A METHOD FOR SEAT OCCUPANCY DETECTION FOR AUTOMOBILE SEATS WITH INTEGRATED HEATING ELEMENTS

Transcription

1 XIX IMEKO World Congress Fundamental and Alied Metrology Setember 6 11, 2009, Lisbon, Portugal A MEHOD FO SEA OCCUPANCY DEECION FO AUOMOBILE SEAS WIH INEGAED HEAING ELEMENS Boby George, Hubert Zangl, homas Bretterklieber and Georg Brasseur Graz University of echnology, Graz, Austria, bgeorge@tugraz.at Abstract Seat occuancy detection systems based on a caacitive or an electric field rincile, develoed for an automobile seat, may not work well for seats with integrated electric heating systems. A new detection method, suitable for this tye of automobile seats, is resented in this aer. In the roosed scheme, the heating element is a art of the sensor system while its oeration as a heating element is ket unaffected. esults obtained from a rototye system demonstrate the ractical validity of the method. Keywords: Seat occuancy detection, caacitance measurement, synchronous detection. 1. INODUCION Passenger safety has considerably imroved with the introduction of airbag systems [1], [2]. Seat occuancy systems rovide information about the resence of a assenger to the airbag control system. he airbag system uses this data for making a decision to trigger the airbag in case of an accident, if the seat is occuied by a assenger and deactivate the firing of the airbags to vacant seats. his revents the wastage of air bags and related costly reair. Fatal injuries have been reorted when an airbag is executed with full ower into a seat occuied by an infant [1]. Seat occuancy systems based on otical, vision, change in ressure or force, caacitive or electric field rinciles have been reorted. Otical and vision based methods may not erform well for changing illumination intensity in the car [3]. hey also have to rocess huge amount of data in order to obtain a final decision. Systems based on ressure or force sensing are directly based on the weight alied in the seat [4]. A belt-tightened emty baby seat or a basket of beer bottles may also be sensed as a human. Caacitive seat occuancy systems mostly rovide a simle and cost-effective solution [5], [6]. However, in most of the existing systems, the electrodes need to be laced in the surface or near the surface of the seat [6], [7]. Conseuently, the overall seat manufacturing rocess is tedious and costly. hus, the otimal location of the sensor electrodes is below the seat foam, fixed to the seat frame. It is found from our investigations that such systems may fail to function well for seats with integrated electric heating elements. his aer rooses a method for seat occuancy detection suitable for automobile seats with electric heating elements. In the new method, the heating element is advantageously used as a art of the sensing system. A synchronous detection method [8] is emloyed to detect changes in caacitances owing to the resence of a assenger. he rincile of seat occuancy sensing, the method of measurement and exerimental set-u with results and conclusions are elucidated in the following sections. 2. MEHOD OF SEA OCCUPANCY DEECION Fig. 1 illustrates the basic rincile of a caacitance based seat occuancy detection using the shielding effect [6] of electric field due to the resence of a human or art of human body within the sensing volume. he transmitter () is excited by a sinusoidal source. he receiver () is connected to a low imedance shunt (A) to ground and its otential is ket nearly eual to ground. here are electric field lines, originating from the transmitter and ending at the receiver as well as near by surfaces at ground otential. An illustration of the useful electric field lines for human roximity sensing is shown in Fig. 1. When a human body aroaches the sensing volume, some of the electric field lines get shielded. his generates a current flow I HG through eceiver () A art of human body A C HG I HG Electric field lines ransmitter () Fig. 1. Princile of seat occuancy detection based on electric field shielding. Parts of the electric field lines from the transmitter to the receiver are shielded by the human body. C HG is the euivalent value of distributed caacitances between the human body and ground. A indicates a low imedance shunt.

2 the distributed caacitance C HG that exists between the human body and ground. he shielding effect essentially reduces the number of field lines that are terminating on the receiver electrode. Hence, when a human occuies a seat, a reduction in the dislacement current received by the receiver is noticed. his feature can be emloyed for seat occuancy detection. A class of seats that incororates heating elements to rovide better comfort to the assenger is already available in the market. A cross-sectional view of the sitting area of such a seat is described in Fig. 2(a). he transmitter and receiver electrodes of the caacitive sensor are also shown in Fig. 2(a). A to view of the same, with a tyical structure of the lacement of the heating element is illustrated in Fig. 2(b). he nodes and are the electrical inut terminals of the heating element. he heating element is nearly at ground otential, hence a major art of the electric field lines emanating from the transmitter are shielded by the heating element itself. hus, when a human comes close to the seat, very low additional shielding is observed and ultimately a reliable detection becomes extremely difficult with the basic method illustrated in Fig. 1. o ower source Soft to cover Heating element Seat foam (a) Fig. 2. A cross-sectional view (a) and to view (b) of the sitting area of the seat with integrated heating elements. he method described above (Fig. 1) works well only when an electric field exists above the seat surface and a human body shields it. For seats with integrated heating elements as in Fig. 2, this condition can be achieved only when the heating element is in a floating condition. In such a condition, the heating element will have an electrical otential, due to caacitive couling with the transmitter electrode, and therefore an electric field will be established around it. An ideal floating condition is difficult as the heating element is connected to a dc ower suly (or a modulated voltage from a dc ower suly). Nearly floating conditions can be reached by inserting inductors in both the inut terminals and (refer to Fig. 2) of the heating element and erforming the measurements at very high freuency. A nearly floating condition can also be reached at a suitable measurement freuency if a arallel resonance condition can be achieved, with the hel of the additional inductors, in the heating element circuit. In such a condition, the heating element reaches its maximum otential due to its caacitive couling with the transmitter electrode. A method for achieving such a condition is exlained next. (b) 2.1. Introducing a resonance in the heating circuit Fig. 3 shows the heating element with two inductors L a and L b connected to its terminals and resectively. C denotes the caacitance between the transmitter and the heating element. A ictorial reresentation of the electric field distribution is illustrated in Fig. 3. his is euivalent to the resence of a distributed caacitance between the heating element and ground. he value of this distributed caacitance deends on the area occuied by the heating element and its distance to the ground lane. For this seat system (according to Fig. 2), the ground lane is in a comaratively far location from the heating element. herefore the distributed caacitance between heating element and ground is mainly deended on the surface area of the wire used for heating element [9] and the medium between the heating element and near by surfaces at ground otential. hus, for such a given comosition, the value of the distributed caacitance remains constant for a vacant seat condition. Let L EQ be the euivalent inductance of the arallel combination of L a and L b and C EQ be the euivalent caacitance of the heating element to ground. he resonance freuency f r is given by f r Plane at ground otential 1 =. (1) 2π L C EQ EQ Electric field Heating L a v element 1 Fig. 3. Heating element of the seat connected to ground through inductors L a and L b. he heating element will be at a certain otential, due to the sinusoidal source, and the couling caacitance C. A ictorial view of the electric field is also shown. he resence of L a and L b hels to achieve a resonance condition but introduces an additional imedance in the heating element circuit. his imedance can be minimised by relacing L a and L b with a Common mode Inductor (CI) as shown in Fig. 4. he CI connected in this manner rovides a low imedance for a differential signal [10] and hence to the ower suly of the heating element. his is, in articular, useful for a heating element connected to a modulated ower suly. In contrast, the CI rovides a large imedance to common mode signals, i.e., the measurement signal Euivalent electrical circuit An euivalent electrical circuit of the entire system is deicted in Fig. 4. he caacitances C H1 and C H2 reresent the euivalent distributed caacitance between the heating C L b

3 element and the ground lane. C is the caacitance between the heating element and the receiver electrode. HE denotes the resistance of the heating element. V S is the dc ower suly for the heating system. he switch S is used to regulate the heating system. he control voltage v c from a temerature regulator controls the switch S. Fig. 5 illustrates a simlified circuit of Fig. 4 at resonance condition (measurement condition) as exlained in section 2.1. he couled circuit is relaced by an euivalent common mode inductance L E, rovided by CI. he caacitance C E reresents the euivalent caacitance between the heating element and near by surfaces at ground otential. his includes a arallel combination of caacitances C H1, C H2 and C. According to the euivalent circuit in Fig. 5, the euation for resonance freuency (1) modifies to f r = 2π L E ( C + C + C + C ) H1 1 H 2. (2) A sinusoidal excitation at freuency f r gives a arallel resonance condition for the circuit shown in Fig. 5. In this condition, the otential of the heating element attains maximum value as the current flow from to ground through L E and C E reaches its minimum. hus, as can be seen in Fig. 4, the receiver () receives the maximum dislacement current. S V S CI C H2 v c A C HE Fig. 4. Euivalent electrical circuit of the system. HE is the resistance of the heating element. C H1 and C H2 reresent the distributed caacitances between the heating element and ground. When a assenger (adult) occuies the seat, the caacitances C H1 and C H2 change. An adult human has good caacitive couling [11] (above 100 F) to ground. hus, the resence of a assenger acts as a caacitive load for the heating element at resonance. Hence, the otential of the heating element reduces as the human body comes closer. his effect reduces the dislacement current received by the receiver. Another imlication of the change of C H1 and C H2 is the shift in the resonance freuency f r. For an adult occuied condition, the caacitances C H1 and C H2 increase. Conseuently, the resonance freuency gets shifted to say f r1. According to (2), f r1 < f r, conseuently the heating C C H1 element circuit is no longer in resonance at f r and thus it will not develo a useful electric field as described in Fig. 3. he change in outut due to such a condition is significant esecially for a heating element circuit with high uality factor. Ultimately, for a system with excitation freuency at f r, considerably lower dislacement current is received by the receiver, when the seat is occuied by an adult assenger. Once an adult assenger occuies a seat, the seat foam gets comressed and the entire or art of the heating element, deending on the weight and osition of the assenger, moves downwards by a few centimetres. hese factors increase the values of the caacitances C and C. Hence, the receiver receives more dislacement current comared to a vacant condition. But this effect is low comared to the effect due to changes in caacitances C H1 and C H2. L E Fig. 5. Simlified electrical euivalent circuit showing the resonance condition. Caacitance C E is the arallel combination of C H1, C H2 and C. L E is the effective inductance offered by the couled inductor (CI) to a common mode signal Measurement Princile A synchronous detection method [8] is used for the measurement of the dislacement current received by the receiver. Fig. 6 indicates a block diagram reresentation of the measurement system. he source rovides a sinusoidal signal with variable freuency. During the Outut Heating element ADC Z S C PGA L E C E Amlifier Mixer Pre-amlifier and LPF Fig. 6. Block diagram reresentation of the measurement setu. During the investigations, the heating element was owered from a dc source V S = 12 V.

4 measurement, the freuency of is set to the resonant freuency f r of the heating element circuit. he dislacement current received by receiver flows to ground through a low imedance shunt Z S. he dro across Z S is amlified before the mixer and Low-Pass Filter (LPF) stages. Offset voltages that aear at the outut of LPF are removed [8] before giving the outut of LPF to a Programmable Gain Amlifier (PGA). he PGA amlifies this offset comensated signal to an otimum voltage for the Analogto-Digital Converter (ADC). he ADC rovides the final outut roortional to the dislacement current received by the receiver. 3. EXPEIMENAL SE-UP AND ESULS A rototye seat occuancy system has been develoed and tested on an automobile seat. he transmitter and receiver electrodes were fabricated using coer sheets with 100 µm thickness. he transmitter and receiver electrodes have identical dimensions with a length of 12 cm and a width of 2 cm. hese electrodes were fixed on an insulating medium with a thickness of 1.5 cm which was firmly fixed to the seat frame. Inductors were inserted between the dc ower suly V S and each of the nodes and of the heating element. he switch S was shunted with a 200 nf caacitor. his makes the switch to be a low imedance ath for high freuencies even if it is in oen condition, while this will not affect its low freuency oeration as a switch. A measurement system, according to section 2.3, develoed for caacitance measurement was used for measuring the voltage roortional to the dislacement current (I D ). he details of this raid rototying system have been reorted in [12]. In the beginning of the measurement, the freuency of the measurement source was varied from 500 khz to 5 MHz and occurrence of a resonance was noticed at 1.5 MHz. he resonance freuency was verified with a Network Analyzer, model 8712E from Agilent echnologies. An excitation at this freuency was used for the investigations. Outut (Normalized) Heating ON ef. C S Adult occuancy Heating OFF Data oints Fig. 7. Outut signals recorded for an adult occuancy, when the heating element was turned ON (left side) and turned OFF (right side) conditions. C S indicates the outut from the sensor at resonance freuency. he reading corresonding to a vacant condition is shown by a reference (ef.) signal. A noticeable change in outut (C S ) was observed for adult occuancy in comarison with the reading obtained for a vacant seat condition (ef.). A tyical change in outut noticed for such a condition was recorded and shown in Fig. 7. Fig. 7 shows the outut for adult occuancy when the heating element was switched ON and switched OFF resectively. he difference in readings between the ON and OFF conditions of the heating element was found to be less than 2 % of the outut noted for an adult occuancy. Change in outut was also noted for conditions such as resence of water bottles, mobile hones, vacant infant seats, and battery owered latos. he changes noted in the outut for the various test cases mentioned above were very low comared to an adult occuancy condition. he results are tabulated and given in able 1. he change in outut is comarable with an adult occuancy condition when a lato connected to ower suly was laced on the seat. his roblem has been tackled by taking another simultaneous measurement, with the same sensor, at a freuency far below f r. his has been successfully verified in the rototye. able 1. Caacitance change noted for different conditions. est conditions Outut (Normalized) Adult occuancy 1.00 Water bottle (5 litres) 0.19 Mobile hone (dialling) 0.01 Infant seat (vacant) 0.06 Infant seat with a dummy infant 0.29 Lato floating 0.06 Vacant seat 0.00 Fig. 8 deicts the outut (C S ) of the sensor at resonance freuency and outut (C P ) at 0.5 MHz for adult occuancy, electrically floating lato and lato (grounded osition as far as the sensor is concerned) owered from main suly resectively. It can be seen that C S shows similar outut for adult occuancy and lato in grounded condition but C P gives large change for adult occuancy and very low change for lato in grounded osition. hus, adult occuancy can be distinguished from a lato in grounded condition. he change in C P during adult occuancy is mainly due to the additional shielding of the electric field owing to the movement of the heating element in down-ward direction. his does not haen for a lato, as its weight is very low comared to a human. Outut C P alone is not sufficient for reliable occuancy detection because it is also sensitive to weight of other objects such as water/beer bottles. hus, C S in combination with C P, i.e. measurement with an excitation freuency eual to f r and another freuency lower than f r rovides sufficient information for reliable seat occuancy detection. 4. CONCLUSIONS A simle and efficient seat occuancy sensing system suitable for automobile seats with an integrated electric heating system is resented. In the resented sensing system, the resence of the seat heating element is advantageously used for occuancy detection. he sensor system introduces a few assive comonents in the heating

5 system and does not affect the standard oeration of it. A rototye detection system has been develoed and tested. est results from the rototye indicate the usefulness of the roosed method [10] M. J. Nave, On modelling the common mode inductor, International Symosium on Electromagnetic Comatibility, , Cherry Hill, NJ, USA, December [11] N. Jonassen, Human body caacitance: Static or dynamic concet?, Proc. Electrical Overstress/Electrostatic Discharge Symosium, 1998, , eno, NV, USA, June [12] D. Hrach, H. Zangl, A. Fuchs and. Bretterklieber, A versatile rototying system for caacitive sensing, Sensors and ransducers Journal, vol. 90, , Aril Caacitance Outut (Normalized) (Normalized) ef. C P 1. Adult occuancy 2. Lato floating 3. Lato grounded C S Data oints Fig. 8. Outut signal obtained for conditions such as an adult occuancy, lato floating and lato grounded conditions resectively. he measurement result at f r (1.5 MHz) is indicated by C S and that at 0.5 MHz is marked by C P. EFEENCES [1] National Highway ransortation and Safety Administration, Federal Motor Vehicle Standard [2] Ching-Yao Chan, A treatise on crash sensing for automotive airbags, IEEE/ASME rans. on Mechatronics, vol. 7, no. 2, , June [3] A. Giralt, M. Devy, M. Antonio, Detection and classification of assenger seat occuancy, IEEE Intelligent vehicles Symosium, , Dearborn, USA, October [4] K. Kasten, A. Stratmann, M. Munz, K. Dirscherl, S. Lamers, ibolt-echnology-a weight sensing system for advanced assenger safety, in Book itled, Advanced Microsystems for automotive alications 2006, Sringer Berlin Heidelberg, , July [5]. Saito, M.Ofuji, K. Jino, M. Sugino, Passenger detecting system and assenger detecting method, U. S. Patent , March [6] B. George, H. Zangl,. Bretterklieber and G. Brasseur, Seat occuancy detection based on caacitive sensing, IEEE rans. on Instr. and Meas., vol. 58, no. 5, [7] J.. Smith, Electric field imaging, PhD dissertation, Massachusetts Institute of echnology, Cambridge, MA, [Online],Available: htt://ubs.media.mit.edu/ubs/aers/99.02.smithhd.df. [8]. Bretterklieber, H. Zangl, M. Motz, D. Hammerschmidt and. Werth, Versatile Sensor front end for lowmodulation caacitive snesors, IEEE I 2 MC 2008, , Britsh Columbia, Canada, May [9] Larry K. Baxter, Caacitive sensors: Design and alications, IEEE ress series on Electronics echnology. Published by the Institute of Electrical and Electronics Engineers, Inc., New York, 1997.