Battery Management Runtime Calculation for Materials Handling Vehicles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The reliance on state-of-charge (SOC) measurements alone is insufficient for accurately conveying the residual charge of batteries in materials handling vehicles, as the state-of-health (SOH) declines over charge/discharge cycles, leading to discrepancies in energy capacity and varying power requirements across different vehicles sharing rechargeable batteries.
Innovation Solution
Implementing a forward-looking remaining runtime calculation using learning algorithms that account for current and voltage characteristics, incorporating Exponentially Weighted Moving Average (EWMA) signal filters to provide more predictable and accurate runtime estimates, adaptable to individual battery duty cycles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If SOC measurement is used to indicate battery charge status, then the indication is simple and easy to implement, but the accuracy of conveying residual charge deteriorates due to SOH decline over charge/discharge cycles
Solution Approach 1:
The system implements a feedback mechanism where the learning algorithm continuously monitors actual runtime data and adjusts the runtime prediction model accordingly. This feedback loop compensates for SOH degradation over time, maintaining accurate runtime predictions despite battery aging, thereby resolving the contradiction between simple SOC indication and accurate residual charge conveyance.
Solution Approach 2:
The battery management system performs self-characterization by automatically learning and storing the unique charge/discharge patterns and voltage characteristics of each battery through the EWMA filtering algorithm. This self-service approach eliminates the need for manual battery testing or complex calibration procedures while achieving high measurement precision.
2Device complexity
If conventional SOC-based runtime calculation is used, then the system is simple to implement, but the runtime prediction accuracy deteriorates due to abrupt drop-offs and non-linearity
Solution Approach 1:
The system dynamically adapts to changing battery conditions by continuously updating the EWMA filtered current and voltage values. This dynamic approach allows the runtime calculation to respond to real-time battery state changes while maintaining smooth, linear predictions without abrupt drop-offs, resolving the contradiction between system simplicity and prediction accuracy.
Solution Approach 2:
The learning algorithm changes the parameters used for runtime calculation from simple SOC-based estimates to EWMA-filtered current and voltage characteristics. This parameter transformation enables more accurate and linear runtime predictions while keeping the overall system architecture relatively simple through algorithmic processing.
3Measurement precision
If individual battery learning algorithms are implemented, then runtime calculation accuracy improves by accounting for unique battery characteristics, but the computational complexity and data processing requirements increase
Solution Approach 1:
Each battery performs self-characterization by automatically learning its own unique charge/discharge patterns through the EWMA filtering algorithm. This self-service mechanism eliminates the need for complex external calibration systems or manual testing procedures, achieving high measurement precision while keeping the system architecture relatively simple through autonomous battery self-learning.
Solution Approach 2:
The system uses the EWMA filtered historical data to create a simplified mathematical model (copy) of each battery's unique characteristics. This copying approach captures essential battery behavior patterns without requiring complex computational models, achieving accurate runtime predictions with manageable algorithmic complexity.
4Adaptability or versatility
If batteries of different capacities are used in the same vehicle fleet, then operational flexibility improves, but the usefulness of SOC indication deteriorates because SOC does not account for varying energy content
Solution Approach 1:
The system dynamically calculates runtime predictions based on the specific battery installed in each vehicle, using EWMA-filtered current and voltage data to adapt to different battery capacities. This dynamic approach preserves operational flexibility while providing accurate, battery-specific runtime information that compensates for varying energy content, resolving the information loss problem.
Data Source
AI summary
Battery management systems, removable battery assemblies with integrated battery management systems, and vehicles are provided for implementing the various runtime calculations disclosed herein. A vehicle, which may be a materials handling vehicle, is provided comprising a drive subsystem, a removable battery assembly, and vehicle control hardware. The battery assembly comprises a battery management system programmed to input or generate a state of charge signal representing a state of charge SOC of the battery assembly, implement a first EWMA signal filter F1 to calculate a succession of contemporary current calculations IC, implement a second EWMA signal filter F2 to calculate a runtime current IR, wherein the runtime current calculation IR comprises the contemporary current calculations IC from the first EWMA signal filter F1, and implement a remaining runtime calculation such that the vehicle control hardware responds functionally to the runtime calculation.