Time Domain Partitioning of Electricity Production Cost Simulations

A Parallel Decomposition Scheme for Solving Long-Horizon Optimal Control Problems

“…Long-horizon optimal control problems (OCPs) arise in model predictive control (MPC) applications such as chemical process systems [1], autonomous vehicle steering [2], and battery systems [3]. They also appear in other application domains such as chemical production planning [4] and electricity production planning [5]. Different decomposition techniques have been reported in the literature to improve computational tractability of these problems including dual decomposition [6], alternating direction method of multipliers [7], dual dynamic programming [8], Gauss-Seidel schemes [9], [10], and parallel Newton schemes [11].…”

Repurposing an energy system optimization model for seasonal power generation planning

“…Figure 2 While this UCM solution is not optimal for the full 30-day horizon, the rolling horizon approach allows for the representation of longer-term problems than the normal day ahead operational planning with a significantly reduced computational time. The approach here is similar to Barrows et al (2014), where the authors divide simulations into shorter, overlapping periods in order to improve computational tractability without creating large discrepancies in the results. The main advantage of this approach over sequential day ahead solutions is the ability to add constraints that affect intertemporal choices beyond standard day ahead operation planning.…”

Electric Grid Expansion Planning with High Levels of Variable Generation

“…In expansion planning, the load duration curve (LDC) is commonly used to model the load levels over a LT, such as 1 week, 1 month, or 1 year. Table 12 shows the required computation time and resources when a different number of load levels are considered in a three-node model [47] under a 20-year planning horizon. Table 12 shows that the number of variables (columns) and the number of constraints (rows) in the optimization problem increase linearly with the number of load blocks, while computation complexity increases nonlinearly with the problem size.…”

“…The Chronology Splitting technique has been used to accelerate ST market simulations through parallel execution. Ancillary techniques for Chronology Splitting, such as time overlap for seamless chronology partition, has been developed to improve the accuracy [47].…”

“…To test Splitting Chronology, two ST simulate cases are created based on the three-node system [47]. The ST simulation utilizes the hourly data over a 5-year horizon to simulate system operation under a market environment.…”