Asenkron motorlu lokomotifler için çekiş sistemi kontrolü

Abstract

Bu tez bir elektrikli lokomotif çekiş sistemi kontrolünde iki kritik kontrolörün tasarımı için yaklaşım geliştirmektedir. Bunlardan ilki elektrikli lokomotiflerde yaygın olarak kullanılan asenkron çekiş motorunun uygulamaya özgü problemlerini göz önüne alarak dayanıklı bir kontrolör tasarımı oluşturmaktır. İkinci bölümde bu amaca yönelik olarak literatürdeki çözümler taranmıştır. Bu bölümde çekiş motoru modellenmiş ve kontrolünün yapılabilmesi için akı gözlemleyicisi tasarlanmıştır. Rotor akısından oryantasyonlu kontrol yaklaşımı benimsenmiş ve bunun için bir akım kontrolörü tasarımı yapılmıştır. Akım kontrolörü olarak çekiş sistemleri dışında kendini kanıtlamış olan kompleks değişkenli akım kontrolörü temel olarak alınmıştır. Bu kontrolörün referans takibi performansı, parametre değişimine duyarsız olması, senkron frekansta bozucu etkileri giderme performansı diğer seçeneklere göre avantajlarıdır. Ancak, özellikle lokomotif çekiş sistemlerinde problem oluşturabilecek düşük frekans bozucu etkileri giderme performansı rakiplerine nazaran kötüdür. Bu problemin giderilmesi için literatürde akım kontrolörünü iki serbestlik dereceli bir kontrolör yapan aktif sönümleme direncinin geri besleme yoluna eklenmesi önerilmiştir. Böylece iç bir kontrol döngüsü ile bozucu etkiler azaltılmaktadır. Fakat, bu çözüm yüksek anahtarlama frekanslı endüstriyel uygulamalarda başarılı sonuçlar verirken, düşük örnekleme zamanına sahip düşük anahtarlama frekansında kararlılık problemlerine yol açabilmektedir. Bu sebeple normal uygulamalarda ihmal edilen gecikmeler ve doğrudan sürekli hal modelleri kullanılarak tasarlanan kontrolörler lokomotif çekiş motoru kontrolünde yetersiz kalmaktadır. Bu problemi aşmak için bu bölümde tüm gecikmeler hesaba katılarak ve ayrık zamanlı model kullanılarak aktif sönümleme direncinin sistem üzerindeki etkisi analiz edilmiştir. Buna göre motor ve evirici parametrelerine bağlı olarak sistemi kararlı kılan aktif sönümleme direnci aralığı belirlenmiştir. Bunun ötesinde bozucu etkileri en aza indirgeyebilecek optimum aktif sönümleme direnci seçimi için motor ve evirici parametrelerine bağlı analitik bir ifade bulunmuştur. Bu bölümde geliştirilen akım kontrolörü için benzetimler yapılmış ve dinamometre sistemi üzerinde deneyler yapılmıştır. Geliştirilen akım kontrolörü referans araç dışında bir şehir içi hafif raylı taşıtta ve bir dizel elektrikli lokomotifte de başarıyla uygulanmıştır. Geliştirilen kontrol yöntemi sahada, gerçek çalışma koşullarında, problemsiz bir şekilde kullanılmaktadır. Tez kapsamında geliştirilen ikinci yaklaşım kayma-kızaklama kontrolüne yöneliktir. Üçüncü bölümde kayma-kızaklama kontrol yöntemlerine dair literatür taraması yapılmıştır. Bu bölümde daha önceden geliştirilmiş ve benzetimleri başarıyla yapılmış bir yöntem hıza bağlı uyarlanabilir bir forma getirilmiştir. Düşük araç hızlarında kayma-kızaklamanın daha kolay oluştuğu bilinen bir gerçektir. Araç hızının değişiminin kayma-kızaklama kontrolü üzerindeki etkisi bu bölümde analiz edilmiştir. Maksimum adezyon kuvvetinin %90'ını elde edebilmenin yüksek hızlarda daha kolay olduğu gösterilmiştir. Motor torkuna verilen sinüzoidal bir tanıma işaretinin, motor hızında aynı frekansta oluşan işarette oluşturduğu faz kayması adezyon karakteristiği ile ilgili bilgi vermektedir. Bu kapsamda sektörde de kullanılmakta olan bu yöntemde araç hızına bağlı olarak istenen faz kayması aralığı uyarlamalı olarak değiştirilmiştir. Yapılan benzetimlerde uyarlanabilir hibrit yöntemin sıklıkla kullanılan diğer yöntemlere ve baz alınan yönteme göre daha iyi sonuçlar verdiği gözlemlenmiştir.

Rail transportation is a preferred transportation method since it is efficient, comfortable, safe and more environmentally friendly. Because of all these advantages, rail vehicles have a long history and new technological developments have been continuously implemented on these vehicles during this time. Thanks to recent technological developments, it is now possible to develop faster, more powerful and more intelligent rail vehicles. One of the most critical technological components of rail vehicles is the traction system. The traction system transforms readily available onboard energy into mechanical energy that provides tractive effort with reference to acceleration/deceleration request of the driver. The purpose of the traction control system is to try to provide the acceleration/deceleration request of driver within the system limits under external disturbances such as sudden change in energy supply or variation in wheel-rail dynamics. The traction control system must also ensure that the wheels do not excessively slip/slide when providing the required traction force/torque to avoid wear of wheels. Since the rapid control of torque is possible in electric machines, use of electric traction systems in rail vehicles is more suitable for vehicle comfort and efficiency regarding effective and high-performance wheel slip-slide control. Induction motors are the most widely used electric motor type in rail vehicle traction systems due to their advantages such as being cheaper and less maintenance compared to other options. At first sight, although traction control seems to be a motor control problem, it contains many issues specific to the application. First, the motor torque should be controlled rapidly in order to recover from a sudden wheel slip/slide condition. Also, torque control should not be affected by these sudden changes. Especially in electric rail vehicles, the vehicle traction system is in interaction with the single-phase power transmission line called catenary. Sudden changes in the catenary voltage should not affect the torque control performance. As a result of rectifying the single-phase voltage, a fluctuation of two times the frequency of the catenary voltage on the DC voltage used to drive the motor is inevitably formed. The disturbance effect from this voltage fluctuation should not affect the torque control. Locomotives have different traction system control problems amongst rail vehicles. Since it is the only vehicle that can provide traction effort in a train configuration, this vehicle is required to provide the maximum traction force and power within limits. That is why locomotives usually have traction motors on all of their axles. Rapid transportation requires higher speeds, which results in sizing traction motors more powerful. Both the high starting force and the high speed are expected from the locomotives, and the electric motors used in these vehicles are used up to their maximum achievable speeds that can only be possible using field weakening techniques. The power electronics switches required for the control of high-powered electric machines have high switching losses. Therefore, the ratio of switching frequencies to the maximum stator frequency in locomotive traction motor control is quite low compared to other standard applications. The traction on all of the axles of the locomotives makes it impossible to measure the speed of the wheel to obtain vehicle speed. When the vehicle speed cannot be measured precisely, the required slip ratio information for slip/slide control cannot be readily obtained. Some speed estimation methods are used vehicle model for road vehicles. However, because the load behind the locomotives is uncertain, speed estimation is not possible by using the vehicle model. For this reason, locomotives, especially in the case of all wheels are slipping/sliding, detecting the slip/slide condition is a problem to be solved. In the wheel slip/slide control, at low speeds, the wheel can shift to an immediate slip state, which is more difficult at high speeds. Therefore, vehicle performance and safety can be increased if the wheel slip-slide controller is designed considering the speed of the vehicle. This thesis develops approaches for the design of two critical controllers in the control of an electric locomotive traction system. The first is to create a robust controller design by considering the application-specific problems of the asynchronous traction motor commonly used in electric locomotives. In the second part, the solutions in the literature are reviewed for this purpose. In this section, traction motor is modeled by using complex vectors, and a flux observer is designed for control. Rotor field oriented control approach is adopted, and a current regulator is designed to control the motor torque. As a current regulator, a complex variable structure which is proven outside the traction applications is chosen as the basis. This controller's reference tracking performance, insensitivity to change of system parameters, and the performance of eliminating disturbance effects on synchronous frequencies are advantages over other current controller topologies. However, the performance of low-frequency disturbance rejection, which is a problem in locomotive traction systems, is worse than its competitors. In order to solve this problem, it has been proposed in the literature to add active damping resistance to the feedback path which makes the current controller a two degree of freedom controller. Hence, this inner control loop reduces the disturbing effects. However, adding active resistance results in stability problems at low switching frequency applications implying low sampling rate while delivering successful results in high switching frequency industrial applications. For this reason, neglected delays in typical applications and controllers designed using direct continuous models are insufficient to control locomotive traction motor. In order to overcome this problem, the effect of active damping resistance on the system was analyzed by taking into account all delays and using exact discrete time model. At first sight, active damping resistance value should be increased to ensure good low- frequency disturbance rejection capability. In reality, the delays in the inner control loop limit the maximum value of active resistance which ensures stability. Accordingly, motor and inverter parameters determine the active damping resistance selection range of the system. Furthermore, an analytical expression based on the motor and inverter parameters is found for the selection of the optimum active damping resistance value which could minimize the disturbing effects. In this section, simulations are carried out for the proposed current controller and experiments were performed with actual traction motor and converter on the dynamometer system. The proposed current controller has also been applied successfully in an urban light rail vehicle and a diesel-electric locomotive. The developed control method is used in the field in real working conditions without any problem. The second approach developed in the thesis is for wheel slip-slide control. In the third chapter, a literature review of slip-slide control methods is given. In this section, a previously developed and simulated method has been introduced into an adaptive form based on vehicle speed. In the method used as a groundwork, the commonly implemented methods in the sector are used in a hybrid manner which uses phase shift and wheel acceleration information. It is a known fact that slip/slide development is more natural at low vehicle speeds. The effect of the change of vehicle speed on slip-slide control is analyzed in this section. It has been shown that achieving 90% of the maximum adhesion force requires looser control goals at higher speeds. A sinusoidal perturbation added on the motor torque results in an identification signal on the motor speed at the same frequency. The phase shift of the motor speed identification signal with respect to motor torque perturbation signal gives a valuable information about the operating point on adhesion characteristics. Therefore, the desired phase shift range can be changed according to the vehicle speed. In the simulations performed, it is observed that the adaptive hybrid method gave better results than the other methods.