IMechatronic Systems, Applications Edited by Annalisa Part 3 docx

MechatronicSystems,Applications34 Zhou, Y., Liu, W., & Huang, P. (2007). Laser-activated RFID-based Indoor Localization System for Mobile Robots, Proceedings of IEEE International Conference on Robotics and Automation, pp. 4600 – 4605, ISBN: 1-4244-0601-3, Roma, Italy, April 2007. Contactsensorforroboticapplication 35 Contactsensorforroboticapplication PetrKrejci X Contact sensor for robotic application Petr Krejci Brno University of Technology Czech Republic Abstract The chapter deals with design of contact force vector sensor. The information about interaction between robotic parts and surroundings is necessary for intelligent control of robot behavior. The simplest example of such interaction is mechanical contact between working part of robot and surroundings. Than the knowledge of contact characteristic is important for robot control. This mechanical contact could be described by vector of contact force which includes information about force magnitude as well as information about orientation and contact point. The information about contact force vector will allow to predict the geometry of object which is in the contact with robots parts and modify robots behaviour. This kind of sensor can be used for instance for control of robotic hand gripping force as well as for detection of collision between robot and surrounding. 1. Introduction The design of contact force sensor was published by Schwarzinger, 1992. This design requires application of 24 strain gauges on active part of sensor. The quantity of strain gauges is sufficient for analytical determination of contact force vector. Demand on small size of sensor for a lot of robotic applications (Grepl, R., Bezdicek, M., Chmelicek, J., Svehlak, M., 2004) disable application of a large number of strain gauges. Quantity of applied strain gauges and their size is limiting factor for using such design in our applications. Our design of contact sensor supposes to use only three strain gauges on active part of sensor. However three strain gauges are not enough for the analytical expression of contact force vector. Due to this fact the neural network is used for force vector identification based on measured deformations of sensor body. The application of three strain gauges and new design will reduce size of sensor but requires a lot of numerical simulations for correct and accurate sensor behaviour. The main advantage of using neural network is in low computational requirements for vector determination. It means fast response of sensor to contact load. The neural network is able to process measured data faster than nonlinear equations for force vector expression in analytical way. The other advantage of our design is in reduced requirements for strain measurement by strain gauges. Generally, the Wheatstone bridge has to be used for strain measurement 3 MechatronicSystems,Applications36 ( H co m el e Fi g 2. B a lo c id e of fo r se n n u w o 3. T h se n he co r T h F E 2 . th i H offman, 1989) f o m plexit y of elec t e ctronics into sen s g . 1. Geometr y of Sensor work i a sic principle of c ations b y three s e ntified b y neur a neural network. r ce vectors corr e n sor was used f o u merical simulat i o rk of neural net w FE model of s h e FE model of s e n sor flan g e. The ad. It is necessa r rect work of se n h ese positions co r E model the force The total strains i s numerical sim u o r each strain g t rical measurin g s or bod y . a) contact sensor a n i ng principle sensor is based s train g au g es on. a l network. Hu g e The trainin g ma t e sponding to se n o r creatin g of tra i i ons with varied w ork. s ensor e nsor (see Fig. 1b ) contact force is r y to hold g eo m n sor. Deformatio n r respond to posi t of 20N was appl in z-direction ar u lation. g au g e. It means unit. This redu c n d FE model of s e on measurin g o Based on these matrix of traini n t rix contains pair n sor body defor i nin g matrix. It i ma g nitude and ) consists amon g simulated as ap p m etr y and dime n n of sensor duri n t ions of strain ga ied on sensor he a e S1= 2.9um/m, that our desi gn c tion will allow e nsor f deformatio n o f d eformations th e ng pairs is necess s of de f ormation s r mation. Finite e i s necessar y to m position of co n others of sensor p lied loads in se l n sions of model w ng load is calcul a a u g es on real se n a d. Results of th e S2=12.6um/m a n n si g nificantl y r e us to build in c b) f sensor bod y i n e contact force v e ar y for proper f u s in three locatio n e lement (FE) m o m ake a lar g e nu m n tact force for p r head, sensor bo d l ected nodes on w ith real struct u a ted in three po s n sor. For verifica t e verification are o n d S3= -22.97um / e duces c ontrol n three e ctor is u nction n s and o del of m ber of r operl y dy and sensor u re for s itions. t ion of o n Fig. / m for Fi g 4. T h co n co n T h pr o su g tr a Fi g g . 2. Sensor defo r Neural netw o h e architecture o f n tains deformat i n tains informati o h e trainin g matri x oj ect. This amo u gg ested sensor. F a inin g matrix. g . 3. Architecture r mation (applied o rk f artificial neura l i ons of sensor b o n about contact f x of 1000 trainin g u nt of trainin g p F or better accur a of used neural n e load of 20 N) (u n l network (ANN ) b od y measured f orce and positio n g pairs was used p airs was used j a c y of ANN as w e twork n it of results are m ) is shown on fi g b y strain g au ge n of contact force for trainin g of A j ust for verificat i w ell as sensor ca n m /m) g . 3. The input e s. The output on sensor head. A NN in first step i on of function a n b y use much g vector vector of this a lit y of g reater Contactsensorforroboticapplication 37 ( H co m el e Fi g 2. B a lo c id e of fo r se n n u w o 3. T h se n he co r T h F E 2 . th i H offman, 1989) f o m plexit y of elec t e ctronics into sen s g . 1. Geometr y of Sensor work i a sic principle of c ations b y three s e ntified b y neur a neural network. r ce vectors corr e n sor was used f o u merical simulat i o rk of neural net w FE model of s h e FE model of s e n sor flan g e. The ad. It is necessa r rect work of se n h ese positions co r E model the force The total strains i s numerical sim u o r each strain g t rical measurin g s or bod y . a) contact sensor a n i ng principle sensor is based s train g au g es on. a l network. Hu g e The trainin g ma t e spondin g to se n o r creatin g of tra i i ons with varied w ork. s ensor e nsor (see Fig. 1b ) contact force is r y to hold g eo m n sor. Deformatio n r respond to posi t of 20N was appl in z-direction ar u lation. g au g e. It means unit. This redu c n d FE model of s e on measurin g o Based on these matrix of traini n t rix contains pair n sor bod y defo r i nin g matrix. It i ma g nitude and ) consists amon g simulated as ap p m etr y and dime n n of sensor duri n t ions of strain ga ied on sensor he a e S1= 2.9um/m, that our desi gn c tion will allow e nsor f deformatio n o f d eformations th e ng pairs is necess s of de f ormation s r mation. Finite e i s necessar y to m position of co n others of sensor p lied loads in se l n sions of model w ng load is calcul a a u g es on real se n a d. Results of th e S2=12.6um/m a n n si g nificantl y r e us to build in c b) f sensor bod y i n e contact force v e ar y for proper f u s in three locatio n e lement (FE) m o m ake a lar g e nu m n tact force for p r head, sensor bo d l ected nodes on w ith real struct u a ted in three po s n sor. For verifica t e verification are o n d S3= -22.97um / e duces c ontrol n three e ctor is u nction n s and o del of m ber of r operl y dy and sensor u re for s itions. t ion of o n Fig. / m for Fi g 4. T h co n co n T h pr o su g tr a Fi g g . 2. Sensor defo r Neural netw o h e architecture o f n tains deformat i n tains informati o h e trainin g matri x oj ect. This amo u gg ested sensor. F a ining matrix. g . 3. Architecture r mation (applied o rk f artificial neura l i ons of sensor b o n about contact f x of 1000 trainin g u nt of trainin g p F or better accur a of used neural n e load of 20 N) (u n l network (ANN ) b od y measured f orce and positio n g pairs was used p airs was used j a c y of ANN as w e twork n it of results are m ) is shown on fi g b y strain g au ge n of contact force for trainin g of A j ust for verificat i w ell as sensor ca n m /m) g . 3. The input e s. The output on sensor head. A NN in first step i on of function a n b y use much g vector vector of this a lit y of g reater MechatronicSystems,Applications38 Fi g Fi g 5. T h T h m o ap S1 = ve c (c o g . 4. Points of mo g . 5. Sensor coor d Verification o h e force of 20N a h e position of th e o del load used f o plied force are 6. = 10.81um/m, S c tor of trained A o ordinate s y stem del load d inate s y stem o f ANN functio n a pplied on senso e force was diffe r o r trainin g matrix 1mm, -3.31mm, 1 S 2=-23.57um/m a A NN. The resu l of sensor is sho w n ality r head was use d r ent than forces creation are sho w 1 5.95mm respect i a nd S3=10.04um l t of contact for w n o n .) d for verification applied for trai n w n on Fig. 4). Th e i vel y . The total s t / m. These strai n ce vector deter m of ANN functi o n in g of ANN (po e x, y and z direc t rains in z-direct i n s were used as m inatio n is in T o nalit y . ints of tion of i on are input T able 1 Contact force coordinates Accuracy [%] Point of FE model load Position of contact force determined by Simulated by ANN x [mm] 6.10 6.03 98.85 y [mm] -3.31 -3.29 99.39 z [mm] 15.95 15.97 99.87 Table 1. Result of verification 6. Experimental verification of sensor functionality The sensor functionality was verified by experimental simulation in laboratory of Mechatronics. During experiment the loads of sensor was applied in several positions of sensor head. Gauging fixture (Fig. 6) was used for sensor positioning. Load was applied by materials testing machine Zwick Z 020-TND (Fig. 7, Fig. 8) where the real load force was measured. The deformation of sensor body was measured by strain gauges through HBM Spider 8 unit which is among other things designed for measuring of deformation by strain gauges. Fig. 6. Gauging fixture Measured deformations was transferred to information about contact force position and magnitude by neural network implemented in Matlab software. The results of experimental verification for selected points are shown in Table 2 for four positions of load force and shows really good accuracy of designed sensor. Contactsensorforroboticapplication 39 Fi g Fi g 5. T h T h m o ap S1 = ve c (c o g . 4. Points of mo g . 5. Sensor coor d Verification o h e force of 20N a h e position of th e o del load used f o plied force are 6. = 10.81um/m, S c tor of trained A o ordinate s y stem del load d inate s y stem o f ANN functio n a pplied on senso e force was diffe r o r trainin g matrix 1mm, -3.31mm, 1 S 2=-23.57um/m a A NN. The resu l of sensor is sho w n ality r head was use d r ent than forces creation are sho w 1 5.95mm respect i a nd S3=10.04um l t of contact for w n o n .) d for verification applied for trai n w n on Fig. 4). Th e i vel y . The total s t / m. These strai n ce vector deter m of ANN functi o n in g of ANN (po e x, y and z direc t rains in z-direct i n s were used as m inatio n is in T o nalit y . ints of tion of i on are input T able 1 Contact force coordinates Accuracy [%] Point of FE model load Position of contact force determined by Simulated by ANN x [mm] 6.10 6.03 98.85 y [mm] -3.31 -3.29 99.39 z [mm] 15.95 15.97 99.87 Table 1. Result of verification 6. Experimental verification of sensor functionality The sensor functionality was verified by experimental simulation in laboratory of Mechatronics. During experiment the loads of sensor was applied in several positions of sensor head. Gauging fixture (Fig. 6) was used for sensor positioning. Load was applied by materials testing machine Zwick Z 020-TND (Fig. 7, Fig. 8) where the real load force was measured. The deformation of sensor body was measured by strain gauges through HBM Spider 8 unit which is among other things designed for measuring of deformation by strain gauges. Fig. 6. Gauging fixture Measured deformations was transferred to information about contact force position and magnitude by neural network implemented in Matlab software. The results of experimental verification for selected points are shown in Table 2 for four positions of load force and shows really good accuracy of designed sensor. MechatronicSystems,Applications40 Ta Fi g Load p 1 2 3 4 ble 2. Results of v g . 7. Testin g mac h p oint Direction [mm] x y z x y z x y z x y z v erificatio n h ine Zwick Z 020 Contact force c o Position of force durin g experiment 0.8 -2.5 20.0 -1.0 -1.9 22.0 -1.0 0.1 22.0 2.0 -5.0 16.0 -TND o ordinates Simulated by ANN 0.81 -2.75 20.62 -0.97 -2.03 24.1 -1.02 0.11 22.91 2.07 -4.52 16.38 Accuracy [%] 98.8 91.0 96.8 97.0 93.6 90.9 98.0 90.9 96.0 96.6 90.4 97.7 Fi g 7. T h te s in c T h to p "s h to p cr i re d T h lo c o p T h di r w h re d T h g . 8. Loaded sens o Optimization h e low sensitivit y s tin g . Therefore c reasin g of sensit i h e Finite eleme n p olo g ical optimi z h ape" optimizati p olo g ical optimi z i terion takes on a d uction). h e sensor bod y i c ated under su p p timization. h is optimization r ections. The firs t h ile second step w d uction of 80% i n h e boundar y con d o r durin g experi m of sensor des of sensor was o b the topolo g ical i vit y for loads a p n t model of sen z ation procedur e on, sometimes r z ation is to find t maximum/min i i s the volume w p posed locatio n was done for t step of optimi z w as done for ra d n Fig. 10. The fi gu d itions used duri n m ental verificati o ign b served in axial d optimizatio n o p plied to sensor i n sor in finite el e e . Topolo g ical o p r eferred to as " he best use of m a i mum value sub je w hich was sub j e c n s of strain g a u two load steps z ation procedure d ial load. The res u u re shows distri b ng optimization p o n of functionalit y d irectio n during o f sensor g eom e n axial direction. e ment software p timization (ref. la y out" optimiz a a terial for a bod y e ct to g iven cons t c ted to optimiz a ug es was exclu - for load force s was done for ax u lt of optimizati o b ution of pseudo d p rocedure are sh o y simulations and e tr y was requir e ANSYS was us e ANSYS) is a f o a tion. The purp y such that an o b t raints (such as v a tion process. V o ded from proc e s oriented in di f ial load of senso o n is shown for v d ensit y in senso r o wn in Fig. 9. sensor e d for e d for o rm of ose of bj ective v olume o lumes e ss of f ferent r head v olume r bod y . Contactsensorforroboticapplication 41 Ta Fi g Load p 1 2 3 4 ble 2. Results of v g . 7. Testin g mac h p oint Direction [mm] x y z x y z x y z x y z v erificatio n h ine Zwick Z 020 Contact force c o Position of force durin g experiment 0.8 -2.5 20.0 -1.0 -1.9 22.0 -1.0 0.1 22.0 2.0 -5.0 16.0 -TND o ordinates Simulated by ANN 0.81 -2.75 20.62 -0.97 -2.03 24.1 -1.02 0.11 22.91 2.07 -4.52 16.38 Accuracy [%] 98.8 91.0 96.8 97.0 93.6 90.9 98.0 90.9 96.0 96.6 90.4 97.7 Fi g 7. T h te s in c T h to p "s h to p cri re d T h lo c o p T h di r w h re d T h g . 8. Loaded sens o Optimization h e low sensitivit y s tin g . Therefore c reasin g of sensit i h e Finite eleme n p olo g ical optimi z h ape" optimizati p olo g ical optimi z i terion takes on a d uction). h e sensor bod y i c ated under su p p timization. h is optimization r ections. The firs t h ile second step w d uction of 80% i n h e boundar y con d o r durin g experi m of sensor des of sensor was o b the topolo g ical i vit y for loads a p n t model of sen z ation procedur e on, sometimes r z ation is to find t maximum/min i i s the volume w p posed locatio n was done for t step of optimi z w as done for ra d n Fig. 10. The fi gu d itions used duri n m ental verificati o ign b served in axial d optimizatio n o p plied to sensor i n sor in finite el e e . Topolo g ical o p r eferred to as " he best use of m a i mum value sub je w hich was sub j e c n s of strain g a u two load steps z ation procedure d ial load. The res u u re shows distri b ng optimization p o n of functionalit y d irectio n during o f sensor g eom e n axial direction. e ment software p timization (ref. la y out" optimiz a a terial for a bod y e ct to given cons t c ted to optimiz a ug es was exclu - for load force s was done for ax u lt of optimizati o b ution of pseudo d p rocedure are sh o y simulations and e tr y was requir e ANSYS was us e ANSYS) is a f o a tion. The purp y such that an o b traints (such as v a tion process. V o ded from proc e s oriented in di f ial load of senso o n is shown for v d ensit y in senso r o wn in Fig. 9. sensor e d for e d for o rm of ose of bj ective v olume o lumes e ss of f ferent r head v olume r bod y . MechatronicSystems,Applications42 Fig. 9. Loads of sensor used in topological optimization procedure Fig. 10. Results of optimization (pseudodensity - red color means that volume will be included in final design, blue color mean that volume will be excluded from final design) Optimized shape of sensor body need to by simplified by reason of good manufacturing. Due to this fact few shapes of cutting was designed with consideration of optimized shape (Fig. 10) and machining. Based on results of structural analysis rectangular shape of cutting with 1 mm hole (Fig. 11b) ) produces the best results in terms of sensitivity. This shape is also suitable for simple machining. Fig. 12 shows prototype of optimized and non- optimized sensor which is made from aluminium alloy. Fi g Fi g 7. 1 St r li n Fi g de g . 11. Optimized s g . 12. Optimized a 1 . Structural an a r uctural anal y sis n ear behaviour of g . 13 for load for c formation of sen s a) s ensors with diff e a nd no n -optimiz a lysis of sensor of optimized sen structure occurs . c e of 140N. This v s or bod y can occ u b) e rent shapes of c u ed sensor protot y prototype sor was done in o . Results of this s i v alue defines up p u rs. u ttin g s y pe o rder to find out i mulation are sh o p er bound of sen s c) load limits wher e o wn in s or limits where e the plastic [...]... Poland, ISBN 83- 60102 -30 -9 Schwarzinger Ch., Supper L & Winsauer H (1992) Strain gauges as sensors for controlling the manipulative robot hand OEDIPUS: RAM vol 8, pp.17-22 46 Mechatronic Systems, Applications Develop a Multiple Interface Based Fire Fighting Robot 47 4 x Develop a Multiple Interface Based Fire Fighting Robot 1Department Ting L Chien1 , Kuo Lan Su2 and Sheng Ven Shiau3 of Electronic... deformation of sens body can occu sor urs 44 Mechatronic Systems, Applications Fig 13 Von-Misses stress (MPa) of sensor for load of 140 N applied in radial direction 8 Verification of optimized sensor functionality Functionality of optimized sensor was also done by two methods Finite elements model of sensor is used for calculation of body deformation caused by specified load in first method The Second method... research plan MSM 0021 630 518 "Simulation modeling of mechatronic systems" and GAČR project no 101/08/0282 11 References Ansys, Inc., Theory Reference, Release 10, Southpointe, 275 Technology Drive, Canonsburg, PA 1 531 7 Grepl, R., Bezdicek, M., Chmelicek, J., Svehlak, M.: Experimental quadruped walking robot: conception, design and control Magazine - Elektronika, 8-9/2004, ISSN 0 033 -2089, Poland 2004 Hoffman,... which receive from the remote supervise computer The program of the client’s user interface is designed by VB 52 Mechatronic Systems, Applications g supervise interfac of the fire fight ce ting robot Fig 6 The remote s he ule e e oller of Th security modu and appliance control module are designed by us The contro the modules is m ese microprocessor (A ATMEL89C2051) In the security m module, it contai... Baldwin Messtechnik, GmbH, Darmstad Krejci, P.: Contact sensor for robotic applications, Engineering Mechanics, Vol.12, (2005), No.A1, pp.257-261, ISSN 1210-2717 Krejci, P., Vlach, R., Grepl, R.: Contact sensor for robotic applications – Verification of Functionality, in conference proceedings: ”Engineering Mechanics 2006”, pp 1821 83, May 2006, Svratka, Czech Republic, ISBN 80-86246-27-2 Krejci, P., Vlach,... functionality of sensor was proofed by numerical Contact sensor for robotic application 45 simulations and also by experimental verification using and simulating real load of sensor prototype Verification was done for optimized and non-optimized prototype of sensor Using only three strain gauges for deformation measurement of sensor body allow us to use SMD electronics parts and build up the unit to hollow... card and driver devices, and it can control two robot arms using the motion control card In the avoidance obstacle function, the robot catches eight IR sensor signals by digital input terminal (in the motion control 50 Mechatronic Systems, Applications card), and measure distance of obstacle using eight ultrasonic sensors via series interface We use microprocessor to drive eight ultrasonic sensors, and... experimental verification of sensor subjected to real load Deformations of sensor body observed by both methods are used as inputs of neural network which produces information about contact force magnitude and coordinates 8.1 Verification of functionality by FEM simulation Sensor functionality was proof by numerical simulation using FE model of sensor Verification was done in same way as procedure... Su2 and Sheng Ven Shiau3 of Electronic Engineering WuFeng Institute of Technology Ming-Hsiung, Chia 621, Taiwan cdl@mail.wfc.edu.tw 2 Department of Electrical Engineering National Yunlin University of Science & Technology Douliou, Yunlin 640, Taiwan,sukl@yuntech.edu.tw 3Graduate school Engineering Science and technology National Yunlin University of Science & Technology Douliou, Yunlin 640, Taiwan g9610808@yuntech.edu.tw... robot can uses two flame sensors to find out fire source by the proposed method, and move to fire source to fight the fire using extinguisher Keywords：fire fighting robot、man-machine interface、touch screen 2 Introduction Home can provide safety, convenience, and efficiency for people in the 21st century An intelligent home system is integrated by many function and systems One of the most important systems . Point of FE model load Position of contact force determined by Simulated by ANN x [mm] 6.10 6. 03 98.85 y [mm] -3. 31 -3. 29 99 .39 z [mm] 15.95 15.97 99.87 Table 1. Result of verification. Point of FE model load Position of contact force determined by Simulated by ANN x [mm] 6.10 6. 03 98.85 y [mm] -3. 31 -3. 29 99 .39 z [mm] 15.95 15.97 99.87 Table 1. Result of verification. requirements for strain measurement by strain gauges. Generally, the Wheatstone bridge has to be used for strain measurement 3 Mechatronic Systems, Applications3 6 ( H co m el e Fi g 2. B a lo c id e of