[Paper Review] Distributed Frequency Control with Operational Constraints, Part II: Network Power Balance

Zhaojian Wang, Feng Liu+3 more|arXiv (Cornell University)|Feb 28, 2017

Frequency Control in Power Systems4 citations

TL;DR

This paper proposes a fully distributed optimal frequency control for multi-area power systems under network power balance, where power mismatches are balanced across the entire system rather than within individual areas. The controller uses primal-dual updates with saturation to enforce generation and load regulation capacity constraints during both transient and steady-state, while ensuring tie-line power flows remain within limits, achieving asymptotic stability via a nonpathological Lyapunov function and validated through simulations on a modified Kundur system.

ABSTRACT

In Part I of this paper we propose a decentralized optimal frequency control of multi-area power system with operational constraints, where the tie-line powers remain unchanged in the steady state and the power mismatch is balanced within individual control areas. In Part II of the paper, we propose a distributed controller for optimal frequency control in the network power balance case, where the power mismatch is balanced over the whole system. With the proposed controller, the tie-line powers remain within the acceptable range at equilibrium, while the regulation capacity constraints are satisfied both at equilibrium and during transient. It is revealed that the closed-loop system with the proposed controller carries out primal-dual updates with saturation for solving an associated optimization problem. To cope with discontinuous dynamics of the closed-loop system, we deploy the invariance principle for nonpathological Lyapunov function to prove its asymptotic stability. Simulation results are provided to show the effectiveness of our controller.

Motivation & Objective

• Address the challenge of optimal frequency control in multi-area power systems where power mismatches are balanced across the entire network, not per area.
• Ensure regulation capacity constraints on generators and controllable loads are satisfied not only at equilibrium but also during transient dynamics.
• Maintain tie-line power flows within acceptable limits to prevent congestion, even under unknown load disturbances.
• Enable fully distributed control using only local and neighboring information, eliminating the need for centralized coordination.
• Achieve optimal economic operation while restoring nominal frequency and ensuring system stability despite discontinuous dynamics from saturation constraints.

Proposed method

• Formulate the optimal frequency control problem as a constrained optimization problem with network-wide power balance and operational constraints.
• Design a distributed controller that implements primal-dual updates with state-dependent saturation to enforce generation, load, and tie-line power limits.
• Construct a nonpathological Lyapunov function to handle discontinuities arising from saturation constraints, enabling stability proof via invariance principles.
• Integrate the controller with physical power system dynamics to form a closed-loop system that converges to the optimal solution.
• Use only local frequency measurements and neighborhood communication—no load measurement required—enabling adaptive response to unknown disturbances.
• Leverage the port-Hamiltonian structure of the system dynamics and the primal-dual algorithm to ensure optimality and stability.

Experimental results

Research questions

• RQ1 How can optimal frequency control be achieved in a multi-area power system when power mismatches are balanced across the entire network rather than within individual control areas?
• RQ2 How can generation and load regulation capacity constraints be enforced not only at equilibrium but also during transient dynamics?
• RQ3 What control architecture enables fully distributed operation with only local and neighboring information while maintaining system stability under saturation constraints?
• RQ4 How can congestion in tie-line power flows be automatically eliminated while ensuring all operational limits are satisfied?
• RQ5 What Lyapunov-based stability proof technique can be applied when the system dynamics become discontinuous due to saturation?

Key findings

• The proposed controller successfully restores nominal frequency across all four control areas in a modified Kundur system after an unknown load disturbance.
• Tie-line power flows remain within their specified limits (e.g., ±65 MW) at equilibrium, with no congestion observed in the baseline scenario.
• When tie-line limits were reduced to ±50 MW, the controller automatically eliminated congestion, maintaining flows within bounds (e.g., P*(4,2) = -49.9 MW).
• The equilibrium point achieved by the distributed controller matched the optimal solution computed via centralized optimization (e.g., P*g = 618 MW in Area 1), confirming optimality.
• Simulation results show that the controller maintains stability and constraint satisfaction even during transients, with mechanical power outputs and controllable loads adjusting within their capacity limits.
• The closed-loop system achieves the same optimal equilibrium as the centralized solution, demonstrating that distributed control can replicate centralized optimal dispatch without coordination.