PDC Sensor for Solid Level - Adaptive Signal Processing and Acoustic Compensation for Ultrasonic Measurement of Powders and Granules
This technical article explores the adaptive signal processing and acoustic compensation techniques for ultrasonic solid level measurement. It covers the algorithms for echo detection in dusty conditions, the compensation for material surface angle, the use of automatic gain control and time-variable gain, and the system's ability to adapt to different material types for reliable level measurement.
The adaptive signal processing for solid level measurement is designed to overcome the challenges of measuring powders and granules in dusty environments. The algorithm continuously monitors the received signal and adjusts the processing parameters based on the current conditions. The first stage is a digital filter that removes the high-frequency noise. The filtered signal is then rectified to extract the echo envelope. The detection algorithm looks for the first significant peak in the envelope that is above the dynamic threshold. The dynamic threshold is calculated based on the average noise floor over the previous measurements. The algorithm also uses a "slope" criterion: the echo from the surface should have a steep rise time. If the rise time is slow, it may indicate a dusty condition or a false echo. The algorithm can also switch to a "multiple-pulse" mode, where several pulses are averaged to improve the signal-to-noise ratio, at the cost of a slower update rate. The adaptive algorithms are implemented in the sensor's microcontroller, with the parameters optimized for the specific application during the commissioning.
PDC Sensor
Acoustic compensation for the material surface angle is important for accurate level measurement in silos. Due to the way solids pile up, the surface is not flat but has a natural angle of repose (typically 10-30 degrees). This means that the ultrasonic beam may not be reflected directly back to the transducer, reducing the echo amplitude and potentially causing a measurement error. To compensate, the sensor uses a combination of techniques. The sensor can be mounted off-center to direct the beam at the center of the pile, but this is not always possible. The algorithm uses the signal amplitude and the echo width to estimate the surface angle. If the amplitude is lower than expected and the echo is wider, it indicates a sloped surface. The algorithm then adjusts the gain or uses a lower threshold to still detect the echo. Some advanced sensors can also use multiple transducers (array) to steer the beam toward the area of the pile, improving the reflection. The compensation ensures that the measurement remains accurate even with varying material pile shapes.
The use of automatic gain control (AGC) and time-variable gain (TVG) is essential for solid level measurement. The AGC adjusts the receiver gain based on the peak amplitude of the previous echo, ensuring that the signal remains within the dynamic range of the receiver. If the echo is weak (e.g., due to a dusty condition), the gain is increased. The TVG is a time-dependent gain that increases the amplification for echoes arriving later, compensating for the signal attenuation over distance. The TVG profile is typically set to a logarithmic curve that matches the attenuation of the acoustic signal in air. For solid applications, the attenuation is higher due to dust, so the TVG curve is steeper. The combination of AGC and TVG ensures that the sensor can detect echoes from both near and far distances, even in challenging conditions. The gain settings are calibrated during installation and can be fine-tuned based on the specific material and environment.
The system's ability to adapt to different material types is a key feature of modern ultrasonic solid level sensors. Different materials have different acoustic properties: plastic pellets are highly reflective, while cement powder is absorptive. The sensor's algorithm needs to be adjusted accordingly. This is often done via a "material selection" menu in the sensor's configuration, where the user chooses the material type from a list. For new or unknown materials, the sensor can perform an "auto-learn" routine, where it measures the echo characteristics (amplitude, rise time, decay) and sets the gain and threshold accordingly. The auto-learn routine requires that the silo is at a known level (e.g., empty or full) to calibrate. Once the material is characterized, the sensor can automatically adjust its settings based on the changing conditions, ensuring reliable measurement throughout the fill cycle.
The integration of adaptive signal processing with acoustic compensation results in a robust level measurement system for solids. The sensor can handle the variability in dust levels, material surface shape, and material type, providing a consistent and reliable level reading. The ongoing development in this field is focused on the use of machine learning to automatically adjust the signal processing parameters based on the history of measurements. This would enable the sensor to learn the patterns of the specific process and optimize its performance over time. The ultrasonic solid level sensor is a key component in automated material handling systems, providing the critical data needed for inventory control and process optimization in industries ranging from food processing to mining and construction.
