Fuel Volume Estimation Using Sensor Fusion Under Measurement Uncertainty

The patent combines multiple data sources (fill level sensor readings, fuel consumption data from volume flow sensors, temperature data, pressure data, and inclination data) into a unified determination system. This merging of multiple measurement approaches allows the system to compensate for individual sensor limitations and achieve higher measurement precision despite the increased complexity of integrating multiple data sources.

The system implements feedback mechanisms where fuel consumption data is continuously fed back to update the fill level determination. By comparing expected fuel volume based on consumption patterns with actual sensor readings, the system can correct for systematic errors and maintain high measurement precision over time, accounting for the complexity of continuous data processing and correction algorithms.

The patent introduces intermediate correction factors that act as mediators between raw sensor data and final fuel volume determination. These correction factors, derived from temperature, pressure, and inclination measurements, compensate for systematic errors without requiring complete redesign of the sensing system. The complexity is managed by using mathematical correction models rather than additional physical sensors.

The system dynamically adjusts determination parameters based on changing environmental conditions. By monitoring temperature, pressure, and inclination parameters, the system modifies its calculation approach to account for thermal expansion, pressure-related deformation, and geometric changes in the fuel container, thereby maintaining reliability across varying operating conditions without requiring a fundamentally complex adaptive system.

The system performs preliminary actions by pre-calculating and storing correction factors, characteristic data, and lookup tables for common operating conditions. This allows the recursive estimation algorithm to quickly retrieve pre-computed values rather than performing complex calculations in real-time, thereby maintaining high measurement precision while reducing the time loss associated with continuous processing.

The patent implements a hybrid approach where full recursive estimation is performed only when necessary, while relying on simplified methods or pre-stored values for routine conditions. This partial application of the computationally intensive algorithm reduces overall processing time while maintaining precision where it matters most, balancing the trade-off between measurement accuracy and time consumption.