induction motor design optimization. The aim of design optimization is to several objective functions F(x) reach their optimum values while keeping others in their acceptable limits. Simulated annealing was choosen as optimization method. This process is based on an analogy from thermodynamics where a system is slowly cooled in order to achieve its lowest energ state. Simulated annealing is an hill climbing iterative search in which sometimes the points corresponding to worse objective function values can be accepted in order to avoiding to trapped in a local minima. The algorithm employs a random search, which not only accepts changes that decrease objective function, f, but also some changes that increase it. The latter are accepted with a probability p = exp(− ΔF T) , where ΔF is the decrease in F and T is an independent control parameter that simulates the impact of thermodynamic temperature in physical annealing. A simple and efficient simulated annealing algorithm was developed to obtain multi–objective design optimization of the induction motor. The accuracy of algorithm was proven by optimizing some well-known optimization test functions. The algorithm was also tested by comparing it with the well known methods such as pattern search and adaptive simulated annealing. Some of the the selected objective functions are temperature, efficiency and weights of active parts such as stator and rotor magnetic circuits and windings. These objectives were optimized under several constraints such as start-up torque and current, nominal and pull off torques, magnetizing current, and power factor. The induction motor's steady state performance was calculated by a design program which was written as a part of this study. Only dominated solutions were accepted. A more efficient, more compact motor design with minimum temperature rise was achieved for constant power.