Obstacleavoidance is an important ability to autonomous navigation robot [11-13]. In recent years, many collision avoidance algorithms have been studied. Some of these popular methods for path planning include Artificial Potential field Method(APM) [14, 15], Vector Field Histogram method(VFH)  and, fuzzy logic control method , neural network method , genetic algorithm  and so on. APM simplifies all the information of obstacles and the destination to a resultant force. Thus, the local details of obstacles will be lost. VFH selects the robot candidate heading by quantifying the obstacle strength value of each angle of the robot, which cannot make the robot go through narrow channels. The obstacleavoidance method based on fuzzy logic control forms certain rules according to the prior knowledge. Fuzzy logic control approach demonstrates good robustness property, real-time performance and less dependence on the environment. But there exists the phenomenon of symmetry which cannot be determined. The obstacleavoidance method based on neural network designs controller according to the position of obstacles. But it needs a lot of time to train the network in offline to find a global optimal path. The obstacleavoidance method based on genetic algorithm divides the robot movement environment into some equal parts and searches an optimal path. Path planning using genetic algorithm is a simple and effective method. But genetic algorithm tends to arise the local-trap problem and premature convergence problem [17-19].
Abstract— In this article, we are interested in the reactive behaviours navigation training of a mobile robot in an unknown environment. The method we will suggest ensures navigation in unknown environments with presence off different obstacles shape and consists in bringing the robot in a goal position, avoiding obstacles and releasing it from the tight corners and deadlock obstacles shape. This is difficult to do manually in the case of FIS with a large rule base. In this framework, we use the reinforcement learning algorithm called FuzzyQ-learning, based on temporal difference prediction method. The application was tested in our experimental PIONEER II platform.
The fuzzy inference system for obstacleavoidance developed in this paper is designed for Naorobot. The robot has been tested in Webots virtual environment and will be tested on the real robot in the future. The Fuzzy OA is almost two times faster than the NAOqi OA and the robot is much more stable. Because the fuzzy inference system is a method that relies on trial an error and experience, the obstacleavoidance algorithm is subject to improvement. Future developments will take into account these results and will add other fuzzy inference systems, such as goal seeking fuzzy system, in order to get more autonomy for Naorobot. In addition to these methods a decision making algorithm will be needed in order to fire the appropriate method. This algorithm is intended to be developed using deep learning, machine learning or neutrosophic logic.
1. Inability to plan under uncertainty of dynamic environments: Conventionally, global planners rely on a complete map of the environment in order to calculate the ideal and collision-free path between the starting point and the ending point prior to execution of the robot. The original plans of those conventional algorithms must be revised accordingly if a dynamic environment is encountered (Dijkstra, 1959; Hart et al., 1968). In practise, environment of robots often includes various hazard sources that robots must avoid, for example landmines, fire in rescue duty, and war enemies. Since it is impossible or expensive to acquire their accurate locations, decision- makers know only their action ranges in most cases (Zhang et al., 2013). Mobile robots must be able to evade both static and moving obstacles (Ferguson et al., 2006). Algorithms such as sampling-based methods (Khaksar et al., 2012) are not suitable for online planning when involving moving obstacles, due to the fact that these methods are designed based on a static environment model. These models are time- consuming when applied to a dynamic environment (involving interpolation cycle during each update, see (Huptych & Röck, 2015)). Therefore, classical path planning methods such as Visibility Graph (Lozano-Perez, 1987), Voronoi Diagrams (Leven & Sharir, 1987), Grids (Weigl et al., 1993), Cell Decomposition (Regli, 2007), Artificial Potential Field (Khatib, 1985), Rule Based methods (Fujimura, 1991) and Rules Learning techniques (Ibrahim & Fernandes, 2004) are not practical (Mohanty & Parhi, 2013). Occasionally, these algorithms are optimized to handle a specific problem at the expense of sacrificing the performance of other parameters such as increasing of the computational cost of the algorithm.
Besides that, obstacleavoidance ability is also popular in security system. For an example, a Intelligent Automation Robot (IA Robot) was developed with intelligent security architecture. In order to navigate IA Robot to complete mission with obstacleavoidance system and security system by using IR sensors, and USB Web-Camera installed in IA Robot. The IA Robot used to navigating and patrolling around the surrounding freely, and will avoid the obstacles if it meets the uncertain obstacles including static and dynamic obstacles. 
The obstacle detection is an important topic and in this paper we have considered a mobile robot equipped with infra-red sensors able to measure the distance between the robot and its environment. This paper is organized into five sections. In section 2 the model of the robot and its equipments are presented. In section 3 the mathematical formulation for pathfollowing and obstacleavoidance is described. Finally, section 5 contains the results of the simulation and experimental tests.
For segmenting continuous-tone gray-scale images, we also proposed an extended discrete-time oscillator network with a degradation (posterization) system . In this system, couplings between neighboring oscillators can change every discrete time. The updating of coupling strengths is based on the dynamics of a posterization system that is modified from a discrete-time system . We demonstrated that the extended dynamic image-segmentation system worked well for continuous-tone gray- scale images .
For a mobile robot to identify where it is or how it got there, or to be able to reason about where it has gone, sensors are necessary. For measuring the distance that wheels have traveled along the ground and for measuring inertial changes and external structure in the environment, the sensors can be flexible and mobile. The sensors can be generally divided into two categories: internal state sensors, and external state sensors. The internal state sensors such as accelerometers and gyroscopes, provide the internal information about the robot’s movements. The external state sensors, such as laser, infrared sensors, sonar, and visual sensors, provide the external information about the environment. The data from internal state sensors can be used to estimate the position of the robot in a 2-dimensional space; however, cumulative error is inevitable. The data from external state sensors can be applied for recognizing a place or a situation, or be used to construct a map of the environment. Laser, infrared, and sonar sensors can obtain distant and directional information about an object. Visual sensors can also provide rich information of the environment, but can be very expensive to process. In most cases, because of the available noises, the sensor readings are imprecise and unreliable. Thus, it is inevitable for the mobile robot navigation algorithm to process the sensor data with noises. Given that neural networks have many processing units, each with primarily local connections, they may provide some degree of robustness or fault tolerance for interpretation of the sensor data .
Aiming at the problem that the artificial potential field method is easy to fall into the local minimum and the target unreachable in the process of obstacleavoidance of robot path planning, a fixed path artificial potential field method is proposed. The algorithm first improves the repulsive field function on the basis of the artificial potential field method and then presents a new method of path planning. When the robot moves to a certain range near the obstacle, let it take the obstacle as the center of the robot to the obstacle distance for the radius to do circular motion. After moving the 1/2 circle, let the robot move in the direction of resultant force. In the simulation experiment, the algorithm can effectively avoid the problem of local minimum and improve the efficiency of the algorithm while solving the problem of unreachable target.
I declare that this report entitled “Design and Construction of a Simple Bipedal Robot with ObstacleAvoidance System” is the result of my own research except as cited in the references. The report has not been accepted for any degree and is not concurrently submitted in candidature of any other degree.
In this section, we present experiments results and algorithm illustration of the presented techniques .When the system starts, servo motor rotates the ultrasonic sensor with five corners as shown in Figure 8, and at every angle the robot calculate the distance of nearest obstacle, and then compares the distances to know where the obstacles are, after that the robot getting its direction by using compass sensor and looking for the desired path by calculating the angle of steering, Figure 7 shows the flowchart of the process.
By identifying the above-mentioned model to detect the recognition performance of the rail, and comparing it with the traditional rail detection algorithm, Figure. 5 is the traditional railway extraction result. It can be seen from the figure that there is no extraction and leakage extraction in the railway extraction due to illumination, which makes the risk area division have a large error, and the proposed depth detection based rail detection algorithm can accurately identify the position of the railway. The dangerous area of the planning office, as shown in Figure. 6, can be completed accurately when iterating to 30,000 times, with an accuracy rate of 99.9%. Therefore, the method of this paper can meet the needs of the unmanned subway and ensure the safe operation of the train.
verify the detection of fault and to use replicas to solve the problems. Each complex goal represents a group of agents that work concurrently to achieve this goal. Exception handling technique is used at this level to guarantee system recovery when an agent fails to carry on a specific action. This second part of our approach is based on the idea of adding specific agents to each goal’s group of agent called the goal agent, we use Petri Nets to give a general model the activities of these agents that can detect, evaluate and solve a fault when it occurs. The whole system is controlled by another agent that is called the system’s supervisor.
Pipe robot is a robot capable of moving inside the pipe. The function of the pipe robot is to moni- tor pipe defects. The pipe robot is designed to move steadily in the center position inside the pipe. HCSR-04 distance sensor required as input to give distance value on microcontroller so that robot keep running stable and ballance, for movement of robot using DC motor. This robot is made with the aim to move autonomously following the pipeline in detecting pipe cracks. Programming on this robot using Artificial Neural Network algorithm with Backpropogation method of network structure, consist of 3 input layer, 3 output layer and 20 hidden layer. Conducted experiments on inputs, hidden layers and outputs with varying amounts to obtain robust network structure of ef- ficient and precise movement of robots in detecting pipe cracks. The result of movement in the robot in the application of Artificial Neural Network algorithm with Backpropogation method is able to move well and more stable. In this case, it uses 2 ultrasonic sensors and 2 motor outputs. The average robot speed movement is 10 cm / sec.
Overview of ABC-FuzzyQlearning algorithm is shown in Fig. 3. This algorithm includes three main phases: employed, onlooker and scout bees [8-9]. The algorithm will be finished when the answer is achieved in each of these phases. In ABS-FQ algorithm, N food sources are considered and each food source shows the values of L result rules in a fuzzy system ! ~ > . Q-Value indicates the nectar amount for each food source 7 ) . The more the Q-Value of a food source shows that this food source generates a stronger fuzzy system to control the agent . The fuzzy system is considered as follows:
Abstract: As a fast growing field, robots are greatly used to achieve the desired task more accurately and mitigate the difficulties in odd environments where human face immense difficulties. In this paper an obstacleavoidancerobot has been designed using basic gates. It can detect the obstacle and directs itself with the help of five sensors. When sensor detects an obstacle it gives the pulse high and vice-versa. A differential drive model has been chosen, which has two wheels and a cluster wheel. Left and right motor are used as a physical machine and it will be controlled by logic; K-map has been used to do it. Basic gates help to execute the equation of motors as well as to make robot faster, precise and efficient. To make more comprehensible comparative time delay estimation has been added in this paper.
Over the course of this project, the team evaluated different sensor options before choosing to work with computer vision in order to develop an obstacle detection system. The team was able to implement a DIY stereo depth mapping system capable of mounting on a small UAV. Through camera calibration and bench testing, a stereo imaging algorithm was developed. The bench testing results showed that the depth algorithm was highly precise, but more tuning and bias-correction must be performed in order to increase the accuracy. The Raspberry Pi was identified as a capable on-board processor, and a payload and mount were designed and built to allow the IRIS quadcopter to perform flight tests with the stereo imaging system. The results showed that more specialized hardware is required for implementation in the field. Finally, the team created recommendations on how to move forward with this project in the future.
Ranging based on the reflection principle of ultrasonic wave propagating in the air has been widely used in modern life, such as car reversing radar, robot automatic obstacleavoidance etc. Aiming at the situation that the blinds have no way to know whether there are obstacles or big safety risks in front of them when they are walking, this paper designed obstacleavoidance gloves for the blinds based on ultrasonic sensors. With Arduino Nano single chip microcomputer as main controller, combined with ultrasonic sensor module, bluetooth module and speaker module, the glove realized the function of ob- stacle detection and alarm. The main working principle is by using the ultra- sonic sensors to transmit and receive ultrasonic, the time difference for transmitting and receiving to detect the distance of obstacles ahead. Besides, by means of voice module output audio signals with different frequency ac- cording to the obstacle distance, the blinds can judge the distance between them to the obstacles based on the sound with different frequencies they heard. In this way, they can make responses in advance to avoid the obstacles ahead and the happening of the risk.
The main contribution of this work is to steer the mobile robot while following the line. The mobile robot also avoids colliding with any obstacles on its path. The fuzzy logic controller was designed for the line following mechanism where it takes the delta value (∆) as an input and provides the left and right robot motor speeds based on the defined fuzzy rules. The proximity sensors are responsible for the collision avoidance technique. If one of these sensors reaches the threshold value, obstacle is close to the robot and thus it needs to adjust its speed and direction. Fig. 9 shows the proximity sensors readings for robots A and B. As shown in Fig.9 (a), the front proximity sensors (PS0 & PS7) for Robot A have values higher than the threshold (1000) at times 4s, 43s, and 178s. This means that the robot has detected an obstacle. At time 4s, Robot A detects the first obstacle and at time 178s detects the second obstacle. At time 43s, Robot A has detected the other mobile robot as a dynamic obstacle.