Parameter Setting Recommender for Unknown Plant Operating States

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for setting adjustable parameters in technical plants and business transactions require domain knowledge and are complex, often relying on manual decisions that are not transparent or efficient, especially when dealing with new operating states or customer-specific adaptations.

Innovation Solution

A computer-implemented recommendation method using a recommender system based on machine learning and collaborative filtering, which processes input data to provide accurate and reliable recommendations for adjustable parameters through representation learning and multi-output Gaussian processes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual decision-making based on domain knowledge is used for setting adjustable parameters, then accuracy of settings can be maintained, but productivity and speed of implementation deteriorate

Engineering Contradiction:
Improveaccuracy of parameter settingsVSAvoidspeed of implementation
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The system enables automatic self-service by using machine learning models to autonomously determine optimal parameter settings without requiring manual expert intervention. The recommender system processes input data and generates settings automatically, allowing the system to serve itself in terms of configuration decisions.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces the mechanical/manual process of expert decision-making with an automated information processing system. Machine learning algorithms substitute for human experts, transforming the manual cognitive process into an automated computational process that maintains accuracy while dramatically improving speed.

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

2Productivity

If automated decision-making systems are implemented for setting adjustable parameters, then productivity and speed improve, but reliability and accuracy may deteriorate due to lack of domain knowledge

Engineering Contradiction:
Improvespeed of implementationVSAvoidaccuracy of settings
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The system performs preliminary action by training machine learning models in advance on historical data and expert knowledge. This pre-training phase captures domain knowledge and patterns, enabling the automated system to make reliable decisions during operation without requiring real-time expert intervention.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements feedback mechanisms where the recommender system's outputs are evaluated and used to continuously improve the underlying machine learning models. This closed-loop approach ensures that the automated system learns from its performance and maintains high reliability over time.

Inventive Principle:
Principle #23Feedback

3Measurement precision

If complex data processing methods are used to handle sparse and high-dimensional data, then measurement precision improves, but device complexity increases

Engineering Contradiction:
Improveaccuracy of recommendationsVSAvoidcomplexity of processing system
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies segmentation by dividing the complex data processing task into distinct modules: data preprocessing, feature extraction, machine learning model processing, and recommendation generation. This modular approach handles sparse and high-dimensional data systematically while keeping each component manageable and well-defined.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3889840A1A recommender system and method for recommending a setting of an adjustable parameter
Publication Date: 2021.10.06 SIEMENS AG
  • EP3889840A1 patent drawingFigure 1~3
  • EP3889840A1 patent drawingFigure 4~6
  • EP3889840A1 patent drawingFigure 7~8

AI summary

The invention refers to a recommender system and method for recommending a setting of an adjustable parameter of an object for an application. For example, the object can be a device of a plant and the application corresponds to an operating state of the plant. The adjustable parameter can be a configuration parameter which requires a certain setting depending on the operating state of the plant. The invention suggests a method based on a machine learning approach to recommend such a setting, even in case of a previously unknown operating state.