大型医疗监护设备电池健康状态检测算法研究

Research on Battery Health Status Detection Algorithm for Large Medical Monitoring Equipment

目的 研究大型医疗监护设备电池健康状态的检测算法,用以检测电池的健康状态,解决电池由于在使用过程中受到温度变化、充放电循环等影响而产生的时变效应和故障多样性等问题。方法 分析设备电池在充电和放电过程中的电压变化情况,提取等压降放电时间、电池内阻和等间隔放电时间序列3种健康因子;将其输入至基于非线性自回归模型神经网络的非线性回归模型中进行训练,评估大型医疗监护设备的电池容量;结合粒子群算法改进反向传播神经网络,检测电池健康状态。结果 实验结果表明:该检测方法的误差较小;提出的3种健康因子与估计大型医疗监护设备电池容量的相关性大于0.95,且电池容量估计结果准确。结论 通过该方法,可及时发现电池问题,提前采取措施,减少因电池故障引起的设备停机时间,降低医疗事故风险。

Objective To research a detection algorithm for the health status of batteries in large medical monitoring equipment,aimed at detecting the health status of batteries and addressing issues such as time-varying effects and fault diversity caused by temperature changes,charging and discharging cycles during use.Methods The voltage variation of the battery during charging and discharging was analyzed,and three health factors such as constant voltage drop discharge time,battery internal resistance and constant interval discharge time series were extracted.It was trained into a nonlinear regression model based on nonlinear autoregressive with exogenous inputs model neural network to estimate the battery capacity of large medical monitoring equipment.The backpropagation neural network was improved by particle swarm optimization algorithm to detect the state of health of the battery.Results The experimental results showed that the detection error of this method was small;The correlation between the three health factors and the estimated battery capacity of large medical monitoring equipment was higher than 0.95,and the estimated battery capacity was accurate.Conclusion Through this method,battery problems can be detected in time,and measures can be taken in advance to reduce equipment downtime caused by battery failures and reduce the risk of medical errors.