Wanming Chen

SP-NN: A novel neural network approach for path planning

In this paper, a neural network approach named shortest path neural networks (SP-NN) is proposed for real-time on-line path planning. Based on grid-based map and mapping this kind of map to neural networks, this proposed method is capable of generating the globally shortest path from the target position to the start position without collision with any obstacles. The dynamics of each neuron is distinctive to other previously presented methods by other researchers and ensures that the generated path is shortest without collision and that the state of neurons varied continuously. Extensive simulations show the efficiency of the presented method.

Environment-Map-free Robot Navigation Based on Wireless Sensor Networks

Its a new and good method to apply wireless sensor networks on robot navigation. We proposed an environment map free navigation for robot based on wireless sensor networks. In this system, the robot did not need to obtain the environment map and can get online navigation. The wireless sensor nodes collected environment and position information and made distributed information fusing to give a path for the robot. The robot then communicated with the wireless sensor network and went through following networks order. To combine the shorter path and the safer path, we proposed a navigation cost method for robots navigation. It makes know by analysis that the environment map free navigation algorithm for robot in wireless sensor networks can achieve an online navigation and get highly precise navigation.

Particle filtering for range-based localization in Wireless Sensor Networks

The paper proposed a range-based particle filtering localization method for static WSN (Wireless Sensor Network). In the method, every blind node in WSN localizes itself through local distance measurements. So the method is parallel. With the development of hardware, RSSI (Received Signal Strength Indicator) measurements were already hardwired in some chips, and then almost no extra hardware was required when the measurement stratagem was adopted. Firstly, the state-space function of localization in a static WSN was proposed in the paper. And the measurement noise model was particularly studied. Because of the non-linear and non-Gaussian model, particle filter was adopted as a suitable localization algorithm for the application. Static nodes bring some new problems to us. So important steps in a particle filter were specifically illustrated in the paper, including: initialization, prediction, sequential importance sampling and the resampling. Software simulations were done in the paper to demonstrate the effect of the proposed method. A blind node was selected as a sample to illustrate the convergence of the particle filter. And some problems emerged in the simulation were also analyzed. After that, it showed a result about localization in a 100 nodes network by the proposed method.

Particle filtering for WSN aided SLAM

The paper proposed a particle filter based and Wireless Sensor Networks (WSN) aided Simultaneous Localization and Mapping (SLAM) strategy. The proposed method aims at solving two troublesome problems in the traditional particle filter based SLAM algorithms. The first problem is high dimension of question space; and the second one is multi-date association. Firstly, the paper analysed the model of the WSN aided SLAM problem. Then noises in the model were analysed. According to analyses, a particle filtering algorithm was developed as the kernel algorithm for data fusion. Detailed procedures of the particle filter were introduced. Besides, all key steps, including initialization, prediction, sequential importance sampling and also resampling, were especially specified. Software simulations were done to analyse and prove the validity and the efficiency of the proposed method. The simulation results supported that the proposed method can diminish dimensions of the SLAM problem and resolve the multi-data association problems. Moreover, through adopting PF algorithm and involving WSN, following advantages were also acquired: firstly, the method could locate blind nodes in WSN with high resolution while no anchor node was available; secondly, it could improve precision of localization and mapping for mobile robots and, especially, it could inhibit error accumulation of dead reckoning without requirement of close-loop.

A Localization Algorithm nin Wireless Sensor Networks Using a Mobile Beacon Node

Wireless sensor networks is one of important information acquisition methods and also has been a promising technique for many applications in recent years. Node localization is one of challenging and fundamental topics in distributed WSNs research. The goal of this paper is to propose an efficient range-free localization algorithm in WSN. We use directional antennas other than omnidirectional antennas. While the mobile beacon node using several directional antennas moves, it periodically broadcasts its packet in K directions. Ordinary sensor nodes are able to estimate their position based on the messages received from only one virtual beacon. We also use discrete imprecision range measurement. It can send messages in different lever of power to obtain the approximate distance between neighbor nodes. Considering WSN is an event-based network, we propose a refinement phase in latter part of this paper. The main advantage of the algorithm is that it is both simple and economical. Our simulation has shown that the given approach is able to provide accurate and robust estimation of sensor location.

A quadtree based neural network approach to real-time path planning

The paper proposed combining traditional quadtrees and framed-quadtrees with the shunting equation based neural network model to improve the efficiency of path planning. The introduction of quadtrees is used for improving the efficiency of the trajectory generation and enlarging the representation capability of maps, especially in sparse environments. And the introduction of framed-quadtree is used for the generation of Euclidean shortest paths. The introduction of quadtrees and framed-quadtrees does not change the structure of the neural network model based on the shunting model; so the stability and the convergence of the neural network were reserved. And the feature that a map can be represented by quadtrees with multi-resolution was betaken to simplify the selection of parameters in the neural network model. Theoretical analyses and Simulation studies of the proposed method were done to demonstrate following conclusions: the Euclidean shortest paths can be generated without collision and without much computational complexity; the improved neural network method does not suffer from undesired local minima; the proposed method can generate shorter collision free trajectory and has bigger representation capabilities of maps. Index Terms - Quadtrees; Framed-quadtrees; Neural

Error analyzing for RSSI-based localization in wireless sensor networks

Analyzing the measuring errors of nodes using received signal strength indications (RSSI) method, such as the parameter error of RSSI theoretical model and the obstacle error between neighbor nodes, this paper proposed a localization algorithm based on analyzing of RSSI errors in wireless sensor networks called ERSSI. Based on the two errors mentioned, it got the online parameter changing method for RSSI theoretical model and reduced the effect of those small obstacles. We show by both theoretical analysis and simulations that this algorithm can get low localization errors when the size of obstacles is small.

Data fusion based on RBF and nonparametric estimation for localization in Wireless Sensor Networks

Localization is one of important functions in Wireless Sensor Networks (WSNs). And Data fusion is commonly regarded as an efficient method that can improve precision of localization. The paper proposed a novel method based on nonparametric estimation techniques and Radial Basis Function (RBF) Neural Networks to decrease the indeterminacy of Time Difference of Arrival (TDOA) and Received Signal Strength Indicator (RSSI) measurements. The different sources of errors for each measurement types cause that the Probability Density Functions (PDFs) of measurements are not completely dependent. So, theoretically, the fusion of the two kinds of measurements could be effective. Nonparametric estimation techniques are introduced to resolve the problem that measurements do not completely submit to a known PDF. And RBF networks can partly eliminate the influence of environments by regulation of weights. The paper theoretically demonstrated that the data fusion based on RBF networks could achieve location estimation with the Minimum Mean Square Error (MMSE). After that, simulation results of the classical linear combination method and the single RBF fusion were compared with the proposed method in the paper to demonstrate that the proposed method can improve precision of localization with a little of increment in complexion and is robust to the variance of environments.

DESIGN AND IMPLEMENTATION OF WIRELESS SENSOR NETWORK FOR ROBOT NAVIGATION

Global Positioning System (GPS) is often used as a main information source for robot localization and navigation. However, it cannot be used in room or in field complex environment because of the bad signal there. To solve this problem, the authors designed and implemented a specific wireless sensor network (WSN) to provide information about the environment and indicate path for robot navigation. A two-stage auto-adaptive route selecting mechanism of the WSN was proposed to facilitate data relaying in localization and the robots navigation. A low complexity localization algorithm was used to localize both the nodes and the robot. An indirect communication method was designed to make the communication between the WSN and the robot possible. In addition, a robot navigation method was proposed based on the wireless sensor network. In this method, the robot did not need to obtain the environment information; the wireless sensor nodes collected and fused the distributed information and then indicate a path for the robot. Experiments showed that the wireless sensor network can result in obstacle avoidance navigation, and can implement the online navigation.

Design of Quadruped Robot Based CPG and Fuzzy Neural Network

The paper proposed a method for a quadruped robot control system based central pattern generator (CPG) and fuzzy neural networks (FNN). The common approach for the control of a quadruped robot includes two methods mainly. One is the CPG that is based the bionics, the other is the dynamic control that is based the model of quadruped robot. The control result of CPG is decided by the gait data of the quadruped and the parameters of the CPG are choosing manually. Modeling a quadruped robot is difficult because it is a high nonlinear system. This paper presents a much simpler method for the control of a quadruped robot. A simple CPG is adopted for a timing oscillator; it generates the motion periodic pattern of legs. The FNN is used to control the joint motion in order to get a desired stable trajectory motion.