Shadowed type 2 fuzzy-based Markov model to predict shortest path with optimized waiting time

Kumar, Pawan; Dudeja, Chanchal

doi:10.1007/s00500-020-05194-y

Cited by 6 publications

(10 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The history of reinforcement learning can be traced back to the 1950s. Its basic idea is to transform the sequential decision-making problem into a Markov model [28]. It establishes the mapping between the environmental state and the state action value function through the interaction between the agent (robot) and the environment.…”

Section: The Improved Q-learning Algorithmmentioning

confidence: 99%

Improved Q-Learning Applied to Dynamic Obstacle Avoidance and Path Planning

2022

View full text Add to dashboard Cite

Due to the complexity of interactive environments, dynamic obstacle avoidance path planning poses a significant challenge to agent mobility. Dynamic path planning is a complex multi-constraint combinatorial optimization problem. Some existing algorithms easily fall into local optimization when solving such problems, leading to defects in convergence speed and accuracy. Reinforcement learning has certain advantages in solving decision sequence problems in complex environments. A Q-learning algorithm is a reinforcement learning method. In order to improve the value evaluation of the algorithm in solving practical problems, this paper introduces the priority weight into the Q-learning algorithm. The improved algorithm is compared with existing algorithms and applied to dynamic obstacle avoidance path planning. Experiments show that the improved algorithm dramatically improves the convergence speed and accuracy and increases the value evaluation. The improved algorithm finds the shortest path of 16 units in 27 seconds.

show abstract

Section: The Improved Q-learning Algorithmmentioning

confidence: 99%

Improved Q-Learning Applied to Dynamic Obstacle Avoidance and Path Planning

2022

View full text Add to dashboard Cite

show abstract

“…Lemma 1: [25] Taking the stochastic system (11). If there exists a Lyapunov function V ( ̟(k), δ k ) :…”

Section: Dmodelling Of Closed-loop Systemmentioning

confidence: 99%

“…for all ̟ ∈ Q 2n ̟ , k > 0. Then the system (11) Lemma 3: [39] ( Abel Lemma-based Finite-sum Inequality) For any integers ι 1 and ι 2 with ι 2 −ι 1 > 1 with a matrix Q > 0 , then the following inequality holds:…”

Section: Dmodelling Of Closed-loop Systemmentioning

confidence: 99%

“…The neural network approach is utilized in this paper to estimate the harmful information delivered by attackers. Before examining the stability of the closed-loop system, it is essential to address the boundedness of neural network estimation error ω(k)in order to avoid the possibility of neural network weight estimations drifting (11). The law of updating was established for the this reason, the neural network weight estimation error dynamics will be given and its corresponding boundedness is proved .…”

Section: Bneural Network Estimation Error Boundednessmentioning

confidence: 99%

“…NCSs have a number of benefits, better maintainability, including lower installation costs and greater flexibility, as well as better maintainability [9]. Because of time delays, packet failures, quantization inaccuracy and communication limits efforts have been paid into control synthesis and stability analysis of NCSs in recent years [10]- [11].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Adaptive event-triggered control for Markov jump systems under hybrid cyber-attacks and its application

Ali

Rajchakit

et al. 2022

Preprint

View full text Add to dashboard Cite

This article investigates event-triggered control (ETC) of discrete-time Markov jump systems (MJSs) subject to hybrid cyber-attacks and unmeasured states. To reduce data transmission load, a mode-dependent event-triggered mechanism (ETM) has been developed. The unmeasured states are estimated using a Luenberger observer. Here deception attacks and DoS attacks are addressed. It is anticipated that input control signals sent across a network are vulnerable to cyber-attacks, in which adversaries could inject fake data into the control signals. So to secure the closed-loop system, sufficient conditions are derived. The solution of set matrix inequalities is used to build observer and controller gains. Finally, inverted pendulum applications are presented to illustrate the effectiveness and feasibility of the obtained theoretical results.

show abstract