图1 斜坡式重力储能系统平面结构

重力储能系统中，码放环节是其重要组成部分，介于码放环节对时效性的需求，传统行吊对重物的位置识别方法不再适用。本文通过深度学习构建质量块的运行路程预测模型，令抓取装置提前动作至指定位置等候，缩短码放环节整体时间。同时，得益于深度学习的引入，可直接根据实际运行数据建立路程预测模型，不需要复杂的数学推导过程，便于实际工程应用。

本文的主体内容分为以下三部分：首先介绍了斜坡式重力储能系统的整体工作流程以及对质量块抓取装置的需求，其次提出使用质量块瞬时加速度计算其运行路程的方法，并根据质量块在缓冲平台的运行路程影响因素分析了该方法的误差来源，最后，提出基于深度神经网络的斜坡式重力储能系统质量块抓取装置控制方法，满足系统对码放环节的时效性需求，提高码放环节的工作效率。

图2 斜坡式重力储能系统立体结构

本文工作对重力储能系统的码放环节进行了详细分析，提出的方法极具推广应用价值。

This paper studies the workflow of the mass block stacking process of the slope-type gravity energy storage system, combines deep learning with the stacking method, and proposes a gripping device control method suitable for slope-type gravity energy storage.

In the gravity energy storage system, the stacking link is an important component. Due to the timeliness requirements of the stacking link, the traditional method of identifying the position of heavy objects by traveling cranes is no longer applicable. This paper uses deep learning to build a prediction model for the running distance of mass blocks, allowing the grabbing device to move to the designated position in advance and shorten the overall time of the stacking process. At the same time, thanks to the introduction of deep learning, the distance prediction model can be established directly based on actual operating data, which does not require complex mathematical derivation processes and is convenient for practical engineering applications.

The main content of this article is divided into three parts: firstly, it introduces the overall workflow of the slope gravity energy storage system and the requirements for the mass block grasping device. Secondly, a method is proposed to calculate the running distance of the mass block by measuring its instantaneous acceleration. The errors of this method are analyzed based on factors affecting the running distance of the mass block on the buffering platform. Finally, a control method for the mass block grasping device of the inclined gravity energy storage system based on deep neural networks is proposed. This method meets the timeliness requirements of the system in the stacking process and improves the efficiency of the stacking process.

This work conducts a detailed analysis of the stacking process of the gravity energy storage system, and the method proposed has great promotion and application value.

支持基金：

论文链接：

http://www.motor-abc.cn/djykzyy/article/abstract/20231105

推荐引用格式：

CHEN Julong, LI Zhen, ZHU Yongqing, LIU Dameng, ZHANG Yu, WANG Zufan, GAO Tian, HAO Zilin, DONG Linlin, ZHAO Haisen. Control Method of Mass Block Grasping Device of Slope Gravity Energy Storage System Based on Deep Neural Network[J].  Electric Machines & Control Application, 2023, 50(11): 37-45.

陈巨龙，硕士，高级工程师，贵州电网有限责任公司电网规划研究中心能源经济研究室副总经理，研究方向新能源并网与新型储能技术。

王祖凡，华北电力大学电气与电子工程学院博士研究生，研究方向为电能转换与高效利用，重力储能技术等。

邮箱：zhaohisen@163.com

欢迎广大优秀学生报考赵海森教授课题组！