Dynamic optimisation of unbalanced distribution network management by model predictive control with Markov reward...

Area of Science:

• Electrical Engineering
• Control Systems
• Renewable Energy Integration

Background:

• Active distribution systems require efficient management to minimize energy losses.
• Integrating renewable energy sources like wind and solar presents challenges in power system stability and efficiency.
• Voltage control assets play a crucial role in maintaining power quality and reducing losses.

Purpose of the Study:

• To develop and evaluate a two-level control system for minimizing total active power losses in an active distribution system.
• To compare the effectiveness of a model-based predictive control (MPC) framework against short-horizon analysis for loss minimization.
• To investigate the impact of varying renewable power generation and battery storage capacities on system performance.

Main Methods:

• Implementation of a two-level control system featuring model-based predictive control (MPC) at the first level.
• Utilization of non-homogeneous and homogeneous Markov reward models for accurate wind and photovoltaic power prediction, respectively.
• Employing an optimization algorithm for voltage control asset management, including voltage regulating transformers, to minimize system losses.

Main Results:

• The MPC framework demonstrated superior performance in minimizing total active power losses compared to short-horizon analysis.
• Long-horizon optimization within the MPC resulted in a significant decrease in active power losses, despite an increase in variables.
• Short-horizon analysis led to a reduction in variables but a compromise in the quality of loss minimization results.

Conclusions:

• The proposed two-level control system, particularly with the MPC framework, effectively minimizes active power losses in active distribution systems.
• The choice of control horizon significantly impacts the trade-off between the number of variables and the achieved loss reduction.
• The study highlights the benefits of predictive control strategies for optimizing the operation of complex grids with renewable energy integration.