【作者】 郑雷；

【导师】 屈卫东； 许伟达；

【作者基本信息】 上海交通大学 ， 控制工程（专业学位）， 2019， 硕士

【摘要】 在经济全球化,工业4.0及可持续发展的影响下,国内各个制造企业都面临着转型升级的巨大挑战,尤其是最近几年,汽车市场中新能源汽车、互联网汽车、以及无人驾驶汽车等很多新产品的出现,使得汽车的更新换代和产品升级周期明显缩短。所以,对于汽车制造企业来说,生产线以及生产线的物流调度需要满足柔性化生产,而汽车的底盘生产线,由于其工艺选装的组合类型多,对柔性化的需求更加迫切。自动导引小车(Automated-Guided Vehicle,AGV),目前在各个制造领域中被广泛运用,用于仓库与现场生产工位的物料运输。为了提高生产物料的供给效率,需要提升AGV的效率,解决满足各类约束条件下的最优路径规划。AGV小车在各类约束下的配送路径问题是典型的NP-Hard问题,传统算法求解此类问题具有效率低、计算代价高、容易陷入局部最优等问题,且订单式的生产模式、物料供给,无法满足生产车型配比变化,生产环境变化等情况。因此,本文重点通过智能型算法,并通过RFID的部署,对现场各类扰动因素进行实时捕获,用于解决动态生产环境下的AGV配送路径问题与底盘生产线集成调度问题。(1)根据汽车底盘生产线的物料配送特点,将AGV小车的行驶总路径长度、载重总量、每个工位都需得到物料配送等作为约束条件,建立路径最短的目标函数,并建立汽车底盘生产线的物料AGV配送模型,通过智能算法解决路径最优配送问题。(2)针对传统订单式的AGV配送方法无法满足现场车型生产配比变化,以及各类生产过程中,不可预估的扰动因素,如:设备突发故障、物料供给缺陷、生产质量问题等,提出了基于在汽车底盘生产线部署RFID的方案,通过RFID采集的数据如节拍、各个工位物料装配合格率、生产停机等信息,为算法提供实时的数据,通过对现场扰动因素的实时捕获,并将这些扰动因素转化为各个工位的时间窗口约束,作为算法的输入,提高算法的全局路径规划问题的能力。(3)随着射频识别技术(RFID)的飞速发展,使得对生产线的实时数据采集变得简单、可靠,通过对RFID硬件、软件部署,实现对汽车底盘生产线的托盘进行识别,通过控制托盘的输送,实现不同车型的不同装配线、工艺分配,实现生产线的柔性化调度,提升生产线的柔性化生产能力。针对柔性化生产的物料调度问题,提出了遗传算法和基于RFID技术的物流小车调度路径优化方案,并通过应用遗传算法对所建模型进行验证。最后,结合RFID技术,在底盘生产线搭建了生产过程扰动实时捕捉系统,实现底盘生产线的智能化监控,通过模拟仿真,验证了在实际生产过程中,基于RFID技术的生产线扰动因素监控,为智能算法提供数据输入、决策依据的方式,对于汽车底盘生产线的物流调度效率提升更具实际指导意义。更多还原

【Abstract】 Under the influence of economic globalization,industry 4.0 and sustainable development.Domestic manufacturing enterprises are facing great challenges of transformation and upgrading.In recent years,the emergence of new energy vehicles,Internet vehicles,driverless vehicles and many other new products in the automobile market,which significantly shortens the cycle of automobile upgrading and product upgrading.Therefore,the logistics scheduling of production line and production line needs to meet the flexible production and the chassis production line of automobile is more urgent because of its complex process and production equipment.Automatic guided vehicle(AGV)is widely used in various manufacturing fields for material transportation in warehouse and field production station.In order to improve the supply efficiency,it is necessary to improve the efficiency of AGV and solve the optimal path planning under various constraints.Path planning is a typical NP hard problem and solved by traditional algorithm with low efficiency,high calculation cost,easy to fall into local optimization and so on.Therefore,this paper focuses on real-time capture of all kinds of disturbance factors through intelligent algorithm and RFID deployment.It is can also solve the AGV distribution path problem and chassis production line integrated scheduling problem in dynamic production environment.(1)According to the material distribution characteristics of automobile chassis production line.Taking the total path length of AGV car,the total load and the material distribution of each station as constraints,the shortest path objective function is established,and the material AGV distribution model of automobile chassis production line is established,and the optimal path distribution problem is solved with intelligent algorithm.(2)In view of the fact that the traditional method can not meet the adjustment of the production proportion of vehicle models.The unpredictable disturbance factors in various production processes,such as unexpected equipment failure,material supply defects,production quality problems.This paper puts forward a scheme based on the deployment of RFID in the automobile chassis production line,and the data collected by RFID,such as speed of the line,qualified assembly of materials in each station,production downtime and other information provide real-time data for the algorithm.Through real-time acquisition of field disturbance factors,these disturbance factors are transformed into time window constraints of each station,as the input of the algorithm,to improve the ability of the algorithm for path planning.(3)With the rapid development of RFID technology,the real-time data collection of the production line becomes simple and reliable.Through the deployment of RFID hardware and software,the pallet of the automobile chassis production line can be identified.Through the control of the transportation of the pallet,different assembly lines and process allocation of different models can be identified,the flexible scheduling of the production line can be realized,and the production line can be improved Flexible production capacity.For the material scheduling problem of multi vehicle co-production,this paper proposes the genetic algorithm and RFID Based Logistics car scheduling path optimization scheme,and verifies the model by using genetic algorithm.Finally,the real-time capture system of production process disturbance is built in the chassis production line based on RFID technology to realize the intelligent monitoring of the chassis production line.Through simulation,it is verified that in the actual production process,the monitoring of production line disturbance factors based on RFID technology and provides algorithm with the decision data input can improve the efficiency of the logistics scheduling.更多还原