Grinding Process Control Using Adaptive Multi-Model Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for controlling grinding processes in grinding devices lack accuracy due to limitations in process modeling, as a single process model cannot adequately describe the effects of actuating actions across various ranges of state variables and disturbance influences.
Innovation Solution
Creating at least two process models with different structures to describe the effects of actuating actions on state variables, allowing for optimized control by selecting the model that provides the best prediction for the current system conditions, and continuously adapting these models based on past and predicted effects of actuating actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a single process model is used to describe the effects of actuating actions on state variables, then the device complexity is reduced, but the manufacturing precision and control accuracy deteriorate because the model cannot adequately describe the process across various ranges of state variables and disturbance influences
Solution Approach 1:
The patent divides the single process model into multiple process models (at least two) with different structures. Each process model is specialized to describe the process behavior under specific ranges of state variables or disturbance conditions. This segmentation allows the system to select the most appropriate model for the current operating conditions, thereby improving control accuracy without requiring one overly complex universal model.
Solution Approach 2:
The patent implements a dynamic model selection mechanism that adapts the process model based on current system conditions. The system continuously monitors state variables and disturbance influences, then dynamically selects or switches between different process models to match the current operating regime. This dynamic adaptation ensures that the most accurate model is always used, improving control precision while maintaining manageable model complexity through specialization.
2Manufacturing precision
If multiple process models with different structures are created to improve accuracy, then the manufacturing precision improves, but the device complexity increases due to the need to create and manage multiple models
Solution Approach 1:
The patent segments the process modeling task into multiple specialized models, each with a different structure optimized for specific operating conditions. This segmentation improves prediction accuracy for each condition while keeping individual model structures relatively simple and manageable, rather than requiring one extremely complex universal model.
Solution Approach 2:
The patent varies the structural parameters of the process models (such as the number of layers, neurons, or model order) to optimize each model for its specific operating range. By changing structural parameters rather than using a fixed complex structure for all models, the system achieves high prediction accuracy across different conditions while maintaining reasonable complexity for each individual model.
3Reliability
If process models are continuously adapted based on past and predicted effects of actuating actions, then the reliability of control improves, but the loss of time for model adaptation and real-time decision-making increases
Solution Approach 1:
The patent performs preliminary adaptation of multiple process models during offline training or between operational cycles, so that when real-time control is needed, the models are already prepared and can be selected based on current conditions without requiring extensive real-time computation. This preliminary preparation reduces the time loss during actual control operations while maintaining high reliability through continuous model improvement.
Solution Approach 2:
The patent implements a feedback mechanism where the system monitors the performance of different process models and uses this information to continuously adapt and improve the models. The feedback loop allows the system to learn from past actuating actions and their effects, progressively improving model accuracy and reliability while optimizing the adaptation process to minimize time loss through efficient learning algorithms.
Data Source
AI summary
A method for controlling a process within a system, in particular a grinding process in a grinding device, comprising the following steps: .detecting 1 of state variables (st) of the system; creating 2 of at least two process models (PM), each describing the effects of actuating actions (at) on the state variables (st) of the system, wherein the structure of the at least two process models (PM) differs from each other; controlling 3 of the process within the system by executing actuating actions (at) under consideration of predefined control objectives and the process model, which currently provides the best prediction for the process running in the system. A device for carrying out the method according to the invention.