Production Network Model for Supply Chain Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Mixed Integer Programming (MIP) models for supply chain optimization are limited by their inability to encode directionality of supply and demand, leading to inefficiencies and slow computation times, especially when dealing with complex problems that require more than two constraints per variable.
Innovation Solution
The introduction of 'production nodes' and directed arcs in a network model, along with a parametrized multi-level search heuristic, allows for faster and more general optimization of supply chain problems by encoding directionality and enabling techniques like backwards and forwards planning, while maintaining solution quality through configurable speed/quality tradeoffs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If standard MIP modelling techniques are used for supply chain optimization, then the model can handle complex constraints, but the computation time becomes excessively long and the directionality of supply and demand cannot be encoded
Solution Approach 1:
The patent replaces the standard MIP optimization approach with a network flow-based system that uses directed arcs and production nodes to model supply chain problems. This substitution enables the system to encode directionality of supply and demand while reducing computation time through specialized network algorithms that are more efficient for this specific problem domain.
Solution Approach 2:
The patent transforms the problem representation by changing key parameters: introducing directionality bits to encode supply/demand flow direction, creating production nodes to represent proportional constraints, and using network flow variables instead of standard MIP variables. These parameter changes enable faster computation while maintaining the ability to handle complex supply chain constraints.
2Adaptability or versatility
If standard MIP models are used, then the model structure is flexible, but it cannot encode directionality required for efficient supply chain planning
Solution Approach 1:
The patent introduces asymmetric directed arcs in the network model, where each arc has a specific direction representing the flow from supply to demand. This asymmetry explicitly encodes the directionality information that is lost in symmetric MIP models, enabling the system to capture the inherent directional nature of supply chain flows while maintaining structural flexibility through the network configuration.
3Productivity
If existing techniques translate MIP models to network models, then the model can be solved faster, but it cannot handle variables with more than two constraints
Solution Approach 1:
The patent segments the network into specialized nodes including production nodes for variables with multiple constraints, flow nodes for standard constraints, and demand nodes. This segmentation allows the system to handle variables with more than two constraints by routing them through production nodes that manage multiple incoming and outgoing arcs, thereby maintaining both speed and constraint handling capability.
Solution Approach 2:
The patent creates a universal network model framework that can handle various constraint types through a single unified structure. Production nodes serve multiple functions by managing proportional constraints and directing flow to multiple destinations, while the overall network structure can accommodate different problem configurations. This multi-functionality enables the system to solve a broader range of supply chain problems efficiently.
4Adaptability or versatility
If general optimization techniques are used, then the solution covers broad problem types, but the computation speed is slow
Solution Approach 1:
The patent applies local quality by creating specialized network components (production nodes, flow nodes, demand nodes) tailored to specific problem requirements. Each node type is optimized for its specific function, allowing the overall system to maintain broad problem coverage while achieving fast computation speeds through the efficient local processing at each node. The network algorithms leverage these localized optimizations to solve complex supply chain problems rapidly.
Data Source
AI summary
Disclosed herein are methods and systems that construct an optimization capability that is able to solve optimization models of supply chain problems faster and with higher quality solutions than prior approaches.