Recently developed advanced metering infrastructure in an intelligent physical system (IPS) sorts large quantities of data available in design, whereas the potential electricity system is implemented for profits and aid in the client’s transition from inactive to an active role. This paper investigates the usage of Deep Reinforcement Learning (DRL), in Intelligent Physical System for Strategic Planning in Enterprise Information (IPSSPEI) has been proposed to achieve the online schedule optimization for building energy management of Enterprise Data in an intelligent grid context. Here, two methods such as, profound R-learning and deep policy gradient, have been proposed to compute and examine the learning procedure which performs several actions simultaneously to overcome the scheduling problems. Hence, this high-dimensional database contains information about the generation and utilization of energy by photovoltaic power cars and smart buildings, with advanced metering infrastructure. Moreover, the electrical energy forecasting approaches could be used to give a real-time input to the consumers, facilitating the better application of electricity in an intelligent physical system to overcome the scheduling problems.

最近在智能物理系统（IPS）中开发的高级计量基础结构可对设计中可用的大量数据进行分类，而实施潜在的电力系统是为了牟利并帮助客户从非活动状态转变为活动状态。本文研究了深度强化学习（DRL）的用途，提出了用于企业信息战略规划的智能物理系统（IPSSPEI），以实现在智能网格环境中对企业数据的建筑能源管理进行在线计划的优化。在此，已经提出了两种方法，例如深层R学习和深策略梯度，以计算和检查同时执行多个动作以克服调度问题的学习过程。因此，这个高维数据库包含有关光伏电动汽车和智能建筑以及先进计量基础设施的能源产生和利用的信息。此外，电能预测方法可用于向消费者提供实时输入，从而有助于在智能物理系统中更好地利用电，从而克服调度问题。