Guo Zhengjiang, Xia Yingkai, Li Jiawei, Liu Jiajun, Xu Kan
College of Engineering, Huazhong Agricultural University, Wuhan 430070, China.
Wuhan Second Ship Design and Research Institute, Wuhan 430205, China.
Sensors (Basel). 2024 Sep 11;24(18):5898. doi: 10.3390/s24185898.
To address the path planning problem for automated guided vehicles (AGVs) in challenging and complex industrial environments, a hybrid optimization approach is proposed, integrating a Kalman filter with grey wolf optimization (GWO), as well as incorporating partially matched crossover (PMX) mutation operations and roulette wheel selection. Paths are first optimized using GWO, then refined with Kalman filter corrections every ten iterations. Moreover, roulette wheel selection guides robust parent path selection, while an elite strategy and partially matched crossover (PMX) with mutation generate diverse offspring. Extensive simulations and experiments were carried out under a densely packed goods scenario and complex indoor layout scenario, within a fully automated warehouse environment. The results showed that this hybrid method not only enhanced the various optimization metrics but also ensured more predictable and collision-free navigation paths, particularly in environments with complex obstacles. These improvements lead to increased operational efficiency and safety, highlighting the method's potential in real-world applications.
为解决自动导引车（AGV）在具有挑战性和复杂性的工业环境中的路径规划问题，提出了一种混合优化方法，该方法将卡尔曼滤波器与灰狼优化算法（GWO）相结合，并引入部分匹配交叉（PMX）变异操作和轮盘赌选择。路径首先使用GWO进行优化，然后每十次迭代使用卡尔曼滤波器校正进行细化。此外，轮盘赌选择指导稳健的父路径选择，而精英策略和带有变异的部分匹配交叉（PMX）生成多样化的后代。在全自动仓库环境中，在货物密集场景和复杂室内布局场景下进行了广泛的模拟和实验。结果表明，这种混合方法不仅提高了各种优化指标，还确保了更可预测且无碰撞的导航路径，特别是在具有复杂障碍物的环境中。这些改进提高了运营效率和安全性，突出了该方法在实际应用中的潜力。
