针对郊狼优化算法(coyote optimization algorithm,COA)存在收敛速度慢、求解精度低、易陷入局部最优的不足,提出一种基于双策略学习机制和自适应混沌变异策略的改进郊狼算法(coyote optimization algorithm based on dual strategy lea...针对郊狼优化算法(coyote optimization algorithm,COA)存在收敛速度慢、求解精度低、易陷入局部最优的不足,提出一种基于双策略学习机制和自适应混沌变异策略的改进郊狼算法(coyote optimization algorithm based on dual strategy learning and adaptive chaotic mutation,DCSCOA)。首先,引入振荡递减因子,以产生具有多样性的个体来增强全局搜索能力;其次,利用双策略学习机制,适度地增强组群头狼的影响,以平衡算法的局部挖掘能力和全局搜索能力,同时提高算法的求解精度和收敛速度;最后,使用自适应混沌变异机制,在算法停滞时产生新个体,以使算法跳出局部最优。通过对20个基本测试函数和11个CEC2017测试函数进行仿真实验,结果验证了改进算法具有更高的求解精度、更快的收敛速度和更强的稳定性。展开更多
In order to raise the detection precision of the extended binary phase shift keying (EBPSK) receiver, a detector based on the improved particle swarm optimization algorithm (IMPSO) and the BP neural network is des...In order to raise the detection precision of the extended binary phase shift keying (EBPSK) receiver, a detector based on the improved particle swarm optimization algorithm (IMPSO) and the BP neural network is designed. First, the characteristics of EBPSK modulated signals and the special filtering mechanism of the impacting filter are demonstrated. Secondly, an improved particle swarm optimization algorithm based on the logistic chaos disturbance operator and the Cauchy mutation operator is proposed, and the EBPSK detector is designed by utilizing the IMPSO-BP neural network. Finally, the simulation of the EBPSK detector based on the MPSO-BP neural network is conducted and the result is compared with that of the adaptive threshold-based decision, the BP neural network, and the PSO-BP detector, respectively. Simulation results show that the detection performance of the EBPSK detector based on the IMPSO-BP neural network is better than those of the other three detectors.展开更多
文摘针对郊狼优化算法(coyote optimization algorithm,COA)存在收敛速度慢、求解精度低、易陷入局部最优的不足,提出一种基于双策略学习机制和自适应混沌变异策略的改进郊狼算法(coyote optimization algorithm based on dual strategy learning and adaptive chaotic mutation,DCSCOA)。首先,引入振荡递减因子,以产生具有多样性的个体来增强全局搜索能力;其次,利用双策略学习机制,适度地增强组群头狼的影响,以平衡算法的局部挖掘能力和全局搜索能力,同时提高算法的求解精度和收敛速度;最后,使用自适应混沌变异机制,在算法停滞时产生新个体,以使算法跳出局部最优。通过对20个基本测试函数和11个CEC2017测试函数进行仿真实验,结果验证了改进算法具有更高的求解精度、更快的收敛速度和更强的稳定性。
基金The National Natural Science Foundation of China (No.60872075)the National High Technology Research and Development Program of China (863 Program) (No. 2008AA01Z227)
文摘In order to raise the detection precision of the extended binary phase shift keying (EBPSK) receiver, a detector based on the improved particle swarm optimization algorithm (IMPSO) and the BP neural network is designed. First, the characteristics of EBPSK modulated signals and the special filtering mechanism of the impacting filter are demonstrated. Secondly, an improved particle swarm optimization algorithm based on the logistic chaos disturbance operator and the Cauchy mutation operator is proposed, and the EBPSK detector is designed by utilizing the IMPSO-BP neural network. Finally, the simulation of the EBPSK detector based on the MPSO-BP neural network is conducted and the result is compared with that of the adaptive threshold-based decision, the BP neural network, and the PSO-BP detector, respectively. Simulation results show that the detection performance of the EBPSK detector based on the IMPSO-BP neural network is better than those of the other three detectors.