Continuous Production Optimization With Backpropagated Prediction Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional continuous production processes rely heavily on human expertise for adjusting control parameters, which limits the ability to accurately capture dependencies between prediction models and relationships between production optimization goals and control parameters.
Innovation Solution
The implementation of a computer-implemented method that utilizes a deep neural network with backpropagation to optimize continuous production processes. This method involves receiving input data, generating prediction models, creating an objective optimization model, and optimizing weights using a loss function to improve production efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If human expertise is used to adjust control parameters, then flexibility and adaptability are maintained, but accuracy in capturing dependencies between prediction models and optimization goals deteriorates
Solution Approach 1:
The patent introduces an intermediary optimization system that includes prediction models, an objective function, and a backpropagation mechanism. This intermediary system mediates between the control parameters and the optimization goals, automatically capturing dependencies through the objective function that quantifies the relationship between predictions and goals, while the backpropagation algorithm automatically adjusts weights to improve accuracy without requiring manual expert intervention.
2Productivity
If manual adjustment of control parameters is used, then system simplicity is maintained, but productivity and timeliness of optimization deteriorate
Solution Approach 1:
The optimization system is designed to be self-service through the backpropagation mechanism that automatically adjusts the weights of prediction models based on the objective function. The system autonomously captures dependencies between predictions and optimization goals, and automatically tunes control parameters without requiring continuous human intervention, thereby improving productivity and timeliness while maintaining appropriate levels of automation.
3Manufacturing precision
If simple prediction models are used, then ease of operation is maintained, but manufacturing precision in achieving optimization goals deteriorates
Solution Approach 1:
The patent implements dynamic prediction models where the weights are not fixed but are automatically adjusted through backpropagation based on the objective function. This dynamic adjustment allows the models to adapt to changing conditions and capture complex dependencies between predictions and optimization goals, improving manufacturing precision while the automation of the adjustment process prevents the operational complexity from becoming unmanageable.
Data Source
AI summary
One embodiment of the invention provides a computer-implemented method for optimization of a continuous production process. The method comprises receiving input data comprising a plurality of datasets each including one or more variables relating to a production equipment involved in the continuous production process. The method further comprises generating different prediction models based on the input data. Each of the prediction models is configured to output a target prediction relating to the production equipment. The method further comprises generating an objective optimization model based on each target prediction output from each of the prediction models. The objective optimization model comprises a deep neural network. The method further comprises generating a loss function corresponding to the objective optimization model, and optimizing weights for parameters of the prediction models using backpropagation of the deep neural network and the loss function, resulting in optimized weights for the parameters of the prediction models.