论文部分内容阅读
为解决控制约束对一类多相流量计系统液位控制造成的影响,提出了一种基于神经动态优化的模型预测控制算法。首先构建多相流量计系统传递函数模型,通过离散化建立相应状态空间模型,进一步地提出含控制约束的模型预测控制问题,然后将带约束的模型预测控制问题转化为带约束的标准二次规划问题,并运用简化对偶神经网络模型进行实时在线优化求解,从而获得系统最优控制序列,该网络模型的神经元个数仅与不等式约束个数相等,与现有文献中相关网络模型相比规模小、计算复杂度低,充分利用神经网络并行处理的优点,以提高模型预测控制的在线优化能力。最后,通过仿真实例验证了算法的有效性和优越性。
In order to solve the influence of control constraints on the level control of a multiphase flowmeter system, a model predictive control algorithm based on neural dynamic optimization is proposed. Firstly, the transfer function model of multiphase flowmeter system was constructed, the corresponding state space model was established by discretization, and then a model predictive control problem with control constraints was proposed. Then, the constrained model predictive control problem was transformed into a constrained standard quadratic programming Problem, and use the simplified dual neural network model for real-time on-line optimization solution to obtain the optimal control sequence of the system. The number of neurons in the network model is only equal to the number of inequality constraints. Compared with the existing network model, Small, low computational complexity, make full use of the advantages of neural network parallel processing to improve the online optimization ability of model predictive control. Finally, the simulation results show the effectiveness and superiority of the algorithm.