Glosarium AI
Kamus lengkap Kecerdasan Buatan
State of Charge (SoC)
The State of Charge represents the current charge level of a battery, expressed as a percentage of its total nominal capacity. It is a crucial state variable for energy storage optimization algorithms.
State of Health (SoH)
The State of Health is an indicator of a battery's current condition compared to its new state, measuring its degradation and its ability to store and deliver energy. It is essential for maintenance planning and lifespan prediction.
Depth of Discharge (DoD)
The Depth of Discharge measures the percentage of the total battery capacity that has been used. Managing the DoD is a key lever to optimize the lifespan of storage systems.
Reinforcement Learning (RL) for storage
Reinforcement Learning applied to energy storage involves training an AI agent to make optimal charge/discharge decisions by maximizing a reward, often based on economic profits or grid efficiency.
Multi-objective optimization
Multi-objective optimization aims to find the best compromise between several conflicting objectives, such as maximizing revenue, minimizing battery degradation, and balancing grid load.
Energy arbitrage
Energy arbitrage is a strategy that consists of storing electricity when prices are low to sell or consume it when prices are high, thus generating profits. Predictive AI algorithms are essential to anticipate price variations.
Battery Fleet Management
Battery Fleet Management refers to the centralized supervision and optimization of a set of storage systems, using AI to coordinate their charge/discharge based on grid needs and the state of each unit.
Battery Digital Twin
A Battery Digital Twin is a dynamic virtual replica of a physical storage system, fed with real-time data and AI models, to simulate its behavior, predict its degradation, and test optimization strategies.
Smart Charging Algorithm
A smart charging algorithm dynamically adjusts a battery's charging profile based on renewable energy production forecasts, electricity prices, and grid demand to maximize self-consumption and minimize costs.
Load Forecasting
Load forecasting uses machine learning models to anticipate short or medium-term future electricity consumption. This information is vital for deciding when to charge or discharge storage systems.
Renewable Smoothing
Renewable smoothing is a storage function that uses batteries to absorb rapid fluctuations in solar or wind production, thereby stabilizing the power injected into the grid through reactive control algorithms.
Demand Response
Demand response is a mechanism by which consumers adjust their consumption in response to price signals or grid constraints. AI-driven batteries can participate by discharging during peak demand periods.
Cell Balancing
Cell balancing is a process aimed at ensuring all cells within a battery pack have a uniform state of charge, to maximize the usable capacity and lifespan of the pack. AI algorithms can optimize this process in real-time.
Stochastic Optimization
Stochastic Optimization is an optimization method that accounts for the uncertainty of input variables, such as renewable energy production or energy prices, to find robust and resilient storage management strategies.
Q-Learning for Optimization
Q-Learning is a model-free reinforcement learning algorithm where an agent learns an optimal policy by evaluating the quality (Q-value) of state-action pairs, applied to determine the best charging/discharging decisions.
AI-Optimized Battery Management System (BMS)
An AI-Optimized Battery Management System integrates machine learning algorithms to refine SoC/SoH monitoring, predict failures, and optimize protection parameters in real-time, surpassing the capabilities of traditional rule-based BMS.
Peak Shaving
Peak Shaving, or peak load shifting, is a strategy that involves using energy stored in batteries to reduce the maximum power drawn from the grid during periods of high demand, in order to limit costs associated with consumption peaks.