Transmembrane pressure difference inference device and diffused air amount control device
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing approaches to diffused air volume control in membrane filtration operations are not always effective in managing transmembrane pressure, leading to potential fouling and inefficiencies, as they do not account for the dynamic nature of membrane conditions.
Innovation Solution
A transmembrane pressure inference device and diffused air volume control system using a regression model based on machine learning to predict and manage transmembrane pressure changes over time, adjusting air diffusion accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Stress or pressure
If the diffused air volume is increased to reduce the increase speed of transmembrane pressure, then the transmembrane pressure increase speed is reduced, but the transmembrane pressure may increase and cause fouling
Solution Approach 1:
The system dynamically adjusts the diffused air volume based on real-time transmembrane pressure data and predictive models, rather than using fixed control rules. The control amount determination section modifies the diffused air volume dynamically to achieve optimal pressure control while preventing fouling
Solution Approach 2:
The system uses transmembrane pressure sensors to continuously monitor pressure changes and feeds this information back to the predictive model. The model then predicts future pressure values and adjusts the diffused air volume accordingly, creating a closed-loop feedback control system that prevents fouling while managing pressure
2Stress or pressure
If the diffused air volume is decreased to reduce transmembrane pressure, then the transmembrane pressure may be reduced, but the increase speed of transmembrane pressure may increase
Solution Approach 1:
The predictive model forecasts future transmembrane pressure values before they actually occur. By predicting the pressure trajectory in advance, the system can adjust the diffused air volume proactively to prevent excessive pressure buildup or rapid pressure increases, rather than reacting after the problem occurs
3Stress or pressure
If typical control approaches are used to manage transmembrane pressure, then pressure control is attempted, but the system does not account for dynamic membrane conditions leading to inefficiencies
Solution Approach 1:
The system changes the control parameters dynamically by using predictive models that account for dynamic membrane conditions. Instead of fixed control rules, the system adjusts diffused air volume based on predicted transmembrane pressure trajectories, allowing optimal control that adapts to changing membrane states and maintains high filtration efficiency
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This system allows for appropriate membrane filtration operations by accurately inferring transmembrane pressure changes, optimizing air diffusion to prevent fouling and enhance efficiency.
Implementation Method 1
an air diffusion device configured to perform air diffusion through a membrane surface of the separation membrane
Data Source
AI summary
To appropriately infer changes over time in transmembrane pressure to perform an appropriate membrane filtration operation, without being bound by typical approach regarding the diffused air volume control. An inference device (2) includes: an input data acquisition section (21) configured to acquire input data derived from operation data that is measured during a membrane filtration operation, the operation data including a membrane filtration pressure and a diffused air volume; and an inference section (23) configured to (i) use a regression model (31) to infer transmembrane pressure-related data after the predetermined time, and (ii) execute an updating process M times (where M is an integer of not less than 2) while changing part of data included in the input data, the updating process being a process of updating the input data N times (where N is an integer of not less than 2) by changing the transmembrane pressure-related data included in the input data to the inferred transmembrane pressure-related data, so as to obtain M inference results on changes over time in the transmembrane pressure in a period up to N×the predetermined time.