基于后验预测分布的机器人焊接质量监控研究北大核心

针对KUKA机器人焊接质量监测工作量大、抽样样本小的特点,提出一种基于后验预测分布的贝叶斯动态监控方法。从历史数据中选择合适的数据,计算先验分布的超参数;再结合当前样本构建服从负二项分布的后验预测分布,实时计算控制限,实现对焊接质量的动态监测。结果表明:该方法优于传统似然估计法,有更强的异常检出力和稳健性。     

Achievable working range for solid all-desiccant air-conditioning systems under specific space comfort requirements

Solid desiccant air-conditioning systems present a promising solution, in terms of performance level and environmental protection, pointing out their potential to be coupled with thermal solar or waste heat energy source. Nevertheless, these systems are characterized by constraints to the load they can satisfy, through the trade-off between the dehumidification cooling capacity and the latent load of the conditioned space. On that level, one has to note that, for steady environmental conditions, the conditioned space does not present unique value of load, but range of load, corresponding to an array of acceptable temperature and (usually relative) humidity values. In this work a methodology is proposed for the definition of the system's achievable working range under specific set of space (comfort) requirements. Through this approach, the systems present greater potential for covering the space requirements, thus presenting more possibilities on a design basis, and more flexible control strategies, as well. The proposed methodology is presented and discussed through the case study of a solar desiccant air-conditioning system coupled to a typical residential building.