State-of-charge estimation

 What needs to be estimated, and why? 

EVs need to know two battery quantities: 

• How much energy is available in the battery pack? 
• How much power is available in the immediate future? 

■ An estimate of energy is most important for EV: Energy tells me how far I can drive. 

■ An estimate of power is most important for HEV: Power tells me whether I can accelerate or accept braking charge. 

■ Both are important for E-REV/PHEV. 

■ To compute energy, we must know (at least) all cell states-of-charge Zk and capacities Qk. 

■ To compute power, we must know (at least) all cell states-of-charge and resistances Rk. 

■ But, we cannot directly measure these parameters—we must estimate them as well. 

■ Available inputs include all cell voltages, pack current, and temperatures of cells or modules.

■ We’ll see that there are both good and poor methods to produce estimates: The poor methods are generally simpler to understand, code, and validate, but yield less accurate results. 

■ The impact of this can be:

•  Abrupt corrections when voltage or current limits exceeded, leading to customer perception of poor drivability, or 
• Over-charge or over-discharge, which damages cells, or 
• Compensating for uncertainty of estimates by over-designing pack. 

■ All of these have costs in dollars, weight and/or volume. 

■ A major premise of this course is that investing in good battery management and control algorithms and electronics capable of implementing the algorithms can reduce pack size and end up with a considerable net savings. 

State of Charge (SoC) Estimation Techniques

State of Charge (SoC) estimation is crucial for the effective management and operation of battery systems in electric vehicles (EVs) and hybrid electric vehicles (HEVs). SoC represents the remaining capacity of the battery, typically expressed as a percentage of the full charge. Accurate SoC estimation helps in predicting the driving range, scheduling charging, and preventing battery damage from overcharging or deep discharging.

Common SoC Estimation Techniques

1. Coulomb Counting (Ah Counting)

• Principle: Measures the charge entering and leaving the battery to estimate the SoC.

• Advantages: Simple implementation, high accuracy over short periods.
• Disadvantages: Accumulative errors due to current sensor drift, requires accurate initial SoC.

1. Open Circuit Voltage (OCV) Method

   • Principle: Uses the relationship between the open circuit voltage and the SoC.
   • Method: Measures the battery's voltage at rest and compares it to a pre-determined OCV-SoC curve.
     • Rest period: Allow the battery to rest for a period to reach equilibrium.
   • Advantages: Direct relationship with SoC, non-intrusive.
   • Disadvantages: Requires rest periods for accurate measurements, influenced by temperature and aging.

2. Kalman Filter

   • Principle: Uses a recursive algorithm to estimate the SoC by combining a battery model with real-time measurements.
   • Method:
     • Predict: Estimate the next state (SoC) based on the current state and the battery model.
     • Update: Correct the prediction with actual measurements (voltage, current).
   • Advantages: Can handle noisy measurements and model uncertainties, adaptive.
   • Disadvantages: Requires a good battery model, complex implementation.

3. Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF)

   • Principle: Extensions of the Kalman filter to handle non-linearities in the battery model.
   • Method:
     • EKF: Linearizes the non-linear model around the current estimate.
     • UKF: Uses a set of sample points to better capture the non-linear behavior.
   • Advantages: Improved accuracy for non-linear systems, robust to model inaccuracies.
   • Disadvantages: More computationally intensive, requires accurate modeling.

4. Impedance Spectroscopy

   • Principle: Analyzes the battery's impedance response over a range of frequencies to estimate SoC.
   • Method: Measures impedance and uses it to infer SoC based on known impedance-SoC characteristics.
   • Advantages: Can provide insights into battery health and SoC, useful for diagnostics.
   • Disadvantages: Requires specialized equipment, sensitive to temperature and aging.

5. Artificial Neural Networks (ANN) and Machine Learning Techniques

   • Principle: Uses data-driven approaches to learn the relationship between battery parameters and SoC.
   • Method: Trains a neural network or other machine learning models on historical data to predict SoC.
   • Advantages: Can capture complex relationships, adaptable to different battery types.
   • Disadvantages: Requires large datasets for training, computationally intensive, can be a black box.

6. Hybrid Methods

   • Principle: Combines multiple SoC estimation techniques to leverage their strengths and mitigate weaknesses.
   • Method: Typically integrates Coulomb counting with OCV or Kalman filters.
   • Advantages: Enhanced accuracy and robustness, compensates for individual method limitations.
   • Disadvantages: Increased complexity, requires careful integration and calibration.

Example: SoC Estimation in Practice

Tesla Model S:

• Primary Methods: Uses a combination of Coulomb counting and Kalman filtering.
• Implementation: The BMS starts with an initial SoC based on OCV, then continuously integrates current (Coulomb counting) and adjusts the SoC estimate using Kalman filters to account for measurement errors and model inaccuracies.
• Benefits: Provides accurate and real-time SoC estimation, improving range prediction and battery management.

Summary

Accurate SoC estimation is critical for the reliable operation of EVs and HEVs. Various techniques, ranging from simple Coulomb counting to advanced Kalman filters and machine learning methods, offer different trade-offs in terms of complexity, accuracy, and computational requirements. Hybrid approaches that combine multiple techniques are often employed to achieve the best performance, ensuring accurate and robust SoC estimation across different operating conditions.

Requirement of Battery Monitoring

Battery monitoring is a crucial aspect of managing and ensuring the optimal performance, safety, and longevity of battery packs in electric vehicles (EVs) and hybrid electric vehicles (HEVs). A Battery Management System (BMS) is typically responsible for this task, providing real-time monitoring and control over various battery parameters.

Key Requirements of Battery Monitoring

1. State of Charge (SoC) Estimation

   • Purpose: Determines the remaining capacity of the battery in percentage.
   • Importance: Helps in range estimation, charge scheduling, and avoiding deep discharge which can harm battery health.
   • Methods: Coulomb counting, voltage-based estimation, and advanced algorithms using Kalman filters.

2. State of Health (SoH) Estimation

   • Purpose: Assesses the overall health and remaining useful life of the battery.
   • Importance: Predicts maintenance needs, replacement schedules, and ensures reliability.
   • Methods: Capacity fade measurement, internal resistance measurement, and machine learning techniques for degradation prediction.

3. Voltage Monitoring

   • Purpose: Measures the voltage of individual cells and the overall battery pack.
   • Importance: Ensures cells are operating within safe voltage ranges, detects over-voltage and under-voltage conditions.
   • Implementation: High-precision voltage sensors connected to each cell.

4. Temperature Monitoring

   • Purpose: Monitors the temperature of individual cells and the battery pack.
   • Importance: Prevents overheating, thermal runaway, and ensures cells operate within optimal temperature ranges.
   • Implementation: Temperature sensors (e.g., thermocouples, RTDs) placed throughout the battery pack.

5. Current Monitoring

   • Purpose: Measures the current flowing in and out of the battery pack.
   • Importance: Detects over-current conditions, helps in SoC estimation, and ensures safe charging and discharging rates.
   • Implementation: Current sensors (e.g., shunt resistors, Hall effect sensors).

6. Cell Balancing

   • Purpose: Ensures uniform charge and discharge across all cells to prevent imbalances.
   • Importance: Extends battery life, improves performance, and prevents overcharging/over-discharging of individual cells.
   • Methods: Passive balancing (resistor-based) and active balancing (capacitor/inductor-based).

7. Safety Features

   • Overcharge Protection: Prevents cells from being charged beyond their maximum voltage limit.
   • Over-discharge Protection: Prevents cells from being discharged below their minimum voltage limit.
   • Short-circuit Protection: Detects and interrupts short-circuit conditions to prevent damage and fire hazards.
   • Thermal Management: Activates cooling systems when temperatures exceed safe thresholds.

8. Data Logging and Communication

   • Purpose: Records operational data and communicates it to the vehicle’s central control unit.
   • Importance: Enables performance analysis, diagnostics, and remote monitoring.
   • Implementation: CAN bus or other communication protocols for real-time data transmission.

9. Firmware Updates and Diagnostics

   • Purpose: Allows for software updates and troubleshooting of the BMS.
   • Importance: Keeps the system updated with the latest algorithms and security patches, and facilitates diagnostics.
   • Methods: Over-the-air (OTA) updates, diagnostic ports for service tools.

Example: Tesla Model S Battery Monitoring System

The Tesla Model S employs a sophisticated BMS to ensure optimal performance and safety of its Li-ion battery pack. Key features include:

• SoC and SoH Estimation: Advanced algorithms for accurate SoC and SoH calculations.
• Voltage and Temperature Sensors: High-precision sensors for monitoring individual cells and overall pack conditions.
• Current Sensing: High-accuracy current sensors for monitoring charging and discharging currents.
• Cell Balancing: Active cell balancing to maintain uniform cell voltages.
• Thermal Management: Liquid cooling system controlled by the BMS to manage cell temperatures.
• Safety Protocols: Comprehensive safety measures including overcharge, over-discharge, and short-circuit protection.
• Communication: CAN bus communication for real-time data transmission to the vehicle’s control unit.
• Firmware Updates: Capability for OTA updates to enhance functionality and security.

Summary

Battery monitoring is essential for the safe, efficient, and reliable operation of EV and HEV battery packs. A well-designed BMS with robust monitoring capabilities ensures that the battery operates within safe parameters, provides accurate information on battery status, and implements necessary safety protocols to protect both the battery and the vehicle.