Device and Method for Determining a Physical Quantity

Simple SummaryContent extracted from patent full text and abstract with AI.

This invention describes a device and method for accurately determining a physical quantity (such as position, temperature, or pressure) from multiple measured values, some of which may contain errors. The method groups the measurements into overlapping subgroups with built-in redundancy, then calculates a reliability score for each group using statistical tests. These group reliability values are used to assign individual evidences to each measured value, which are combined to provide an overall evidence score for every measurement. Based on these evidences, the system either weights or excludes unreliable data, enabling a more accurate and robust calculation of the desired physical quantity, even when multiple measurement errors exist simultaneously.

Use CasesContent extracted from patent full text and abstract with AI.

• Satellite-based navigation systems (e.g., GPS, Galileo) for robust position determination under error-prone conditions.
• Industrial process control where sensor readings (temperature, pressure, flow) may be affected by noise or outliers.
• Data acquisition systems in scientific experiments requiring high-accuracy computation from multiple, noisy sensors.
• Structural health monitoring (bridges, buildings), where multiple sensors measure physical parameters subject to faults or bias.
• Quality control and calibration of measurement equipment with redundant sensor arrays.
• Automotive and aerospace systems that rely on redundant sensing for safety-critical applications (e.g., autonomous driving, avionics).

BenefitsContent extracted from patent full text and abstract with AI.

• Enables detection and exclusion of multiple simultaneous faulty or biased measurements, not just single outliers.
• Does not require fixed, empirically determined thresholds for outlier rejection, increasing flexibility and adaptability.
• Allows weighted use of measurements rather than strict inclusion/exclusion, preserving valuable information from partially corrupted data.
• Improves overall accuracy and reliability of physical quantity determination in systems with redundant measurements.
• Applicable to a wide range of physical measurements (position, temperature, pressure, voltage, etc.), making the method highly versatile.
• Supports both hardware and software implementation, and can be integrated into existing measurement and control systems.

Technical Classifications (CPCs)

Inventors & Applicants

Patent Abstract

In a method and a device for determining a physical quantity from a number of measured values containing errors, first of all a grouping of the number of measured values containing errors into a plurality of subgroups of measured values is executed, wherein each subgroup comprises a redundancy, so that more measured values than the number of measured quantities are contained in each subgroup. Hereupon, a reliability quantity for each subgroup is calculated based on the redundancy contained in the subgroup. In a further process an individual evidence is allocated to the measured values containing errors of each subgroup based on the reliability quantity for the respective subgroups. An evidence determiner then determines one overall evidence each for each measured value containing errors based on the individual evidence quantities for a respective measured value. Hereupon, a processor calculates the physical quantity using at least some of the measured values containing errors and at least some of the overall evidences.