Continuous Fuel Cell Performance Monitoring Methods

Sensor and method for monitoring gas quality of a fuel reformer in a fuel cell system or power generation system. The sensor is a miniature fuel cell that estimates hydrogen content of fuel by measuring voltage. Placing sensors at the reformer inlet and outlet, and maintaining constant temperature, allows measuring voltage difference to assess reformer efficiency.

Source

Predictive fuel cell management for integrated hydrogen-electric systems like aircraft. It optimizes the number of fuel cells online at any given time to avoid wasting energy or damaging cells. The system uses a controller to monitor aircraft flight conditions and predict the power requirements for each phase of flight. Based on this data, it activates or deactivates fuel cells to match the power needs without overloading the fuel cell stack.

Source

Monitoring and assessing the health of fuel cells and electric motors in a fuel cell powered electric aircraft to improve reliability, safety, and efficiency. The method involves comparing current fuel cell and motor performance during flight with prior data to identify deviations. These differences are transformed into health indicators using algorithms. The indicators are displayed to the pilot or maintenance personnel for assessment and warnings. The goal is to provide real-time health monitoring using flight data that can be used for trend analysis and predictive maintenance.

Source

Diagnostic system to determine the cathode gas quality in a fuel cell stack to improve diagnosis of cathode gas contamination impact on fuel cell performance. The diagnostic system involves connecting a separate diagnostic fuel cell to the cathode and anode supply lines of the main fuel cell stack. Measuring the diagnostic cell's performance with the main stack's cathode and anode gases allows comparing against nominal values to detect contamination. If contamination is detected, actions are initiated to reduce it in the main stack. This prevents unnecessary restoration actions on the main stack that could damage components.

Source

Artificial intelligence fuel cell system that learns and analyzes historical fuel cell data to predict and control performance, diagnose cell degradation, and optimize operation. The system collects fuel cell operating data at regular intervals, trains a predictive model using machine learning, compares the model to real-time data, distinguishes between temporary and irreversible degradation, and generates control signals based on the diagnosis.

Source

Monitoring fuel cell performance in a fuel cell stack using embedded sensors and tone generation. A sensor circuit inside the stack detects low performing cells. When it finds one, it sends a signal to a tone generator. The generator switches a load into and out of the cell group at a frequency. A voltage sensor detects the stack voltage with the load. A tone decoder analyzes the frequency to determine that low performing cells were found. This allows monitoring every cell without external connections.

Source

Predicting fuel cell stack performance and adjusting parameters for optimization by using sensors to measure operating conditions and calculate parameters like current density and stack temperature. The performance metrics are determined based on the sensor inputs and used to optimize stack parameters like maximum current output, scheduling of current requests, and voltage suppression. This allows adaptive stack control that improves efficiency by tailoring stack behavior to the current operating conditions.

Source

Diagnosing fuel cell deterioration without adding extra hardware or degrading performance. The method involves measuring the alternating current (AC) impedance of the fuel cell in a frequency range by controlling the frequency of the current drawn from the cell and analyzing the pulse component of the output current. This allows diagnosing cell degradation without additional equipment or affecting cell efficiency compared to normal operation.

Source

Monitoring fuel cell stack performance to adjust or tune the performance of a fuel cell stack. The technique involves inducing a condition in a first fuel cell to reduce fuel flow, then measuring its voltage output compared to another cell. If the first cell's voltage drops below the other cell's, it indicates potential stack performance issues. This allows proactive adjustment of operating conditions to prevent stack issues. The technique can apply to fuel cell systems like DMFC with low fuel flow, carbon bubbles, or water droplets.

Source

Predicting hydrogen production in a fuel cell system using a mathematical model that adapts over time based on feedback from the fuel cell system. The model generates an indication of hydrogen production by the reformer in the fuel cell system. During continuous operation, the model is adapted based on feedback from the fuel cell system and used to control the fuel cell system. This allows hydrogen production to be predicted without direct sensing, which is challenging and expensive.

Source

Estimating and controlling states of a fuel cell system that cannot be directly measured due to cost or practicality limitations by using a mathematical model and feedback from the system. The technique involves creating a model that relates the directly measurable states to the unmeasured ones. The model is used with feedback from the system to estimate the unmeasured states.

Source

Estimating the output characteristic of a fuel cell over time to improve efficiency and optimize operation as the cell ages. The method involves estimating a basic output characteristic based on fuel pressure and cell temperature, then deriving the actual output characteristic using measured current and voltage. This allows accurate estimation of the cell's output behavior as it degrades over time. A controller uses the estimated characteristic to set target output and adjust cell voltage for efficient operation.

Source

Monitoring the status of individual cells in a fuel cell stack without using A/D converters. The monitoring system compares the voltage of each cell directly to a reference voltage using a comparator. This eliminates the need for converting the analog cell voltages to digital and reduces the number of components compared to using A/D converters. The comparator outputs a digital signal, like 1 or 0, indicating if the cell voltage is below or above the reference. The digital signals are used to monitor the cells without requiring complex signal processing and converting large numbers of analog voltages.

Source

Accurately estimating the remaining driving range of a fuel cell vehicle taking into account the energy needed to restart the fuel cell system when the vehicle stops. The method involves calculating the predicted driving range by subtracting the startup energy from the remaining fuel amount in the tank. This provides a more realistic estimation compared to just using the remaining fuel alone.

Source