Automation Component Parameterization Using Neural Network Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The commissioning of automation components requires manual parameterization by customers, which can lead to malfunctions and is not adaptable to different plant types, especially for pneumatic systems with complex control algorithms.
Innovation Solution
An automated method using a machine learning module with a pre-trained neural network to test and adjust basic parameterization based on measured data, allowing for autonomous optimization of parameter settings and error pattern recognition, applicable to various automation systems without relying on customer-provided data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual parameterization by customer is used, then component can be adapted to intended purpose, but customer requires know-how and manual effort is needed
Solution Approach 1:
The component performs self-parameterization by automatically determining its own parameters through test runs and evaluation algorithms, eliminating the need for customer know-how and manual configuration efforts
Solution Approach 2:
The component automatically performs parameter determination and optimization before actual operation begins, completing the parameterization process in advance through automated test runs and evaluation
2Extent of automation
If configuration software with autotuning mode is used, then certain system parameters can be determined automatically, but customer must still determine and transmit all necessary data locally and manually
Solution Approach 1:
The component automatically collects its own operational data during test runs and uses this data for self-parameterization, eliminating the need for customers to manually collect and transmit data
Solution Approach 2:
The component performs test runs, evaluates the results through algorithms, and uses this feedback to automatically optimize its parameters without requiring external data input from the customer
3Adaptability or versatility
If local data collection is used, then parameterization can be specific to the plant, but data or information obtained locally cannot be used globally for other plants
Solution Approach 1:
The automated parameterization method and algorithms developed can be universally applied across different plants and components of the same type, allowing knowledge transfer and consistent optimization without re-collecting data at each location
4Ease of operation
If incorrect configuration of control parameters is used, then component can be operated, but malfunctions may occur with potential damage to the plant
Solution Approach 1:
The component performs automated parameter optimization and validation through test runs before actual operation begins, ensuring correct parameters are established in advance to prevent malfunctions
Solution Approach 2:
The component continuously monitors its operation and uses evaluation algorithms to detect and correct parameter deviations, ensuring reliable operation through automated feedback control
Data Source
AI summary
A method for testing a basic parameterization of a component in an automation system is provided. The method includes: starting a test run of the component in the automation system with the basic parameterization, measuring of a measured value data record during trial operation, access to a machine learning module comprising a pre-trained neural network, wherein the pre-trained neural network is pre-trained to calculate a target parameterization for the respective component for a measured value data set, wherein the basic parameterization is compared with the calculated target parameterization and in the event of deviation a result message for adapting the basic parameterization is provided, and receipt of the provided result message for adaptation of the basic parameterization.