基于多模型切换的全地域发射平台随动系统多工况优化控制研究

In view of the current situation of the whole-region launch platform, the operating conditions usually change drastically. The traditional control system cannot perform optimization control according to different working conditions. In order to achieve the best control effect, this project proposes multiple operating conditions optimized control methods based on a multiple model switching strategy. On the basis of previous research, based on fuzzy clustering and multi-chaotic neural network method, the prediction model of the all-terrain launching platform follow-up system with parameters can be adjusted online. Based on this, the multiple operating states under the changing conditions of the system are analyzed. In each state, the model of the system is established for each typical work area. Then the underlying model predictive control strategy is designed for each model, and the upper layer supervised hierarchical adaptive controller is designed, and the adaptive switching optimization system is implemented by the underlying control strategy. The shafting torsional impact experiment platform is built and based on the experimental platform, the dynamic characteristics experiment of the whole-region launch platform under the different torsional impact conditions are carried out. Furthermore, the validity of the dynamic model and the controllability of the whole-region launch platform are verified by the experimental results. According to the strong nonlinear and complexity of the whole-region launch platform servo system, the adaptive control strategy of non-dependent model is adopted and the effectiveness of the control strategy based on a multiple model switching strategy is verified by simulation and experiment results. This project provides a new method for suppressing the torsional impact load based on the whole-region launch platform, which provides a theoretical basis for the engineering application of the whole-region launch platform system.

针对目前全地域发射平台随动控制系统无法根据不同工况进行自适应优化控制，为实现最佳的控制效果问题，本项目提出基于多模型切换的全地域发射平台多工况优化控制方法。在前期研究基础上，首先基于模糊聚类及多混沌神经网络方法建立参数可在线调整的全地域发射平台随动系统的预测模型，在此基础上分析系统不同工况等多个运行状态，在每个状态下针对各个典型的工作区分别建立系统的模型；然后为各个模型设计底层多模态模型预测控制策略，最后设计上层监督分层自适应控制器，通过对底层控制策略进行自适应切换优化系统，以达到系统平稳运行的目的。开展多工况下的全地域发射平台动态特性实验研究，验证所建立的模型的有效性；针对系统的强非线性和复杂性，采用不依赖模型的多模型切换鲁棒自适应预测控制策略，并利用仿真和实验验证该控制策略的有效性。上述理论与方法探索研究及原理试验装置的建立，将为全地域发射平台随动系统工程应用提供新的技术实现方式。

全地域发射平台随动系统是一个包含了多个子系统且各子系统之间存在着多物理过程、多参量复合耦合关系的复杂机电系统。这种现象的长期存在会增加机电系统的损耗，使系统无规则运动，以致在复杂的战争环境下出现重大故障。因此本项目对系统中这种多物理过程、多参数耦合现象深层次的机理研究并加以控制对提高高精度武器发射平台伺服系统的运行性能是十分必要的。本项目针对这些问题，开展武器发射平台伺服系统建模与控制研究，主要研究内容和结果有：1)根据全地域发射平台随动系统运行过程中的动态特性以及性能要求，建立全地域发射平台随动系统多模态切换动态模型；2）基于切换理论研究了全地域发射平台随动系统过渡过程的优化控制问题，对全地域发射平台随动系统运行过程能量消耗、切换点选择等目标函数进行优化，以提高系统的整体性能；3）兼顾环境因素和物理约束下系统多模型自适应切换控制策略研究。4）设计了全地域发射平台随动系统系统轨迹跟踪优化算法的仿真系统，并开展了项目的实验研究。以上基于多模型切换的自适应模型预测优化控制将为复杂工况下的全地域发射平台随动系统提供新的模式，以达到系统平稳运行并节能的目的。木项目深入探索了基于多模型切换的自适应模型预测优化融合方法，形成了一套基于基于模型的非线性最优控制策略的的机电传动系统控制器设计方法、理论体系和评价指标，对于提高全地域发射平台这一类机电传动系统的动态性能与稳定性具有极其重要的理论意义和实际应用价值。