High-precision Electromagnetic Flowmeter with Empty Pipe Detection via Complex Programmable Logic Device-based Waveform Recognition

Summary

This study improves electromagnetic flowmeter accuracy by optimizing excitation waveforms, applying multi-stage filtering, and using Complex Programmable Logic Device (CPLD)-based rectification. A novel waveform-based empty pipe detection method enhances reliability. Experiments show 0.1% accuracy within 0.1-15 m/s, validating industrial applicability.

Abstract

Traditional electromagnetic flowmeters are inherently prone to external interference and uneven velocity distribution during measurement, which severely limits their accuracy. In this study, an improved method is proposed, which optimizes the excitation drive waveform, performs multiple filtering and amplification of the electrode input, and uses a Complex Programmable Logic Device to achieve rapid switching between positive and negative induction signals. This enables smooth rectification and, in combination with software filtering techniques, achieves highly precise performance. Additionally, empty pipe detection is realized by recognizing excitation waveform and input waveform patterns.

Experimental verification shows that the designed electromagnetic flowmeter achieves an accuracy of 0.1% within a flow velocity range of 0.1-15 m/s, with system repeatability errors of less than 1%. The results validate the effectiveness of the proposed method in high-precision flow measurement. The study demonstrates that high-precision detection can be achieved with minimal additional cost, which is important for industry applications.

Introduction

Electromagnetic flowmeters are flow measurement instruments that operate based on Faraday's law of electromagnetic induction. Compared with traditional mechanical flowmeters, electromagnetic flowmeters exhibit superior adaptability to various media and have lower requirements for straight pipe sections¹. When fluid passes through the pipeline, the electromagnetic flowmeter generates a magnetic field and measures the induced voltage difference in the fluid to calculate the flow velocity². Electromagnetic flowmeters are particularly suitable for complex environments like those in the chemical and petroleum industries^3,4,5. However, due to their operation in harsh environments, the accuracy of electromagnetic flowmeters is easily affected by external interference⁶, necessitating advancements in detection technologies to improve accuracy⁷.

Accuracy can be improved in several ways. Optimizing the electrode shape has been shown to effectively enhance precision⁸, and optimizing the excitation coil's magnetic field can significantly improve flow measurement accuracy while maintaining magnetic field uniformity⁹. Additionally, improvements in drive waveforms, such as using dual-frequency driving, can effectively boost precision¹⁰. However, these methods still face issues of insufficient adaptability and limited flexibility when dealing with dynamic changes in complex environments.

To improve the performance of electromagnetic flowmeters in complex environments, this study implements two key enhancements aimed at improving accuracy and stability. First, a multi-stage step waveform drive is implemented to suppress high-order harmonics and optimize excitation waveforms. Second, signal processing is enhanced through a combination of Complex Programmable Logic Device (CPLD)-based hardware filtering, rectification, and software-based filtering techniques.

The step waveform drive controlled by the analog switch effectively suppresses high-order harmonics that typically arise in traditional methods. By adjusting the current step amplitude and switching timing, the excitation waveform is optimized, reducing interference with the electrodes. Additionally, after undergoing multi-stage amplification and band-pass filtering, the signal is effectively denoised and its strength is enhanced. Furthermore, the positive and negative half-cycle signals are separated and recombined to ensure signal stability, leading to improved measurement accuracy. The integration of these two enhancements significantly boosts the precision and anti-interference capability of the flowmeter, making it more reliable in complex industrial environments.

In industrial applications, pipelines may not always be fully filled with fluid. If the fluid level falls below the measurement electrodes, the electromagnetic flowmeter cannot provide valid flow velocity readings, making empty pipe detection a critical aspect of system reliability. Traditional empty pipe detection methods primarily rely on conductivity variations, but these are highly susceptible to changes in fluid composition and concentration, leading to instability under dynamic conditions.

To address these challenges, alternative detection strategies have been explored. A method based on electrode capacitance variation has been proposed¹¹, but its performance deteriorates when fluid properties change or when external interference is present. Similarly, an approach utilizing interference amplitude variations has been introduced¹²; yet its threshold-based detection mechanism is significantly influenced by the type of liquid, limiting its adaptability. These limitations underscore the need for a more robust and adaptive solution.

In this study, a waveform-based empty pipe detection method is also proposed, leveraging the correlation between excitation waveforms and signal processing mechanisms to analyze waveform characteristics. This method effectively improves detection accuracy by eliminating dependencies on amplitude variations or conductivity fluctuations. More importantly, it enhances stability and reliability, particularly in complex industrial environments where fluid properties and external disturbances frequently change.

In summary, this study presents a high-precision electromagnetic flow measurement method that enhances accuracy and stability in complex environments. The proposed method integrates a multi-stage amplification and filtering process with an optimized excitation waveform and CPLD-based rectification to effectively suppress high-order harmonics and reduce noise interference. Additionally, software-based filtering techniques are incorporated to further refine the signal, enhancing measurement stability and reducing the impact of external disturbances. Furthermore, an empty pipe detection approach based on waveform pattern recognition is introduced, providing enhanced detection reliability compared to conventional amplitude- or conductivity-based methods.

It is worth noting that velocity non-uniformity in pipelines can introduce significant measurement errors¹³. Therefore, this study assumes a uniform velocity distribution as a prerequisite to ensure high-precision flow measurement. Experimental results demonstrate that the proposed approach achieves a measurement accuracy of 0.1% within a velocity range of 0.1-15 m/s, with a repeatability error of less than 1%. These findings validate the effectiveness of the proposed methodology and offer a promising solution for high-precision industrial flow measurement applications. Future research will focus on further evaluating the method's adaptability to varying fluid properties and external disturbances to enhance its robustness in real-world environments.

Protocol

1. Processing of electrode-induced voltage

1. Take the induced electromotive force from both sides of the sensor as the input signal (Figure 1A).
   ​NOTE: The original voltage signal is extremely weak and severely contaminated by noise, exhibiting a low signal-to-noise ratio (SNR).
2. Apply a 10x differential amplifier to amplify the signal (Figure 1B).
3. Proceed to feed the signal into an active second-order bandpass filter comprising cascaded high-pass and low-pass filter stages. Initially, remove low-frequency components through the high-pass filter, then channel the filtered output via a coupling capacitor to the subsequent low-pass filter stage. At this phase, suppress residual high-frequency noise, with the resultant output waveform illustrated in Figure 1C.
4. Amplify the denoised signal using an inverting amplifier, as illustrated in Figure 1D.
5. Implement a gain of -1 through an inverting amplifier to convert the negative-polarity signal into positive polarity while preserving its amplitude unchanged.
6. Direct the positive and negative half-cycle signals (Figure 1E) to the two channels of the analog switch, respectively, and simultaneously input both signals into the comparator.
   1. Process the two output signals generated by the comparator using a CPLD to detect pipeline vacancy status and determine fluid flow direction.
   2. Utilize CPLD to control the channels of the analog switch, employing zero-crossing detection to precisely regulate switching timing and thereby introduce only minimal delay (Figure 1F).
7. After gating via an analog switch, feed the signal into the third-stage signal amplifier.
8. Apply an integrating low-pass filter to process the signal, then transmit the processed signal (Figure 1G) to the microcontroller unit (MCU) for subsequent computational operations.

2. Implemented schematic and working principle

1. Position the signal amplifier as illustrated in Figure 2 to amplify the signal by a factor of 10.
2. Connect the signal amplifier to the band-pass filter.
3. Connect the secondary amplifier to the output of the band-pass filter. Directly buffer the positive half-cycle signal for output while routing the negative half-cycle signal through an inverter prior to its input into the analog switch.
4. Configure two comparators beneath the analog switch. Transmit the comparator output signals to the CPLD, and utilize the CPLD to control the analog switch's on/off states based on sequential logic.
5. After undergoing secondary filtering, input the rectified signal from the analog switch output into the variable-gain amplifier.
6. Route the processed signal through the low-pass filter into the analog-to-digital (AD) conversion channel of the processor.

3. Forward and reverse flow determination

1. As illustrated in Figure 3A, observe that the forward flow mode is characterized by the falling edge of the excitation signal corresponding to the low-level forward conduction signal.
2. Observe that the reverse flow pattern illustrated in Figure 3B manifests as a temporal correspondence between the falling edge of the excitation signal and the activation of the high-level forward conduction signal.
3. Employ a CPLD to differentiate two characteristic signal patterns, thereby achieving precise discrimination between forward and reverse water flow.

4. Linearity correction

1. Apply the piecewise linear correction method to rectify the input signal using the following mathematical expression of the correction function:
   
   Where y is the corrected flow rate, f is the flow rate generated by the standard instrument, n is the number of segments, k_i is the correction coefficient for the i-th interval, and x_i is the upper boundary value of the i-th interval.
2. Derive the correction coefficient formula based on the least squares method using the linear regression slope formula, using the following mathematical expression:
   
   Where k is the correction coefficient, n is the number of data points, x_i is the flow rate measured by the experimental instrument, y_i and is the flow rate generated by the standard instrument.

Results

To ensure standardized experimental conditions and the reliability of results, the experiment utilizes the hydraulic pump shown in Figure 4 to generate a stable standard water flow as the experimental environment. The water flow generated by this hydraulic pump can be approximated as a constant-velocity stream due to its stable power output characteristics, thereby meeting the experimental requirements for uniform fluid delivery. The standard instrument used ...

Discussion

There are various implementations of excitation waveforms in electromagnetic flowmeters, among which square wave excitation and step wave excitation are two commonly used types. Square wave excitation is widely adopted due to its simplicity in implementation¹⁵. However, this method is prone to inducing eddy current effects during the transient phase of excitation switching, which negatively impacts the stability of the measurement signal¹⁶. Additionally, the issue of zero-p...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Electromagnetic flowmeter	ABB	ABB-DN50	As a standard instrument, it is compared with the instrument in this article.
Electromagnetic flowmeter sensor	ABB	ABB-DN50	Used for collecting induced electromotive force.