TL;DR
This paper presents a novel method for identifying tire pressure in real-time using only 3-axis accelerometers embedded within an "intelligent tire." By analyzing the characteristic frequency of the rim's vibration in the circumferential direction, the researchers demonstrate a reliable and accurate way to estimate tire pressure, offering a simpler and more integrated solution compared to traditional Tire Pressure Monitoring Systems (TPMS).
What Problem Does This Paper Address?
Accurate tire pressure is crucial for vehicle safety, performance, and fuel efficiency. Traditional Tire Pressure Monitoring Systems (TPMS) often rely on dedicated pressure sensors, which can add complexity and cost, especially when integrating with advanced vehicle systems like ABS or ESP, or with the concept of "intelligent tires" that aim to provide comprehensive tire status information. While accelerometers are increasingly used in intelligent tires to monitor parameters like deformation and forces, their application for direct tire pressure identification has been less explored. This paper aims to bridge this gap by proposing a method to leverage existing accelerometer data for precise, real-time tire pressure monitoring, thereby reducing the need for additional dedicated sensors.
Methods
The core of the proposed method lies in analyzing the vibration characteristics of the wheel rim. The researchers developed an intelligent tire prototype equipped with three 3-axis accelerometers: one on the rim well (Accelerometer C) and two on the tire tread and sidewall (Accelerometers A and B).
- Data Acquisition: Road tests were conducted with the intelligent tire prototype mounted on a vehicle. Data was collected at various tire pressures (150 kPa, 250 kPa, 350 kPa) across different road surfaces and speeds.
- Vibration Analysis: The vibration acceleration data from the accelerometers was processed using the Fast Fourier Transform (FFT) to obtain amplitude-frequency curves.
- Focus on Rim Vibration: While accelerometers on the tire tread and sidewall showed little distinction across different pressures, the accelerometer on the rim well (Accelerometer C) revealed a key insight: the peak frequency in the circumferential direction shifted with tire pressure.
- Characteristic Frequency Identification: The researchers identified a characteristic frequency range (35 Hz to 80 Hz) where this pressure-dependent shift was most prominent. They developed a method to extract this peak frequency, accounting for vehicle speed variations by using a weighted average of amplitude-frequency curves.
- Relationship Modeling: A mathematical model (Equation 7) was established to correlate the identified characteristic frequency with tire pressure, using a least-squares fitting approach.
- Verification: The developed scheme was verified with new experimental data at different tire pressures (200 kPa and 300 kPa), comparing the identified pressure with the actual pressure.
Datasets
The study utilized data collected from road tests performed with a prototype intelligent tire. The experiments involved driving a vehicle on a ~2 km route with varied road surfaces (asphalt, cement, damaged, snow-covered) under sunny, cold conditions (-20 °C). Data was acquired at three distinct tire pressures: 150 kPa, 250 kPa, and 350 kPa. For verification, additional tests were conducted at 200 kPa and 300 kPa. No pre-existing public datasets were used; the data was generated specifically for this research.
Key Insights and Ideas
- Circumferential Rim Vibration is Key: The most significant finding is that the vibration of the rim in the circumferential direction is highly sensitive to tire pressure, while tire tread and sidewall vibrations are less so. This suggests that the rim's structural resonance is directly influenced by the tire's internal pressure.
- Frequency Shift as a Pressure Indicator: The characteristic frequency of this rim vibration shifts predictably with tire pressure. As pressure increases, the peak frequency also increases within a certain range. This provides a direct, measurable link to tire pressure.
- Simplification through Accelerometers: The paper demonstrates that existing 3-axis accelerometers, often already considered for other intelligent tire functions, can be repurposed for accurate tire pressure monitoring, potentially eliminating the need for dedicated pressure sensors.
- Speed Independence of Characteristic Frequency: Crucially, the identified characteristic frequency itself is largely independent of vehicle speed, making the identification robust across various driving conditions. While peak amplitude is speed-dependent, the frequency at which it occurs is not.
Key Results
- The proposed method successfully identified tire pressure with high accuracy.
- At 200 kPa, the identified tire pressure had an error of 2.31% compared to the actual value, with 85% of identification results having an error below 5%.
- At 300 kPa, the identified tire pressure had an error of 0.13%, with 83% of results having an error below 5%.
- The characteristic frequencies identified for 150 kPa, 250 kPa, and 350 kPa were approximately 45.4 Hz, 50.2 Hz, and 53.2 Hz, respectively.
- The relationship between characteristic frequency and tire pressure was modeled and verified, showing a consistent trend.
Limitations and Future Work
- Limited Pressure Range: The study focused on a specific range of tire pressures (150-350 kPa). The behavior of the characteristic frequency at very low or very high pressures, or outside this tested range, is not fully explored.
- Accuracy of Instantaneous Extraction: The paper acknowledges that instantaneous peak-frequency extraction can be prone to errors, which are normally distributed. To mitigate this, they suggest averaging several continuous peak frequencies, implying that a more stable real-time estimate would involve temporal smoothing.
- Future Work: The authors suggest that this method can be integrated into future multi-functional intelligent tires, potentially replacing dedicated TPMS sensors. Further research could explore the behavior of this method under more extreme conditions (temperature, load) and investigate its integration into advanced vehicle control systems. The potential for using this technique to identify other tire parameters beyond pressure could also be a future avenue.
Summary updated