Critical Network Research Articles

A Fast Optimal Coordination Method for Multiagent in Complex Environment

Facing the implementation problems such like low growth reward, long training time, and poor stability of the multiagent learning methods when dealing with complex environment and more agents, this paper proposes a fast optimal coordination method for multiagent in complex environment (FOC-MACE). Firstly, the environment exploration strategy is introduced into the policy network based on the MADDPG method for higher growth rewards. Then, the parallel computing technology is adopted in the critic network, in purpose to effectively reduce the training time. These tactics together are beneficial to enhance the stability of multiagent learning. Lastly, the optimal resource allocation is carried out to realize optimal coevolution of the multiagents and further improve the learning ability of the agents’ group. To verify the effectiveness of our proposal, the FOC-MACE is compared with several advanced methods at current stage in the MPE environment. Three different experiments prove that by using our method, the growth reward is increased by up to 37.1%, the training is speed up significantly, and the stability of the method, which represented by standardized variance, is also improved. In addition, this paper validated the fast optimal coordination method for multiagent systems in the context of UAV scenarios, demonstrating the practical performance of the approach. Through comprehensive experiments and scenario validations, the study successfully confirmed the effectiveness of the proposed fast optimal coordination method for multiagent systems in complex environments.

Sliding-mode surface-based approximate optimal control for nonlinear multiplayer Stackelberg-Nash games via adaptive dynamic programming

This paper studies the sliding-mode surface (SMS)-based approximate optimal control issue for a class of nonlinear multiplayer Stackelberg-Nash games (MSNGs). First, considering different roles of the players in MSNGs, a hierarchical decision-making process is expressed as designing different cost functions for the leader and the followers. Therein, the leader makes its decision preferentially with consideration of all followers, while each follower responds optimally to leader’s strategy simultaneously via Nash games. Then, based on the sliding mode control technology, a novel robust control policy is proposed to solve the coupling Hamilton–Jacobi equations of nonlinear MSNGs. Subsequently, SMS-based critic network is utilized to approximate optimal control inputs for all players. Based on the Lyapunov stability theory, it is strictly proven that all signals of the closed-loop nonlinear systems are uniformly ultimately bounded. Finally, the effectiveness of the proposed method is demonstrated via a simulation example.

Enhancing Heterogeneous Network Performance: Advanced Content Popularity Prediction and Efficient Caching

With the popularity of smart devices and the growth of high-bandwidth applications, the wireless industry is facing an increased surge in data traffic. This challenge highlights the limitations of traditional edge-caching solutions, especially in terms of content-caching effectiveness and network-communication latency. To address this problem, we investigated efficient caching strategies in heterogeneous network environments. The caching decision process becomes more complex due to the heterogeneity of the network environment, as well as due to the diversity of user behaviors and content requests. To address the problem of increased system latency due to the dynamically changing nature of content popularity and limited cache capacity, we propose a novel content placement strategy, the long-short-term-memory–content-population-prediction model, to capture the correlation of request patterns between different contents and the periodicity in the time domain, in order to improve the accuracy of the prediction of content popularity. Then, to address the heterogeneity of heterogeneous network environments, we propose an efficient content delivery strategy: the multi-intelligent critical collaborative caching policy. This strategy models the edge-caching problem in heterogeneous scenarios as a Markov decision process using multi-base-station-environment information. In order to fully utilize the multi-intelligence information, we have improved the actor–critic approach by integrating the attention mechanism into a neural network. Whereas the actor network is responsible for making decisions based on local information, the critic network evaluates and enhances the actor’s performance. We conducted extensive simulations, and the results showed that the Long Short Term Memory content population prediction model was more advantageous, in terms of content-popularity-prediction accuracy, with a 28.61% improvement in prediction error, compared to several other existing methods. The proposed multi-intelligence actor–critic collaborative caching policy algorithm improved the cache-hit-rate metric by up to 32.3% and reduced the system latency by 1.6%, demonstrating the feasibility and effectiveness of the algorithm.

Variation in risk-adjusted cardiac intensive care unit (CICU) length of stay and the association with in-hospital mortality: An analysis from the Critical Care Cardiology Trials Network (CCCTN) registry

BackgroundPrevious studies have suggested that there is wide variability in cardiac intensive care unit (CICU) length of stay (LOS); however, these studies are limited by the absence of detailed risk assessment at the time of admission. Thus, we evaluated inter-hospital differences in CICU LOS, and the association between LOS and in-hospital mortality. MethodsUsing data from the Critical Care Cardiology Trials Network (CCCTN) registry, we included 22,862 admissions between 2017 and 2022 from 35 primarily tertiary and quaternary CICUs that captured consecutive admissions in annual two-month snapshots. The primary analysis compared inter-hospital differences in CICU LOS, as well as the association between CICU LOS and all-cause in-hospital mortality using a Fine and Gray competing risk model. ResultsThe overall median CICU LOS was 2.2 (1.1-4.8) days, and the median hospital LOS was 5.9 (2.8-12.3) days. Admissions in the longest tertile of LOS tended to be younger with higher rates of pre-existing comorbidities, and had higher Sequential Organ Failure Assessment (SOFA) scores, as well as higher rates of mechanical ventilation, intravenous vasopressor use, mechanical circulatory support, and renal replacement therapy. Unadjusted all-cause in-hospital mortality was 9.3%, 6.7%, and 13.4% in the lowest, intermediate, and highest CICU LOS tertiles. In a competing risk analysis, individual patient CICU LOS was correlated (r2=0.31) with a higher risk of 30-day in-hospital mortality. The relationship remained significant in admissions with heart failure, ST-elevation myocardial infarction and non-ST segment elevation myocardial infarction. ConclusionsIn a large registry of academic CICUs, we observed significant variation in CICU LOS and report that LOS is independently associated with all-cause in-hospital mortality. These findings could potentially be used to improve CICU resource utilization planning and refine risk prognostication in critically ill cardiovascular patients.

Optimized leader‐follower consensus control using combination of reinforcement learning and sliding mode mechanism for multiple robot manipulator system

AbstractThis article is to develop an optimized leader‐follower consensus control for multiple robot manipulator system by combining sliding mode control (SMC) and reinforcement learning (RL). The SMC mechanism aims to steer both position and velocity states of multiple manipulator reaching the predefined trajectory. And the RL is designed and performed under identifier‐critic‐actor architecture for the achievement of optimized control performance. Compared to traditional optimal control, this proposed method has two main advantages: (i) the RL updating laws for training both the actor and critic networks are simpler; (ii) the optimized control can not require the complete dynamic knowledge because the adaptive identifier is designed into the RL learning. Consequently, this optimized control method can smoothly steer the multiple robot manipulator system to achieve the leader‐follower consensus. Finally, feasibility of this control method is verified through both theory and simulation.

An efficient and robust gradient reinforcement learning: Deep comparative policy

Recently, actor-critic architectures such as deep deterministic policy gradient (DDPG) are able to understand higher-level concepts for searching rich reward, and generate complex actions in continuous action space, and widely used in practical applications. However, when action space is limited and has dynamic hard margins, training DDPG can be problematic and inefficiency. Since real-world actuators always have margins and interferences, after initialization, the actor network is likely to be stuck at a local optimal point on action space margin: actor gradient orients to the outside of action space but actuators stop at the margin. If the hard margins are complex, dynamic and unknown to the DDPG agent, it is unable to use penalty functions to recover from local optimum. If we enlarge the random process for local exploration, the training could be in potential risk of failure. Therefore, simply relying on gradient of critic network to train the actor network is not a robust method in real environment. To solve this problem, in this paper we modify DDPG to deep comparative policy (DCP). Rather than leveraging critic-to-actor gradient, the core training process of DCP is regulated by a T-fold compare among random proposed adjacent actions. The performance of DDPG, DCP and related algorithms are tested and compared in two experiments. Our results show that, DCP is effective, efficient and qualified to perform all tasks that DDPG can perform. More importantly, DCP is less likely to be influenced by the action space margins, DCP can provide more safety in avoiding training failure and local optimum, and gain more robustness in applications with dynamic hard margins in the action space. Another advantage is that, complex penalty for margin touching detection is not required, the reward function can always be brief and short.

Weather-related fragility modelling of critical infrastructure: a power and railway case study

Climate change has led to more frequent and severe extreme weather events, which impact critical infrastructure networks such as railway and power systems. Although infrastructure networks are interdependent, the analysis to understand the impact of weather events on infrastructure systems is usually performed in sector-specific silos. A methodology to examine how the same weather events affect different infrastructure sectors is presented, in order to understand cross-sectoral impact of extreme weather for interconnected regional infrastructure. Fragility modelling was used to examine the impact of temperature and rainfall on power and rail system failures using the West Midlands (in the UK) as a case study. The results demonstrated that the impact of temperature was broadly consistent across both infrastructure networks, showing less impact until specific upper and lower thresholds are passed; these thresholds were found to be similar for the different infrastructure networks evaluated, but railway infrastructure was found to be impacted more by lower temperatures. A growing correlation between the number of faults on power and railway systems was also found for both rainfall and temperature, indicating the value in coordinating preparation and planning efforts. For infrastructure operators and owners, regional resilience forums and other decision makers, this study provides an approach to assess the regional impact of extreme weather across multiple infrastructure sectors. The results give useful insights to inform the allocation of resources in response to extreme weather events.

Data for critical infrastructure network modelling of natural hazard impacts: Needs and influence on model characteristics

Natural hazards impact interdependent infrastructure networks that keep modern society functional. While a variety of modelling approaches are available to represent critical infrastructure networks (CINs) on different scales and analyse the impacts of natural hazards, a recurring challenge for all modelling approaches is the availability and accessibility of sufficiently high-quality input and validation data. The resulting data gaps often require modellers to assume specific technical parameters, functional relationships, and system behaviours. In other cases, expert knowledge from one sector is extrapolated to other sectoral structures or even cross-sectorally applied to fill data gaps. The uncertainties introduced by these assumptions and extrapolations and their influence on the quality of modelling outcomes are often poorly understood and difficult to capture, thereby eroding the reliability of these models to guide resilience enhancements. Additionally, ways of overcoming the data availability challenges in CIN modelling, with respect to each modelling purpose, remain an open question. To address these challenges, a generic modelling workflow is derived from existing modelling approaches to examine model definition and validations, as well as the six CIN modelling stages, including mapping of infrastructure assets, quantification of dependencies, assessment of natural hazard impacts, response & recovery, quantification of CI services, and adaptation measures. The data requirements of each stage were systematically defined, and the literature on potential sources was reviewed to enhance data collection and raise awareness of potential pitfalls. The application of the derived workflow funnels into a framework to assess data availability challenges. This is shown through three case studies, taking into account their different modelling purposes: hazard hotspot assessments, hazard risk management, and sectoral adaptation. Based on the three model purpose types provided, a framework is suggested to explore the implications of data scarcity for certain data types, as well as their reasons and consequences for CIN model reliability. Finally, a discussion on overcoming the challenges of data scarcity is presented.

Controller-driven vector autoregression model for predicting content popularity in programmable named data networking devices.

Named Data Networking (NDN) has emerged as a promising network architecture for content delivery in edge infrastructures, primarily due to its name-based routing and integrated in-network caching. Despite these advantages, sub-optimal performance often results from the decentralized decision-making processes of caching devices. This article introduces a paradigm shift by implementing a Software Defined Networking (SDN) controller to optimize the placement of highly popular content in NDN nodes. The optimization process considers critical networking factors, including network congestion, security, topology modification, and flowrules alterations, which are essential for shaping content caching strategies. The article presents a novel content caching framework, Popularity-aware Caching in Popular Programmable NDN nodes (PaCPn). Employing a multi-variant vector autoregression (VAR) model driven by an SDN controller, PaCPn periodically updates content popularity based on time-series data, including 'request rates' and 'past popularity'. It also introduces a controller-driven heuristic algorithm that evaluates the proximity of caching points to consumers, considering factors such as 'distance cost,' 'delivery time,' and the specific 'status of the requested content'. PaCPn utilizes customized DATA named packets to ensure the source stores content with a valid residual freshness period while preventing intermediate nodes from caching it. The experimental results demonstrate significant improvements achieved by the proposed technique PaCPn compared to existing schemes. Specifically, the technique enhances cache hit rates by 20% across various metrics, including cache size, Zipf parameter, and exchanged traffic within edge infrastructure. Moreover, it reduces content retrieval delays by 28%, considering metrics such as cache capacity, the number of consumers, and network throughput. This research advances NDN content caching and offers potential optimizations for edge infrastructures.

Clinical Full-Time Equivalent in PICUs: Survey of the U.S. Pediatric Critical Care Chiefs Network, 2020-2022.

To inform workforce planning for pediatric critical care (PCC) physicians, it is important to understand current staffing models and the spectrum of clinical responsibilities of physicians. Our objective was to describe the expected workload associated with a clinical full-time equivalent (cFTE) in PICUs across the U.S. Pediatric Critical Care Chiefs Network (PC3N). Cross-sectional survey. PICUs participating in the PC3N. PICU division chiefs or designees participating in the PC3N from 2020 to 2022. None. A series of three surveys were used to capture unit characteristics and clinical responsibilities for an estimated 1.0 cFTE intensivist. Out of a total of 156 PICUs in the PC3N, the response rate was 46 (30%) to all three distributed surveys. Respondents used one of four models to describe the construction of a cFTE-total clinical hours, total clinical shifts, total weeks of service, or % full-time equivalent. Results were stratified by unit size. The model used for construction of a cFTE did not vary significantly by the total number of faculty nor the total number of beds. The median (interquartile range) of clinical responsibilities annually for a 1.0 cFTE were: total clinical hours 1750 (1483-1858), total clinical shifts 142 (129-177); total weeks of service 13.0 (11.3-16.0); and total night shifts 52 (36-60). When stratified by unit size, larger units had fewer nights or overnight hours, but covered more beds per shift. This survey of the PC3N (2020-2022) provides the most contemporary description of clinical responsibilities associated with a cFTE physician in PCC. A 1.0 cFTE varies depending on unit size. There is no correlation between the model used to construct a cFTE and the associated clinical responsibilities.

MrERF039 transcription factor plays an active role in the cold response of Medicago ruthenica as a sugar molecular switch.

Cold stress severely restricts plant development, causing significant agricultural losses. We found a critical transcription factor network in Medicago ruthenica was involved in plant adaptation to low-temperature. APETALA2/ethylene responsive factor (AP2/ERF) transcription factor MrERF039 was transcriptionally induced by cold stress in M. ruthenica. Overexpression of MrERF039 significantly increased the glucose and maltose content, thereby improving the tolerance of M. ruthenica. MrERF039 could bind to the DRE cis-acting element in the MrCAS15A promoter. Additionally, the methyl group of the 14th amino acid in MrERF039 was required for binding. Transcriptome analysis showed that MrERF039 acted as a sugar molecular switch, regulating numerous sugar transporters and sugar metabolism-related genes. In addition, we found that MrERF039 could directly regulate β-amylase gene, UDP glycosyltransferase gene, and C2H2 zinc finger protein gene expression. In conclusion, these findings suggest that high expression of MrERF039 can significantly improve the cold tolerance of M. ruthenica root tissues during cold acclimation. Our results provide a new theoretical basis and candidate genes for breeding new legume forage varieties with high resistance.

A new Q‐function structure for model‐free adaptive optimal tracking control with asymmetric constrained inputs

SummaryThis article aims to design a model‐free adaptive tracking controller for discrete‐time nonlinear systems with unknown dynamics and asymmetric control constraints. First, a new Q‐function structure is designed by introducing the control input into the tracking error of the next moment, in order to eliminate the final tracking error, avoid the steady control, and ignore the discount factor. Second, via system transformation, a general performance index is developed to overcome the challenge caused by asymmetric constraints of implicit control inputs. By this operation, the constrained tracking problem is converted to an unconstrained optimal tracking problem without the traditional nonquadratic performance function that is only applicable to explicit control inputs. Then, a value‐iteration‐based Q‐learning (VIQL) algorithm is derived to seek the optimal Q‐function and the optimal control policy by using offline data rather than the mathematical model. Next, the convergence, monotonicity, and stability properties of VIQL are investigated to demonstrate that the iterative Q‐function sequence can converge to the optimal Q‐function under ideal conditions. To realize the VIQL algorithm, the critic neural network is employed to approximate the Q‐function. Finally, simulation results and comparative experiments are conducted to demonstrate the validity and effectiveness of the present VIQL scheme.

Research Center and Olympics as a strategy for science education in the centers Educates More of Maranhão

This article addresses the implementation of the Research and Olympics Center (NPO) in the Educa Mais Centers in the State of Maranhão, part of the Full-Time Public Education Networks. The main objective was to analyze the results obtained by students and teachers after the introduction of the Center, focusing on participation and awards in knowledge olympiads and scientific events. The data collected revealed a significant increase in the engagement of Educa Mais Centers in challenging academic activities after the implementation of the NPO. The significant participation of students and teachers in Olympiads and scientific events indicated a clear interest in deepening knowledge and expanding educational horizons. Also noteworthy is the considerable increase in the number of awards, indicating not only recognition of merit, but also encouraging participation in other activities, with different methodologies. This positive scenario suggests that the Research and Olympics Center not only introduces an innovative approach to teaching, but also promotes a school culture that values academic excellence. The focus on research and participation in Olympiads not only challenges students intellectually, but also develops fundamental skills and competencies such as critical thinking, problem solving, networking and teamwork. When considering the longevity and effectiveness of this implementation, constant monitoring of results is essential, identifying areas for improvement and necessary adjustments. The dissemination of these positive results not only celebrates the success of the initiative, but also serves as an inspiring model for other educational institutions seeking to improve educational standards and encourage student participation in enriching activities beyond the conventional curriculum. This study highlights the importance of innovative pedagogical strategies to promote the academic and scientific development of students, contributing to the formation of a more capable generation to face future challenges.

Min–max adaptive dynamic programming for zero-sum differential games

In this paper, a min–max adaptive dynamic programming approach is studied for a zero-sum differential game problem with unknown nonlinear dynamics. The unknown nonlinear dynamics is learned through the online Recurrent Neural Network (RNN). The error between the learned RNN model and the dynamics of the system is proved to converge to zero. The policy iteration algorithm is utilised to solve the Hamilton–Jacobi–Isaacs (HJI) equation associated with the differential game. Furthermore, the value function in the HJI equation is approximated through a critical Neural Network with its weights and activation functions updated online iteratively. The Uniform Ultimate Bounded stability of the closed-loop system is proved based on the Lyapunov theory. Finally, the effectiveness of the proposed solution method is demonstrated by applying it to three examples.

Hierarchical Sliding-Mode Surface-Based Adaptive Actor-Critic Optimal Control for Switched Nonlinear Systems With Unknown Perturbation.

This article studies the hierarchical sliding-mode surface (HSMS)-based adaptive optimal control problem for a class of switched continuous-time (CT) nonlinear systems with unknown perturbation under an actor-critic (AC) neural networks (NNs) architecture. First, a novel perturbation observer with a nested parameter adaptive law is designed to estimate the unknown perturbation. Then, by constructing an especial cost function related to HSMS, the original control issue is further converted into the problem of finding a series of optimal control policies. The solution to the HJB equation is identified by the HSMS-based AC NNs, where the actor and critic updating laws are developed to implement the reinforcement learning (RL) strategy simultaneously. The critic update law is designed via the gradient descent approach and the principle of standardization, such that the persistence of excitation (PE) condition is no longer needed. Based on the Lyapunov stability theory, all the signals of the closed-loop switched nonlinear systems are strictly proved to be bounded in the sense of uniformly ultimate boundedness (UUB). Finally, the simulation results are presented to verify the validity of the proposed adaptive optimal control scheme.

Modes of Death in Patients with Cardiogenic Shock in the Cardiac Intensive Care Unit: A Report from the Critical Care Cardiology Trials Network

BackgroundThere are limited data on how patients with cardiogenic shock (CS) die. MethodsThe Critical Care Cardiology Trials Network is a research network of cardiac intensive care units coordinated by the Thrombolysis In Myocardial Infarction (TIMI) Study Group (Boston, MA). Using standardized definitions, site investigators classified direct modes of in-hospital death for CS admissions (October 2021 to September 2022). Mutually exclusive categories included 4 modes of cardiovascular death and 4 modes of noncardiovascular death. Subgroups defined by CS type, preceding cardiac arrest (CA), use of temporary mechanical circulatory support (tMCS), and transition to comfort measures were evaluated. ResultsAmong 1068 CS cases, 337 (31.6%) died during the index hospitalization. Overall, the mode of death was cardiovascular in 82.2%. Persistent CS was the dominant specific mode of death (66.5%), followed by arrhythmia (12.8%), anoxic brain injury (6.2%), and respiratory failure (4.5%). Patients with preceding CA were more likely to die from anoxic brain injury (17.1% vs 0.9%; P < .001) or arrhythmia (21.6% vs 8.4%; P < .001). Patients managed with tMCS were more likely to die from persistent shock (P < .01), both cardiogenic (73.5% vs 62.0%) and noncardiogenic (6.1% vs 2.9%). ConclusionsMost deaths in CS are related to direct cardiovascular causes, particularly persistent CS. However, there is important heterogeneity across subgroups defined by preceding CA and the use of tMCS.

Abstract TMP11: Cerebral Visual Impairment in Pediatric Stroke Patients Assessed by Virtual Reality Testing

Introduction: Pediatric stroke is one of the leading causes of disability in children, many of whom incur brain injury affecting critical visual anatomical networks. Dedicated ophthalmologic examinations that detect cerebral visual impairment (CVI) - the leading cause of congenital visual impairment in developed countries - are not standard of care and can delay diagnosis and interventions. There is limited CVI data in pediatric stroke patients, and in this work we present data defining visual processing deficits in pediatric stroke patients. Hypothesis: We hypothesize that pediatric stroke patients will (1) demonstrate greater visual-spatial processing impairments than controls and (2) demonstrate increasing performance impairment as a function of increasing visual task demands. Methods: Using a virtual reality (VR) based visual search task system, we characterized significant deficits in higher-order visual processing of pediatric stroke patients (n = 17) as compared to healthy controls (n = 13). Stroke patients suffered from a perinatal (n = 8) or pediatric (n = 9) stroke - confirmed with brain MRI - affecting various vascular territories of the anterior and posterior circulations. The task required searching for on-screen targets with increasing visual complexity. Metrics such as reaction time, gaze error, shifted ellipse area, success rate, and pursuit saccade ratio were analyzed using unpaired group-level t-tests and considered significant if p &lt; 0.05. Results: Stroke patients took longer to identify the target (p = 0.001), were less accurate in finding it (p = 0.007), and exhibited greater impairment in performance with an increasing number of distractors compared to healthy children. Groups had comparable task compliance, as indexed by an off-screen count metric. Conclusion: This original work demonstrates that pediatric stroke patients, particularly those with strokes affecting the middle cerebral artery and posterior circulation vascular territories, exhibit impaired higher-order visual processing. This highlights the urgent need for standardized clinical testing to provide quantitative performance measures of cortical visual processing. This will allow timely detection and implementation of interventional therapies.

Zero-Sum Differential Game-Based Fault-Tolerant Control for a Class of Affine Nonlinear Systems.

This article proposes a zero-sum differential game-based control scheme for a class of affine nonlinear control systems with actuator faults, including loss of control, loss of effectiveness, and bias faults. In the control design, the bias faults and the control signal are chosen as the two opposite sides. The Nash equilibrium is achieved while the optimal control signal and the upper bound of bias faults can be derived from the Hamilton-Jacobi-Isaacs (HJI) equation. An adaptive dynamic programming (ADP) method is employed to estimate the weight of the critic network. The developed zero-sum differential game-based control scheme can ensure the system stability and optimal performance. Two simulation results on a numerical system and a rigid spacecraft model illustrate the effectiveness of the proposed control scheme.

A Critical Network Governance Issues in Social Sciences Discipline: A Literature Review

Some scholars have underlined network governance issues in theoretical, conceptual, operational, and practical terms. However, it is still requested to core map this issue, as past scholars have lacked attention. Therefore, this paper explores network governance’s pivotal issues and proposes a theoretical model. A systematic literature review was conducted using Scopus database sources, and NVivo 12 Plus software was used as an assistive qualitative data analysis tool. This study revealed several viewpoints. First, network governance has received much attention over the last decade. Second, some essential core key terms are linked to network governance issues, such as social, public, government, policy, actor, political, institutional, and organizational. In addition, we proposed a theoretical model that can be used as a recommendation for future studies.

An improved fuzzy inference strategy using reinforcement learning for trajectory-tracking of a mobile robot under a varying slip ratio

AbstractIn this study, a fuzzy reinforcement learning control (FRLC) is proposed to achieve trajectory tracking of a differential drive mobile robot (DDMR). The proposed FRLC approach designs fuzzy membership functions to fuzzify the relative position and heading between the current position and a prescribed trajectory. Instead of fuzzy inference rules, the relationship between the fuzzy inputs and actuator voltage outputs is built using a reinforcement learning (RL) agent. Herein, the deep deterministic policy gradient (DDPG) methodology consisted of actor and critic neural networks is employed in the RL agent. Simulations are conducted with considering varying slip ratio disturbances, different initial positions, and two different trajectories in the testing environment. In the meantime, a comparison with the classical DDPG model is presented. The results show that the proposed FRLC is capable of successfully tracking different trajectories under varying slip ratio disturbances as well as having performance superiority to the classical DDPG model. Moreover, experimental results validate that the proposed FRLC is also applicable to real mobile robots.