Distributed Autonomous Control for Load Balancing in Heterogeneous Power Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods struggle to efficiently manage and control large-scale systems, particularly in the infrastructure and IT fields, due to challenges in real-time unit commitment and load balancing among heterogeneous components, especially when dealing with unexpected external disturbances and energy efficiency.
Innovation Solution
A system and method that utilize function blocks with activation/shutdown control units, which store and use evaluation functions to determine the activation or shutdown of components based on their performance indices and the performance indices of related blocks, allowing for real-time load balancing and optimal resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If offline scheduling based on deterministic mathematical programming is used to solve the unit commitment problem, then optimal allocation of generation amount and load can be achieved, but the system cannot deal with unexpected external disturbances in real time
Solution Approach 1:
The patent transforms the static offline scheduling system into a dynamic real-time control system. Each power plant is equipped with a control unit that continuously monitors system state and autonomously adjusts generation output based on current conditions, enabling the system to adapt dynamically to external disturbances while maintaining optimal allocation through continuous evaluation of performance indices.
Solution Approach 2:
The patent implements a decentralized autonomous control mechanism where each power plant's control unit independently makes decisions about generation adjustment based on locally stored evaluation functions and real-time system state information. This self-service approach eliminates the need for centralized real-time computation while achieving optimal load allocation through autonomous agents that cooperate to satisfy overall system demands.
2Quantity of substance
If the system scale is increased to include innumerable control objects, then the system can handle larger power networks, but conventional control methods become inadequate for real-time management
Solution Approach 1:
The patent divides the large-scale power network into autonomous segments, with each power plant operating as an independent controlled unit. Each segment has its own control unit that makes local decisions, eliminating the computational burden of centralized control over innumerable objects. This segmentation allows the system to scale to large networks while maintaining real-time control efficiency through distributed autonomous decision-making.
Solution Approach 2:
The patent implements dynamic autonomous control at each segment level, where control units continuously adapt generation output based on real-time evaluation of performance indices and system state. This dynamic approach enables efficient real-time management of large-scale networks by distributing computational tasks across numerous independent agents rather than requiring centralized processing of all control objects simultaneously.
3Ease of operation
If conventional load balancing methods are used that balance internal system state, then control policy is clear for uniform systems, but it is not clear how to balance in mixed-machine systems with different types of machines
Solution Approach 1:
The patent assigns different evaluation functions tailored to the specific characteristics of each power plant type. Instead of applying a uniform control policy to all machines, the system recognizes local differences in machine performance, efficiency curves, and operational constraints. Each control unit uses its own customized evaluation function that reflects the specific qualities of its associated power plant, enabling optimal control of mixed-machine systems while maintaining clear operational policies for each machine type.
Solution Approach 2:
The patent transforms the control approach by changing from fixed control policies to dynamic parameter-based evaluation functions. These evaluation functions take as inputs various operational parameters specific to each machine type (efficiency curves, cost functions, environmental constraints) and dynamically determine optimal generation output. This parameter-driven approach allows the same control framework to adapt to different machine types without requiring separate control policies for each.
Data Source
AI summary
Provided is a system including a plurality of correlated function blocks, for executing load balancing among the function blocks. The plurality of function blocks each include an activation/shutdown control unit to determine whether to activate/shut down its own function block based on information of an evaluation function of its own function block and information of an evaluation function of another function block that has a relation to its own function block, and the activation/shutdown control unit determines whether to activate/shut down its own function block by using an amount relevant to the evaluation function of its own function block in a state where the evaluation function of its own function block takes a desired value, and using a state of the evaluation function of the another function block having a relation to its own function block that is equivalent to, or related to, the relevant amount.