In this paper, a new implementation of genetic algorithms (GAs) is developed for the machine scheduling problem, which is abundant among the modern manufacturing systems. The performance measure of early and tardy com...In this paper, a new implementation of genetic algorithms (GAs) is developed for the machine scheduling problem, which is abundant among the modern manufacturing systems. The performance measure of early and tardy completion of jobs is very natural as one's aim, which is usually to minimize simultaneously both earliness and tardiness of all jobs. As the problem is NP-hard and no effective algorithms exist, we propose a hybrid genetic algorithms approach to deal with it. We adjust the crossover and mutation probabilities by fuzzy logic controller whereas the hybrid genetic algorithm does not require preliminary experiments to determine probabilities for genetic operators. The experimental results show the effectiveness of the GAs method proposed in the paper.展开更多
为了提升锂离子电池组均衡系统的性能,提出了一种基于模糊自适应模型预测控制(fuzzy adaptive model predictive control,FAMPC)的模块化均衡系统。首先,由改进的buck-boost电路和反激变压器组成双层均衡拓扑结构;其次,以不同电池剩余容...为了提升锂离子电池组均衡系统的性能,提出了一种基于模糊自适应模型预测控制(fuzzy adaptive model predictive control,FAMPC)的模块化均衡系统。首先,由改进的buck-boost电路和反激变压器组成双层均衡拓扑结构;其次,以不同电池剩余容量(state of charge,SOC)的状态作为模糊逻辑算法的输入,对均衡电流的约束条件进行调节;再次,基于FAMPC均衡控制方法,直接利用开关管的占空比作为系统输入;最后,在改变电池组状态并不使用额外电流控制机制的情况下进行仿真实验。结果表明,与传统的模糊控制方法相比,所提系统在正常条件下均衡速度提高了约24.51%,在电池低SOC的极端条件下均衡速度可以进一步提高至34.48%。所提系统将模糊算法提供的稳定性与模型预测控制算法的快速性相结合,保证了电池组更安全稳定的运行,可为电池组性能提升研究提供参考。展开更多
随着电力物联网技术的快速发展,建设能源互联网具有重大意义。电力物联终端设备的识别认证是保障能源互联网安全稳定运行的基础。为实现海量电力终端设备信息高效采集与安全认证,研究提出一种面向电力物联网的RFID(radio frequency iden...随着电力物联网技术的快速发展,建设能源互联网具有重大意义。电力物联终端设备的识别认证是保障能源互联网安全稳定运行的基础。为实现海量电力终端设备信息高效采集与安全认证,研究提出一种面向电力物联网的RFID(radio frequency identification)认证方案,该方案利用RFID技术,基于国密SM3和SM4设计算法,实现了阅读器与电力设备之间的相互认证,保障了电力通信数据的传输安全,降低设备标签的计算复杂度。安全性分析表明,该方案满足不可追踪性、抗重放攻击、抗去同步攻击、抗拒绝服务攻击等安全特性,BAN逻辑分析进一步表明该方案满足相互认证性。性能分析表明,该方案在标签计算量、存储量、通信量及数据库搜索效率方面具有较好的性能优势。展开更多
Task scheduling for electro-magnetic detection satellite is a typical combinatorial optimization problem. The count of constraints that need to be taken into account is of large scale. An algorithm combined integer pr...Task scheduling for electro-magnetic detection satellite is a typical combinatorial optimization problem. The count of constraints that need to be taken into account is of large scale. An algorithm combined integer programming with constraint programming is presented. This algorithm is deployed in this problem through two steps. The first step is to decompose the original problem into master and sub-problem using the logic-based Benders decomposition; then a circus combines master and sub-problem solving process together, and the connection between them is general Benders cut. This hybrid algorithm is tested by a set of derived experiments. The result is compared with corresponding outcomes generated by the strength Pareto evolutionary algorithm and the pure constraint programming solver GECODE, which is an open source software. These tests and comparisons yield promising effect.展开更多
文摘In this paper, a new implementation of genetic algorithms (GAs) is developed for the machine scheduling problem, which is abundant among the modern manufacturing systems. The performance measure of early and tardy completion of jobs is very natural as one's aim, which is usually to minimize simultaneously both earliness and tardiness of all jobs. As the problem is NP-hard and no effective algorithms exist, we propose a hybrid genetic algorithms approach to deal with it. We adjust the crossover and mutation probabilities by fuzzy logic controller whereas the hybrid genetic algorithm does not require preliminary experiments to determine probabilities for genetic operators. The experimental results show the effectiveness of the GAs method proposed in the paper.
文摘为了提升锂离子电池组均衡系统的性能,提出了一种基于模糊自适应模型预测控制(fuzzy adaptive model predictive control,FAMPC)的模块化均衡系统。首先,由改进的buck-boost电路和反激变压器组成双层均衡拓扑结构;其次,以不同电池剩余容量(state of charge,SOC)的状态作为模糊逻辑算法的输入,对均衡电流的约束条件进行调节;再次,基于FAMPC均衡控制方法,直接利用开关管的占空比作为系统输入;最后,在改变电池组状态并不使用额外电流控制机制的情况下进行仿真实验。结果表明,与传统的模糊控制方法相比,所提系统在正常条件下均衡速度提高了约24.51%,在电池低SOC的极端条件下均衡速度可以进一步提高至34.48%。所提系统将模糊算法提供的稳定性与模型预测控制算法的快速性相结合,保证了电池组更安全稳定的运行,可为电池组性能提升研究提供参考。
文摘随着电力物联网技术的快速发展,建设能源互联网具有重大意义。电力物联终端设备的识别认证是保障能源互联网安全稳定运行的基础。为实现海量电力终端设备信息高效采集与安全认证,研究提出一种面向电力物联网的RFID(radio frequency identification)认证方案,该方案利用RFID技术,基于国密SM3和SM4设计算法,实现了阅读器与电力设备之间的相互认证,保障了电力通信数据的传输安全,降低设备标签的计算复杂度。安全性分析表明,该方案满足不可追踪性、抗重放攻击、抗去同步攻击、抗拒绝服务攻击等安全特性,BAN逻辑分析进一步表明该方案满足相互认证性。性能分析表明,该方案在标签计算量、存储量、通信量及数据库搜索效率方面具有较好的性能优势。
基金supported by the National Security Fundamental Research Foundation of China (61361)the National Natural Science Foundation of China (61104180)
文摘Task scheduling for electro-magnetic detection satellite is a typical combinatorial optimization problem. The count of constraints that need to be taken into account is of large scale. An algorithm combined integer programming with constraint programming is presented. This algorithm is deployed in this problem through two steps. The first step is to decompose the original problem into master and sub-problem using the logic-based Benders decomposition; then a circus combines master and sub-problem solving process together, and the connection between them is general Benders cut. This hybrid algorithm is tested by a set of derived experiments. The result is compared with corresponding outcomes generated by the strength Pareto evolutionary algorithm and the pure constraint programming solver GECODE, which is an open source software. These tests and comparisons yield promising effect.
