Multi-Model Predictive Controller for Insulin Delivery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current artificial pancreas systems lack effective control methods to manage abrupt and slow variations in glucose dynamics, relying on limited control algorithms that struggle with constant or slowly changing glucose levels, and do not adequately handle rapid changes.
Innovation Solution
A closed-loop system that includes a medication delivery device and a controller with control logic to calculate a bolus delivery schedule, comprising an initial and delayed bolus delivery amount based on pre-meal glucose levels, using insulin and pramlintide, and employing a multi-model predictive controller algorithm to adapt to changing glucose conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional control algorithms are used in artificial pancreas systems, then the system can operate with simple control logic, but the system cannot effectively handle abrupt variations in glucose dynamics
Solution Approach 1:
The control algorithm transitions from static to dynamic by incorporating multiple models that can adapt to changing glucose dynamics. The system switches between different models based on the current glucose state, enabling it to respond to both abrupt and slow variations in glucose levels without requiring overly complex control logic.
Solution Approach 2:
The system changes control parameters dynamically by selecting different models with varying parameters based on the current glucose situation. This allows the algorithm to adjust its behavior to match the specific glucose dynamics being experienced, improving adaptability while maintaining manageable complexity through parameter selection rather than structural complexity.
2Reliability
If a single fixed insulin model is used, then the control algorithm is simple to implement, but the model cannot accurately represent varying insulin dynamics in different glucose conditions
Solution Approach 1:
The single insulin model is segmented into multiple models, each representing different insulin dynamics under specific glucose conditions. This segmentation allows the system to select the appropriate model for the current situation, improving accuracy without requiring a single overly complex model that attempts to represent all possible conditions simultaneously.
Solution Approach 2:
The insulin model transitions from a fixed, static representation to a dynamic multi-model structure that can adapt to changing glucose conditions. The system dynamically selects which insulin model to use based on the current glucose state, ensuring accurate representation of insulin dynamics across varying conditions while maintaining implementation simplicity through modular model selection.
3Speed
If rapid response to glucose changes is implemented, then glycemic control improves, but the risk of oscillations and instability increases
Solution Approach 1:
The control algorithm dynamically adjusts its response characteristics based on the current glucose situation. By selecting appropriate models for different glucose states, the system can respond rapidly to acute changes when necessary while maintaining stability during steady-state conditions, avoiding the oscillations that would result from a consistently aggressive control approach.
Solution Approach 2:
The system changes control parameters based on the selected model, allowing rapid response when glucose dynamics require it while maintaining stability during normal conditions. This parameter adaptation enables the system to achieve both fast response and stability by adjusting its behavior to match the specific glucose situation rather than using fixed aggressive control parameters.
Data Source
AI summary
The present disclosure relates to systems and methods for controlling physiological glucose concentrations in a patient.