For jammer formations, due to advanced digital signal processing and precise synchronization, suppressing and penetrating netted radar systems face more challenges than ever. In this letter, a deep reinforcement learning-based joint path planning and jamming power allocation optimization (DRL-JPPAO) method is proposed. DRL-JPPAO models the penetration operation by a Markov Decision Process and utilizes the Proximal Policy Optimization algorithm to learn optimal sequential actions, which consist of waypoints and jamming power allocation matrices, to solve it. The learning process is guided by a reward function that depends on the success or failure of penetration and the distance to the target. Empirical results with a newly developed environment show that DRL-JPPAO can learn optimal paths and jamming power allocation strategies jointly, accomplishing the penetration operation effectively and elegantly.

In a mechanical interface, the contact surface topography has an important influence on the contact stiffness. In the contact processes of asperities, elastic-plastic change can lead to discontinuity and lack of smoothness at a critical contact point. The result is a large difference between the elastic-plastic deformation and the actual asperity deformation. Based on Hertz contact theory, the heights of asperities on a rough surface obey a Gaussian distribution. To take into consideration the continuity of elastic-plastic asperity deformation, we divide the elastic-plastic deformation into three stages: pre-elastic-plastic, mid-elastic-plastic, and post-elastic-plastic deformation. This establishes an elastic-plastic contact model of asperity at a continuous critical point. The contact model of a single asperity fits well with the Kogut–Etsion model and the Zhao–Maietta–Chang model, and the variation trend is consistent. At a lower plastic index, the present model coincides with classical models of contact area and contact load. At a higher plastic index, the simulation results of the present model differ from the Greenwood–Williamson model and the Chang–Etsion–Bogy model but are similar to results from the Kogut–Etsion and Zhao–Maietta–Chang models. This study provides a more accurate microscopic contact model for rough surfaces and a theoretical framework for interface design and analysis.

In order to investigate the relationship between rock microfracture mechanism and acoustic emission (AE) signal characteristic parameters under split loads, the MTS322 servo-controlled rock mechanical test system was employed to carry out the Brazilian split tests on granite, marble, sandstone, and limestone, while FEI Quanta-200 scanning electron microscope system was employed to carry out the analysis of fracture morphology. The results indicate that different scales of mineral particle, mineral composition, and discontinuity have influence on the fracture characteristics of rock, as well as the b-value. The peak frequency distribution of the AE signal has obvious zonal features, and these distinct peak frequencies of four types of rock fall mostly in ranges of 0–100 kHz, 100–300 kHz, and above 300 kHz. Due to the different rock properties and mineral compositions, the proportions of peak frequencies in these intervals are also different among the four rocks, which are also acting on the b-value. In addition, for granite, the peak frequencies of AE signals are mostly distributed above 300 kHz for granite, marble, and limestone, which mainly derive from the internal fracture of k-feldspar minerals; for marble, the AE signals with peak frequency are mostly distributed in over 300 kHz, which mainly derive from the internal fracture of dolomite minerals and calcite minerals; AE signals for sandstone are mostly distributed in the range of 0–100 kHz, which mainly derive from the internal fracture of quartz minerals; for limestone, the AE signals with peak frequency are mostly distributed in over 300 kHz, which mainly derive from the internal fracture of granular-calcite minerals. The relationship between acoustic emission signal frequency of rock fracture and the fracture scale is constructed through experiments, which is of great help for in-depth understanding of the scaling relationship of rock fracture.

Reinforcement learning has recently made great progress in various challenging domains such as board game of Go and MOBA game of StarCraft II. Policy gradient based reinforcement learning method has become the mainstream due to its effectiveness and simplicity both in discrete and continuous scenarios. However, policy gradient methods commonly involve function approximation and work in an on-policy fashion, which leads to high variance and low sample efficiency. This paper introduces a novel policy gradient method to improve the sample efficiency via a pair of trajectory based prioritized replay buffers and reduce the variance in training with a target network whose weights are updated in a “soft” manner. We evaluate our method on the reinforcement learning suit of Open AI Gym tasks, and the results show that the proposed method can learn more steadily and achieve higher performance than existing methods.

The exponential explosion of joint actions and massive data collection are two main challenges in multiagent reinforcement learning algorithms with centralized training. To overcome these problems, in this paper, we propose a model‐free and fully decentralized actor‐critic multiagent reinforcement learning algorithm based on message diffusion. To this end, the agents are assumed to be placed in a time‐varying communication network. Each agent makes limited observations regarding the global state and joint actions; therefore, it needs to obtain and share information with others over the network. In the proposed algorithm, agents hold local estimations of the global state and joint actions and update them with local observations and the messages received from neighbors. Under the hypothesis of the global value decomposition, the gradient of the global objective function to an individual agent is derived. The convergence of the proposed algorithm with linear function approximation is guaranteed according to the stochastic approximation theory. In the experiments, the proposed algorithm was applied to a passive location task multiagent environment and achieved superior performance compared to state‐of‐the‐art algorithms.

Passive location systems receive electromagnetic waves at one or multiple base stations to locate the transmitters, which are widely used in security fields. However, the geometric configurations of stations can greatly affect the positioning precision. In the literature, the geometry of the passive location system is mainly designed based on empirical models. These empirical models, being hard to track the sophisticated electromagnetic environment in the real world, result in suboptimal geometric configurations and low positioning precision. In order to master the characteristics of complicated electromagnetic environments to improve positioning performance, this paper proposes a novel geometry optimization method based on multiagent reinforcement learning. In the proposed method, agents learn to optimize the geometry cooperatively by factorizing team value function into agentwise value functions. To facilitate cooperation and deal with data transmission challenges, a constraint is imposed on the data sent from the central station to vice stations to ensure conciseness and effectiveness of communications. According to the empirical results under direct position determination systems, agents can find better geometric configurations than the existing methods in complicated electromagnetic environments.