Instrumentation & Measurement Magazine 26-4 - 16

instrumentationand
measurementsystems continued
buffer to temporarily store samples and as internal pool for local
processing and data analysis. On the same 0.8 mm rigid
PCB, two analog to digital converters (ADC) (ADS131M06)
are included to support high speed analog sensors, such as
microphones. Moreover, an internal 6-axes (3D accelerometer
and gyroscope) IMU sampled at 100 Hz complete the main
board together with a power optimized clustered power supply
management. The energy harvesting circuit, placed on a
separate block, includes a 360 mW flexible solar panel with a
dimension of 7 × 15 cm. A Lithium battery of 1.85 Wh is used as
energy buffer, capable of supplying the sensor node for a minimum
of 6.5 days in normal operational conditions.
The three sensor PCBs are designed to be flexible and robust
to the elements, mounted on a 0.2 mm substrate, are
incapsulated on a 3D-printed waterproof enclosure, which
provide low aerodynamic impact with a maximum thickness
of 4 mm. The 40 supported barometers are the ST LPS27HHW.
This commercial and cost-effective sensor is the best-performing
water-resistant MEMS on the market, and it also includes a
temperature sensor to enable custom and on-site calibrations.
The LPS27HHW is an ultra-compact piezoresistive barometer.
It can measure from 260 to 1260 hPa absolute pressure range,
with a relative/absolute accuracy of ±2.5/±100 Pa, configured
with a sampling rate of 100 Hz. Aerosense further supports
five differential pressure sensors, allowing for true-differential
measurements and in-situ calibrations. The Pewatron 52-series
was selected due to its good accuracy and stability, ±0.5 of
the full scale. Moreover, it offers a compact solution with only
3 mm thickness and a multi-order compensation for offset
correction, sensitivity, thermal error and a non-linearity correction
algorithm. The sampling rate is hard-coded at 1 kHz
with 16-bits resolution. Lastly, an array of 10 MEMS microphones
is realized using the Vesper VM2020, selected due to its
low power consumption, only 446 μW, and an acoustic overload
point of 152 dBm that prevents the microphone saturation
in case of strong wind blowing directly on the sensing element.
Each microphone is sampled at 16 kHz with a resolution of 24
bit, generating a data-flow of 3.8 Mbps. Technical details regarding
the sensor requirements, geometrical characteristics,
metrological characterizations, and calibration have been investigated
in previous works [7], [10].
The power consumption at different operational settings
has been experimentally measured to estimate the Aerosense
battery lifetime [7]. We consider a 10-minute sampling interval
with a period of 120 minutes, for a corresponding duty
cycle of 8.3%. The system energy was measured outdoors at
various distances from the base station, ranging from 25 to
438 m. In these conditions, the power consumption during
operations can vary between 120 mW to 230 mW depending
on the distance between the base station and the sensor (variable
due to the intrinsic sinusoidal blade altitude oscillation).
The main contribution is given by the BLE, the barometers
16
that need up to 65 mW, microphones (32 mW including the
analog to digital conversion and the non-volatile memory),
and the CC2553P SoC (18 mW). From our measurements, the
average power consumption in active mode is 142 mW, while
considering a 8.3% duty cycle and a micro-W power consumption
in sleep mode, and the sensor node power stands
at 12 mW in average.
Field Demonstration of the
Measurement System
The system has been evaluated on an Aventa AV-7 wind turbine,
which has a rated power of 6.5 kW and blade length of
6.5 m. The Aerosense system was tested and improved during
two measurement campaigns with harsh weather conditions:
a first one with microphones and absolute pressure sensors
during the summer, with storms and air temperature reaching
40 °C, and a second one with the differential pressure sensors
and the absolute pressure sensors during winter, with snow
and temperature below -5 °C (Fig. 3a). The Aerosense system
was able to acquire field data from an operating wind turbine
during a period of more than 30 days in variable and harsh
weather conditions. Following this period, the node was removed
for maintenance and upgrading. It was installed in
a couple of hours, including the cleaning of the blade, the
positioning of the sensors and the filming of the system to determine
the position of the sensors.
The quick installation was made possible thanks to the 3Dprinted
enclosure which includes a robust adhesive tape that is
dedicated for wind turbine blades. After the installation of the
different parts and the protection of the cables (Fig. 3b), photos
of the Aerosense sensor node were taken and were used to
reconstruct the 3D shape of the blade section with the sensor
node using photogrammetry (Fig. 3c) [11]. Based on known
distances on the enclosures and the blade, the average error
was estimated at less than 0.5 mm. With the exact positions of
the sensors, it is then possible to quantify the motions of the
blade and integrate the pressure along the blade section to
measure the aerodynamic force.
To obtain accurate results, pressure measurements were
corrected using the IMU measurements, to remove the influence
of the hydrostatic pressure (due to the difference of height
during a blade rotation) and of the accelerations of the blades.
Moreover, the inertial data was subsequently used to infer the
local angular position of the blade, or to detect steady states
with absence of wind.
Fig. 4 shows the time-resolved measurements collected in
December, 2022 during an acquisition window of 10 s. It is possible
to see the oscillations due to blade rotation, in this case at
a rotational speed of 54 rpm. The oscillatory part is generated
by the gravitational force, with the sensor pointing upwards
or downwards depending on the blade position. On the other
hand, the constant component is given by the centripetal
IEEE Instrumentation & Measurement Magazine
June 2023
Instrumentation & Measurement Magazine 26-4

Table of Contents for the Digital Edition of Instrumentation & Measurement Magazine 26-4
