Yol Taşıtı Boyuna Dinamiği Kural Tabanlı Kontrolcüleri

thumbnail.default.placeholder
Tarih
2012-06-22
Yazarlar
Tunacı, Semih
Süreli Yayın başlığı
Süreli Yayın ISSN
Cilt Başlığı
Yayınevi
Fen Bilimleri Enstitüsü
Institute of Science and Technology
Özet
Bu tez çalışmasında, yol taşıtı boyuna dinamiği için kural tabanlı kontrol sistemleri tasarlanmıştır. Yol taşıtı boyuna dinamiği sistemi olarak hem CarSim hem MATLAB/Simulink yazılımları kullanılmıştır. Başarılı bir taşıt dinamiği simulasyon programı olan CarSim yazılımının, varsayılan B segmenti 5 kapılı aracı kullanılmıştır. MATLAB/Simulink programında ise boyuna taşıt dinamiği pnömatik tekerlek modeli, süspansiyon sistemi ihmal edilmiş ön ve arka yük transferi olan asılı kütle modeli ve yol taşıtlarının maruz kaldıkları direnç kuvvetleri yapıları ile modellenmiştir. MATLAB/Simulink modelinde kullanılan parametrelerin değerleri CarSim yazılımındaki varsayılan B segmenti 5 kapılı aracın parametreleri olarak alınmıştır. Oluşturulacak kontrol sisteminde, eyleyici bloğa karşılık gelen fren sistemi, hem birinci dereceden bir dinamik sistem olarak, hem de Simulink/SimHydraulics kullanılarak fren sisteminin hidrolik silindir, selenoid valfler ve eyleyicileri gibi hidrolik bileşenlerinin de modellendiği bir sistem olarak oluşturulmuştur. Fren sisteminde ayrıca master silindirden birincil ve ikincil basınç hatları olarak çıkan, ön ve arka tekerlek basınç hatları ve fren oranlaması da tanımlanmıştır. Yol taşıtları boyuna dinamiği kural tabanlı kontrol sistemlerinin kontrol parametresi olan tekerlek relatif boyuna kayma değişkeninin referans değerleri, pnömatik tekerleğin üreteceği kuvvetin kayma değeri ile doğru orantılı olduğu aralıkta belirlenmiştir. Kural tabanlı kontrolcüden hidrolik fren sistemine gidecek kontrol sinyali, referans kayma değerleri kullanılarak mantıksal işlevlerle üretilmiştir. ABS kontrol sistemi, CarSim ve MATLAB/Simulink ortamlarındaki taşıt modelleri için 50 km/saat ilk hızlı olarak yol sürtünme katsayısı 1 ve 0.5 olan iki farklı eğimsiz ve düz bir yol ortamında, sürücünün ani frenleme girdisinin olduğu frenleme senaryoları için koşturuldu. Bunun yanında aynı dinamik sistemler aynı senaryolar için aç kapa kontrolcülü ve birinci dereceden eyleyici dinamiği ile modellenen fren yapısı ile tekrar koşturularak kural tabanlı kontrolcü ve hidrolik fren sistemi ile karşılaştırıldı. Sonuç kayma değerlerinin her iki senaryo için de istenen aralıkta olduğu değerlendirmesi yapıldı. Çekiş kontrol sistemi için de yine CarSim ve MATLAB/Simulink ortamlarındaki boyuna taşıt dinamikleri kullanıldı. Sistemin test edilmesi için ilk hızı olmayan taşıt, CarSim yazılımında kullanılan söz konusu aracın motorundan elde edilen tork ile sürtünme katsayısı 1 ve 0.5 olan eğimsiz ve düz bir yoldaki ani hızlanma senaryoları için koşturuldu. Sonuçlar aç kapa kontrolcü sistem çıktıları ile karşılaştırıldı ve yol taşıtının boyuna dinamiğinin iyileştirilmesinde amaçlanan kayma değerinin amaç aralıkta tutulabildiği değerlendirilmesi yapıldı.
The scope of this thesis covers a rule-based control systems designed for road vehicle longitudinal dynamics. Road vehicle dynamics modeled by using CarSim software and MATLAB/Simulink tools. As a road vehicle the default B class 5 door hatchback vehicle of the CarSim software is choosen. Road vehicle longitudinal dynamics model parameter in MATLAB/Simulink same as the parameter values with CarSim one. Hydraulic brake systems also defined in two ways. The brake system defined as a first order dynamic system in fist way. So that hydraulic brake system can be explained by using only one a parameter is time constant of this system. Secondly more realistic brake cycle modeled in SimHydrolics is hydraulic system modeling tool of MATLAB/Simulink. In SimHydraulics hydraulic brake system of a vehicle can be defined with its components of single acting hydraulic cylinder, flow contol valves and their actuators and brake fluid. Back movement of single acting hydraulic system composed with a spring system which is very close to the real system’s sealing ring. Hydaulic brake cycle built for every individual tyre and every cycle composed of one single acting hydraulic cylinder. Two solenoid 2 way valve and valve actuator. First solenoid 2 ways valve placed on the pressure line and it is named pressure valve. The other valve is placed on the reservoir line which named by reservoir valve. Both valves have two states are on or off. First state lets the fluid pass. In second state valves does not let the fluid pass. Pressure of the cylinder increased when pressure valve is open, reservoir valve is close. Pressure released when pressure valve is close and reservoir valve is open. Pressure is holded when both the valves are close. On the other hand, master cylinder separates the brake pressure as a primery line and secondary line in real braking systems. This situation also defined in the brake system using gain blocks. That one out of five of the brake pressure applied to the secondary line which is rear brake pressure. Moreover, pressure and braking torque relation defined linearly. Brake torque of the front tire [Nm] is 250 times greater than pressure [MPa]. Torque pressure relation of the rear tire obtained by using the gain of 150. Vehicle dynamics of the CarSim presented by CarSim-S-Function block in MATLAB/Simulink. Its datas used by MATLAB/Simulink in the control system. For ABS control structure CarSim sends every wheels’ longitudinal and vehicle longitudinal speed and driver pressure. CarSim block input is controlled pressure values for every individual wheel. CarSim sent only driven wheels’s longitudinal and vehicle longitudinal speed for traction control system. It took pressure values for driven wheels. MATLAB/Simulink model of the road vehicle longitudinal dynamics composed of two main topics. Fist title is pneumatic tire modeling. Pneumatic tire produces longitudinal force by slip and tire vertical load. To define pneumatic tire longitudinal force for all surfaces only one set of slip and vertical force dataset is enough. Because using similarity relation tire longitudinal force can obtaine for all of road surface friction coefficient according to Pacejka H. [1]. Force definition of the pneumatic tire slip and vertical load datasets are taken from CarSim 185/65 R15 type of tire for when tire road friction coefficient 1. After the tire force definition, two wheels dynamic equation is composed for acceleration and deceleration states according to moment balance at the pneumatic wheel center. Other vehicle longitudinal dynamics modeling section is definition of sprung mass of the vehicle. This definition includes front rear axle dynamical load transfer and resistance forces and moments of the road vehicle. Vehicle sprung mass longitudinal dynamics defined by using dynamical load transfer relation of the front and rear axle. Suspension springs and dampers ignored in Simulink vehicle model. Road resistance forces consist of four main types. Acceleration resistance force, aerodynamic resistance force, tyre rolling resistance moment and resistance force of the road inclination are all. First three resistance forces modeled in this study. Road inclination resistance force did not needed. Because all the simulations occure at the flat surfaces. Aerodynamic forces modeled with aerodynamic resistance coefficient, vehicle frontal area, air density and vehicle longitudinal speed parameters. Tire rolling resistance moment modeled with static rolling resistance coefficient, velocity related rolling resistance coefficient, velocity and tire vertical load. Both ABS and traction control systems uses tire longitudinal slip values as a control variable. Longitudinal relative slip value is obtained wheel rotational speed, wheel effective rolling radius is vehicle loaded radius and vehicle longitudinal speed. Slip formula edited in control studies to obtain symmetrical values of interval for acceleration and deceleration conditions. After that longitudinal relative slip takes velue [0, -1] for deceleration state. -1 means %100 slip which is wheel lock up condition. On the other hand, relative slip takes value [0, 1] interval for acceleration state. 1 means that %100 slip which is wheel is rotating and vehicle is not moving. Rule based ABS control system run for CarSim vehicle dynamics system and MATLAB/Simulink vehicle dynamics system. Two scenarios given for deceleration situation of the road vehicle in two different road coefficient of friction. Fist, vehicle initial speed 50 kmph and driver spike braking pressure up 15 MPa in 0.1 sec at flat and straight road which friction coefficient 1. For this scenario rule based controller refecence slip values choosen as maximum 10% slip and minimum 8% slip. Because the slip values must be in the tire longitudinal force and slip increasing interval. As a result, rule based controller system worked properly and stated every tires’ slip values in the tolerance band. For CarSim and MATLAB/Simulink vehicle dynamic control systems gave similar results. Not controlled system output also given for this scenario and it is seen that slip values went to -1 and vehicle stopping distance increased. Second scenario carried out using same as the first braking scenario for rule based ABS control system. Only the difference is road friction coefficient is 0.5. Accordingly, reference slip values also changed to maximum 4.5% and minimum 5.5%. As a result, rule based controller system worked properly and placed slip values between maximum and minimum values. Also CarSim and MATLAB/Simulink vehicle dynamics systems gave similar results. In ABS scenario 2 also not controlled system’s slip values went to the -1 and stopping distance were increased. Rule based traction control system also run for CarSim vehicle dynamics system and MATLAB/Simulink vehicle dynamics system. Two scenarios given road vehicle acceleration. Friction coefficients are 1 and 0.5 for first and second scenarios repectively. Acceleration torque obtained from CarSim vehicle engine for full throttle in 0.1 second at first gear. This traction moment measured at the wheel and applied to the MATLAB/Simulink vehicle dynamics. First acceleration scenario occurred at flat and straight road which friction of coefficient is 1. By the way, rule based controller reference slip values set to maximum 9.5% and minimum 8.5%. As a result, rule based controller system worked properly and stated driven tires’ slip values in the tolerance band. For CarSim and MATLAB/Simulink vehicle dynamic control systems gave similar results. Not controlled system output also given for this scenario and it is seen that slip values limit to 1 and vehicle acceleration performance was decreased. Second acceleration scenario carried out using same as the first scenario for rule based traction control system. Only difference is road coefficient of friction which is 0.5. Accordingly, reference slip values also changed to maximum 4.5% and minimum 5.5%. As a result, rule based controller system worked properly and placed slip values between maximum and minimum values. CarSim and MATLAB/Simulink vehicle dynamics systems gave similar results. In TCS scenario 2 not controlled system’s slip values limit to the -1 and vehicle acceleration performance were decreased. As a result, vehicle dynamics simulation software CarSim which, highly recommended throught the world is used for vehicle longitudinal control study, rule based anti lock braking system and traction control system. Hydraulic modulators and brake system of the vehicle defined by their hydraulic components. Reference slip values successfully obtained for spike braking and rapid acceleration situations at the flat and straight road conditions.
Açıklama
Tez (Yüksek Lisans) -- İstanbul Teknik Üniversitesi, Fen Bilimleri Enstitüsü, 2012
Thesis (M.Sc.) -- İstanbul Technical University, Institute of Science and Technology, 2012
Anahtar kelimeler
boyuna taşıt dinamiği, abs, çekiş kontrol sistemi, hidrolik modülatör, kural tabanlı kontrolcü, vehicle longitudinal dynamics, abs, traction control system, hydraulic modulator, rule based controller
Alıntı