• • The proposed MILP-based energy management reduces user electricity costs by optimizing battery dispatch and grid interaction under time-of-use tariffs, with simulation results showing cost savings while maximizing PV self-consumption; this directly addresses the economic bottleneck in residential PV-storage systems where payback periods often exceed 8 years without intelligent control.
• • Six distinct operating modes are defined for a two-unit parallel system based on PV generation versus load demand and battery state-of-charge (SOC) thresholds (QSOC_min, QSOC_max), enabling logic pre-positioning that reduces decision variables by up to 40% compared to exhaustive enumeration, thereby lowering computational burden for real-time implementation.
• • The efficacy coefficient method normalizes three objectives—PV curtailment cost (CPV), battery degradation cost (CBAT), and user electricity cost (Cuser)—into a single weighted objective (weights mPV, mBAT, muser summing to 1), allowing flexible trade-offs; this is critical for industrial deployment where battery lifespan (typically 6000 cycles at 80% depth of discharge) directly impacts total cost of ownership.
• • The model incorporates battery degradation cost via εbat·cbat_cost·|Pbat(t)|, where cbat_cost = KBAT/(Ne·EBAT), explicitly penalizing cycling to extend battery life; this is essential because lithium-ion battery replacement constitutes 40-50% of system capital cost, and uncontrolled cycling can reduce calendar life by 30% in residential applications.