The functions of the Battery Management System (BMS) consist of: Sensing and High voltage control (Measure voltage, temperature, and current, control contactors, pre-charge control, detection of ground fault, thermal management). Protection of over voltage, Under voltage, high current, high temperature, short circuit. Communication of available energy estimation, data recording and reporting. Performance management (SOC, power limits computation, cell balancing and equalizing. Diagnosis and Prognosis: Detection of faults, SOH estimation and SOL estimation.
The BMS project included the following:
Battery Performance Management: SOC estimation using Coulomb counting and OCV. Dashboard from the real world data collected into the cloud with many thousands of vehicles for both the distribution of SOC and temperatures across the fleet. Run clustering algorithm . Detect anomalies.
Power capability limitation (charge and discharge) estimating based on the high frequency resistance real time parameter identification. Part of the development work involved: 1- Modeling of the battery cell using 2 RC’s, OCV and high frequency resistance, 2- Running accelerated aging experiments (battery cycler) 3- Regular battery cell testing to identify the values of the circuit components at different temperature and SOC.
Battery Thermal modeling: The use of battery modeling to predict battery temperatures at different location and detect and anomalies.
Diagnosis and Prognosis: SOH of the battery pack based on capacity estimation. Using the combination of coulomb integration and OCV. Remaining useful life prediction based on integration of vehicle usage information and aging models in the cloud (digital twin) where end of life is defined as the battery capacity loses 20% of its nominal value.
Electric Drive Health management
The Electric Drive (ED) consists of the motor, power electronics, gears and sensors. In this project, the ED performance and health are monitored. Data is collected, and visualized. Predictive data analytics of the health & performance indicators of the ED’s of a fleet of vehicles are implemented in the cloud.
Algorithms for the detection and location of the degraded components are developed, evaluated, and validated.
A Layered Architecture for Enabling Virtual Energy Storage in Power Distribution Grids
In this project, we develop a closed-form solution for a game-theoretic decision problem under certain practical assumptions. Based on the closed-form solution, an algorithm is established in which the Distribution System Operator (DSO), as the leader of the game, signals a unit energy price that induces the consumers to reshape the power demand to match the power supply. The consumers while competing for power resources in a Nash game setting, adjust their individual power demand in such a way as to maximize their corresponding objective function under the system constraints.
A generalized model of a virtual storage that expresses the game theoretic strategy has been described. The algorithm that has been developed in this project was tested and evaluated using a simplified electrical distribution system case study. Based on the results of the evaluation, the tested algorithm has been applied to a more realistic example of an electrical distribution system (that incorporates cable losses) with deferrable loads.
Health Monitoring, Failure Prognosis and Predictive Maintenance of Autonomous Vehicles Fleet
In this project we develop, evaluate and validate advanced on-board & off-board algorithms for the health monitoring, failure prognosis and predictive maintenance of the mechanical parts of autonomous vehicles chassis (braking, steering and suspension components and subsystems) based on acoustics, vibration, energy and performance information. Autonomous vehicles are of great need for diagnosis, prognosis and health monitoring of mechanical components since there is no driver to feel vibrations in the vehicle (e.g. via steering wheel) or to hear vehicle noises (e.g. wearing out bearings). Noises and vibrations can often be used as precursors to mechanical failures.
This IOT system will improve up time, reliability, and ensure safety of the autonomous mobility fleet as well as reduce the cost of operation.