Battery storage systems have been widely used in various residential, commercial, and industrial applications. Examples include uninterruptible power supply, robots, vehicles, photovoltaic systems, and more. However, unexpected premature battery failures (which are key issues in mission-critical applications) cannot be easily and accurately identified without conducting thorough offline battery characterization - going through charging and discharging processes. Apart from interrupting the normal operation of the batteries, the processes are also lengthy and energy inefficient. In order to ensure the serviceability of the system, batteries are typically replaced after a few years of service based on the site condition, maintenance practice, and experience of the operators. The present technology applies an energy recycling technology and makes use of artificial intelligence to process the sampled battery information for conducting a real-time estimation of the state-of-charge (SOC) and state-of-health (SOH) of batteries. In addition to consuming a very small amount of energy, the estimated SOC and SOH can also assist operators to profile battery utilization to extend the life expectancy of batteries. The technology also allows operators to monitor remotely the battery state and health, and conduct replacement exercise at the optimal time. As the batteries are to be replaced near the end of their life cycle, it will also reduce undesired electronic waste.
Technology Features, Specifications and Advantages
The present technology has two innovative elements. They are (i) artificial intelligence (AI) for battery condition estimation, and (ii) energy recycling technique to minimize energy wastage during battery testing. These two elements allow operator to test battery with different charging and discharging profiles suitable for real applications. Based on the battery voltage and current measurement, the battery's remaining lifetime can be predicted through state-of-charge (SOC) and state-of-health (SOH) such that battery can be replaced near the end of its life cycle. This can also reduce unnecessary electronic wastage. Moreover, the recycling technique is able to solve thermal and power dissipation issues during the test. The conventional methods dissipate energy and the battery need to be charged after conducting the test, which usually takes hours to complete the charging process. Special features of the present technology include:
- Faster turnaround time for battery diagnostic - on average, the technology utilizes around 3 mins to complete the whole test, while the traditional method requires over hours
- Ideally, no energy is needed to perform condition monitoring of batteries
- The AI algorithm takes both charging and discharging characteristics in estimating the remaining lifetime of the batteries
- Large-signal characterization is conducted
The present technology can be applied to various residential, commercial, and industrial applications. It forms a part of the battery storage system, such as uninterruptible power supply, robots, vehicles, photovoltaic systems, and more. The algorithm can be integrated into an existing charger so that the lifetime of the batteries can be estimated. The algorithm is applicable for both lithium batteries and lead-acid batteries.
The present technology can
- Estimate the remaining useful life of batteries
- Reduce electronic waste
- Conduct remote condition monitoring of batteries with Internet-of-Things (IoT).