Membrane Fouling Prediction Using Normalized Operating Indicators
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current prediction systems for membrane performance and replacement are unreliable due to variations in environmental conditions, membrane ageing, and changes in operating configurations, requiring reconfiguration of models with new hardware, and lack standardization in membrane properties.
Innovation Solution
A method for predicting membrane replacement and cleaning dates using normalized operating indicators generated through a learned normalization model, independent of environmental and operational conditions, by applying regression on sensor data to standardize indicators for fouling and ageing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If prediction models are based on manufacturer data and hardware configurations, then initial prediction accuracy is improved, but model reliability deteriorates when membrane types or operating stages are changed
Solution Approach 1:
The patent transforms raw sensor data into normalized operating indicators by applying learned normalization models that adjust for different membrane types, operating stages, and environmental conditions. This parameter transformation allows the same prediction model to accurately predict maintenance needs across diverse hardware configurations without requiring model reconfiguration, thereby maintaining both initial accuracy and long-term reliability.
Solution Approach 2:
The patent introduces normalized operating indicators as intermediary variables between raw sensor measurements and prediction outputs. These indicators serve as a standardized language that translates diverse hardware-specific data into a universal format, enabling the prediction model to work consistently across different membrane types and operating conditions without direct dependence on hardware-specific parameters.
2Adaptability or versatility
If prediction systems account for multiple factors influencing membrane service life, then prediction comprehensive is improved, but system complexity increases
Solution Approach 1:
The patent segments the complex prediction problem into distinct components: (1) collecting raw sensor data, (2) transforming data into normalized operating indicators through learned models, and (3) generating predictions based on standardized indicators. This segmentation allows each component to be optimized independently and simplifies the overall system by separating the complexity of data transformation from the prediction logic.
Solution Approach 2:
The normalized operating indicators serve multiple functions simultaneously: they standardize data from different hardware configurations, compensate for environmental variations, and provide a unified input format for predictions across different membrane types and operating stages. This multi-functionality reduces system complexity by eliminating the need for separate handling of each factor.
3Adaptability or versatility
If models are reconfigured when hardware changes are made, then adaptation to new conditions is improved, but loss of time increases
Solution Approach 1:
The patent performs preliminary action by pre-training learned normalization models on historical data from various membrane types and operating conditions before they are needed for prediction. These pre-trained models can immediately process data from new hardware configurations without requiring real-time reconfiguration, as the normalization relationships are already established from prior learning.
Solution Approach 2:
The patent creates a universal prediction model that copies the prediction logic across different hardware configurations through normalized indicators. Instead of creating separate models for each membrane type or operating stage, the system uses a single model that processes standardized indicator data, eliminating the need for time-consuming model reconfiguration when hardware changes occur.
4Reliability
If standardized indicators are generated independent of environmental conditions, then prediction reliability is improved, but measurement precision requirements increase
Solution Approach 1:
The patent applies learned normalization models that transform raw sensor measurements into normalized operating indicators by adjusting for environmental conditions such as temperature and humidity. This parameter transformation does not require ultra-precise sensor data; instead, it uses the learned relationships between sensor readings and environmental factors to compensate for variations, thereby achieving reliable predictions with standard measurement precision.
Data Source
AI summary
The present invention relates to a method for automated data processing to assess the state of multiple filtration membranes used in liquid filtration. The method involves receiving data from state sensors positioned within or near a set of membranes, which process incoming water into permeate and concentrate flows. This data, collected as time series at predefined frequencies, pertains to external physical parameters. An operating indicator is determined from this data, forming a second time series. Both the first and second time series are recorded as point clouds over a specified acquisition period. An intermediate operating indicator is generated, representing the state of new, clean, or cleaned membranes, using a learned normalization model. Finally, a normalized operating indicator is produced, characterizing membrane fouling and aging, independent of environmental variations, and forming a third time series. This method enhances the accuracy of membrane state assessment in filtration systems.