In this paper, an approximate online equilibrium solution is developed for N -player nonzero-sum (NZS) game systems with completely unknown dynamics. First, a model identifier based on three-layer neural network (NN) established to reconstruct the NZS games systems. Moreover, weight vector updated experience replay technique which can relax traditional persistence of excitation condition simplified recorded data. Then, single-network adaptive dynamic programming (ADP) algorithm proposed each...

In this paper, the H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">∞</sub> optimal control problem for a class of continuous-time nonlinear systems is investigated using event-triggered method. First, formulated as two-player zero-sum (ZS) differential game. Then, an adaptive triggering condition derived ZS game with policy and time-triggered disturbance policy. The controller updated only when not satisfied. Therefore, communication between...

In this paper, the infinite-horizon robust optimal control problem for a class of continuous-time uncertain nonlinear systems is investigated by using data-based adaptive critic designs. The neural network identification scheme combined with traditional technique, in order to design under environment. First, controller original system specified cost function established adding feedback gain nominal system. Then, identifier employed reconstruct unknown dynamics stability analysis. Hence,...

In this paper, the robust control problem for a class of continuous-time nonlinear system with unmatched uncertainties is investigated using an event-based method. First, transformed into corresponding optimal augmented and appropriate cost function. Under mechanism, we prove that solution can asymptotically stabilize uncertain adaptive triggering condition. That is, designed controller to original system. Note updated only when condition satisfied, which save communication resources between...

This paper investigates the automatic exploration problem under unknown environment, which is key point of applying robotic system to some social tasks. The solution this via stacking decision rules impossible cover various environments and sensor properties. Learning-based control methods are adaptive for these scenarios. However, damaged by low learning efficiency awkward transferability from simulation reality. In paper, we construct a general framework decomposing process into decision,...

t his paper investigates the vision- based autonomous driving with deep learning and reinforcement methods.Different from end-to-end method, our method breaks vision-based lateral control system down into a perception module module.The which is on multi-task neural network first takes driver-view image as its input predicts track features.The then makes decision these features.In order to improve data efficiency, we propose visual TORCS (VTORCS), environment open racing car simulator...

Autonomous driving decision-making is a great challenge due to the complexity and uncertainty of traffic environment. Combined with rule-based constraints, Deep Q-Network (DQN) based method applied for autonomous lane change task in this study. Through combination high-level lateral low-level trajectory modification, safe efficient behavior can be achieved. With setting our state representation reward function, trained agent able take appropriate actions real-world-like simulator. The...

The Practical Byzantine Fault Tolerance algorithm (PBFT)has been highly applied in consortium blockchain systems, however, this kind of consensus can hardly identify and remove faulty nodes time, also vulnerable to many attacks against the primary node PBFT. equality members' discourse rights is inapplicable some real scenarios where dominating members are likely have a larger voting process. To address these problems, paper presents Reputation-based (RBFT)algorithm that incorporates...

This paper is concerned about the nonlinear optimization problem of nonzero-sum (NZS) games with unknown drift dynamics. The data-based integral reinforcement learning (IRL) method proposed to approximate Nash equilibrium NZS iteratively. Furthermore, we prove that IRL equivalent model-based policy iteration algorithm, which guarantees convergence method. For implementation purpose, a single-critic neural network structure for given. To enhance application capability method, design updating...

Self-supervised depth estimation draws a lot of attention recently as it can promote the 3D sensing capa-bilities self-driving vehicles. However, intrinsically relies upon photometric consistency assumption, which hardly holds during nighttime. Although various supervised night-time image enhancement methods have been proposed, their generalization performance in challenging driving scenarios is not satisfactory. To this end, we propose first method that jointly learns nighttime enhancer and...

Named data networking (NDN) enables fast and efficient content dissemination in mission-critical unmanned aerial vehicle ad hoc networks (UAANETs); however, its in-network caching mechanism brings a new security challenge: poisoning. Poisoned can contaminate the cache on routers isolate valid from network, leading to performance degradation or denial of service. To mitigate such attacks enhance network-layer trust NDN-based UAANETs, this article proposes novel systematic framework that...

Stable carbocations such as tritylium ions have been widely explored organic Lewis acid catalysts and reagents in synthesis. However, achieving asymmetric carbocation catalysis remains elusive ever since they were first identified over one century ago. The challenges mainly come from their limited compatibility, scarcity of chiral carbocations, well the extremely low barrier to racemization carbenium ions. We reported here a latent concept for catalysis. In this strategy, trityl phosphate is...

As job insecurity becomes increasingly common, seeking its palliatives has become a hot topic for scholars, especially high-speed railway drivers who are vital the development of China's railway. Researches have demonstrated that, organizational support is valuable psychosocial resource that can alleviate individual negative behavior, yet buffering effect between and safety performance attracted little attention. In this study, in field was identified as supervisory coworker safety. Using...

Sum of squares (SOS) polynomials have provided a computationally tractable way to deal with inequality constraints appearing in many control problems. It can also act as an approximator the framework adaptive dynamic programming. In this paper, approximate solution optimal polynomial nonlinear systems is proposed. Under given attenuation coefficient, Hamilton-Jacobi-Isaacs equation relaxed optimization problem set inequalities. After applying policy iteration technique and constraining...

Block copolymer polymersomes offer considerable access for applications in a variety of fields; however, the traditional cosolvent self-assembly method can only produce at low solids content (typically <1%). Recently, an situ growth method, termed polymerization-induced (PISA), has been developed to allow preparation high (10–50%). Synthesis and block copolymers occur simultaneously PISA, therefore, morphological evolution occurs throughout polymerization. It is highly desirable provide...

Deep reinforcement learning (DRL), combining the perception capability of deep (DL) and decision-making (RL) [1], has been widely investigated for autonomous driving tasks. In this letter, we would like to discuss impact different types state input on performance DRL-based lane change decision-making.

Generalizing policies to unseen scenarios remains a critical challenge in visual reinforcement learning, where agents often overfit the specific observations of training environment. In environments, distracting pixels may lead extract representations containing task-irrelevant information. As result, deviate from optimal behaviors learned during training, thereby hindering generalization.To address this issue, we propose Salience-Invariant Consistent Policy Learning (SCPL) algorithm, an...