PLC Control Code Feature Weighting for ML-Based Program Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial automation environments struggle to effectively integrate and train machine learning models to generate control programs, as existing methods are inefficient and time-consuming, and historical data is not utilized to optimize control programs in near-real time.
Innovation Solution
A system and method for integrating machine learning models into industrial automation environments, where a design component calculates derivative values, ranks and weights them, and generates feature vectors to train the model, enabling targeted optimization of control programs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If machine learning models are integrated into industrial automation environments to optimize control programs, then productivity and performance improvement is achieved, but device complexity and implementation difficulty increases
Solution Approach 1:
The patent introduces an intermediary system comprising a data processing component and a machine learning component that bridges the gap between existing control systems and machine learning models. This intermediary handles data extraction, feature engineering, and model training automatically, reducing the complexity burden on the overall system while enabling productivity improvements through ML-driven control program optimization.
Solution Approach 2:
The machine learning model is designed to automatically train and optimize control programs using historical operational data without requiring manual intervention. The system performs self-service by automatically extracting features from control code and operational data, training the model, and generating optimized control programs, thereby improving productivity while minimizing the need for complex manual configuration and maintenance.
2Adaptability or versatility
If manual editing of control programs is performed in response to operational data, then adaptability is improved, but loss of time and efficiency deteriorates
Solution Approach 1:
The patent implements a feedback mechanism where the machine learning model continuously learns from operational data and automatically generates optimized control programs. The system extracts features from both control code and operational data, trains the model on this feedback loop, and automatically produces adapted control programs, thereby achieving high adaptability without the time loss associated with manual editing.
Solution Approach 2:
The patent replaces the manual mechanical process of editing control programs with an automated machine learning system. The ML model automatically analyzes operational data, extracts relevant features, and generates optimized control code, substituting the time-consuming manual editing process with an automated intelligent system that maintains high adaptability while dramatically reducing time loss.
3Manufacturing precision
If historical data is utilized to train machine learning models, then manufacturing precision is improved, but device complexity increases
Solution Approach 1:
The patent segments the complex data processing task into distinct components: a data processing component that extracts and prepares features from historical operational data and control code, and a machine learning component that consumes these features for training. This segmentation reduces the apparent complexity by dividing the system into specialized modules, each handling specific tasks, while enabling high manufacturing precision through comprehensive utilization of historical data.
Solution Approach 2:
The patent performs preliminary action by automatically extracting and preparing features from historical operational data and control code before feeding them to the machine learning model. The data processing component pre-processes the data, identifies relevant features, and structures them appropriately, thereby reducing the complexity of the training process while maximizing the precision improvements gained from historical data utilization.
Data Source
AI summary
Various embodiments of the present technology generally relate to industrial automation environments. More specifically, embodiments include systems and methods for training a machine learning model for implementation in the environment. In some embodiments, an application receives a data set that comprises a control program configured for implementation by a Programmable Logic Controller (PLC). The application processes the data set and calculates derivative values based on the data set. The design application identifies types for individual ones of the feature vectors and ranks the feature vectors based on their types. The design application weights the feature vectors based on their ranks. The design application generates feature vectors that comprise the derivative values and supplies the weighted feature vectors to the machine learning model for training. The application receives a machine learning training output generated by processing the weighted feature vectors.