A 3D Printed Power-Split Device for Testing Energy Management ...

3rd IFAC IFAC Workshop Workshop on on Internet Internet Based Based Control Control Education Education 3rd 3rd IFAC Workshop Internet Based Control Education November 4-6, Brescia, Italy 3rd IFAC Workshop on Internet Based Control Education November 4-6, 2015. 2015.on Brescia, Italy 3rd IFAC Workshop on Internet Based Control Education November 4-6, 2015. Brescia, Italy Available online at www.sciencedirect.com November 4-6, 2015. Brescia, Italy November 4-6, 2015. Brescia, Italy

ScienceDirect IFAC-PapersOnLine 48-29 (2015) 164–169

A 3D Printed Power-Split Device for A Power-Split Device for A 3D 3D Printed Printed for Testing Energy Power-Split ManagementDevice Strategies Testing Energy Management Strategies Testing EnergytoManagement Strategies Applied Hybrid Vehicles Applied to Hybrid Vehicles Applied to Hybrid Vehicles ∗ ∗∗ ∗

Jose Jose Jose Jose Jose

L. L. L. L. L.

Torres Torres ∗∗∗ Torres Torres Torres ∗

Francisco Jose L. Blanco ∗ Francisco M. M. Arrabal Arrabal∗ ∗∗ ∗∗ Jose L. Blanco ∗ ∗ Francisco M. Arrabal L. Blanco ∗ ∗∗ ∗∗ Jose Francisco M. Arrabal Antonio Gimenez Antonio Gimenez Francisco M. Arrabal∗∗ Jose Jose L. L. Blanco Blanco ∗ Antonio Gimenez ∗ Antonio Gimenez Antonio Gimenez ∗ ∗ Dpto. de Mecanica, Universidad de Dpto. de Mecanica, Universidad de Almeria Almeria -- CIESOL. CIESOL. Campus Campus de de ∗ ∗ Dpto. de Mecanica, Universidad de Almeria - CIESOL. Campus de ∗Excelencia Dpto. de Mecanica, Universidad de Almeria CIESOL. Campus Internacional Agroalimentario, ceiA3, crta. Sacramento Excelencia Internacional Agroalimentario, ceiA3, crta. Sacramento Dpto. de Mecanica, Universidad de Almeria - CIESOL. Campus de de Excelencia Internacional Agroalimentario, ceiA3, crta. Sacramento Excelencia Internacional Agroalimentario, ceiA3, crta. Sacramento s/n, 04120 Almeria, Spain (e-mail: [email protected]). s/n, 04120 Almeria, Spain (e-mail: [email protected]). Excelencia Internacional Agroalimentario, ceiA3, crta. Sacramento ∗∗ s/n, Almeria, Spain [email protected]). ∗∗ s/n, 04120 04120 Almeria, Spain (e-mail: (e-mail: [email protected]). Dpto. de Fisica Fisica Quimica, Universidad de Almeria, Almeria, crta. crta. de yy Quimica, Universidad de s/n, 04120 Almeria, Spain (e-mail: [email protected]). ∗∗ Dpto. ∗∗ Dpto. de Fisica yy Almeria, Quimica, Universidad [email protected]). Almeria, crta. ∗∗ Dpto. de Fisica Quimica, Universidad Almeria, Sacramento s/n, 04120 Spain (e-mail: Sacramento s/n, 04120y Almeria, (e-mail:de Dpto. de Fisica Quimica, Spain Universidad [email protected]). Almeria, crta. crta. Sacramento s/n, 04120 Almeria, Spain (e-mail: [email protected]). Sacramento s/n, 04120 Almeria, Spain (e-mail: [email protected]). Sacramento s/n, 04120 Almeria, Spain (e-mail: [email protected]). Abstract: Abstract: This This work work presents presents a a testbed testbed emulating emulating the the hybrid hybrid electric electric vehicles vehicles powertrain powertrain with with Abstract: This work presents a testbed emulating the hybrid electric vehicles powertrain with Abstract: This work presents a testbed emulating the hybrid electric vehicles powertrain with both teaching and research purposes. The core of this testbed is a 3D printed epicycloidal both teaching research purposes. core ofthe thishybrid testbed is a vehicles 3D printed epicycloidal Abstract: Thisand work presents a testbedThe emulating electric powertrain with both teaching and research purposes. The core of this testbed isequipment aa 3D printed epicycloidal both teaching and The of testbed epicycloidal gearset actuating a split Its design and are as gearset actuating asresearch a power power purposes. split device. device. Itscore design and hardware hardware are explained explained as both teaching andas research purposes. The core of this this testbed is isequipment a 3D 3D printed printed epicycloidal gearset actuating as a power split device. Its design and hardware equipment are explained as gearset actuating as a power powerFrom split an device. Its design design and hardware equipment arean explained as well as as its its workingas principle. From an educational point of hardware view, this this equipment system states states an interesting well working principle. educational point of view, system interesting gearset actuating a split device. Its and are explained as well as its working principle. From an educational point of view, this system states an interesting well as working principle. From an point view, states an control problem and, at exemplifies the operation of this of with control problem and, at the the same same time, exemplifies the of operation ofsystem this kind kind of machines machines with well as its its working principle. From time, an educational educational point of view, this this system states an interesting interesting control problem and, at the same time, exemplifies the operation of this kind of machines with control problem and, at the same time, exemplifies the operation of this kind of machines the scope of a motivating application. Consequently, a teaching methodology comprising this the scope of a motivating teaching methodology this control problem and, at theapplication. same time, Consequently, exemplifies theaoperation of this kind ofcomprising machines with with the scope of a motivating application. Consequently, a teaching methodology comprising this the scope of a motivating application. Consequently, a teaching methodology comprising this testbed is proposed. In addition, the challenging nature of the system encourage the development testbed is proposed. In addition, the challenging nature aofteaching the system encourage the development the scope of a motivating application. Consequently, methodology comprising this testbed is proposed. In addition, the system encourage the development testbed is In the challenging nature of the encourage the of optimization optimization techniques aimedthe at challenging reducing the thenature overallof system energy consumption. Results of of of techniques aimed at reducing overall energy consumption. Results testbed is proposed. proposed. In addition, addition, the challenging nature ofsystem the system system encourage the development development of optimization techniques aimed at reducing the overall system energy consumption. Results of of optimization techniques aimed at reducing the overall system energy consumption. Results of a preliminary experiment are satisfactory addressed. As a consequence, the presented testbed a preliminary satisfactory addressed. Assystem a consequence, the presented testbed of optimizationexperiment techniques are aimed at reducing the overall energy consumption. Results of a preliminary experiment are satisfactory addressed. As a consequence, the presented testbed a addressed. As the proposed a for teaching benchmarking new is preliminary proposed as as experiment a remote remote lab labare forsatisfactory teaching and and benchmarking new control control strategies. strategies. ais preliminary experiment are satisfactory addressed. As aa consequence, consequence, the presented presented testbed testbed is proposed as a remote lab for teaching and benchmarking new control strategies. is proposed as a remote lab for teaching and benchmarking new control strategies. is proposed as a remote lab for teaching and benchmarking new control strategies. © 2015, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved. Keywords: Keywords: Hybrid Hybrid vehicles, vehicles, Motor Motor control, control, Differential Differential gears, gears, Computer Computer interfaces, interfaces, Energy Energy Keywords: Hybrid vehicles, Motor control, Differential gears, Computer interfaces, Keywords: Hybrid vehicles, Motor control, Differential gears, Computer interfaces, Energy Energy management system. management system. Keywords: Hybrid vehicles, Motor control, Differential gears, Computer interfaces, Energy management system. management system. management system. 1. CO 1. INTRODUCTION INTRODUCTION CO22 emissions emissions reduction reduction and and energy energy saving, saving, the the managemanage1. INTRODUCTION INTRODUCTION CO emissions reduction and energy saving, the manage2 1. CO emissions reduction and energy saving, the ment of PSDs require accurate control strategies, see 2 ment of PSDs require accurate control strategies, see Liu Liu 1. INTRODUCTION CO reduction and energy saving, the managemanage2 emissions ment of PSDs require accurate control strategies, see Liu ment of PSDs require accurate control strategies, Liu and Peng (2006). This problem has been studied during and Peng (2006). This problem has been studied during ment of PSDs require accurate control strategies, see see Liu and Peng (2006). This problem has been studied during Environmental cost of mobility is becoming a global proband Peng (2006). This problem has been studied during the two last decades, see Brahma et al. (2000); Lin et al. Environmental cost of mobility is becoming a global prob- and the two last decades, see Brahma et al. (2000); Lin et al. Peng (2006). This problem has been studied during Environmental cost of mobility is becoming a global probthe two last decades, see Brahma et al. (2000); Lin et al. Environmental cost of mobility is becoming a global problem in recent years. Electric vehicles are called to be a the two last decades, see Brahma et al. (2000); Lin et al. (2003) for an example. More recently, in Borhan et lem in recent years. are called to probbe a the Environmental cost ofElectric mobilityvehicles is becoming a global (2003) for an example. More recently, in Borhan et two last decades, see Brahma et al. (2000); Lin et al. lem inalternative recent years. years. Electric vehicles vehicles are called called to be be a (2012), for an example. More recently, in Borhan et al. lem recent Electric are to goodin to conventional conventional cars powered powered by internal internal (2003) for an example. More recently, in Borhan et al. a model predictive control strategy is employed good to cars by lem inalternative recent years. Electric vehicles are called to be aa (2003) (2012), a model predictive control strategy is employed (2003) for an example. More recently, in Borhan et al. good alternative to conventional conventional cars powered by by internal internal (2012), a model predictive control strategy is employed good alternative to cars powered combustion engines. However, some considerations have to (2012), a model predictive control strategy is employed for this purpose, with satisfactory results. Currently, this combustion engines. However, some considerations have to for good alternative to conventional cars powered by internal this purpose, with satisfactory results. Currently, this (2012), a model predictive control strategy is employed combustion engines. However, some considerations have to this purpose, with satisfactory results. Currently, this combustion engines. However, some considerations have be account. At scale, the for this with satisfactory results. Currently, problem still aa challenging field. In be taken taken into into account. At high high scale, the CO CO2 reduction reduction combustion engines. However, some considerations have to to for problem still remains remains being challenging research field.this In for this purpose, purpose, withbeing satisfactory results.research Currently, this reduction be ataken taken into account. account. At high scale, the the CO222 is problem still remains being a challenging research field. In be into At high scale, CO reduction as consequence of the use of electric vehicles strictly problem still remains being a challenging research field. In Sciarretta et al. (2014), a benchmark is presented dealing as a consequence of the use of electric vehicles is strictly be taken into account. At high scale, the CO2 reduction problem Sciarretta et al. (2014), a benchmark is presented dealing still remains being a challenging research field. In as a consequence of the use of electricinvehicles is strictly et al. (2014), aa benchmark is presented dealing as a of of is affected by the the technology technology employed the generation generation of Sciarretta Sciarretta et al. (2014), benchmark is presented dealing with this topic. affected by the of as a consequence consequence of the the use useemployed of electric electricinvehicles vehicles is strictly strictly with this topic. Sciarretta et al. (2014), a benchmark is presented dealing affected by the technology employed in the of generation of affected by the employed in generation of electricity. the autonomy full-electric with this this topic. topic. electricity. Moreover, the limited limited autonomy full-electric affected byMoreover, the technology technology employed in the the of generation of with with this topic. The literature electricity. Moreover, the limited autonomy of full-electric The literature review review carried carried out out in in this this work work reveals reveals electricity. Moreover, the limited autonomy of full-electric vehicles may suppose a handicap to this alternative, see vehicles may supposethe a handicap to this alternative, see The electricity. Moreover, limited autonomy of full-electric literature review carried out in this work reveals The literature review carried out in this work reveals that most of the related studies are addressed exclusively vehicles may suppose a handicap to this alternative, see that most of the related studies are addressed exclusively vehicles may suppose a handicap to this alternative, see The literature review carried out in this work reveals Torres et al. (2014) for a comparative study. As an interTorres et al. (2014) for a comparative study. As an intervehicles may suppose a handicap to this alternative, see that most of the related studies are addressed exclusively that most most of the the related related studies are addressed exclusively through simulations. Among the are commercial tools focused Torres et etterm, al. (2014) (2014) forElectric a comparative comparative study. As As anplaying inter- that through simulations. Among the commercial tools focused Torres al. for a study. an interof studies addressed exclusively mediate Hybrid Vehicles (HEVs) are mediateetterm, HybridforElectric Vehicles (HEVs) areanplaying Torres al. (2014) a comparative study. As inter- through simulations. Among the commercial tools focused through simulations. Among commercial tools on ADVISOR, presented in et mediate term, Hybrid Electric Vehicles (HEVs) are playing on this this subject, subject, ADVISOR, presented in Markel Markel et al. al. mediate term, Hybrid Electric (HEVs) are simulations. Among the the commercial tools focused focused a nowadays reducing the of a key key role role nowadays aimed atVehicles reducing the emissions emissions of through mediate term, Hybrid aimed Electricat Vehicles (HEVs) are playing playing on this subject, ADVISOR, presented in Markel et al. on this subject, ADVISOR, presented in Markel et al. (2002), arises as one of the most frequently employed. The a key role nowadays aimed at reducing the emissions of (2002), arises as one of the most frequently employed. The a key role nowadays aimed at reducing the emissions of on this subject, ADVISOR, presented in Markel et al. CO as a consequence of human and goods transportation, 2 CO a consequence of human and goodsthe transportation, a key nowadays aimed at reducing emissions of (2002), 2 asrole arises as one of the most frequently employed. The (2002), arises as one of the most frequently employed. The investment needed for real experiment is quite expensive. CO as a consequence of human and goods transportation, 2 investment needed experiment is quite expensive. CO as a aand consequence of human human andthe goods transportation, arises as onefor of real the most frequently employed. The see 22Liu Liu and Peng (2008). (2008). Among the benefits of the the hyhy- (2002), see Peng Among benefits of CO as consequence of and goods transportation, for experiment is expensive. investment needed for real real this experiment is quite quite expensive. This is the theneeded idea beyond beyond this work. On On the one one hand, see Liu Liu and and Peng Peng (2008). Among the benefits of the the hy- investment This is idea work. the hand, see (2008). benefits of needed for real experiment is quite expensive. bridization vehicles it not count with bridization ofPeng vehicles it is is Among not only onlythe count with propulsion propulsion see Liu andof (2008). Among the benefits of the hyhy- investment This is the idea beyond this work. On the one hand, This is the idea beyond this work. On the one hand, having a prototype of a PSD enlarges the number of bridizationusing of vehicles vehicles it is issources not only only count with propulsion having a prototype of a PSD enlarges the number of bridization of it not count with propulsion This is the idea beyond this work. On the one hand, machines different of energy but also being machines using different of count energywith but propulsion also being having bridization of vehicles it issources not only aa prototype of aa PSD enlarges the number of having prototype of PSD enlarges the number of experiments aimed at validating new control strategies. On machines using different sources of energy but also being experiments aimed at validating new control strategies. On machines using different sources of energy but also being having a prototype of a PSD enlarges the number of able to operate near the optimal point of each one, as able to operate near thesources optimal each as experiments machines using different of point energyofbut alsoone, being aimed at validating new control strategies. On experiments aimed at validating new control strategies. On the other hand, the PSD is a clear example of a planetary able to operate near the optimal point of each one, as the other hand, theatPSD is a clear a planetary able to to operate operate nearet the optimal point of each each one, one, as aimed validating newexample control of strategies. On discussed in Torres Torres etthe al.optimal (2014). point Mechanically, this as is experiments discussed in al. (2014). Mechanically, this is able near of the other hand, the PSD is aas clear example of a planetary the other hand, the PSD is clear example gearset which can be used complementary testbed discussed in Torres et aal. (2014). Mechanically, this is gearset used aa complementary testbed discussed in Torres (2014). is otherwhich hand,can the be PSD is aaas clear example of of a a planetary planetary achieved power-split device This achieved by by means ofet power-split device (PSD). (PSD).this This discussed in means Torres of et aal. al. (2014). Mechanically, Mechanically, this is the gearset which can be used as a complementary testbed gearset which can be used as a complementary testbed for subjects such as Theory of Machines and Mechanisms, achieved by means of a power-split device (PSD). This for subjects such as Theory of Machines and Mechanisms, achieved by means of a power-split device (PSD). This gearset which can be used as a complementary testbed element comprises the core of the powertrain in a HEV. In element comprises of the powertrain a HEV. In for achieved by meansthe of core a power-split device in (PSD). This subjects such as Theory of Machines and Mechanisms, for subjects such as Theory of Machines and Mechanisms, since most students have difficulty visualizing the gears element comprises the core of the powertrain in a HEV. In since most students have difficulty visualizing the gears element comprises the core of the powertrain in a HEV. In for subjects such as Theory of Machines and Mechanisms, most cases, these kind of devices are composed by a planemost cases, these kind devices composedinby a planeelement comprises the of core of theare powertrain a HEV. In since most students have difficulty visualizing the gears since most students have difficulty visualizing the motion in these systems. At the same time, this prototype most cases, these kind of devices are composed by a planemotion in these systems. At the same time, this prototype mostgear cases, these kind of devices devices are composed composed clutches, by aa planeplanegears tary gear setthese and a akind series of electro-mechanical electro-mechanical clutches, as since most students have difficulty visualizing the gears tary set and series of as most cases, of are by motion in these systems. At the same time, this prototype motion in these systems. At the same time, this prototype exemplifies the systems. use of of this this kind of mechanism mechanism to a a realrealtary gear set and a series of electro-mechanical clutches, as exemplifies the use of to tary gear and a of clutches, as in these At kind the same time, this prototype described in al. Other systems described in Lin Lin et al. (2003). (2003). Other cases, cases, these these systems tary gear set set and et a series series of electro-mechanical electro-mechanical clutches, as motion exemplifies the use of this kind of mechanism to a realexemplifies the use of this kind of mechanism to a realworld problem, which, moreover, results attractive for described in Lin et al. (2003). Other cases, these systems world problem, which, moreover, results attractive for described in Lin et al. (2003). Other cases, these systems exemplifies the use of this kind of mechanism to a realinclude continuously variable transmission (CVT) applyinclude continuously applyworld described in Lin et al.variable (2003). transmission Other cases, (CVT) these systems problem, which, moreover, results attractive for world problem, which, moreover, results attractive for engineering students from different disciplines. It is also include continuously variable transmission (CVT) applyengineering students from different disciplines. It is also include continuously variable transmission (CVT) applyworld problem, which, moreover, results attractive for ing for both cars, see Miller (2006) and motorcycles, see ing for both cars, seevariable Miller (2006) and motorcycles, see engineering include continuously transmission (CVT) applystudents from different disciplines. It is also engineering students from different disciplines. It is also worth to remark that planetary gear sets are been recently ing for both cars, see Miller (2006) and motorcycles, see worth to remark that planetary gear sets are been recently ing for both cars, see Miller (2006) and motorcycles, see Chung and Hung (2014). Yeo(2006) et al. al. and (2006) In Yeo Yeo et et see al. engineering students from different disciplines. It is also Chung Yeo et (2006) In al. ing for and bothHung cars, (2014). see Miller motorcycles, worth to to remark remark that planetary planetary gear sets are are been recently worth that sets employed in of applications, i.e. variable Chung and and Hung (2014). (2014). Yeo et et al. al. (2006) (2006) In Yeo Yeo et al. al. employed in research research of new new gear applications, i.e. recently variable Chung Hung Yeo In et to remark that planetary gear sets are been been recently (2006), this is interesting agent in (2006), and this device device is considered considered asal.an an(2006) interesting agent in worth Chung Hung (2014). Yeo etas In Yeo et al. employed in research of new applications, i.e. variable employed in research of new applications, i.e. variable stiffness actuators for service robotics, as presented in (2006), this device is considered as an interesting agent in stiffness actuators for service robotics, as presented in (2006), this device is considered as an interesting agent in employed in research of new applications, i.e. variable the recovery of energy during decelerations. Regarding the the recovery of energy during decelerations. Regarding (2006), this device is considered as an interesting agentthe in stiffness actuators for service robotics, as presented in stiffness actuators for service robotics, as presented in the recovery of energy during decelerations. Regarding the the recovery of energy during decelerations. Regarding the stiffness actuators for service robotics, as presented in the recovery of energy during decelerations. Regarding the Copyright © 2015, 2015 164 2405-8963 © IFAC (International Federation of Automatic Control) Copyright 2015 IFAC IFAC 164 Hosting by Elsevier Ltd. All rights reserved. Copyright © 2015 IFAC 164 Copyright © 2015 IFAC 164 Peer review under responsibility of International Federation of Automatic Copyright © 2015 IFAC 164Control. 10.1016/j.ifacol.2015.11.231

IFAC IBCE 2015 November 4-6, 2015. Brescia, Italy

Jose L. Torres et al. / IFAC-PapersOnLine 48-29 (2015) 164–169

165

Fig. 1. Exploded view of the planetary gear system.The angular of the sun gear, the carrier and the ring are represented by ω1 , ω2a and ω1 respectively. Groothuis et al. (2014); Kim et al. (2010), or mobile robotics, see Lee and Choi (2012) for an example. The theoretical approach employed in this work for managing our PSD prototype is based on the use of nomographs. This technique, applied to HEVs, will be discussed in depth, as well as the use and construction of a nomograph. It is straightforward to implement graphically in the GUI developed in this work as SCADA tool for managing the complete testbed this way of graphically represent the kinematic relations involved in the attached problem. This paper is structured as follows: First, technical aspects of the prototype developed in this work are explored in section 2, along with its working principle. Then, a teaching exercise is proposed in section 3. Afterwards, in section 4 an experiment is performed as an example of the 3D printed power-split device (3D-PSD) operation is addressed, and preliminary results are presented. Finally, conclusions of this work are presented in section 5.

teeth each. The sun gear (7) is connected to the motorgenerator MG1, placed in opposition to the main case by means of a metallic axle. Consequently, the number of teeth of the sun gear is Zs =20. Both motor-generators, from now on referred to as DC-motors (MG1 and MG2), are connected to the ground by two supports made on stainless steel. Thus, a two degrees of freedom planetary gear system is obtained, whose outputs are the angular speeds of the carrier and sun gear, and the input is the desired angular speed of the ring gear when operating in automatic mode. Fig. 2 shows the real testbed. The two DC-motors employed labeled MG1 and MG2 correspond with the DCmotors of the manufacturer Pololu with a reduction gear of 100:1 and a 64 pulses per revolution (ppr) encoders. The encoder which registers the ring speed (ENC ), is an optical

T-Sen

2. THE 3D-PSD In this section, we present the 3D-PSD in detail, its working principle and the targets behind the prototype. The core of the system is an epicycloidal gearset. According to Fig. 1, the ring gear (1) is a Zr =40 teeth inner gear with an external diameter of 144 mm. This component is attached to a main case (2) which is inserted into a bearing placed on a bracket (3) fixed to the ground. A pulley (4) which is also attached to the case allows to transfer the motion of the ring gear to a parallel axle. The transmission is carried out by means of a belt and another pulley, performing a 44/14 ratio. This axle allows to easily introduce loads into the systems as well as to measure the output angular speed (ω3 ) by means of a rotatory incremental encoder. The carrier link (5) consists of another 3D printed part which is connected to the motor-generator MG2. This link contains the axles of the three planet gears (6), of Zp =10 165

V-Sen

Enc DAQ BAT

C-Sen

Mot Cont

DAQ MG2

MG1

ENC

Fig. 2. Complete system and hardware description.

IFAC IBCE 2015 166 November 4-6, 2015. Brescia, Italy


rotary encoder ISC3004, by Hedss, and performs 1200 ppr. In order to read both the encoders of each DC-motor and the ISC3004 encoder, a PhidgetEncoder Highspeed 4-Input Data Acquisition Board (Enc DAQ) is connected to a laptop via USB. The laptop also controls the two DCmotors simultaneously by means of a PhidgetMotorControl HC (Mot Cont). A 12V/10Ah battery (BAT ) model 6-CNFJ-10 by Lead Crystal actuates as power supply. In order to measure the total energy consumed by the system, two additional sensors are employed. The first one is a Phidgets precision voltage sensor (V-Sen) which registers the battery voltage. The second one is an Phidgets 30 A current sensor AC/DC (C-Sen) which measures the current passing from the battery to the motors controller. Signals from both sensors as well as the battery temperature registered by the Phidgets precision temperature sensor (T-Sens) are processed by PhidgetInterfaceKit 8/8/8 Data Adquisition Board (DAQ) connected to the laptop via USB. The software architecture developed for this purpose is based on the APIs provided by Phidgets, upon a .NET framework. 2.1 Working Principle Following the discussion in Torres et al. (2014), the main objective of a PSD focuses on combining two or more sources of motion. In this work, two motors are connected to the sun gear and the carrier respectively as exposed before. This configuration is based on a simplified version of commercial hybrid vehicles powertrain. For the sake of simplification, the internal combustion engine (ICE) and the clutches typically employed in these kind of systems are removed. Moreover, generator (which usually operates connected to the ICE) is treated as a simple DC-motor. From now on, this component will be referred to as MG2, and will be related to the carrier angular speed, ω2a . On the other hand, the motor connected to the sun gear is another identical DC-motor referred to as MG1, with an angular speed ω1 . Finally, the system output is the ring angular speed, ω3 , when operating in manual mode. The kinematic relation between these three velocities is expressed as follows: ω3 − ω2a ρ= (1) ω1 − ω2a Where ρ = −Zs /Zr represents the apparent gear ratio, that is, the gear ratio when considering the system as an ordinary gear train (the carrier link is blocked). Hence, the ring angular speed is obtained as: 1 1−ρ ω3 = ω1 · − ω2a · (2) 1−ρ ρ Since this equation counts with two independent variables, each operating point could be obtained, in theory, from and infinite number of combinations of ω1 and ω2a . According to this criterion, this system is considered as a Single Input Multiple Output (SIMO) system. With the motion of the planetary gear train fully defined, it becomes important to understand the torque requirements of the system. This has traditionally been achieved by a static force analysis of a specific gear train. While this is a valid technique, it requires the selection of a specific planetary arrangement, and it is an error-prone 166

ω1,T1

ω2b,T2b

ω3, T3

Sun

Carrier

Ring

Zr

Zs

Fig. 3. Nomograph employed in modeling our power-split device. and extensive computational process. Instead, it is simpler to use the principle ot energy conservation. The energy balance equation for the general planetary gear train can be written as: (3) T1 ω1 + T2b ω2a + T3 ω3 = 0 Where Ti represent the torques applied to each branch of the gear train. The careful selection of two specific cases for ω1 and ω3 will quickly yield equations completely defining the torques on the three branches of the gear train. The first case to examine is the instance when the gear train is moving as solid axle, that is ω1 = ω2a = ω3 = ω. In this case, Equiation (3) can be rewritten as: (4) T1 ω + T2b ω + T3 ω = 0 Collecting terms and assuming w is non-zero, it can be quickly deduced that: (5) T1 + T2b + T3 = 0 This is first of two equation that will define the torque requirements of the planetary gear train. The second case of interest is that of zero speed at the carrier. Using (2) and substituting zero for ω2a . Hence: (6) ω3 = ρω1 Making this substituion and substituing zero for ω2a in (3): 1 (7) T1 ( ω 3 ) + T 3 ω 3 = 0 ρ Again, collecting terms and rearranging into a more convenient form, the second torque governing equation is found to be: 1 (8) T3 = −T1 ρ Using this equation, along with (5), one can fully characterize the torque requirements at any two branches of the gear train, given the torque at the remaining branch. Once this analysis is completed, the system has been reduced to the solution of the equations (2), (5), and (8), involving a total of seven variables. While this seems to imply that the designer has free choice of any four variables, the torque equations are independent of rotational speed. This means that the designer must select one torque, along with either three speeds or two speeds and a gear ratio. With so many variables being selected by the engineer, it become important to be able to visualize the response of the design to changes in any ot these variables. A suitable approach when dealing with this kind of problems consist of elaborating a nomographs, as depicted in Fig. 3.


+

-

+

-

V1

PI1

MG1

ω1

ENC1

ω2(ref)

+

-

V2

PI2

MG2

ENC2

ENC

167

3. PROPOSED TEACHING METHODS Nomograph

ω3(ref), T3

Optimization

ω1(ref)


ω2b ω3

Fig. 4. Control loop of the entire system. This procedure leads to an intuitive interface for a first understanding of the attached problem. 2.2 Overall plant control strategy The control problem may be represented by the control loop depicted in Fig. 4, where the optimization problem attends to the criterion of motors operation point. It has been stated that there are infinite operating points that may satisfy a determined output requirement. However, it is convenient to discuss about some issues: • Kinematic constraints that has to be accomplished according to (2). • Torque and speed limitations of the motors. • Efficiency of the complete plant.

Whereas the two first constraints are mandatory, the level of efficiency of the overall plant depends on the selected operating point of each DC-motor analyzed independently. This issue comprises an interesting optimization problem. As mentioned in the introduction of this work, high efforts are made in researching new optimization and control rules in order to maximize the efficiency of HEV. In our 3DPSD, this efficiency mainly depends on the performance of the DC-motors. Hence, the energy management strategy to considers must include two layers. The first one may be considered as the high-level control layer. This layer is the responsible of selecting an optimal setpoint to each DCmotor which satisfies the ring angular speed requirements. At this stage, the efficiency of each DC-motor under this conditions is considered. Afterwards, a low-level control loop has to be implemented in order to each DC-motor reaches the speed demanded in the high-level layer.

The testbed developed in this work presents a problem of control suitable for undergraduate students. This problem deals with two topics which are studied in depth in any engineering curriculum, such as electric motors and epicycloidal gears. Whereas the utilization of DC-motors is frequently employed as laboratory material, it is not so common to count with a completely monitorized planetary gearset. Moreover, the combination of these elements presented in this work are related to an interesting field of study, as the case of hybrid vehicles, boosting the students’ motivation. As mentioned in previous sections, the proposed control problem may be solved, from a kinematic point of view, by a broad number of solutions. However, when considering the energy consumption, an optimization problem arises. Hence, it maybe performed a benchmark in which the students, in addition to solve the kinematic problem controlling properly the DC-motors, are encouraged to minimize the overall system energy consumption. All these features make our proposed 3D-PSD a good tool for teaching several engineering subjects. Even when its overall cost is not excessive, this prototype will be treated as a remote laboratory in a first stage. Other remote labs have successfully been employed in related courses, as the quadruple-tank model presented in Pasamontes et al. (2012). A schedule of the tasks to deal with when using the 3D-PSD in a teaching exercise is presented next: 3.1 Characterize the DC-motors. The students are encouraged to obtain their transfer functions from the constants provided by the manufacturer, along with real input-output data are provided as groundtruth. Parameter identification techniques maybe applied additionally. 3.2 Control each motor independently. Once the DC-motors are modeled correctly, any valid controller may be developed. For the sake of simplicity, a PI controller is proposed. Its constants should be obtained according to a known technique, as the Ziegler-Nichols method employed in Section 2.3. 3.3 Verify the ring speed in open loop mode.

2.3 DC-motors parameters and control According to the data provided by the manufacturer, it has been considered two identical DC-motors featuring 11000 rpm, 300 mA free-run, 0.3 kg · cm and a stall is = 5 A at a rated voltage of vin = 12 V . The resistive term of the motor, which can also be measured experimentally, is Rm = 2.4 Ω. Neglecting the motor armature inductance, the models employed in this work are simplified to first order transfer functions. In order to control each motor independently, PI controllers are designed by using the Ziegler Nichols method in closed loop, in which the proportional constant is obtained from the critic constant that converts the plant output to an oscillating signal, and the critic time is considered the period of such oscillations. 167

At this stage, the students are able to set any arbitrary speed to each DC-Motor. This may be performed in the real system as follows: (i) the controller constants are introduced to the SCADA tool, (ii) the operating mode option in the SCADA tool is switched to manual and (iii) the sliders corresponding to each DC-motor are taken to the position corresponding to the desired velocity. The depending velocity, that is, the ring angular speed, is then read in the SCADA tool. This velocity is compared to that obtained by the students by hand calculations following the recommendations presented in Section 2.1. 3.4 Correct the phenomena that make mismatch the results The system considered is a clear example in which the backlash of the gearings has to be taken into account.

IFAC IBCE 2015 168 November 4-6, 2015. Brescia, Italy


3.7 Perform an energy consumption analysis. In each complete cycle experiment, dataset registering all the signals involved are grabbed. With this information, the students are able to analyze the operating point of each DC-motor during the entire cycle, as well as the amount of energy provided by the battery. Then, results are compared among the participants, and new control strategies can be addressed in order to the reduce the energy consumption without loosing velocity and acceleration performance.

(a) Software interface main menu and angular velocities.

(b) Power consumptions.

Fig. 5. Example experiment and preliminary results. Moreover, some additional issues must be analyzed, such as signals noise or system delays. 3.5 Set a rule for close the loop of the complete system. This task corresponds with the criterion to be stablished for achieving the ring angular velocity from the respective velocities of MG1 and MG2. At this stage, the mode option is switched to automatic in the SCADA tool, and the sliders belonging to the DC-motors remain inactive. Instead, a third slider bar allows to set the desired angular speed of the system reference. The DC-motors velocities, as well as their power consumption an the total power provided by the battery are shown in each time step. 3.6 Test the control strategy in a normalized cycle. At this stage, the high level control law should have been validated for discrete and manually set ring gear velocities references. Then, in this task, the students are encouraged to test their proposed control laws during a complete cycle. For this purpose, with the mode option switched to automatic in the SCADA tool, a text file corresponding to desired cycle must be loaded. The cycles are selected from a database, and represent real standard driving which are adjusted to be performed in the 3D-PSD. The main analysis to carry out is to check how the ring speed fit to the required velocity. 168

This practice takes into account some of the paradigms stated by the Bologna Process, which was implemented in all the Engineering degrees of the University of Almeria in 2010. Concretely, the proposed practical activity is a clear example of Problem Based Learning (PBL). This activity is scheduled to be carried out during the second semester of the third academic year of the Industrial Engineering degree, in the subject of Computer Control (6 ECTS). This subject is mandatory for the specialty in Electronics and optional for the specialties in Chemistry and Mechanics. At this stage, all the students have coursed the subjects of Theory of Machines and Mechanisms and Industrial Automation, so they are familiarized with the mechanical problem involving the planetary gearset and have an appropriate knowledge of Control Theory. The proposed methodology must accomplish two hours of autonomous work focused at the characterization of the motors and the design of a controller. Afterwards, a practical session must be carried out so that the student may verify and correct their models. This task corresponds to a laboratory session of two hours, in which groups of three students will operate the testbed in turns of 10 minutes each group. Then, in another session of autonomous work with duration of 2 hours, the students must define a control law for performing the entire closed-loop system, and test the results for a normalized cycle through simulations. Finally, in a second laboratory session, the students will test their control strategy in the real testbed, and their results in term of consumption and performance will be compared among the other groups. 4. PRELIMINARY EXPERIMENTS AND RESULTS DISCUSSION In order to demonstrate the operation of the 3D-PSD, a preliminary experiment is performed. As a first criterion of decision for determine the operating point of the DCmotors, the angular speed of the sun gear (ω1 ) is imposed to remain proportional to the carrier link angular speed (ω2b ). Thus, the control law follows an arbitrary rule such us ω1 = 0.3ω2b . As a consequence the value of the ring angular speed ω3 is determined according to Equation 2. This strategy does not take into account any optimization law. However, it is possible to evaluate the accuracy of the low-level control layer in which each DC-motor is controlled in order to perform a determined ring angular velocity. Moreover, preliminary energy considerations may be addressed. The input for this experiment is based on normalized profiles of velocities typically employed in the automotive sector, such as the New European Driving Cycle (NEDC). Fig. 5 shows the results of this experiment. In Fig. 3.6, the reference and the ring angular velocity



are represented as well as the angular velocities of MG1 (ω1 ) and MG2 (ω2b ) along with their respective setpoints. On the other hand, Fig. 3.6 show the values of voltage and current which are necessary to compute the power consumed by each DC-motor as well as power provided by the battery. In order to develop a remote lab from the presented testbed, some guidelines has been considered. For this purpose, the student will be assigned a schedule table provided by the professors, performing an authorization system based on a specifically developed software programmed in .NET. In addition, data base is designed so that contains all necessary information namely the students’ name, access date, season time, user name and password. Firstly the student executes the application and a form containing basic fields to get access to remote lab must be correctly fulfilled. If all conditions are satisfied, a TCP/IP connection with a specific socket among the client application and the server raises up. Finally, the client application is connected by an asynchronous mode that means a string is sent to the server when occurs an event. The server is also a daemon application with a select case function which reads the string sent and proceeds according to that string. Both client and server are programmed in .NET. 5. CONCLUSIONS In this work, a testbed to be employed as a remote lab is presented. This testbed, based on a epicycloidal gearset referred to as 3D-PSD equipped with a series of sensors and actuactors has been explained in detail. The theoretical framework and the 3D-PSD working principle has been related to the key points to cover when using this testbed with teaching purposes. Afterwards, an example of experiment has been performed. Results show the correct operation of all the sensors and actuators employed in the testbed. As a future work, the testbed will be connected to a server in order to allow its remote operation. Moreover, more sophisticated control lows will be employed, taking into account the DC-motors efficiency with the aim to be extrapolated to real hybrid electric vehicles. 6. ACKNOWLEDGMENTS This work has been partially funded by the Spanish “Ministry of Economy and Competitiveness” under the contracts DPI 2011-22513 and DPI2014-56364-C2-1-R as well as by the Andalusian Regional Government grant programs FPDU 2009 and the excellence project P12FQM-2668, co-funded by the European Union through the European Regional Development Fund (ERDF). REFERENCES Borhan, H., Vahidi, A., Phillips, A., Kuang, M., Kolmanovsky, I., and Di Cairano, S. (2012). Mpc-based energy management of a power-split hybrid electric vehicle. Control Systems Technology, IEEE Transactions on, 20(3), 593–603. doi:10.1109/TCST.2011.2134852. Brahma, A., Guezennec, Y., and Rizzoni, G. (2000). Optimal energy management in series hybrid electric vehicles. In American Control Conference, 2000. Proceedings of the 2000, volume 1, 60–64. IEEE. 169

169

Chung, C.T. and Hung, Y.H. (2014). Energy improvement and performance evaluation of a novel full hybrid electric motorcycle with power split e-cvt. Energy Conversion and Management, 86(0), 216 – 225. doi: http://dx.doi.org/10.1016/j.enconman.2014.04.043. Groothuis, S., Rusticelli, G., Zucchelli, A., Stramigioli, S., and Carloni, R. (2014). The variable stiffness actuator vsaut-ii: Mechanical design, modeling, and identification. Mechatronics, IEEE/ASME Transactions on, 19(2), 589–597. doi:10.1109/TMECH.2013.2251894. Kim, B.S., Song, J.B., and Park, J.J. (2010). A serial-type dual actuator unit with planetary gear train: Basic design and applications. Mechatronics, IEEE/ASME Transactions on, 15(1), 108–116. doi: 10.1109/TMECH.2009.2019639. Lee, H. and Choi, Y. (2012). A new actuator system using dual-motors and a planetary gear. Mechatronics, IEEE/ASME Transactions on, 17(1), 192–197. doi: 10.1109/TMECH.2011.2165221. Lin, C.C., Peng, H., Grizzle, J., and Kang, J.M. (2003). Power management strategy for a parallel hybrid electric truck. Control Systems Technology, IEEE Transactions on, 11(6), 839–849. doi:10.1109/TCST.2003.815606. Liu, J. and Peng, H. (2006). Control optimization for a power-split hybrid vehicle. In American Control Conference, 2006, 6 pp.–. doi:10.1109/ACC.2006.1655400. Liu, J. and Peng, H. (2008). Modeling and control of a power-split hybrid vehicle. Control Systems Technology, IEEE Transactions on, 16(6), 1242–1251. doi: 10.1109/TCST.2008.919447. Markel, T., Brooker, A., Hendricks, T., Johnson, V., Kelly, K., Kramer, B., OKeefe, M., Sprik, S., and Wipke, K. (2002). Advisor: a systems analysis tool for advanced vehicle modeling. Journal of Power Sources, 110(2), 255 – 266. doi:http://dx.doi.org/10.1016/S03787753(02)00189-1. Miller, J. (2006). Hybrid electric vehicle propulsion system architectures of the e-cvt type. Power Electronics, IEEE Transactions on, 21(3), 756–767. doi: 10.1109/TPEL.2006.872372. ´ Pasamontes, M., Alvarez, J.D., Guzm´an, J.L., and Berenguel, M. (2012). Learning switching control: A tank level-control exercise. IEEE Transactions on Education, 55(2), 226–232. doi:10.1109/TE.2011.2162239. Sciarretta, A., Serrao, L., Dewangan, P., Tona, P., Bergshoeff, E., Bordons, C., Charmpa, L., Elbert, P., Eriksson, L., Hofman, T., Hubacher, M., Isenegger, P., Lacandia, F., Laveau, A., Li, H., Marcos, D., Nesch, T., Onori, S., Pisu, P., Rios, J., Silvas, E., Sivertsson, M., Tribioli, L., van der Hoeven, A.J., and Wu, M. (2014). A control benchmark on the energy management of a plug-in hybrid electric vehicle. Control Engineering Practice, 29(0), 287 – 298. doi: http://dx.doi.org/10.1016/j.conengprac.2013.11.020. Torres, J.L., Gonzalez, R., Gimenez, A., and Lopez, J. (2014). Energy management strategy for plug-in hybrid electric vehicles. A comparative study. Applied Energy, 113, 816–824. doi:10.1016/j.apenergy.2013.08.007. Yeo, H., Hwang, S., and Kim, H. (2006). Regenerative braking algorithm for a hybrid electric vehicle with cvt ratio control. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, 220(11), 1589–1600.