Fluid Transport Control with Self-Learning Input Signal Selection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing control methods for fluid transport systems, such as HVAC and district heating networks, face challenges due to varying installation conditions and require manual configuration, leading to inefficiencies and high operational costs.

Innovation Solution

A self-learning control process that automatically selects a subset of input signals based on performance indicators, reducing the need for manual configuration and enabling efficient control across different types of fluid transport systems.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If Model Predictive Control (MPC) framework is used to improve control quality, then control performance is improved, but manual configuration complexity increases due to need for system-specific models

Engineering Contradiction:
Improvecontrol performanceVSAvoidmanual configuration complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The control system automatically generates system-specific models by collecting and analyzing operational data from the fluid transport system itself, eliminating the need for manual expert configuration. The system performs self-commissioning by autonomously identifying system characteristics and creating customized control models during normal operation.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system collects and stores operational data in advance during normal system operation, preparing the dataset needed for model generation before control optimization is required. This preliminary data accumulation enables rapid model creation without interrupting system operation or requiring manual intervention.

Inventive Principle:
Principle #10Preliminary action

2Reliability

If reinforcement learning is applied to improve control quality, then control performance is improved, but training time increases significantly

Engineering Contradiction:
Improvecontrol performanceVSAvoidtraining time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system collects operational data in advance during normal system operation, preparing the training dataset before reinforcement learning is applied. This preliminary data accumulation eliminates the need for time-consuming training periods during system commissioning, as the model can be rapidly generated from pre-collected data.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system replaces iterative reinforcement learning training with direct model generation from collected operational data. Instead of requiring multiple training iterations, the model is created directly from the accumulated dataset, significantly reducing the time required to achieve optimal control performance.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

3Ease of operation

If generic models are used to reduce configuration effort, then ease of operation is improved, but adaptability to individual installations deteriorates

Engineering Contradiction:
Improveconfiguration effortVSAvoidadaptability to individual installations
Core Design Contradiction:
Ease of operationVSAdaptability or versatility

Solution Approach 1:

The system creates customized control models specific to each individual installation by analyzing local operational data from that particular system. Each system receives a tailored model that reflects its unique characteristics, ensuring optimal performance for that specific installation rather than using a one-size-fits-all generic model.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The system autonomously adapts to individual installations by automatically collecting and analyzing its own operational data to generate customized control models. This self-commissioning capability eliminates the need for manual configuration while ensuring each system receives a model optimized for its specific characteristics.

Inventive Principle:
Principle #25Self-service

4Reliability

If all available input signals are used in self-learning control to improve control quality, then control performance is improved, but processing time increases

Engineering Contradiction:
Improvecontrol performanceVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system extracts and selects only the most relevant input signals from the available dataset for use in model generation and control. By identifying and removing redundant or less important signals, the system reduces processing requirements while maintaining control performance based on the essential information contained in the selected subset of input signals.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP3916492B1Method and system for controlling a fluid transport system
Publication Date: 2025.07.02 GRUNDFOS HLDG
  • EP3916492B1 patent drawingFigure 1
  • EP3916492B1 patent drawingFigure 2~3
  • EP3916492B1 patent drawingFigure 4

AI summary

A computer-implemented method for controlling operation of a fluid transport system, by a applying a self-learning control process, the method comprising: receiving obtained values of a plurality of input signals during operation of the fluid transport system during a first period of time, wherein operation of the fluid transport system during the first period of time is controlled by a predetermined control process, automatically selecting a subset of the plurality of input signals based on the received obtained values of the plurality of input signals, receiving obtained values of at least the selected subset of input signals during operation of the fluid transport system during a second period of time, wherein operation of the fluid transport system during the second period of time is controlled by applying the self-learning control process, wherein the self-learning control process is configured to control operation of the fluid transport system based only on the selected subset of input signals, and wherein applying the self-learning control process comprises updating the self-learning control process based on the received obtained values of the selected subset of the input signals and based on at least an approximation of a performance indicator function.