Adaptive Fill Level Measurement Using Dynamic Model Weighting

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Classic averaging methods in level measurement technologies often lead to lag or inertia, causing delays in responding to changes in fill levels, which can result in overfilling or dry-running issues, especially when switching between filling and emptying a tank.

Innovation Solution

A fill level measuring device that uses a sensor device and a control unit to record measured values and update multiple independent mathematical models using Kalman filters, weighting and mixing them based on their predictive quality to accurately determine the fill level, thereby reducing noise and improving measurement accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If averaging methods are used to suppress noise in measured values, then measurement noise is reduced, but response time increases causing lag in detecting fill level changes

Engineering Contradiction:
Improvenoise suppressionVSAvoidresponse time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent implements dynamic adaptation of the averaging window size based on the detected situation (filling, emptying, or resting state). During filling or emptying operations, a smaller averaging window is used to maintain fast response, while during resting states, a larger window provides better noise suppression. This dynamic adjustment resolves the contradiction by making the noise suppression capability adaptive to operational requirements.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes the parameter of averaging window size according to the operational state. By detecting whether the tank is filling, emptying, or at rest, the system automatically adjusts the number of measured values to average, thereby optimizing the balance between noise suppression and response time for each operational phase.

Inventive Principle:
Principle #35Parameter changes

2Ease of manufacture

If a fixed averaging window is used, then implementation is simple, but accuracy decreases during transitions between filling and emptying

Engineering Contradiction:
Improveimplementation simplicityVSAvoidfill level detection accuracy
Core Design Contradiction:
Ease of manufactureVSMeasurement precision

Solution Approach 1:

The system transitions from a static fixed averaging window to a dynamic adaptive window that changes based on operational state. By implementing situation detection (filling/emptying/resting) and automatically adjusting the averaging parameters, the system maintains high accuracy during transitions while preserving relatively simple implementation through automated state-based control.

Inventive Principle:
Principle #15Dynamics

3Adaptability or versatility

If manual parameterization of averaging is required, then flexibility is provided, but operational complexity increases

Engineering Contradiction:
Improveparameter flexibilityVSAvoidparameter configuration complexity
Core Design Contradiction:
Adaptability or versatilityVSEase of operation

Solution Approach 1:

The system implements self-service by automatically detecting operational states (filling, emptying, resting) and autonomously adjusting averaging parameters without requiring manual intervention. This eliminates the need for users to manually parameterize averaging settings while maintaining the flexibility to adapt to different operational conditions, thereby resolving the contradiction between adaptability and ease of operation.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP2869042B1Model-based noise suppression for fill level measuring devices
Publication Date: 2016.08.10 VEGA GRIESHABER GMBH & CO
  • EP2869042B1 patent drawingFigure 1~2
  • EP2869042B1 patent drawingFigure 3~4
  • EP2869042B1 patent drawingFigure 5

AI summary

Model-based noise reduction is used in level gauges, employing independent mathematical models whose accuracy is calculated by comparison with measured values. The models are then weighted and combined, or a single model is selected to calculate the level at a later time. This results in effective noise reduction.