Gazebo Simulator Research Articles

Development of an Algorithm for Determining the Class of an Obstacle in the Process of its Interaction with the Manipulator

The article discusses the problem of determining the nature of an obstacle using a force sensor for terrain mapping. Three classes of objects have been introduced: the first class is "static" obstacles, the second class is "moving" obstacles, the third class is "unstable" obstacles. To conduct experiments, a simulation system was built in the Gazebo modeling environment, including a manipulator with the ability to horizontally move a probe with a force sensor and a set of obstacles corresponding to three classes of obstacles. All experiments were carried out using the ROS framework, the Gazebo simulator and the Movelt manipulation software. Processing of signals from the force sensor showed that the sig­nals for the three classes of obstacles are different, which makes it possible to identify the class based on sensor readings. Based on the analysis of data received from the sensor, rules were formed that make it possible to determine the class of an obstacle during its interaction with the manipulator in real time. To determine the obstacle class, the values of the force sensor signals are considered, smoothed by the moving average method. According to the rules discussed in the article, an algorithm was developed to determine the class of obstacle, which was implemented in the Gazebo modeling environment. Testing of the algorithm showed the stability of its operation under the given experimental conditions. Successful determination of the obstacle class will allow one not only to construct a map of the area, but also to prevent damage to the device when exposed to a "static" object with excessive force. The presented approach can be expanded to narrower classes of obstacles that characterize objects that can be encountered by a mobile device when mapping a forest area.

Agent path planning based on adaptive polymorphic ant colony optimization

In the path planning of intelligent agents, ant colony algorithm is a popular path solving strategy and has been widely used. However, the traditional ant colony algorithm has problems of local optimum and redundant turning points. The adaptive polymorphic ant colony algorithm greatly improves the search and convergence speed through multi-group division and cooperation mechanism, which helps to enhance the global search ability and avoid falling into the local optimal solution. This paper proposes an improved pheromone update strategy and path selection record table construction to further improve the accuracy of path planning. Finally, the path is processed by cubic B-Spline smoothing curve to effectively reduce the turning points and realize the smoothing of the path. After MATLAB and ROS (Robot Operating System)-Gazebo simulation verification, the results show that the algorithm has good feasibility in complex environments. In summary, the improved adaptive polymorphic ant colony algorithm in this paper brings significant optimization and improvement to the global search of intelligent agents.

A Novel Fuzzy Image-Based UAV Landing Using RGBD Data and Visual SLAM

In this work, an innovative perception-guided approach is proposed for landing zone detection and realization of Unmanned Aerial Vehicles (UAVs) operating in unstructured environments ridden with obstacles. To accommodate secure landing, two well-established tools, namely fuzzy systems and visual Simultaneous Localization and Mapping (vSLAM), are implemented into the landing pipeline. Firstly, colored images and point clouds acquired by a visual sensory device are processed to serve as characterizing maps that acquire information about flatness, steepness, inclination, and depth variation. By leveraging these images, a novel fuzzy map infers the areas for risk-free landing on which the UAV can safely land. Subsequently, the vSLAM system is employed to estimate the platform’s pose and an additional set of point clouds. The vSLAM point clouds presented in the corresponding keyframe are projected back onto the image plane on which a threshold fuzzy landing score map is applied. In other words, this binary image serves as a mask for the re-projected vSLAM world points to identify the best subset for landing. Once these image points are identified, their corresponding world points are located, and among them, the center of the cluster with the largest area is chosen as the point to land. Depending on the UAV’s size, four synthesis points are added to the vSLAM point cloud to execute the image-based visual servoing landing using image moment features. The effectiveness of the landing package is assessed through the ROS Gazebo simulation environment, where comparisons are made with a state-of-the-art landing site detection method.

Neural Network Algorithm for Training a Mobile Robot in the Task of Following a Target

The article is devoted to the research, development and implementation of an artificial neural network with reinforcement for controlling a mobile robot with a differential drive to solve the task of following a target. The authors of the study present a detailed description of the architecture of the neural network, its training using a deep deterministic gradient policy algorithm, and integration with the robot control system. A distributed neural network structure was selected for specialized generation of controls, for which the authors chose the angular and linear speed of the robot. A system of rules has been synthesized that takes into account changes in the distance between the robot and the target object and tracking of the heading angle in real time. A mathematical model of a robot with a differential drive is considered, on the basis of which a simulation program is implemented for intermediate training of a neural network, as a result of which the initial weighting coefficients are formed. Using this program, you can get by with less energy and time costs at the initial stage of the study. Experiments to test the effectiveness of the developed neural network were carried out in the Gazebo simulation environment using the ROS2 communication interface. The article describes the process of integrating a neural network with a robot control system in a simulation environment, as well as test results and analysis of the obtained data. The results of experiments are presented for three scenarios in which the initial position of the robot is the same, the remaining parameters are generated according to the rules described by the authors. The effectiveness of using such a solution for trajectory planning has been confirmed. The research contributes to the field of autonomous systems and demonstrates the potential of artificial reinforcement neural networks in robotics.

Dynamic event-triggered neuroadaptive fault-tolerant control of quadrotor UAV with a novel cosine kernel

The fault-tolerant control (FTC) for trajectory tracking of a quadrotor unmanned aerial vehicle (UAV) has attracted researchers. The non-linear model of the UAV, coupled with model uncertainties, external disturbances, and actuator failures, requires the function approximation of the lumped non-linearity for controller design. One of the most efficient ways to approximate non-linearity is using radial basis function neural networks (RBFNNs). To date, RBFNNs have been formulated, trained, and used directly for function approximation, which requires considerable computation to derive the control laws. The proposed control and parameter estimation laws approximate the non-linearity by using RBFNNs indirectly. The proposed laws with virtual parameter estimation do not require the actual formulation of RBFNNs and their weights, thus saving on computational resources. For kernel optimization, the Gaussian kernels with exponential terms in RBFNNs are replaced by cosine kernels with algebraic terms, which shows faster convergence as per simulation results. To save on communication bandwidth, static event-triggering communication mechanisms (SECM) and dynamic event-triggering mechanisms (DECM) have been proposed. As DECM works on dynamically changing variables, it saves more communication bandwidth, as tested in simulation. Lyapunov stability analysis proves that errors are uniformly ultimately bounded (UUB). The performance of the proposed algorithm has been tested through numerical simulations, which show superior performance when compared with similar studies. The proposed algorithm has been validated in a real-time Gazebo simulator.

Enhanced Visual SLAM for Collision-Free Driving with Lightweight Autonomous Cars.

The paper presents a vision-based obstacle avoidance strategy for lightweight self-driving cars that can be run on a CPU-only device using a single RGB-D camera. The method consists of two steps: visual perception and path planning. The visual perception part uses ORBSLAM3 enhanced with optical flow to estimate the car's poses and extract rich texture information from the scene. In the path planning phase, the proposed method employs a method combining a control Lyapunov function and control barrier function in the form of a quadratic program (CLF-CBF-QP) together with an obstacle shape reconstruction process (SRP) to plan safe and stable trajectories. To validate the performance and robustness of the proposed method, simulation experiments were conducted with a car in various complex indoor environments using the Gazebo simulation environment. The proposed method can effectively avoid obstacles in the scenes. The proposed algorithm outperforms benchmark algorithms in achieving more stable and shorter trajectories across multiple simulated scenes.

Risk Assessment and Motion Planning for MAVs in Dynamic Uncertain Environments

Risk assessment to quantify the danger associated with a planned trajectory is critical for micro aerial vehicles (MAVs) navigating in dynamic uncertain environments. Existing works usually perform risk assessment by reasoning the occupancy status of the MAV’s surrounding space which only incorporates the position information of the MAV and the obstacles in the environment. In this paper, we further consider the MAV’s motion direction in risk assessment to reflect the fact that the obstacles in front of the MAV pose a higher risk while those behind pose a lower risk. In particular, we rely on a particle-based dynamic map which consists of a large number of particles to represent the local environment. The risk is defined to evaluate the safety level of a subspace in the map during some time interval and assessed by reasoning the occurrence of particles in the subspace. Those particles around the MAV are assigned different weights taking into account their relative positions to the MAV and its motion direction. We then incorporate the proposed risk assessment method into MAV motion planning by minimizing both the path length and the associated risk to achieve safer navigation. We compared our method with several state-of-the-art approaches in PX4+Gazebo simulations and real-world experiments. The results show that our method can achieve a 15% higher collision avoidance rate and a 20% lower flight risk in various environments with static and dynamic obstacles.

A deep residual reinforcement learning algorithm based on Soft Actor-Critic for autonomous navigation

The problem of autonomous navigation has attracted significant attention from robotics research community in the last few decades. In this paper, we address the problem of low data utilization due to the large amount of episode experience value data. A maximum entropy algorithm based on prioritized experience replay (Learning Good Experience based on Soft Actor-Criti, LGE-SAC) is proposed to quickly reproduce past good experience episodes. As the deep reinforcement learning method is susceptible to failure to plan ahead and explore the target position in a long sequence environment, a deep Residual Soft Actor-Critic (RSAC) is proposed to alleviate this problem. The reinforcement learning policy is fused with the Artificial Potential Field method to improve the generalization ability of the proposed algorithm, thus improving robot adaptation in new test environments. In order to validate the effectiveness of the proposed algorithm, we conducted simulation experiments in Gazebo simulator environment and real experiments on a Turtlebot3 robot equipped with LiDAR sensor. Simulation and experiment results show that the proposed algorithm effectively avoids obstacles and succeeds in reaching the goal compared to other obstacle avoidance algorithms. In comparison with the Artificial Potential Field method, the planning success rate of the proposed RSAC algorithm in the test environment is increased by 30%, and at the same time, the number of planning steps is reduced by half, and the generalization ability is improved.

Enhancing Visual Homing in Robotics: A Study on Blockchain Integration and Consensus Algorithms

Creating an immutable repository for vital robot and environmental data, ensuring long-term accessibility, and functioning in the absence of GPS or mapping are crucial for visual homing navigation systems. We focus on the intersection of blockchain and robotics, particularly in visual homing. Our research involves an in-depth analysis of various blockchain consensus mechanisms, highlighting their suitability for visual homing applications. The heart of blockchain functionality lies in its consensus mechanism, which facilitates agreement among network nodes. In our first study part, we conduct a comprehensive comparative analysis of key consensus algorithms, emphasizing visual homing's decentralization, fault tolerance, latency, and throughput requirements. This analysis serves as a valuable reference for researchers and developers, emphasizing the importance of aligning the chosen consensus mechanism with specific blockchain application needs. The second part of our work involves extensive experiments exploring the connection between blockchain and visual homing. We assess prominent consensus mechanisms like Proof of Work (PoW), Proof of Stake (PoS), Delegated Proof of Stake (DPoS), and Proof of Authority (PoA) within a virtual environment in Gazebo, leveraging Wide Area Visual Navigation (WAVN). Our research implementation is grounded in the ROS framework and the Gazebo simulation environment.

Dynamic Bandwidth Allocation for Collaborative Multi-Robot Systems Based on Task Execution Measures

Multi-robot systems (MRSs) is a growing field of research that focuses on the collaboration of multiple robots to achieve a common global objective. Managing these systems poses several challenges, including coordination, task allocation, and communication. Among these challenges, a major area of focus is devising an effective communication scheme that ensures robots’ cooperation and adapts to varying conditions during task execution. In this paper, we develop a novel communication management framework tailored for MRSs, specifically addressing dynamic bandwidth distribution in networked teleoperated robotic systems. The algorithm is combined with semi-autonomous formation control based on the Artificial Potential Fields (APF) algorithm, which allows each individual robot to avoid local obstacles autonomously and tries to maintain a desired formation with its neighbors, while the operator is in charge of high-level control only. Common Dynamic Bandwidth Allocation (DBA) algorithms allocate bandwidth to different units based on network conditions and requirements. On the other hand, our proposed DBA scheme dynamically distributes the available bandwidth on communication streams based on factors related to task execution and system performance. In specific, bandwidth is allocated in a way that adapts to changes occurring in the system’s environment and its internal state, including the effect of the autonomous action taken by the path planner on the MRS and the performance of the controller of each individual robot. By addressing the limitations of existing approaches through shaping the communication behavior of the MRS based on performance measures, our proposed algorithm offers a promising solution for improving the performance and efficiency of MRSs. The proposed scheme is tested through simulations on a group of six unmanned aerial vehicles (UAVs) in the Robot Operating System (ROS)-Gazebo simulation environment. The obtained results show the scheme’s capability for enhancing the robotic system’s performance while significantly reducing bandwidth consumption. Experimental testing on two mobile robots further demonstrates the effectiveness of the proposed scheme.

Design and Flight Simulation Verification of the Dragonfly eVTOL Aircraft

Recently, electric vertical take-off and landing (eVTOL) aircraft have become a top priority for urban air transportation due to their ability to overcome urban ground traffic congestion. In this research, a new type of scaled lift–cruise ‘Dragonfly’ has been designed. The ‘Dragonfly’ combines the characteristics of an octocopter and a fixed-wing aircraft. Compared with the same type of eVTOL aircraft, it has a longer wingspan and a more stable aircraft structure, it can not only take off and land vertically without the need for a runway, but also fly quickly in a straight line and hover in mid-air. In order to ensure the success of the flight test, it was also simulated in this paper. A simulation scenario highly fitting with the flight test environment of eVTOL is designed in the Gazebo simulation platform, and then combined with the PX4 flight control platform, the system SITL of the constructed aircraft simulation model is carried out on the Gazebo platform, Finally, simulation flight test data for accurate analysis are obtained, the accuracy and stability of the control algorithm are fed back, and scientific support for the follow-up ‘Dragonfly’ aircraft hardware-in-the-loop simulation and physical flight test is provided.

Mobile Robot Navigation Based on Noisy N-Step Dueling Double Deep Q-Network and Prioritized Experience Replay

Effective real-time autonomous navigation for mobile robots in static and dynamic environments has become a challenging and active research topic. Although the simultaneous localization and mapping (SLAM) algorithm offers a solution, it often heavily relies on complex global and local maps, resulting in significant computational demands, slower convergence rates, and prolonged training times. In response to these challenges, this paper presents a novel algorithm called PER-n2D3QN, which integrates prioritized experience replay, a noisy network with factorized Gaussian noise, n-step learning, and a dueling structure into a double deep Q-network. This combination enhances the efficiency of experience replay, facilitates exploration, and provides more accurate Q-value estimates, thereby significantly improving the performance of autonomous navigation for mobile robots. To further bolster the stability and robustness, meaningful improvements, such as target “soft” updates and the gradient clipping mechanism, are employed. Additionally, a novel and powerful target-oriented reshaping reward function is designed to expedite learning. The proposed model is validated through extensive experiments using the robot operating system (ROS) and Gazebo simulation environment. Furthermore, to more specifically reflect the complexity of the simulation environment, this paper presents a quantitative analysis of the simulation environment. The experimental results demonstrate that PER-n2D3QN exhibits heightened accuracy, accelerated convergence rates, and enhanced robustness in both static and dynamic scenarios.

Enabling Building Information Model-driven human-robot collaborative construction workflows with closed-loop digital twins

The introduction of assistive construction robots can significantly alleviate physical demands on construction workers while enhancing both the productivity and safety of construction projects. Leveraging a Building Information Model (BIM) offers a natural and promising approach to driving robotic construction workflows. However, because of uncertainties inherent in construction sites, such as discrepancies between the as-designed and as-built components, robots cannot solely rely on a BIM to plan and perform field construction work. Human workers are adept at improvising alternative plans with their creativity and experience and thus can assist robots in overcoming uncertainties and performing construction work successfully. In such scenarios, it is critical to continuously update the BIM as work processes unfold so that it includes as-built information for the ensuing construction and maintenance tasks. This research introduces an interactive closed-loop digital twin framework that integrates a BIM into human-robot collaborative construction workflows. The robot’s functions are primarily driven by the BIM, but it adaptively adjusts its plans based on actual site conditions, while the human co-worker oversees and supervises the process. When necessary, the human co-worker intervenes in the robot’s plan by changing the task sequence or workspace geometry or requesting a new motion plan to help the robot overcome the encountered uncertainties. A drywall installation case study is conducted to verify the proposed workflow. In addition, experiments are carried out to evaluate the system performance using an industrial robotic arm in a research laboratory setting that mimics a construction site and in the Gazebo simulation. Integrating the flexibility of human workers and the autonomy and accuracy afforded by the BIM, the proposed framework offers significant promise of increasing the robustness of construction robots in the performance of field construction work.

Path planning of mobile robot based on improved TD3 algorithm in dynamic environment

This paper proposes an improved TD3 (Twin Delayed Deep Deterministic Policy Gradient) algorithm to address the flaws of low success rate and slow training speed, when using the original TD3 algorithm in mobile robot path planning in dynamic environment. Firstly, prioritized experience replay and transfer learning are introduced to enhance the learning efficiency, where the probability of beneficial experiences being sampled in the experience pool is increased, and the pre-trained model is applied in an obstacle-free environment as the initial model for training in a dynamic environment. Secondly, dynamic delay update strategy is devised and OU noise is added to improve the success rate of path planning, where the probability of missing high-quality value estimate is reduced through changing the delay update interval dynamically, and the correlated exploration of the mobile robot inertial navigation system in the dynamic environment is temporally improved. The algorithm is tested by simulation where the Turtlebot3 robot model as a training object, the ROS melodic operating system and Gazebo simulation software as an experimental environment. Meanwhile, the result shows that the improved TD3 algorithm has a 16.6 % increase in success rate and a 23.5 % reduction in algorithm training time. A generalization experiment was designed finally, and it indicates that superior generation performance has been acquired in mobile robot path planning with continuous action spaces through the improved TD3 algorithm.

AUV Obstacle Avoidance Framework Based on Event-Triggered Reinforcement Learning

Autonomous Underwater Vehicles (AUVs), as a member of the unmanned intelligent ocean vehicle group, can replace human beings to complete dangerous tasks in the ocean. It is of great significance to apply reinforcement learning (RL) to AUVs to realize intelligent control. This paper proposes an AUV obstacle avoidance framework based on event-triggered reinforcement learning. Firstly, an environment perception model is designed to judge the relative position relationship between the AUV and all unknown obstacles and known targets. Secondly, considering that the detection range of AUVs is limited, and the proposed method needs to deal with unknown static obstacles and unknown dynamic obstacles at the same time, two different event-triggered mechanisms are designed. Soft actor–critic (SAC) with a non-policy sampling method is used. Then, improved reinforcement learning and the event-triggered mechanism are combined in this paper. Finally, a simulation experiment of the obstacle avoidance task is carried out on the Gazebo simulation platform. Results show that the proposed method can obtain higher rewards and complete tasks successfully. At the same time, the trajectory and the distance between each obstacle confirm that the AUV can reach the target well while maintaining a safe distance from static and dynamic obstacles.

Formation Control of Multiple Autonomous Mobile Robots Using Turkish Natural Language Processing

People use natural language to express their thoughts and wishes. As robots reside in various human environments, such as homes, offices, and hospitals, the need for human–robot communication is increasing. One of the best ways to achieve this communication is the use of natural languages. Natural language processing (NLP) is the most important approach enabling robots to understand natural languages and improve human–robot interaction. Also, due to this need, the amount of research on NLP has increased considerably in recent years. In this study, commands were given to a multiple-mobile-robot system using the Turkish natural language, and the robots were required to fulfill these orders. Turkish is classified as an agglutinative language. In agglutinative languages, words combine different morphemes, each carrying a specific meaning, to create complex words. Turkish exhibits this characteristic by adding various suffixes to a root or base form to convey grammatical relationships, tense, aspect, mood, and other semantic nuances. Since the Turkish language has an agglutinative structure, it is very difficult to decode its sentence structure in a way that robots can understand. Parsing of a given command, path planning, path tracking, and formation control were carried out. In the path-planning phase, the A* algorithm was used to find the optimal path, and a PID controller was used to follow the generated path with minimum error. A leader–follower approach was used to control multiple robots. A platoon formation was chosen as the multi-robot formation. The proposed method was validated on a known map containing obstacles, demonstrating the system’s ability to navigate the robots to the desired locations while maintaining the specified formation. This study used Turtlebot3 robots within the Gazebo simulation environment, providing a controlled and replicable setting for comprehensive experimentation. The results affirm the feasibility and effectiveness of employing NLP techniques for the formation control of multiple mobile robots, offering a robust and effective method for further research and development on human–robot interaction.

Path planning for robot search task using belief criteria decision-making

PurposeThis study aims to realize efficient, fast and safe robot search task, the belief criteria decision-making approach is proposed to solve the object search task with an uncertain location.Design/methodology/approachThe study formulates the robot search task as a partially observable Markov decision process, uses the semantic information to evaluate the belief state and designs the belief criteria decision-making approach. A cost function considering a trade-off among belief state, path length and movement effort is modelled to select the next best location in path planning.FindingsThe semantic information is successfully modelled and propagated, which can represent the belief of finding object. The belief criteria decision-making (BCDM) approach is evaluated in both Gazebo simulation platform and physical experiments. Compared to greedy, uniform and random methods, the performance index of path length and execution time is superior by BCDM approach.Originality/valueThe prior knowledge of robot working environment, especially semantic information, can be used for path planning to achieve efficient task execution in path length and execution time. The modelling and updating of environment information can lead a promising research topic to realize a more intelligent decision-making method for object search task.

An Improved Trilateral Localization Technique Fusing Extended Kalman Filter for Mobile Construction Robot

Semi-open and chaotic environments of building sites are considered primary challenges for the localization of mobile construction robots. To mitigate environmental limitations, an improved trilateral localization technique based on artificial landmarks fusing the extended Kalman filters (EKFs) is proposed in this paper. The reflective intensity of the onboard laser is employed to identify artificial landmarks arranged in the ongoing construction environment. A trilateral positioning algorithm is then adopted and evaluated based on artificial landmarks. Multi-sensor fusion, combined with the EKF, is included to improve the positioning accuracy and reliability of the robot in complex conditions. We constructed validation scenarios in the Gazebo simulation environment to verify the required localization functionality. Concurrently, we established simulated testing environments in real-world settings, where the practicality of the proposed technique was validated through the fitting of ideal and actual localization trajectories. The effectiveness of the proposed technique was corroborated through comparative experimental results.

Adaptive Intelligent Minimum Parameter Singularity Free Sliding Mode Controller Design for Quadrotor

This paper presents a singularity free fast terminal sliding mode control (SFFTSMC) for position and attitude tracking of a quadrotor with unknown dynamics. The contribution of this work is threefold, namely: devising an SFFTSMC strategy using an auxiliary function to eliminate the singularity in the control law, a cost-saving minimum parameter update scheme using a fully connected recurrent neural network (FCRNN) to handle the unknown dynamics of quadrotor and development of an adaptive control law to compensate for the loss of effectiveness of the actuator and saturation effects, considered explicitly, without the requirement of any fault detection and diagnosis (FDD) unit. The proposed approach offers a direct singularity free control action in the presence of unknown dynamics, actuator faults, and saturation. Overall closed-loop stability is guaranteed in finite-time using Lyapunov stability theory and update laws for minimum parameters of FCRNN and fault compensator are derived using the same. Extensive simulations are presented for the trajectory tracking tasks using the proposed method and compared with the existing piece-wise fast terminal sliding mode controller (PFTSMC) and Radial basis function network (RBFN) based fault compensation approach. The proposed method is also validated in the Gazebo simulator via Pixhawk autopilot to demonstrate the feasibility in real-time. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —This research is motivated by a cost-effective control scheme for quadrotors in cases of unknown/uncertain dynamics, external disturbances, actuator faults, and saturation. Precise quadrotor dynamics are difficult to obtain in real time. Also, there may be parametric uncertainties in quadrotor parameters like mass, inertia, and aerodynamic coefficients. External disturbances like wind gusts are always present in outdoor environments, and it may be the case that quadrotor motors lose their effectiveness due to defective motors or propeller damage while flying. To address these issues, first, it requires a precise control scheme to control the quadrotor while guaranteeing stability to ensure the safety of the quadrotor itself and people in the surrounding environment. Second, the approach should not be computationally heavy, which may sluggish the overall response of the quadrotor. Thus, we propose a control scheme to address these two issues simultaneously using a new fast terminal sliding mode control scheme augmented with FCRNN by updating minimum parameters instead of all the weight parameters of FCRNN. A compensator is introduced to mitigate the effect of unexpected actuator fault and saturation. The proposed approach can be helpful in various quadrotor applications, like pesticide spraying in agriculture, payload transportation, search and rescue, construction, etc., where unknown quadrotor dynamics and unknown payload situations occur.

Gait Learning for Hexapod Robot Facing Rough Terrain Based on Dueling-DQN Algorithm

In the handling of dangerous goods in explosive environments, robots are increasingly being used instead of human operators. Robots designed for operation in explosive environments are mostly equipped with tracked structures, which, due to their limited terrain adaptability, struggle to movement rugged landscapes. Hexapod robots, with their excellent maneuverability and adaptability, possess advantages in completing hazardous material handling tasks in such rugged terrains. One current challenge lies in enabling hexapod robots to autonomously adjust their gaits to cope with rugged terrain. This paper proposes a gait learning method based on the Dueling Deep Q-Network (Dueling-DQN) algorithm to address the gait adjustment problem of hexapod robots in sloped, terraced, and rugged terrain. The method combines lidar data and reinforcement learning to extract features from the lidar data to determine terrain types and foot coordinates. Finally, the Dueling-DQN algorithm and redundant phase strategy are employed to facilitate the motion of hexapod robots in these three types of rugged terrain. Simulation and prototype experiments are conducted to evaluate the Dueling-DQN algorithm's performance in terms of rewards and stability margins for the three types of terrain. During algorithm training on sloped, terraced, and rugged terrain, stable rewards and positive stability margins are achieved after approximately 490 iterations. The effectiveness and feasibility of the proposed method are further validated through Gazebo simulation and prototype experiments. In the context of movement in rugged terrain within explosive environments, the gait learning method based on the Dueling-DQN algorithm offers valuable insights into the control of hexapod robots.