Motion Planning for Autonomous Vehicle Based on Radial Basis Function Neural Network in Unstructured Environment

Abstract

The autonomous vehicle is an automated system equipped with features like environment perception, decision-making, motion planning, and control and execution technology. Navigating in an unstructured and complex environment is a huge challenge for autonomous vehicles, due to the irregular shape of road, the requirement of real-time planning, and the nonholonomic constraints of vehicle. This paper presents a motion planning method, based on the Radial Basis Function (RBF) neural network, to guide the autonomous vehicle in unstructured environments. The proposed algorithm extracts the drivable region from the perception grid map based on the global path, which is available in the road network. The sample points are randomly selected in the drivable region, and a gradient descent method is used to train the RBF network. The parameters of the motion-planning algorithm are verified through the simulation and experiment. It is observed that the proposed approach produces a flexible, smooth, and safe path that can fit any road shape. The method is implemented on autonomous vehicle and verified against many outdoor scenes; furthermore, a comparison of proposed method with the existing well-known Rapidly-exploring Random Tree (RRT) method is presented. The experimental results show that the proposed method is highly effective in planning the vehicle path and offers better motion quality.

Neuromorphic Hardware Accelerated Lane Detection System

“…Inside each neuron, there is a local memory, which stores a learned lane pattern in advance and calculates the distance at every each time a component of the input vector is received. By using this distance, the neuron outputs the result of classification based on k-nearest neighbor (KNN) [4] or radial basis function (RBF) [5], [6]. This series of proCopyright c 2017 The Institute of Electronics, Information and Communication Engineers Table 1 Comparison between Von Neumann architecture and neuromorphic architecture Fig.…”

An Intelligent Approach to Educational Data: Performance Comparison of the Multilayer Perceptron and the Radial Basis Function Artificial Neural Networks

“…Basically, the RBFANN was offered as an alternative to the MLPANN for analyzing complex models (Luo & Unbehauen, 1999) since it was shown that the RBFANN can be implemented with increased input dimensions (Wilamowski & Jaeger, 1996). The RBFANN includes two additional advantages: its training process is faster than the conventional back propagation neural network; and it more robust to the complex problems associated with active (nonstationary) inputs (Chen, Zhao, Liang, & Mei, 2014).…”

“…Furthermore, the base function depends on the smaller value of d to obtain a smaller width in the RBFANN (Chen et al, 2014). …”

“…These vectors are matched with the radial basis functions in each hidden node. In addition, vector y, which is a linear combination of the final output, is yielded (Chen et al, 2014) after which the y output can be obtained as follows: , where w i denotes connection weights between the hidden and output layers and w 0 is the bias. The notations x 1 , x 2 ,......x n represent the number of input nodes in the output layer, whereas Ø i (x) is the radial function in the hidden layer and M is the total number of nodes in this layer.…”

“…The notations x 1 , x 2 ,......x n represent the number of input nodes in the output layer, whereas Ø i (x) is the radial function in the hidden layer and M is the total number of nodes in this layer. Basically, a radial basis function is a multi-dimensional function that describes the distance between a given input vector and a pre-defined center vector (Chen et al, 2014).…”