Mobile robot GPS/DR integrated navigation positioning technique research

Dead reckoning accuracy determines the operation quality of pipeline mobile robot to a certain extent. With the progress of the industry, the research on the dead reckoning system of pipeline robot is more and more in‐depth, but how to improve dead reckoning accuracy is still a difficult problem for pipeline mobile robot. In this paper, to overcome the sensor defects of the traditional dead reckoning system, an error backpropagation neural network (BPNN) is introduced to compensate the sensor error caused by static drift and poor dynamic response. To reduce the interference of noise information, the covariance matching technique based on the Adaptive Neuro‐Fuzzy Inference System is used to estimate the measurement noise information. And then the two are combined in the algorithm based on the extended Kalman Filter (EKF) framework for data fusion. An EKF algorithm based on adaptive neural fuzzy combined with BPNN is proposed. Finally, a complete three‐dimensional dead reckoning system is established by coordinate transformation. The motion experiment of the quadruped wall‐climbing robot in the pipeline proves that the proximity of the calculated trajectory is not more than 6.00%, and the lowest is 1.07%, which effectively improves the reliability, robustness, and positioning accuracy of the dead reckoning.

It is well known that GPS can be used for localization and navigation in outside environment. Standard DGPS can be used to get an accuracy of under one meter. The precision improves but the cost is very high. The cheep single frequency GPS receiver is used in this paper to provide the positioning information. This paper proposes a new navigation data fusion method using neural network. DR (dead reckoning) navigation system can provide precise short term navigation information. But the error can accumulate over time without limitation. A BP neural network is employed to predict next sampling time GPS output and a new Kalman filter based data fusion method is proposed to do the navigation data fusion with GPS/DR system. Simulation is conducted to validate the proposed fusion method. The result shows the potential of this fusion method for outside used mobile robot navigation.

A fault-tolerant dead reckoning system is presented for a modular vehicle, which consists of one chassis unit and several wheel units. The units locally estimate the vehicle position based on their own internal sensors. The local estimates are exchanged among the units via an inter-communication system, and they are fused in a decentralized manner. The units can then determine the vehicle position accurately. The decentralized dead reckoning algorithm is formulated based on the information filter and the covariance Intersection method. For enhancing the reliability of the dead reckoning a multi-model based fault detection and diagnosis (FDD) of the internal sensors is incorporated into the dead reckoning system. The units diagnose their sensors with the FDD system, and they apply only the normal sensors for the vehicle localization. In this paper two fault modes (hard fault and noise fault modes) of the sensors are handled; on the hard fault the sensor output is stuck at a constant value. On the noise fault it is disturbed by a large noise. The FDD algorithm is based on the variable structure interacting multiple-model estimator. The fault-tolerant dead reckoning algorithm was implemented on our indoor test-vehicle, which consists of one chassis unit and four wheel units. Experimental results show that our dead reckoning provided better localization accuracy than the conventional one (i.e., the dead reckoning without sensor FDD system) did even though the sensors partially failed.

Abstract：On the basis of the basic principles of weighted fusion, Kalman filtering and BP neural networks, the basic principles of information fusion methods used in integrated navigation systems are expounded. Through the analysis of the basic principles, the association of information fusion methods commonly used in integrated navigation systems and information failure modes is obtained: the information fault mode of weighted fusion method The model is closely related to the specific weight allocation method, which depends on the fault mode of the sensor or sub-system in which the weight is dominant; the information fault mode of the Kalman filtering information fusion method is a continuous mutation fault corresponding to the nonlinear time interval of the system; the information fault mode of the BP neural network method is gradual with time. The information failure mode of the BP neural network method is a slowly varying fault that gradually accumulates over time. Starting from the complexity associated with the information fusion method and the information failure mode, it is pointed out that in order to systematically express the relationship between the information fusion method and the information failure mode, further research can be carried out.

A decentralized data association is presented for 3D dead reckoning in a modular omnidirectional vehicle. Wheel units assembled into the vehicle partially estimates the vehicle position based on their own onboard sensors. The partial estimates are exchanged among the units and fused in a decentralized manner. Then the units can precisely identify the vehicle position. The dead reckoning algorithm is formulated based on the information filter and the Covariance Intersection to effectively integrate the partial estimates without taking into consideration their correlation. The unscented transformation is also exploited to predict the vehicle position without linearizing the nonlinear model related to the vehicle motion. The proposed method brings highly modularity and extendibility in the dead reckoning system. Experimental results validate that the proposed dead reckoning provides better positioning accuracy than the conventional Kalman filter-based method does.

Generally, the mobile robots are researched and used in many applications. The fundamental method of navigation is localization, which includes aspects such as sensing, mapping, localization, planning and control. In order to we build the navigation system for mobile robot operating in unmodified environments, we must know the mobile robot’s position in reference frame. Typically reference frame is a Cartesian or we can specify the landmark as reference position. The dead reckoning based on odometry is widely to identify and classify landmarks. Dead reckoning systems estimate the position of the mobile robot with respect to an initial point by measuring linear and angular velocities or accelerations. There are also hybrid models, which have sensor systems of both dead reckoning and landmark recognition. Dead reckoning is the estimate of current position based on a previously determined position or estimated speeds from a known past position over time. This can also determine the future position by projecting an ordered course and speed of advance from a known present position. A disadvantage of dead reckoning is that new positions are calculated solely from previous positions. The errors of the process are cumulative, so the error in the position fix grows with time (Hans-Joachim von der Hardt & Didier Wolf Rene Husson, 1996; Hye Ri Park et al., 2009). Odometry is widely used in mobile robots to estimate their position relative to reference frame. This method estimates the position over time with errors accumulation due to the integration of velocity measurements (Cristina Tar in Sauer and Hermann Brugger, 2001; Houshangi, N. and Azizi, F., 2006; Agostino Martinelli, 2002). The navigation system for mobile robot based on odometry has typically problem with inaccuracy. Because of the odometry problems navigation systems for mobile robot application cannot work solely on dead reckoning. This is the reason that the navigation system needs the addition of sensory information. Additional sensors can be defined as information provided to assist the accomplishment of a task by a system. Practically, sensor fusion refers to the integration process where the different sources of sensory information are combined into one representational format. Typically in navigation systems sensory information can be fused from complementary sensors, redundant sensors or even from a single sensor over a period of time. The advantage of sensory fusion provides uncertainty and noise reduction, failure toleration and extended flexibility of sensor ability.

A dead reckoning system for a wheeled mobile robot was designed, and the method for robot’s pose estimation in the 3D environments was presented on the basis of its rigid-body kinematic equations. After analyzing the locomotion architecture of mobile robot and the principle of proprioceptive sensors, the kinematics model of mobile robot was built to realize the relative localization. Considering that the research on dead reckoning of mobile robot was confined to the 2 dimensional planes, the locomotion of mobile robot in the 3 coordinate axis direction was thought over in order to estimate its pose on uneven terrain. Because the computing method in a plane is rather mature, the calculation in height direction is emphatically represented as a key issue. With experimental results obtained by simulation program and robot platform, the position of mobile robot can be reliably estimated and the localization precision can be effectively improved, so the effectiveness of this dead reckoning system is demonstrated.

The obstacle avoidance and navigation are important tasks for a mobile robot in applications such as industry, military, exploration and automated vehicles. This paper presents an implementation of mapping and navigation of autonomous mobile robot using Robot Operating System (ROS) and Orbbec Astra camera. In recent days, Lidar is used for the mapping and navigation problem by sensing the objects which are above floor level by using only a single horizontal scanning line and this may result in inaccurate generation of map leads to collision during autonomous navigation of mobile robots. In order to overcome limitation of Lidar, Orbbec Astra (Kinect) camera is used for mapping of indoor environment which can detect objects which are above and below the floor level and navigation of mobile robot from start to target location without colliding with obstacles. Experimental results show the map of the indoor environment generated using Orbbec Astra camera is matched with real indoor environment, and generated map is tested experimentally to navigate the mobile robot from start to target location without collision with obstacles.KeywordsMobile robotsMappingNavigationOrbbec Astra cameraROS

In this paper, position and orientation estimation with high accuracy based on GPS and encoders for a four-wheel-steering vehicle (4WS) mobile robot is addressed. An architecture of position and orientation estimation is proposed, which consists of two Extended Kalman Filters and a processing unit of Runga-Kutta-based dead reckoning. The first EKF fuses data from six encoders to estimate the velocity of vehicle and sideslip angle. The second EKF is applied to the estimation of position and orientation based on the measurement from precise GPS data and output from first EKF. To obtain better accuracy of estimation, an arbitrator is designed to switch between EKF2 and dead reckoning. The results and analysis of experiments are presented to show the effectiveness of the proposed approach.

BP neural network is a multilayer feedforward network for training according to the error back-propagation algorithm , its main advantage is the strong non-linear mapping ability , but the training of BP neural network is easy to fall into local minima, and slow convergence speed. In this paper, through the Levenberg-Marquardt algorithm to improve the BP neural network , using LM algorithm terative results dynamically adjust the convergence direction of neural network,so that each iteration error are decreased, and the convergence speed is fast.Through example analysis, show that the improved BP neural network utility and effectiveness in enterprise performance level evaluation.

Reinforcement learning (RL) is suitable for navigation of a mobile robot. We overcame some difficulties of RL which are large computational cost and determination of parameter values for RL with the help of a genetic algorithm (GA) and method of parameter prediction based on results of GA and complexity measure. As a result of these proposals, we succeeded in navigating the real robot practically. In our previous studies, we just one kind of mobile robot, which has three wheels. Our RL method can decrease the computational cost for learning of navigation and development of mobile robots, provided the skill obtained by RL for one mobile robot can be transferred to other mobile robots.To verify the generalization capability of RL in navigation of a mobile robot, the present paper proposes to transfer the skill obtained by RL to a different kind of a mobile robot. We carried out the experiment and we succeeded in transferring the skill obtained by RL to a different mobile robot.

Localization of a vehicle is a key component for driving assistance or autonomous navigation. In this work, we propose a visual positioning system (VPS) for vehicle or mobile robot navigation. Different from general landmark-based or model-based approaches, which rely on some predefined known landmarks or a priori information about the environment, no assumptions on the prior knowledge of the scene are made. A stereo-based vision system is built for both extracting feature correspondences and recovering 3-D information of the scene from image sequences. Relative positions of the camera motion are then estimated by registering the 3-D feature points from two consecutive image frames. Localization of the vehicle is finally given by the reference to its initial position. I. INTRODUCTION Localization and tracking of a vehicle are major compo- nents for providing positions, directions and travel informa- tion to the driver. They also serve as key technologies for building autonomous navigation systems. Many approaches have been proposed for locating a mobile robot or a ve- hicle based on various techniques. The most commonly used methods include dead-reckoning techniques, navigation using active beacons, landmark-based navigation, map-based navigation, global positioning system (GPS), and vision- based positioning. Dead-reckoning is a procedure for determining the present location of a vehicle by advancing some previous position through known path and velocity information over a given period of time. Since an odometer and optical sensors for wheel direction detection are easily installed on a vehicle, dead-reckoning is usually a less expensive method for vehicle localization. However, the integration of incremental motion information over time will lead to accumulation of errors. In the case of long distance travel, accumulation of orientation error will make position errors diverge as the increase of driving time. Active beacons such as laser, sonar or radio can be used as media for vehicle or mobile robot navigation. These approaches use triangulation to measure the distance between a number of beacons and the mobile platform and then deter- mine the current location. The problems associated with this technique are the inaccuracy of the distance measurement caused by time-delay of the signals, and the installation and maintenance cost of a large number of beacons required for an area. Satellite based differential global positioning system (DGPS) is probably the most advanced and accurate method for identifying the position and orientation of an object. However, DGPS does not work well if the satellite signals are blocked. This situation commonly happens in the indoor environments, or urban areas with tall buildings, etc. In this work, we propose a visual positioning system (VPS) for vehicle or mobile robot navigation. Different from general landmark-based or model-based approaches which rely on some predefined known landmarks or ap rioriinfor- mation about the environment (e.g., 3-D models of buildings, objects, etc.), no assumptions on the prior knowledge of the scene are made. A stereo-based vision system is built for both extracting feature correspondences and recovering 3- D information of the scene from image sequences. A robust feature tracking method based on simultaneously considering the inter-frame images from the same camera and the stereo image pair from different cameras at a fixed time instant is developed. The relative position of the camera motion is then estimated by registering the 3-D feature points from two consecutive image frames. Our system includes the following basic modules:

In order to discover the fault with roller bearing in time, a new fault diagnosis method based on Empirical mode decomposition (EMD) and BP neural network is put forward in the paper. First, we get the fault signal through experiments. Then we use EMD to decompose the vibration signal into a series of single signals. We can extract main fault information from the single signals. The kurtosis coefficient of the single signals forms a feature vector which is used as the input data of the BP neural network. The trained BP neural network can be used for fault identification. Through analyzing, BP neural network can distinguish the fault into normal state, inner race fault, outer race fault. The results show that this method can gain very stable classification performance and good computational efficiency.

A damage detection method using BP neural network based on a novel damage index, the correlation characteristic of the acceleration response, is proposed, and is evaluated through the FEM simulation and experiment verification. On the basis of achievements in existence, the feasibility of using the correlation characteristic as damage index is validated theoretically. The damage detection for a simple-supported beam using the proposed method was FEM simulated. The results showed that the trained BP neural network can correctly detect the location and extent of damages in both single damage case and multi-damage case. A model test of a reinforced concrete simple-supported beam was performed to verify the validity and efficiency of the damage detection method. From the results of the model test, it is shown that the trained BP neural network can correctly locate the damage mostly detect the extent of damage. A conclusion is given that the novel damage detection method using the correlation characteristic of the acceleration response as damage index is feasible and efficient.

The development of spatial location coding: Place learning and dead reckoning in the second and third years