Design of a wireless power transfer system for assisted living applications

Advances in material science and semiconductor technology have enabled a variety of inventions to be implemented in electronic systems and devices used in the medical, telecommunications, and consumer electronics sectors. In this paper, a wireless charging system is described as a wearable body heater that uses a chair as a transmitter (Tx). This system incorporates the widely accepted Qi wireless charging standard. Alignment conditions of a linear three-element coil arrangement and a 3 × 3 coil matrix array are investigated using voltage induced in a coil as a performance indicator. The efﬁciency obtained is demonstrated to be up to 80% for a voltage of over 6.5 Volts and a power transfer of over 5 Watts. Our results and proposed approach can be useful for many applications. This is because the wireless charging system described herein can help design seating areas for the elderly and disabled, commercial systems, consumer electronics, medical devices, electronic textiles (e-textiles), and other electronic systems and devices.


I . I N T R O D U C T I O N
Smart clothing is a new trend in wearable technology aimed at improving human life and satisfaction. Such clothing is made using combinations of micro-electronics embedded in fabrics to create functional and fashionable solutions that meet people's everyday needs. An area of interest is assisted living (AL), which takes into consideration the delicate needs of elderly individuals. In many countries, such as the USA and European Union (EU) countries, modern medicine has increased the longevity of people's lives. However, such countries are also experiencing a decreasing birth-rate that is leading to a disproportionate amount of people aged 65 and above when compared to the younger working population [1]. With current trends, it is estimated that, in the next 30 years, the elderly (i.e. individuals aged 65 or older) will outnumber their younger counterparts in the EU, for example, two to one [2]. This phenomenon will create a higher demand for AL solutions that can be used to provide longterm care to the aging population [3].
The primary objective of AL technologies is to improve lives by providing security, comfort, and independence [4]. These technologies can be applied to different environments, from homes and offices to outdoor public facilities [5][6][7][8][9][10][11][12][13][14][15][16]. However, devices used in an AL system requires a continuous source of power which can be provided by supportive batteries (e.g. lithium-ion batteries) [17]. However, currently available batteries cannot always meet the demands of the devices that make up an AL system [18]. That is, prolonged use of an AL system's devices eventually requires battery recharging and replacement, which can be a tedious, expensive, and complicated task for the elderly and other individuals [15]. For example, replacing batteries in biomedical implants is expensive and a complex procedure. Research into energy harvesting [19][20][21], wireless charging [22][23][24][25][26], battery and power management [27][28][29] has been conducted for AL systems to improve the functionality and efficiency of device batteries.
Wireless charging offers a great solution to the problems with batteries as described above. For example, this way of charging batteries can be more convenient to a user of an AL system than wired charging and some early examples have been demonstrated [18], [30], and [31]. To address safety concerns of the technology, regulations and standards of implementing wireless power transfer (WPT) for charging should also be considered [32][33][34][35][36][37][38]. Currently, two main commercial WPT standards exist: (i) the Qi standard created by the wireless power consortium (WPC) and (ii) the A4WP standard the air fuel alliance (AFA). Both standards enable wireless charging techniques that use magnetic induction to transfer power from a transmitter (Tx) to a corresponding receiver (Rx) within a target device. WPC's Qi and AFA's A4WP standards are discussed extensively in [39][40][41]. The two standards meet the international and regional regulations on safety and electromagnetic interference set up by government and health agencies [32]. Safety considerations are made by selecting a non-radiative WPT system design described in [42][43][44][45][46][47][48][49][50]. More specifically, the research presented in this paper is directed to exploring the design of a WPT system for wirelessly charging a wearable heating belt.
The challenges faced by these wirelessly charged devices include: (i) electromagnetic safety levels that lead to regulatory restrictions, which require a low power source with higher transmission efficiency [47]; (ii) limited range of operation, which leads to positioning issues of the transmitter and receiver [48]; and (iii) problems with integrating the transmitter and receiver in a seamless way to the user's everyday environment [49,50].
Positioning issues have also been researched and overcome using different techniques in some works in [52], [55], [56], and [57]. In [52], primary coils were embedded into a desk surface which significantly improved the charging area. Further, [55] used a sequence of switchable couples of coils to adopt a transmitting link side that retains its coupling coefficient for a continuously moving receiver. In [56], a similar concept to [52] was adopted using magnetically coupled resonant wireless power, while [57] used a bowl-shaped transmitter coil for free positioning.
In our proposed design as reported in this paper, which is based on the widely accepted Qi wireless charging standard, the transmitter of the WPT system is integrated within a backrest of a chair, while the WPT system's receiver is integrated in the user's belt together with a novel far-infrared heating element (see Figs 1(a) and 1(b)). The belt can be charged while its user is seated in the chair. This paper also investigates and examines alignment and power transfer conditions of the charging area on a chair's backrest, considering different coil arrangements and arrays. The WPT system proposed can also be used in multiple charging scenarios, including AL environments. In this way, the proposed WPT charging and heating system can be seamlessly integrated into the user's life with ease and comfort.
The major factors that affect the performance of the WPT system described herein are: (i) distance and misalignment between the transmitter and receiver; and (ii) power transfer between the transmitter and receiver. Both are attributable to inductive power transfer, which is specified by the Qi standard [39]. Table 1 shows the initial design constraints of the system.
When the separation distance of the coils for the system is smaller than 1/10 th of the wavelength of magnetic field, an Equivalent Lumped Parameter Circuit Model of the system can be made as further described in [58].
The first constraint requires that the receiver to be near the transmitter of the WPT system. More specifically, and as shown in Fig. 1(b), the belt to be charged (which includes  the receiver) must be close to the chair's backrest (which includes the transmitter). Furthermore, the second constraint requires power transfer in our WPT system to occur at any sitting position. The proposed WPT system overcomes the constraints described above by using an array of coils to be embedded in the backrest of the chair (previously shown in Fig. 1) and as similarly used in [52] and [56]. Thus our proposed design as described in Fig. 2 can increase the coverage area for WPT. The main design constraints of the coil size of the system include the physical dimensions of the belt (for the receiver) and load (li-ion battery pack). The battery has a nominal input voltage of 7 V and an input power of 5 W as specified in [59].
Our design process is shown in Fig. 3 and physical dimensions of the coils were used as initial input parameters to calculate the required self-inductance of the coils and mutual inductance by the theory reported [60], which was then optimized based on the power transfer efficiency (PTE). The receiver coil had restrictions due to the belt size while there were no size constraints for the transmitter, other than it needed to fit within a chair. (1) The circuit was then tuned to obtain the resonant capacitors for the primary and secondary coils by following [61]. These parameters were then used in the simulation environment for further optimization and where the Keysight software Advanced Design System (ADS) was employed.

A) Transmitter parameters
The transmitting unit is based on the WPC1.1 Qi specification with a 5W power level [62]. A WPC A29-type coil was used and characterized with parameters as in [63]. The transmitter coil is fabricated from litz wire with nylon spinning having 180 strands of no. 40 AWG (0.08 mm diameter). The measured inductance and quality factor of the transmitter coil were 10 mH and 90, respectively. A ferrite sheet was placed underneath the coil to shield the electronics in the base of the charger [64].

B) Receiver parameters
The design constraints for the individual receiving coil (32 × 48 mm 2 ) was more due to the limited dimensions of the textile belt. During the initial design phase, efforts were made to ensure that the employed receiver coil was able to provide a suitable mutual inductance [65] and thus provide sufficient coupling; i.e. the coil had an inductance of 15 mH and a quality factor of 60 ( Fig. 4 illustrates the design process in terms of size, load resistance and coupling distance). Table 2 shows the performance data of the transmitter and receiver modules.

I I I . W I R E L E S S P O W E R D E S I G N
The wireless system transmits power from a transmitter embedded in a chair then a belt containing the receiver unit receives and harvests the power. The harvested power is used to provide a regulated supply voltage suitable for charging the lithium-ion battery. The connection between the battery and the receiver has been implemented on a breadboard. Voltmeters and ammeters were connected to allow power consumption to be measured, as shown in Fig. 4.

I V . S Y S T E M M O D E L I N G A N D S I M U L A T I O N
Modeling the wireless power system (transmitter and receiver modules) through a simulation tool is an important part of the design before implementing the physical system. This kind of simulation-based approach can help to estimate and quantify expected outcomes of the real time experimentation. The design of a wireless power transfer system for assisted living applications 43 simulation software Keysight ADS was used to obtain the power output at the receiver for any coupling condition using coil types of the chosen transmitter and receiver modules.
The transmitter and receiver modules operated using inductive coupling regulated by the Qi standard which operates at frequencies between 100 and 200 kHz. The transmitter module is made up of an oscillator capable of producing the required resonant frequency. An amplifier then drives alternating current into the primary coil. The receiver converts the signal to DC through rectifying circuits [39].
The combination of the BQ500215EVM-648 wireless power transmitter and the BQ51025EVM-649 receiver shown in Fig. 8 are expected to give an output of 7 V at a maximum current of 1 A.

A) Coupling factor and coil modeling
Coupling factor quantifies the strength of magnetic field generated by the transmitter coil and it varies between 0 and 1. It must be noted that, when k ¼ 1 this represents perfect coupling. Coupling is also highly dependent on the distance between the coils as will be shown in the paper.
From [66] and [67] the inductance of the transmitter and receiver modules were obtained respectively to calculate the coupling coefficient, k, before calculating the mutual inductance between the coils, M. Also, the coupling factor can be measured from an existing system as a relative open loop voltage, u, it can be given as follows from (5a) in [61]: where U 1 and U 2 are the voltages applied to coil, L 11 is the selfinductance of the transmitter coil, and L 22 is the self-inductance of the receiver coil. These parameters are also related to the general set of equations for coupled inductors as seen in [68]: Inserting equation (2) into equations (3) and (4) we get the following: These equations show that the voltage can be expressed in the form of flux density. This condition is possible when the transmitting and receiving coils are coupled strongly with each other, in other words, a very small distance is separating them [69]. The overall measured inductance is given by L S in [61], which is:

B) Resonant capacitors
The receiver circuit has two resonant capacitors C s , and a series resonant capacitor and C sp . These capacitors along with the receiver coil make up a dual resonant circuit. The two capacitors are used to tune the receiver coil and by the Qi standard, should be sized correctly based on the WPC specification [62], and where the receiver coil is then placed on the spacer, and L rx is measured with a stimulus of 1-V RMS and 100 kHz. These capacitors can be further calculated using the 100 kHz resonant frequency: and C sp is the secondary resonance considering 1 MHz: The quality factor must also be .77 after tuning and is given by: where R is the resistance of the coil. Table 3 below shows data calculated for the circuit modeling.

C) PTE simulations
Figure 5(a) shows the circuit schematic of the modeled wireless power system while Fig. 5(b) shows the ADS schematic of the system. This simulation model is expanded on and based on the reported findings from [70]. The source is an oscillator which generates the AC signal at 100 kHz that flows into the transmitter coil, L p generates an oscillating magnetic field, C p is the resonant capacitor calculated from equation (8) that generates a resonance for L p with a coil resistance of R equation (10). Through the mutual induction tool from ADS, energy from the magnetic field induces AC current in the receiver coil, L s . At the receiver the conventional bridge diodes made of Schottky diodes, convert the signal to DC. C s is the series resonant capacitor calculated in equation (8) while C sp is the parallel resonant capacitor calculated in equation (9) for the receiver coil under test. This receiver model is based on the Qi specification in [63] for type A5. Figure 6 shows the rectified voltage of a 15 V load to be between 6.7 and 7 V. The current is between 500 mA (0.5 A) and 460 mA (0.46 A). The output power was tabulated for a load resistance ranging from 3 to 200 V.
The lithium ion battery used in our system has a high resistance at full charge. Also, to measure the impedance, a battery must be at least 50% charged in practice. A completely drained or nearly empty battery has a high internal resistance. As the battery reaches 50% state-of-charge (SOC), the resistance drops and then increases again towards full charge [60]. Table 4 Fig design of a wireless power transfer system for assisted living applications 45 shows the output efficiency of the system when charging the battery for increasing resistance as SOC increases towards full charge at 200 V. At a low charge, the system has higher and over time the efficiency drops as the battery continues to charge.

D) Misalignment calculations
Simulations for different alignment conditions were done in MATLAB using equation (22) from [71]. The translational misalignment, d tm was simulated at a varying distance of separation, d r between the coils as shown in Fig. 7. Figure 8 shows the plotted power efficiency against a changing transition misalignment, d tm going from 0 to 7 cm at a varying distance of separation, d r , ranging from 0 to 4 cm. The simulations suggest high coupling at smaller d tm and d r distances and tend to zero as the coils are further separated as a result of a lowered coupling factor. The analysis suggests that a misalignment up to about 3 cm will not affect the performance of the proposed system (since PTE is well above 10%) as this is usually the normal sitting position of the user.

V . E X P E R I M E N T A L P R O C E D U R E
The amount of power transferred with the system varied depending on the load resistance connected to the receiver. Moreover, a range of dummy loads was tested to understand the battery's impedance since the battery's charging and discharging activity is a chemical process.

A) WPT efficiency
The first test procedure is to carry out measurements to determine the power efficiency of the WPT system. The connection was done between the wireless power system consisting of a
dummy load, a receiver and transmitter. The methods, steps, and relevant documentation are described in [66]. These steps serve as a precaution against wrong connections which could lead to components blowing up.

B) Three-Element Array
The next step was to replace the single coil with an array of three overlapping coils; i.e. a 3 × 1 array. The coils overlap each other to achieve uniform distribution of the magnetic field, without fields canceling each other [39]. The 3-coil array is the 760308106 wireless charging coil from Wurth Electronic made of ferrite core material. The middle coil has an inductance of 11.5 and 12.5 mH for the side coils. The coils were connected in parallel to the transmitter and tested with the receiver without the battery, as described in Fig. 9.  design of a wireless power transfer system for assisted living applications 47

C) 3 3 3 Coil array
In this set-up a couple of rows were added to the 3 × 1 array (as described above) to increase the free positioning range on the backrest. Only the primary coils next to the receiver activated the power transfer process. This coil arrangement was made of three single 3 × 1 coil array resulting in a 3 × 3 coil arrangement array as described in Fig. 10.

A) WPT efficiency measurement
The wireless PTE of the system was calculated by measuring the input power at the transmitter and the output power at the receiver as expressed by: h = P RX /P TX 100 %, . where h is the efficiency, P RX is the output power in watts and P TX   is the input power in watts. The efficiency of the system went up to 80% and decreased as the load resistance increased. Figure 11 shows the efficiency plot of the simulated and measured device. From this plot we are certain the WPT system works properly.

B) Power measurement
An open circuit test was done to estimate the batteries SOC, which is the maximum possible charge inside a battery at any one time. It mainly involves measuring the two terminals of a battery in open-circuit as described in [72] to obtain the voltage per cell. The battery was charged using an industrial charger and results from this were compared to the charging process done by the wireless power system, shown in Table 5, Fig. 12.
When the battery starts to charge, the voltage starts to rise quickly while the current is constant and then when charging progresses towards saturation, the current starts to drop while the voltage remains constant until terminated at stage 3. Meanwhile the charge capacity increases as shown in Fig. 13.    design of a wireless power transfer system for assisted living applications 49

C) Discharge testing
It was observed the custom-made battery pack was able to regulate the current going into it. This safety feature was necessary as the battery was used to power a wearable and textile-based commercial product. This led to the significant drop of charge current as compared to the charger which employs a voltage regulator circuitry.
The WPT system cannot raise the voltage per cell higher as the charge current becomes too low resulting in lower saturation which can led to a lower discharge time of 2.5 h compared to 5 h seen when the battery reaches a higher capacity   from the use of an industrial charger. Table 6 shows the voltage discharged after 3 hours when charged with the original charger and WPT system.

D) Alignment testing
One of the main issues of concern is the systems flexibility particularly relating to alignment. Coupling should occur at any part of the system regardless of the sitting position. A possible solution is the use of an array of coils at the transmitter. The single transmitter coil was replaced by a 3-coil array as shown in Fig. 14, Table 7.

V I I . 3 3 3 C O I L A R R A Y I M P L E M E N T A T I O N
The voltage induced at the secondary coil using a 3 × 3 primary coil array was compared to that of a primary 3 × 1 coil array as shown in Figs 14 and 16, this was done to determine the dependence of the flux linkage. The primary coil array was connected in such a way to have an approximate inductance value of 10 uH as specified in [66]. The inner diameter and outer diameter of a single coil element within the array was 44.8 and 15.5 mm, respectively. It has 13 turns and a wire diameter of 1.19 mm. The gap between coils within the array was 1.85 mm and a single 3 × 1 array (3 elements) has a dimension of 94.7 mm × 53.35 mm while the complete 3 × 3 array (9 elements) is 94.7 mm × 162 mm. The 3 × 3 array was found to have an increased coverage area of 187 cm 2 from that of 28.62 cm 2 . Figure 14(b) shows the coverage area as a function of voltage recorded for the 3 × 1 array, the concentration, and charge area is seen to have increased in Fig. 16(b). The induced voltage at the receiver was measured by moving the coil center at grid intervals of 17 cm × 11 cm and spaced at 1 cm apart. A similar technique was used in [73]. Fig. 15 shows the coil arrangement under the charge area.
High and low peaks for the induced voltage can be observed in Fig. 16. The presence of low peaks indicates a low charge area due to the fact that the coil array elements do not completely overlap each other or are not completely close enough to ensure uniform coverage. Thus, improvements in the transfer efficiency could be possible by using a more compact array.

V I I I . C O N C L U S I O N A N D F U T U R E W O R K
A WPT system was designed for a novel wearable heating system that provides warmth to its user. The wireless system consists of an array of coils at the transmitter and a single coil receiver that powers polymer resistors that generate heat when driven by applied current. This provides comfort and heat therapy for the wearer of the belt. The flux distribution was also investigated among two array configurations by moving the receiver across the array surfaces and recording the induced voltage. This helped determine the alignment conditions of the set of the arrays to achieve a considerable area of operation. Results in the paper demonstrated that a significantly increased charge area was made possible by the employed 3 × 3 coil array. Further improvements for the system can be made by designing an accompanying voltage regulator integrated circuit such as LM317 (variable regulator) or any fixed regulator such as LM7808 or LM7812 for better output voltage stability.

A C K N O W L E D G E M E N T
This work was supported in part by the Horizon 2020 European Project CSA-EU (709372) and Scottish Enterprise. The authors would like to indicate that the work is only the authors views and that H2020 is not responsible for any information contained in the paper. The authors would also like to acknowledge the financial support of the UK Engineering & Physical Sciences Research Council (EPSRC) through the programme grant entitled Sonopill (EP/K034537/2). This work was partially supported by the H2020-MSCA-IF-2018 grant #840854 ViSionRF.  Dr. Podilchak is a registered professional engineer (P.Eng.) and has had industrial experience as a computer programmer, and has designed 24 and 77 GHz automotive radar systems with Samsung and Magna Electronics. Recent industry experience also includes the design of highfrequency surface-wave radar systems, professional software design and implementation for measurements in anechoic chambers for the Canadian Department of National Defence and the SLOWPOKE Nuclear Reactor Facility. Dr. Podilchak has also designed new compact multiple-input-multiple-output (MIMO) antennas for wideband military communications, highly compact circularly polarized antennas for microsatellites with COM DEV International, as well as new wireless power transmission systems for Samsung. His research interests include surface waves, leaky-wave antennas, metasurfaces, UWB antennas, phased arrays, and CMOS integrated circuits. Dr  He is URSI fellow and chair of URSI Commission D: Electronics and Photonics. He is IEEE senior member. His research interests include energy harvesting and wireless power transmission, RFID technology, active antennas and phased array antennas, inkjet and 3D printed electronics, millimeter wave systems. He has published more than 200 papers in peer-reviewed journals and international conferences.

R E F E R E N C E S
IEEE and university awards such as Engineering Prize award (HWU) and Best PhD Thesis award (SDSMT).
Dr. Anagnostou serves or has served as associate editor for the IEEE Transactions on Antennas and Propagation, the IET Microwaves, Antennas and Propagation, IEEE Access, and as guest editor for the IEEE Antennas and Wireless Propagation Letters (two special issues), the Hindawi Intl' Journal of Antennas and Propagation (two special issues), and the MDPI Technologies (one special issue). He is a member of the IEEE-APS Educational Committee, a member of the Technical Program Committee (TPC) and session chair for IEEE-APS and EuCAP Intl' Symposia, and a reviewer for international journals such as Nature and IEEE Transactions. He is a member of Eta Kappa Nu (HKN) Honor Society, ASEE, and a professional engineer (PE) of the Technical Chamber of Greece (TEE).