Central Plant Optimization With Schedule-Aware Binary Pruning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Central plant systems face challenges in optimizing thermal energy load distribution across subplants to minimize energy consumption and operating costs, particularly in managing device operation states and constraints effectively.
Innovation Solution
A control system with high and low-level optimization modules, a constraint modifier, and a binary optimization modifier is implemented to determine optimal subplant load allocations and operating states, using user inputs for minimum on and off schedules to adjust constraints and generate simplified solutions for binary optimization processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If traditional optimization methods are used for central plant load distribution, then the optimization process is simpler, but energy consumption and operating costs cannot be minimized effectively due to inadequate constraint management
Solution Approach 1:
The optimization system is divided into two distinct modules: a high-level optimization module that performs initial load distribution across subplants, and a low-level optimization module that refines device operating states within each subplant. This segmentation allows each module to focus on specific aspects of optimization, improving overall energy minimization while keeping individual module complexities manageable.
Solution Approach 2:
A constraint modifier acts as an intermediary component between the optimization modules and the binary optimization process. It modifies constraints based on minimum on and off schedules, translating operational requirements into optimized constraint parameters. This intermediary enables effective constraint management without requiring the optimization algorithms to directly handle complex scheduling logic.
2Productivity
If device operating constraints are strictly enforced without modification, then operational reliability is maintained, but optimization flexibility is reduced preventing energy minimization
Solution Approach 1:
The constraint modifier dynamically adjusts optimization constraints based on minimum on and off schedules derived from operational requirements. Rather than using fixed constraints, the system adapts constraint parameters in real-time, allowing the optimization process to explore more efficient operating points while still respecting fundamental operational limits. This dynamic approach maintains reliability by honoring minimum schedules while improving optimization efficiency.
3Ease of operation
If binary optimization is performed without simplified solutions, then complete device selection is achieved, but computational time and complexity increase significantly
Solution Approach 1:
The system performs preliminary optimization actions through the high-level optimization module before executing the computationally intensive binary optimization. The high-level module establishes initial load allocations and identifies promising device combinations, creating a refined search space for the subsequent binary optimization. This preliminary action significantly reduces the computational burden and time required for the complete device selection process.
Solution Approach 2:
The low-level optimization module applies localized optimization within each subplant, focusing computational resources on specific device selections rather than optimizing the entire central plant simultaneously. By dividing the binary optimization into smaller, subplant-specific problems, the system achieves complete device selection with reduced computational complexity and faster execution time.
Data Source
AI summary
A control system for a central plant having subplants including devices operating to serve energy loads of a building. The system includes a high level optimization module that performs high level optimization of thermal loads subject to constraints to generate subplant load allocations. The control system includes a low level optimization module that performs low level optimization of the subplant load allocations to determine operating states for the devices. The control system includes a constraint modifier that modifies the constraints for the high level optimization module based on equipment schedules. The control system also includes a binary optimization modifier including a pruner module that receives the minimum off schedule to determine adjusted branches and a seeder module that receives the minimum on schedule to determine a starting node for use in binary optimization performed by the low optimization module.