Textile Machine Parameter Management Beyond Existing Solution Space
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for optimizing textile machine parameters in spinning mills are limited by their inability to determine new settings outside the existing solution space, leading to inefficiencies in production quality, raw material usage, waste reduction, and cost optimization.
Innovation Solution
A computer system and method utilizing a neural network combined with Case-Based Reasoning (CBR) and mathematical control, which receives operational data from spinning mills, stores it in databases, and uses algorithms for supervised, unsupervised, and deep learning to determine adapted machine parameters, allowing for optimal allocation of resources and predictive maintenance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional optimization methods are used to determine machine parameters, then the solution remains within the existing solution space, but production quality and efficiency cannot be improved beyond current limits
Solution Approach 1:
The patent transforms the discrete parameter optimization problem into a continuous parameter space by introducing virtual machines with continuously variable parameters. This allows the system to explore parameter values beyond the discrete set of existing machine configurations, enabling determination of optimal parameters that were previously inaccessible. The neural network learns from both real and virtual machine data to generalize parameter optimization across the continuous space.
Solution Approach 2:
The patent introduces virtual machines as intermediary entities between real machines and the optimization objective. These virtual machines serve as a bridge, allowing the system to explore parameter spaces beyond physical machine limitations. The virtual machines encode desired parameter combinations that may not exist in the physical fleet, enabling the optimization to guide real machine adjustments toward optimal configurations.
2Reliability
If more machine parameters are optimized simultaneously, then production quality and resource efficiency improve, but the complexity of the optimization system increases
Solution Approach 1:
The patent merges multiple optimization objectives (production quality, raw material usage, waste reduction, cost minimization) into a single unified neural network model. Instead of implementing separate optimization systems for each objective, the patent integrates them all into one comprehensive system that processes multiple inputs and produces coordinated parameter adjustments, thereby reducing overall system complexity while achieving multi-objective optimization.
Solution Approach 2:
The neural network is designed as a universal optimizer that handles multiple machine types, multiple parameters, and multiple optimization objectives simultaneously. The system uses a standardized interface and unified mathematical framework that can accommodate different machine configurations and optimization goals without requiring separate specialized systems, thus managing complexity through generalization.
3Adaptability or versatility
If real machines are used for optimization, then the solution is constrained to existing machine configurations, but if virtual machines are introduced, then new parameter settings can be determined
Solution Approach 1:
The patent performs preliminary optimization in the virtual machine space before implementing changes on real machines. The neural network first determines optimal parameters using virtual machines that can explore any parameter combination, then validates and transfers these optimized settings to real machines. This preliminary exploration in the virtual domain enables the system to identify optimal configurations without being constrained by existing real machine limitations.
Solution Approach 2:
The patent creates virtual copies (virtual machines) of real machines that replicate their operational characteristics but allow unrestricted parameter exploration. These virtual copies serve as digital twins for optimization purposes, enabling the system to test and determine optimal parameters in a risk-free environment before applying changes to physical equipment, thus maintaining reliability while achieving adaptability.
Data Source
AI summary
A textile mill system and associated method include a plurality of spinning mills each having textile machines. A computer system determines adapted machine parameters for the textile machines and processes within the spinning mills. The computer system includes a receiving and transmitting section configured to receive operational information from the spinning mills and the textile machines, and a first database configured to store the received operational information. A processing section includes an optimizer section with a neural network, wherein the neural network uses the operational information stored in the first database with processes for or derived from supervised or unsupervised machine or deep learning to determine the adapted machine parameters.