大型矿用电动轮自卸车动力系统优化控制方法研究

With the rapid development in economy, there is a growing demand for materials and energy in industry. Hundred-ton level electric wheels dump truck is important transportation for mining, which is a nonlinear system including time-varying, uncertainty and interconnection. According to the special controlled object and its operating environment, researches will be presented in this project for improving the vector control performance of its induction motor, synthetic control performance of its power system and its robust stability. (1) a open-loop intelligent identification system for induction motor is established，and rotor parameters intelligent estimation and rotor flux online compensation of the induction motor under low speed and large torque working condition are achieved by using online dynamic estimation method combining with intelligent compensation scheme based on the open-loop identification data, and a kind of intelligent vector decoupling control algorithm is designed adapted to the mining working condition. (2) The power system of electric wheels dump truck is a nonlinear control system. In order to analyse the dynamic performance and robustness of electric wheels dump truck, considering the uncertainty and nonlinearity of its power system, the control mathematical model of its power system is established. (3) Because mix H-two/H-infinity control can takes account of robust stability and robustness, in combining the direct feedback linearization with the mixed H-two/H-infinity control theory to design the synthetic controller for the power system, the specifications of the system are transformed to that of a standard H-two/H-infinity control problem, and a nonlinear H-two/H-infinity synthetic control law is developed for the power system. (4) A multi-objective optimization of mix H-two/H-infinity controller is design by means of differential evolution algorithm(DE). The research on this project has great significance for improving automation technology of our mining transport equipment.

百吨级电动轮自卸车是矿山开采过程中的主要运输设备，是一个具有时变、不确定性、关联耦合的非线性系统。针对矿用兆瓦级电动轮自卸车这一特殊控制对象及其运行环境，为提高电动轮自卸车牵引电机矢量控制性能和整车动力系统综合控制性能和鲁棒稳定性，本项目主要研究：(1)采用动态在线检测估计与开环工艺数据智能补偿相结合的方案，实现动态低速大扭矩工况下牵引感应电机转子参数智能估计与转子磁链智能在线补偿，在此基础上，设计出一种适应于矿山工况的智能矢量解耦控制算法；(2)充分考虑不确定性、非线性等因素，分析建立电动轮自卸车整车动力系统数学模型；(3)应用H2/H∞理论方法设计动力系统非线性H2/H∞综合控制律，应用差分进化算法实现对H2/H∞控制器多目标优化设计，达到提高电动轮自卸车动力系统动态性能、鲁棒性，以及降低动力系统整体能耗提高运行效益的目的。本项目的研究对于提升我国工矿运输装备自动化水平具有重要意义。

电动轮自卸车的电源由车载大功率柴油发电机组提供，容量有限，运行工况恶劣，工艺参数具有非线性，且波动大，对动力系统运行稳定性和鲁棒性要求更高。本项目从数理统计的角度出发，将不断变化的转子工艺参数视为随机变量，提出了一种基于随机概率分布的转子工艺参数样本的分布拟合方法，为牵引电机转子电阻的在线估计提供有效的实验数据；提出了基于动态学习粒子群（DPSO-LS）的逆变器驱动永磁同步电机系统参数估计方法，将逆变器、电气系统参数和永磁同步电机看作统一整体建模，并将参数辨识问题作为优化问题进行寻优估计，设计了动态学习型粒子群算法（DPSO-LS）对参数估计器进行寻优，所提参数估计方法能提高永磁同步电机系统电气参数估计和逆变器非线性因素扰动电压估计精度，并能有效跟踪相关电气参数变化；提出了一种新的动态差分进化与自适应突变算子算法，用于永磁同步电动机的多参数同步估计。数值仿真结果表明：以上提出的理论方法为电动轮自卸车动力系统参数辨识和建模提供了新的方法。矿用电动轮自卸车的交流传动系统是一个复杂的多能域系统，提出了基于键合图理论的复杂多能域系统的建模方法；将电动轮自卸车柴油发电机组转速H2/H∞控制器优化转化为多目标优化问题，提出了基于改进的多目标差分进化算法（IMOSADE）的H2/H∞控制器优化设计理论方法，较好得兼顾了动力系统跟踪性能和抗扰性能。