In the field of calculating the attack area of air-to-air missiles in modern air combat scenarios,the limitations of existing research,including real-time calculation,accuracy efficiency trade-off,and the absence of t...In the field of calculating the attack area of air-to-air missiles in modern air combat scenarios,the limitations of existing research,including real-time calculation,accuracy efficiency trade-off,and the absence of the three-dimensional attack area model,restrict their practical applications.To address these issues,an improved backtracking algorithm is proposed to improve calculation efficiency.A significant reduction in solution time and maintenance of accuracy in the three-dimensional attack area are achieved by using the proposed algorithm.Furthermore,the age-layered population structure genetic programming(ALPS-GP)algorithm is introduced to determine an analytical polynomial model of the three-dimensional attack area,considering real-time requirements.The accuracy of the polynomial model is enhanced through the coefficient correction using an improved gradient descent algorithm.The study reveals a remarkable combination of high accuracy and efficient real-time computation,with a mean error of 91.89 m using the analytical polynomial model of the three-dimensional attack area solved in just 10^(-4)s,thus meeting the requirements of real-time combat scenarios.展开更多
A novel space-borne antenna adaptive anti-jamming method based on the genetic algorithm （GA）, which is combined with gradient-like reproduction operators is presented, to search for the best weight for pattern synth...A novel space-borne antenna adaptive anti-jamming method based on the genetic algorithm （GA）, which is combined with gradient-like reproduction operators is presented, to search for the best weight for pattern synthesis in radio frequency （RF）. Combined, the GA＇s the capability of the whole searching is, but not limited by selection of the initial parameter, with the gradient algorithm＇s advantage of fast searching. The proposed method requires a smaller sized initial population and lower computational complexity. Therefore, it is flexible to implement this method in the real-time systems. By using the proposed algorithm, the designer can efficiently control both main-lobe shaping and side-lobe level. Simulation results based on the spot survey data show that the algorithm proposed is efficient and feasible.展开更多
为适应大容量同步发电机组并网点母线电压波动增加对自动电压调节器(automatic voltage regulator,AVR)系统响应能力的更高要求,提出一种基于含探索网络的双延迟深度确定性策略梯度(twin delayed deep deterministic policy gradient wi...为适应大容量同步发电机组并网点母线电压波动增加对自动电压调节器(automatic voltage regulator,AVR)系统响应能力的更高要求,提出一种基于含探索网络的双延迟深度确定性策略梯度(twin delayed deep deterministic policy gradient with Explorer network,TD3EN)算法的同步发电机励磁电压控制方法。首先,通过传递函数对同步发电机励磁调压子系统进行建模;然后建立TD3EN算法探索网络、动作网络和评价网络,并设置相应参数;接着利用TD3EN算法训练智能体,通过探索网络探索动作空间,并根据评价网络更新动作网络参数,使其为AVR提供控制信号;将训练完成的智能体接入AVR系统,实现对发电机机端电压的控制。仿真结果表明,所提方法提高了AVR系统响应调节指令和应对电压暂降的能力。展开更多
针对深度确定性策略梯度(deep deterministic policy gradient,DDPG)算法在一些大状态空间任务中存在学习效果不佳及波动较大等问题,提出一种基于渐近式k-means聚类算法的多行动者深度确定性策略梯度(multi-actor deep deterministic po...针对深度确定性策略梯度(deep deterministic policy gradient,DDPG)算法在一些大状态空间任务中存在学习效果不佳及波动较大等问题,提出一种基于渐近式k-means聚类算法的多行动者深度确定性策略梯度(multi-actor deep deterministic policy gradient based on progressive k-means clustering,MDDPG-PK-Means)算法.在训练过程中,对每一时间步下的状态进行动作选择时,根据k-means算法判别结果辅佐行动者网络的决策,同时随训练时间步的增加,逐渐增加k-means算法类簇中心的个数.将MDDPG-PK-Means算法应用于MuJoCo仿真平台上,实验结果表明,与DDPG等算法相比,MDDPG-PK-Means算法在大多数连续任务中都具有更好的效果.展开更多
基金National Natural Science Foundation of China(62373187)Forward-looking Layout Special Projects(ILA220591A22)。
文摘In the field of calculating the attack area of air-to-air missiles in modern air combat scenarios,the limitations of existing research,including real-time calculation,accuracy efficiency trade-off,and the absence of the three-dimensional attack area model,restrict their practical applications.To address these issues,an improved backtracking algorithm is proposed to improve calculation efficiency.A significant reduction in solution time and maintenance of accuracy in the three-dimensional attack area are achieved by using the proposed algorithm.Furthermore,the age-layered population structure genetic programming(ALPS-GP)algorithm is introduced to determine an analytical polynomial model of the three-dimensional attack area,considering real-time requirements.The accuracy of the polynomial model is enhanced through the coefficient correction using an improved gradient descent algorithm.The study reveals a remarkable combination of high accuracy and efficient real-time computation,with a mean error of 91.89 m using the analytical polynomial model of the three-dimensional attack area solved in just 10^(-4)s,thus meeting the requirements of real-time combat scenarios.
基金the National Natural Science Foundation of China (60502045).
文摘A novel space-borne antenna adaptive anti-jamming method based on the genetic algorithm （GA）, which is combined with gradient-like reproduction operators is presented, to search for the best weight for pattern synthesis in radio frequency （RF）. Combined, the GA＇s the capability of the whole searching is, but not limited by selection of the initial parameter, with the gradient algorithm＇s advantage of fast searching. The proposed method requires a smaller sized initial population and lower computational complexity. Therefore, it is flexible to implement this method in the real-time systems. By using the proposed algorithm, the designer can efficiently control both main-lobe shaping and side-lobe level. Simulation results based on the spot survey data show that the algorithm proposed is efficient and feasible.
文摘为适应大容量同步发电机组并网点母线电压波动增加对自动电压调节器(automatic voltage regulator,AVR)系统响应能力的更高要求,提出一种基于含探索网络的双延迟深度确定性策略梯度(twin delayed deep deterministic policy gradient with Explorer network,TD3EN)算法的同步发电机励磁电压控制方法。首先,通过传递函数对同步发电机励磁调压子系统进行建模;然后建立TD3EN算法探索网络、动作网络和评价网络,并设置相应参数;接着利用TD3EN算法训练智能体,通过探索网络探索动作空间,并根据评价网络更新动作网络参数,使其为AVR提供控制信号;将训练完成的智能体接入AVR系统,实现对发电机机端电压的控制。仿真结果表明,所提方法提高了AVR系统响应调节指令和应对电压暂降的能力。
文摘针对深度确定性策略梯度(deep deterministic policy gradient,DDPG)算法在一些大状态空间任务中存在学习效果不佳及波动较大等问题,提出一种基于渐近式k-means聚类算法的多行动者深度确定性策略梯度(multi-actor deep deterministic policy gradient based on progressive k-means clustering,MDDPG-PK-Means)算法.在训练过程中,对每一时间步下的状态进行动作选择时,根据k-means算法判别结果辅佐行动者网络的决策,同时随训练时间步的增加,逐渐增加k-means算法类簇中心的个数.将MDDPG-PK-Means算法应用于MuJoCo仿真平台上,实验结果表明,与DDPG等算法相比,MDDPG-PK-Means算法在大多数连续任务中都具有更好的效果.
