This work is concerned with identification and nonlinear predictive control method for MIMO Hammerstein systems with constraints. Firstly, an identification method based on steady-state responses and sub-model method ...This work is concerned with identification and nonlinear predictive control method for MIMO Hammerstein systems with constraints. Firstly, an identification method based on steady-state responses and sub-model method is introduced to MIMO Hammerstein system. A modified version of artificial bee colony algorithm is proposed to improve the prediction ability of Hammerstein model. Next, a computationally efficient nonlinear model predictive control algorithm(MGPC) is developed to deal with constrained problem of MIMO system. The identification process and performance of MGPC are shown. Numerical results about a polymerization reactor validate the effectiveness of the proposed method and the comparisons show that MGPC has a better performance than QDMC and basic GPC.展开更多
A decentralized adaptive neural network sliding mode position/force control scheme is proposed for constrained reconfigurable manipulators. Different from the decentralized control strategy in multi-manipulator cooper...A decentralized adaptive neural network sliding mode position/force control scheme is proposed for constrained reconfigurable manipulators. Different from the decentralized control strategy in multi-manipulator cooperation, the proposed decentralized position/force control scheme can be applied to series constrained reconfigurable manipulators. By multiplying each row of Jacobian matrix in the dynamics by contact force vector, the converted joint torque is obtained. Furthermore, using desired information of other joints instead of their actual values, the dynamics can be represented as a set of interconnected subsystems by model decomposition technique. An adaptive neural network controller is introduced to approximate the unknown dynamics of subsystem. The interconnection and the whole error term are removed by employing an adaptive sliding mode term. And then, the Lyapunov stability theory guarantees the stability of the closed-loop system. Finally, two reconfigurable manipulators with different configurations are employed to show the effectiveness of the proposed decentralized position/force control scheme.展开更多
Aiming at a class of nonlinear systems with multiple equilibrium points, we present a dual-mode model predictive control algorithm with extended terminal constraint set combined with control invariant set and gain sch...Aiming at a class of nonlinear systems with multiple equilibrium points, we present a dual-mode model predictive control algorithm with extended terminal constraint set combined with control invariant set and gain schedule. Local LQR control laws and the corresponding maximum control invariant sets can be designed for finite equilibrium points. It is guaranteed that control invariant sets are overlapped each other. The union of the control invariant sets is treated as the terminal constraint set of predictive control. The feasibility and stability of the novel dual-mode model predictive control are investigated with both variable and fixed horizon. Because of the introduction of extended terminal constrained set, the feasibility of optimization can be guaranteed with short prediction horizon. In this way, the size of the optimization problem is reduced so it is computationally efficient. Finally, a simulation example illustrating the algorithm is presented.展开更多
To study the uncertain optimization problems on implementation schedule, time-cost trade-off and quality in enterprise resource planning （ERP） implementation, combined with program evaluation and review technique （...To study the uncertain optimization problems on implementation schedule, time-cost trade-off and quality in enterprise resource planning （ERP） implementation, combined with program evaluation and review technique （PERT）, some optimization models are proposed, which include the implementation schedule model, the timecost trade-off model, the quality model, and the implementation time-cost-quality synthetic optimization model. A PERT-embedded genetic algorithm （GA） based on stochastic simulation technique is introduced to the optimization models solution. Finally, an example is presented to show that the models and algorithm are reasonable and effective, which can offer a reliable quantitative decision method for ERP implementation.展开更多
大规模开发和利用风能有利于实现电力系统清洁低碳转型,是实现国家“碳达峰、碳中和”战略目标的重要技术手段,但风电出力的强不确定性对电力系统区域间可用输电能力(available transfer capability,ATC)评估带来了全新的挑战,传统用于...大规模开发和利用风能有利于实现电力系统清洁低碳转型,是实现国家“碳达峰、碳中和”战略目标的重要技术手段,但风电出力的强不确定性对电力系统区域间可用输电能力(available transfer capability,ATC)评估带来了全新的挑战,传统用于求解计及风电出力不确定性的概率ATC评估模型在计算效率和计算精度方面均存在一定的不足。为此,该文提出一种基于多项式混沌展开(polynomialchaos expansion,PCE)的电力系统概率ATC评估方法,该方法首先构建基于机会约束的电力系统概率ATC评估模型;然后,根据风电出力预测误差的概率分布特征,选择对应的正交多项式为基函数以近似风电出力预测误差及电力网络中与之相关联的其他随机变量;进一步,借助Galerkin投影和基于一阶矩、二阶矩的机会约束转化方法,将所构建的机会约束模型的概率约束转化为确定性约束,实现基于机会约束的概率ATC评估模型向易于求解的确定性优化模型的转化;进而,将概率ATC评估模型的求解问题转化为ATC的最优多项式逼近系数的求解问题,根据求得的最优多项式逼近系数和选取的基函数计算电力系统ATC的概率分布特征;最后,通过修改后的PJM-5节点测试系统、IEEE-118节点测试系统及吉林西部电网实际算例验证了所提基于多项式混沌展开的电力系统概率ATC评估方法的准确性和有效性。展开更多
离线强化学习通过减小分布偏移实现了习得策略向行为策略的逼近,但离线经验缓存的数据分布往往会直接影响习得策略的质量.通过优化采样模型来改善强化学习智能体的训练效果,提出两种离线优先采样模型:基于时序差分误差的采样模型和基于...离线强化学习通过减小分布偏移实现了习得策略向行为策略的逼近,但离线经验缓存的数据分布往往会直接影响习得策略的质量.通过优化采样模型来改善强化学习智能体的训练效果,提出两种离线优先采样模型:基于时序差分误差的采样模型和基于鞅的采样模型.基于时序差分误差的采样模型可以使智能体更多地学习值估计不准确的经验数据,通过估计更准确的值函数来应对可能出现的分布外状态.基于鞅的采样模型可以使智能体更多地学习对策略优化有利的正样本,减少负样本对值函数迭代的影响.进一步,将所提离线优先采样模型分别与批约束深度Q学习(Batch-constrained deep Q-learning,BCQ)相结合,提出基于时序差分误差的优先BCQ和基于鞅的优先BCQ.D4RL和Torcs数据集上的实验结果表明:所提离线优先采样模型可以有针对性地选择有利于值函数估计或策略优化的经验数据,获得更高的回报.展开更多
基金Projects(61573052,61273132)supported by the National Natural Science Foundation of China
文摘This work is concerned with identification and nonlinear predictive control method for MIMO Hammerstein systems with constraints. Firstly, an identification method based on steady-state responses and sub-model method is introduced to MIMO Hammerstein system. A modified version of artificial bee colony algorithm is proposed to improve the prediction ability of Hammerstein model. Next, a computationally efficient nonlinear model predictive control algorithm(MGPC) is developed to deal with constrained problem of MIMO system. The identification process and performance of MGPC are shown. Numerical results about a polymerization reactor validate the effectiveness of the proposed method and the comparisons show that MGPC has a better performance than QDMC and basic GPC.
基金Project(61374051,61603387)supported by the National Natural Science Foundation of ChinaProjects(20150520112JH,20160414033GH)supported by the Scientific and Technological Development Plan in Jilin Province of ChinaProject(20150102)supported by Opening Funding of State Key Laboratory of Management and Control for Complex Systems,China
文摘A decentralized adaptive neural network sliding mode position/force control scheme is proposed for constrained reconfigurable manipulators. Different from the decentralized control strategy in multi-manipulator cooperation, the proposed decentralized position/force control scheme can be applied to series constrained reconfigurable manipulators. By multiplying each row of Jacobian matrix in the dynamics by contact force vector, the converted joint torque is obtained. Furthermore, using desired information of other joints instead of their actual values, the dynamics can be represented as a set of interconnected subsystems by model decomposition technique. An adaptive neural network controller is introduced to approximate the unknown dynamics of subsystem. The interconnection and the whole error term are removed by employing an adaptive sliding mode term. And then, the Lyapunov stability theory guarantees the stability of the closed-loop system. Finally, two reconfigurable manipulators with different configurations are employed to show the effectiveness of the proposed decentralized position/force control scheme.
基金Supported by National Natural Science Foundation of P. R. China (60474051, 60534020)Development Program of Shanghai Science and Technology Department (04DZ11008)the Program for New Century Excellent Talents in Universities of P. R. China (NCET)
文摘Aiming at a class of nonlinear systems with multiple equilibrium points, we present a dual-mode model predictive control algorithm with extended terminal constraint set combined with control invariant set and gain schedule. Local LQR control laws and the corresponding maximum control invariant sets can be designed for finite equilibrium points. It is guaranteed that control invariant sets are overlapped each other. The union of the control invariant sets is treated as the terminal constraint set of predictive control. The feasibility and stability of the novel dual-mode model predictive control are investigated with both variable and fixed horizon. Because of the introduction of extended terminal constrained set, the feasibility of optimization can be guaranteed with short prediction horizon. In this way, the size of the optimization problem is reduced so it is computationally efficient. Finally, a simulation example illustrating the algorithm is presented.
基金the National High-Tech. R & D Program for CIMS, China (2003AA413210).
文摘To study the uncertain optimization problems on implementation schedule, time-cost trade-off and quality in enterprise resource planning （ERP） implementation, combined with program evaluation and review technique （PERT）, some optimization models are proposed, which include the implementation schedule model, the timecost trade-off model, the quality model, and the implementation time-cost-quality synthetic optimization model. A PERT-embedded genetic algorithm （GA） based on stochastic simulation technique is introduced to the optimization models solution. Finally, an example is presented to show that the models and algorithm are reasonable and effective, which can offer a reliable quantitative decision method for ERP implementation.
文摘大规模开发和利用风能有利于实现电力系统清洁低碳转型,是实现国家“碳达峰、碳中和”战略目标的重要技术手段,但风电出力的强不确定性对电力系统区域间可用输电能力(available transfer capability,ATC)评估带来了全新的挑战,传统用于求解计及风电出力不确定性的概率ATC评估模型在计算效率和计算精度方面均存在一定的不足。为此,该文提出一种基于多项式混沌展开(polynomialchaos expansion,PCE)的电力系统概率ATC评估方法,该方法首先构建基于机会约束的电力系统概率ATC评估模型;然后,根据风电出力预测误差的概率分布特征,选择对应的正交多项式为基函数以近似风电出力预测误差及电力网络中与之相关联的其他随机变量;进一步,借助Galerkin投影和基于一阶矩、二阶矩的机会约束转化方法,将所构建的机会约束模型的概率约束转化为确定性约束,实现基于机会约束的概率ATC评估模型向易于求解的确定性优化模型的转化;进而,将概率ATC评估模型的求解问题转化为ATC的最优多项式逼近系数的求解问题,根据求得的最优多项式逼近系数和选取的基函数计算电力系统ATC的概率分布特征;最后,通过修改后的PJM-5节点测试系统、IEEE-118节点测试系统及吉林西部电网实际算例验证了所提基于多项式混沌展开的电力系统概率ATC评估方法的准确性和有效性。
文摘离线强化学习通过减小分布偏移实现了习得策略向行为策略的逼近,但离线经验缓存的数据分布往往会直接影响习得策略的质量.通过优化采样模型来改善强化学习智能体的训练效果,提出两种离线优先采样模型:基于时序差分误差的采样模型和基于鞅的采样模型.基于时序差分误差的采样模型可以使智能体更多地学习值估计不准确的经验数据,通过估计更准确的值函数来应对可能出现的分布外状态.基于鞅的采样模型可以使智能体更多地学习对策略优化有利的正样本,减少负样本对值函数迭代的影响.进一步,将所提离线优先采样模型分别与批约束深度Q学习(Batch-constrained deep Q-learning,BCQ)相结合,提出基于时序差分误差的优先BCQ和基于鞅的优先BCQ.D4RL和Torcs数据集上的实验结果表明:所提离线优先采样模型可以有针对性地选择有利于值函数估计或策略优化的经验数据,获得更高的回报.
