首页|基于MPC控制和改进极限学习机的企业用车轨迹跟踪模型

基于MPC控制和改进极限学习机的企业用车轨迹跟踪模型

扫码查看
为了 了解企业用车情况,掌握用车轨迹,提出基于MPC控制和改进极限学习机的企业用车轨迹跟踪模型.采集和预处理车辆行驶的基础信息;利用粒子群优化算法优化改进ELM模型,通过DNN模型和改进ELM模型输出具备不同优点的车辆行驶信息;使用融合层融合不同特征车辆行驶信息,使用输出层输出车辆行驶的横纵向预测位置;采用MPC模型控制器对车辆行驶预测位置进行线性化与离散化处理;利用模型预测控制算法计算车辆跟踪周期内的时域输入序列,并将该序列内的首个元素作为控制输入量,反复循环后实现车辆轨迹跟踪.实验结果表明,该模型的车辆信息均方误差波动区间为0.006~0.013,泛化能力好;模型的跟踪精度高达100%,可靠性较强.
Enterprise Vehicle Trajectory Tracking Model Based on MPC Control and Improved Extreme Learning Machine
In order to understand the situation of enterprise vehicles,more accurately grasp the car trajectory,this paper studies the enterprise car trajectory tracking model based on MPC control and improved limit learning machine.It collects and prepro-cesses the basic information of vehicle driving;optimizes and improves ELM model using particle group optimization algorithm,outputs vehicle driving information with different advantages through DNN model and the improved ELM model.It fuses dif-ferent characteristic vehicle driving information using fusion layer,and the output layer is used to output the horizontal and lon-gitudinal predicted position of the vehicle.The MPC model controller is used to linearize and discretize the vehicle prediction position.The time domain input sequence is calculated during the vehicle tracking cycle using the model prediction control algo-rithm,and the first element within the sequence is used as the control input to realize the vehicle track tracking after repeated cycles.The experimental results show that the mean square error fluctuation range of the model is 0.006~0.013 with good generalization ability;the tracking accuracy is up to 100%,and has strong reliability.

MPC controlextreme learning machineenterprise vehicletrack trackingdiscretization

于光宗、祝存平、高攀、刘育平、张斐、邢玉娟

展开 >

国网甘肃省电力公司,后勤工作部,甘肃,兰州 730050

国网思极飞天(兰州)云数科技有限公司,甘肃,兰州 730050

MPC控制 极限学习机 企业用车 轨迹跟踪 离散化

2024

微型电脑应用
上海市微型电脑应用学会

微型电脑应用

CSTPCD
影响因子:0.359
ISSN:1007-757X
年,卷(期):2024.40(8)