基于协同神经网络的智能车行驶路径记忆算法优化

Optimization of Intelligent Vehicle Driving Path Memory Algorithms Based on Collaborative Neural Network

传统智能车行驶路径记忆算法由于路径信息计算过程复杂,导致算法的效率较低,为此提出基于协同神经网络的智能车行驶路径记忆算法优化研究。使用传感器对智能车行驶路径图像信息进行采集,利用图像二值化方法对图像边缘信息进行获取,以其为基础,通过曲率计算公式对智能车行驶路径的曲率与转角信息进行计算,从而得到智能车行驶路径完整信息,采用协同神经网络最终完成对智能车行驶路径的记忆。通过实验得到,采用优化智能车行驶路径记忆算法的智能车在低速试跑时使用时间比采用传统算法的智能车少了1.8h,在竞速赛跑时使用时间比采用传统算法的智能车少了2.1h,充分说明优化的智能车行驶路径记忆算法具备更高的效率。

Traditional intelligent vehicle path memory algorithm is inefficient because of the complexity of the calculation process of path information. Therefore, an optimization study of intelligent vehicle path memory algorithm based on Collaborative neural network is proposed. Sensors are used to collect image information of intelligent vehicle driving path, image edge information is acquired by image binarization method. Based on this, curvature calculation formula is used to calculate curvature and corner information of intelligent vehicle driving path, so as to obtain complete information of intelligent vehicle driving path. Collaborative neural network is used to memorize intelligent vehicle driving path. This completes the memory of intelligent vehicle driving path. The experiment results show that the intelligent vehicle using the optimized intelligent vehicle path memory algorithm spends 1.8 hours less than the intelligent vehicle using the traditional algorithm in the low-speed test run, and 2.1 hours less than the intelligent vehicle using the traditional algorithm in the race, which fully shows that the optimized intelligent vehicle path memory algorithm has higher efficiency.