论文部分内容阅读
提出一种神经网络和粒子群算法相结合的移动机器人路径规划方法。采用小波网络和RBF网络相结合的四层神经网络结构,克服了传统神经网络方法进行路径规划时对每个障碍均设计一些特定的隐节点,当障碍较多且环境动态时,网络结构庞大且神经元的阈值随时间的变化而需要不断改变的缺点。利用粒子群对神经网络的参数进行训练,在规定的代数内对网络参数优化,使得机器人在移动过程中能够快速响应环境的变化。通过对移动机器人在动、静态不同环境下的仿真实验,证明了方法的有效性。
A path planning method for mobile robot based on neural network and particle swarm optimization is proposed. Adopting the combination of wavelet network and RBF network, the four-layer neural network structure overcomes the traditional neural network method to design some specific hidden nodes for each obstacle during path planning. When there are more obstacles and the environment is dynamic, the network structure is huge The neuron’s threshold changes over time and needs to constantly change. Particle swarm optimization is used to train the parameters of neural network, and the network parameters are optimized within the given algebra so that the robot can respond quickly to changes in the environment during the movement. The simulation experiments of mobile robot under different dynamic and static conditions show that the method is effective.