状态预测模型中的多步预测算法

Multi-step Prediction Algorithm for State Prediction Model

一般系统的状态受多个因素的影响,而基于动态贝叶斯网的状态预测模型就能够较准确地描述系统状态和影响因素之间的关系。针对此模型,提出推理宽度的概念以减少推理过程中的数据量,并利用时间片扩充办法来对状态进行多步预测。

The state of the system is generally affected by many random factors. The state prediction model based on dynamic bayesian network can more accurately describe the relationship between the system state and the influencing factors. According to this model, the span of the reasoning is proposed to reduce the amount of data in the reasoning process. Meanwhile the multi-step state prediction algorithm is achieved by extending time-slice.