The unpredictable dynamic events in smart factory seriously influence the scheduling schemes and production efficiency. To minimize the total tardiness of orders, this paper proposes a Deep Reinforcement Learning (DRL) method to solve the Dynamic Flexible Job Shop Scheduling Problem (DFJSP) with random job arrival. In the scheduling process, the intelligent agent can select the operations to be processed on the available machines according to the job shop state at each scheduling point by transforming DFJSP into a Markov Decision Process (MDP). For reflecting the job shop environment, eight state features are extracted, and six composite dispatching rules are designed as the action space. Moreover, a reward function is developed to enhance the learning efficiency and solution quality of the intelligent agent, the Soft-max selection strategy is applied to balance exploration and exploitation. The Dueling Double Deep Q-Network (Dueling DDQN) algorithm is introduced to train the intelligent agent, which effectively alleviates the overestimation problem. Numerical simulation results show that the proposed Dueling DDQN method effectively solves the DFJSP. Compared with existing scheduling strategies, the proposed approach not only exhibits superior performance but also maintains its advantages among different scale instances.
AbstractIn recent years, eco‐driving for hybrid electric vehicles (HEVs) has been studied with the emerging connected and automated vehicle (CAV) technologies to improve the mobility, fuel economy and safety of HEVs. This paper develops a hierarchical eco‐driving control strategy for HEV platoons consisting of CAVs and human‐driven vehicles (HDVs) to improve fuel economy at signalized intersections. For each platoon, the speed trajectories of CAVs are optimized in the upper layer controller using model predictive control (MPC) to minimize the total fuel consumption of the whole platoon. For each HEV, the optimal power split between the engine and the battery is obtained in the lower layer controller using adaptive equivalent consumption minimization strategy (A‐ECMS). The time‐varying powertrain efficiencies of HEVs are explicitly considered in the speed trajectory optimization. At last, simulation studies are conducted using MATLAB and VISSIM to evaluate the performances of the strategy in mixed traffic scenarios and different CAV penetration rates. Simulation results indicate that compared with the single vehicle control strategy, the proposed strategy can improve the average fuel economy by up to 8.34% and considering the time‐varying powertrain efficiencies of HEVs in the optimization can further reduce the fuel consumption by up to 1.23%.
Intelligent manufacturing requires industrialized software bases.Deep application of industrial software is the key to ac-celerating the digital transformation of the manufacturing industry.The Manufacturing Execution System(MES)plays an increas-ingly important role in workshop production.In the future development,intelligent manufacturing puts forward the demand for modularization,componentization,and platformization of the MES system.Based on microservices,first of all,the business functions of the MES system are divided into microservices,and the manufacturing execution system framework based on microser-vices architecture is designed.Then the binding mechanism of the registration center and the micro service library is proposed,which can effectively respond to the changing and dynamic changes of the business process of the enterprise.On this basis,how to select the optimal microservices and business process microservices required by the MES system from the microservices library is further studied,which plays a reference role for the development of MES system based on microservices.
随着机器人技术的发展,关节型工业机器人凭借效率高、精度高及良好的环境适应性与工作稳定性优点而被广泛地应用于自动化工业之中.轨迹规划是提升工业机器人作业质量的有效途径.为全面了解现有轨迹规划研究方法,对各领域轨迹规划及优化研究进展进行全面综述.首先,简述了轨迹规划的流程并根据各种轨迹规划的原理进行分类,介绍了一般轨迹规划中多项式插值法和B样条插值法的研究现状;然后,针对能量、时间及冲击等不同因素的优化研究进行详细的归纳,并指出现有轨迹寻优算法的具体思路;最后,结合实际明确动态避障、强化学习下的轨迹规划、轨迹误差补偿与预测及多机器人协作轨迹规划将是今后的研究重点.
非标自动化设备项目初期的设计环节在整个项目管理过程中占据重要地位,对其设计时间的精准预测是有效实施项目调度的基础.基于非标设备设计和制造特点,选择客户协同程度、造型难度、设计相似程度、订单规模和团队参与程度5个因素,探究其对设计时间的影响.考虑非标产品样本数量较少的特点,提出一种遗传算法与向量机回归(GA-SVR)混合方法,即在向量机回归模型(SVR)预测方法的基础上,结合十倍交叉验证与遗传算法对预测模型进行参数寻优.最后,选择高斯过程回归(GPR)方法进行对比,从预测精度和稳定性两个方面验证了 GA-SVR混合方法的有效性.
