Operation Plan Decision System Using Time Cross-Section Division
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large-scale and complicated mixed integer linear programming (MILP) models for energy-saving and cost-saving operation plans in industrial plants require extensive calculation time and often fail to obtain an optimum solution due to exponential search time and difficulty in identifying constraint violations.
Innovation Solution
The method involves using a constraint violation minimization model to derive a feasible solution, updating candidates using a time cross-section division model, and selecting the optimum solution from a candidate list, which reduces calculation time and identifies constraint violations quickly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the MILP method is used to optimize operation plans for energy saving and cost saving, then the optimum solution can be obtained, but the calculation time period becomes excessively long and the model becomes intractable when scaled up
Solution Approach 1:
The patent divides the large-scale MILP model into multiple smaller sub-models based on time cross-sections (e.g., hourly, daily, or weekly time periods). Each sub-model represents a specific time period and can be solved independently and quickly. The solutions are then aggregated to form the overall optimum operation plan, dramatically reducing total calculation time while maintaining solution optimality.
Solution Approach 2:
The patent performs preliminary actions by pre-calculating and storing operational parameters, equipment characteristics, and constraint conditions for different time periods before the actual optimization runs. This pre-processing allows the optimization system to focus only on variable parameters during execution, reducing real-time calculation requirements and enabling faster solution generation.
2Adaptability or versatility
If the number of equipment and complexity of the model increases to handle large-scale supply-demand cooperation, then the optimization can cover more factories and byproducts, but the search time increases exponentially and the model becomes intractable
Solution Approach 1:
The patent segments the optimization model by time cross-sections, dividing the entire optimization problem into multiple smaller sub-problems that can be solved independently. This segmentation allows the system to handle large numbers of equipment and factories across multiple time periods without experiencing exponential search time increases, as each time period is optimized separately using efficient algorithms.
Solution Approach 2:
The patent employs dynamic optimization techniques that adapt the model structure and solution methods based on the specific time period and operational conditions. The system dynamically adjusts optimization parameters, constraint priorities, and calculation methods according to real-time data and historical patterns, enabling efficient handling of varying scales and complexities without fixed computational overhead.
3Reliability
If the MILP method is used to solve large-scale optimization models, then the optimization can handle complex constraint conditions, but the identification of constraint violations requires vast amounts of man-hours and professional knowledge
Solution Approach 1:
The patent implements comprehensive feedback mechanisms that automatically detect, report, and analyze constraint violations as they occur during optimization. The system provides real-time feedback on which constraints are violated, by how much, and suggests corrective actions or alternative solutions. This automated feedback loop eliminates the need for manual analysis of constraint violations, reducing the required professional knowledge and time while maintaining high constraint satisfaction.
Solution Approach 2:
The optimization system performs self-diagnosis and self-correction by automatically identifying constraint violations, analyzing their causes, and generating corrective operation plans without requiring external expert intervention. The system uses built-in diagnostic tools and algorithms to self-service the identification and resolution of constraint issues, dramatically reducing the manual effort and specialized knowledge needed to handle complex optimization problems.
Data Source
AI summary
An operation plan decision method includes deriving an feasible solution by using a constraint violation minimization model, updating candidates for an optimum solution and adding the updated candidates to a candidate list by taking the derived feasible solution, as an initial value of a candidate for the optimum solution, and by using a time cross-section division model that is obtained by dividing an optimization model for each time cross-section, and selecting the optimum solution from the candidate list to which the updated candidates are added.