当前位置：首页 > 过刊浏览->2025年第51卷第6期
基于Smith预测算法大功率石油钻机飞轮储能调峰控制
Peak shaving control of flywheel energy storage for high-power oil drilling rig based on Smith prediction algorithm
浏览(434) 下载(1)
- DOI:
- 10.3969/j.issn.1001-2206.2025.06.015
- 作者:
- 石小利
SHI Xiaoli
- 作者单位:
- 西安宝美电气工业有限公司, 陕西西安 710065
Bomay Electric Industry Co., Ltd., Xi'an 710065, China
- 关键词:
- 大功率负载; Smith预测; 石油钻机飞轮; 储能调峰控制; 单神经元; PSD控制器
high-power load;Smith prediction;oil drilling rig flywheel;energy storage peak shaving control;single neuron;PSD controller
- 摘要:
- 针对石油钻机大功率负载具有强非线性和响应滞后特性，容易因间歇性波动引发电网冲击问题，提出了基于Smith预测算法的飞轮储能调峰控制方法。大功率负载石油钻机飞轮储能调峰系统由制动取能电机、飞轮储能装置、升压整流系统和调峰控制系统构成，通过电力电子变换装置实现电能与动能的双向转换。利用Smith预测算法构建飞轮储能调峰系统的传递函数，通过设置滞后时间和滞后特性参数共同作用于滞后补偿过程，实现对大功率负载波动的前馈预测与动态调整，消除滞后对飞轮储能调峰控制的影响。将Smith预测算法与单神经元PSD控制器结合，设计新型复合控制器；将飞轮储能调峰控制误差、误差变化率、误差累积和作为输入状态量；采用有监督Hebb学习算法更新加权系数，动态调整控制增益，从而提升石油钻机飞轮储能调峰控制系统的响应速度和稳态性能。实验结果表明，该方法能有效降低钻机运行对电网的冲击，平抑功率波动，提升飞轮储能系统的稳定性。
With strong nonlinearity and large lag characteristics of high-power loads on oil drilling rigs, power grid impact is easily caused by intermittent fluctuations. In response to this, a peak shaving control method of flywheel energy storage is proposed based on the Smith prediction algorithm. The flywheel energy storage peak shaving system for high-power load oil drilling rig consists of a braking energy harvesting motor, flywheel energy storage device, boost rectifier system, and peak shaving control system. It achieves bidirectional conversion of electrical energy and kinetic energy through a power electronic conversion device. The Smith prediction algorithm is employed to construct the transfer function of the flywheel energy storage peak shaving system. By setting lag time and lag characteristic parameters to jointly act on the lag compensation process, the feedforward prediction and dynamic adjustment of highpower load fluctuations are achieved, thus eliminating the impact of lag on flywheel energy storage peak shaving control. Combining the Smith prediction algorithm with a single neuron PSD controller, a new composite controller is designed. The flywheel energy storage peak shaving control error, error change rate, and error accumulation are taken as input state variables. The supervised Hebb learning algorithm is applied to update weighting coefficients and dynamically adjust control gains, thereby improving the robustness of flywheel energy storage peak shaving control in oil drilling rigs. The experimental results show that this method can effectively reduce the impact of drilling rig operation on the power grid, smooth out power fluctuations, and improve the stability of the flywheel energy storage system.