Linearization of Nonlinear Conditions in Autonomous Vehicle Control Systems

对工业4.0、无人驾驶汽车、飞行汽车等下一代制造和运输手段日益增长的兴趣和需求，以及将无人机作为一种交付选择带到我们家门口的日益增长的愿望，继续为学术界和工业界带来具有挑战性的研究问题。无人机、无人驾驶飞行器（UAV）、自动导引车（AGV）、无人驾驶汽车和飞机等自动驾驶车辆（AV）都具有三个共同特征。它们必须：i）按时完成从起点到目的地的旅程; ii）没有碰撞和冲突;以及ii）使用最少的能源。工业工程和运营研究社区提供了几种解决方案技术来解决车辆路线和排序问题。另一方面，来自航空航天和控制领域的研究人员已经独立于路径规划问题解决了AV控制系统中的碰撞和冲突避免问题。很少有研究人员将能源消耗问题作为车辆运营规划过程的一部分。 虽然在与AV控制领域相关的这三个特征方面已经进行了大量的研究工作，但由于经常面临拥堵和僵局，组织在运营层面仍然面临巨大挑战，特别是在航空运输和AGV运营方面。我们的研究结果表明，在AV控制系统的主要问题是缺乏一个统一的数学模型，同时考虑所有的影响因素的方法。由于燃油消耗和避碰问题的非线性特征，以往的研究者未能将联合收割机的能耗、避碰和路径规划目标结合在同一线性规划模型中。 因此，在未来五年内，我们计划研究各种物理条件的线性化，这些物理条件在各种交通系统中是固有的非线性现象，以便开发统一的线性规划模型来解决AV控制系统。在我们的例子中，我们将专注于开发方法的空中交通管理（ATM）问题，这是一类特殊的车辆路径问题。由于安全和成本问题以及对环境的影响，避免碰撞和冲突以及燃料消耗问题对航空运输业至关重要。虽然我们当前的重点将是碰撞和冲突避免和燃料消耗问题的ATM问题，重要的是要注意，所有其他类型的AV操作规划问题，需要明确考虑避免碰撞和能源管理在规划阶段，以奠定成功实施的基础。因此，可以肯定地说，拟议研究计划的成果有可能彻底改变下一代AV控制系统，如ATM，AGV，UAV，无人机和无人驾驶车辆系统。

Growing interests and the needs for the next generation manufacturing and transportation means such as Industry 4.0, driverless vehicles, flying cars and ever growing desire to bring drones to our doors as a delivery option continue bringing challenging research questions to the academia and the industry. Autonomous vehicles (AVs) such as drones, unmanned aerial vehicles (UAVs), Automated Guided Vehicles (AGVs), Driverless cars and Airplanes all share three common characteristics. They are i) required to complete a journey from an origin to destination on time; ii) free of collision and conflict; and ii) using minimum energy. Industrial Engineering and Operations Research communities have provided several solution techniques to tackle vehicle routing and sequencing problems. On the other hand, researchers from aerospace and control filed have tackled collision and conflict avoidance issues in AV control systems independently from the path-planning problems. Very few researchers have looked into energy consumption problem as part of the vehicle operations planning process. While abundance of research work has been produced in these three traits related to AV control areas, organizations continue facing great challenges at the operation level particularly in air-transportation and AGV operations due to frequently faced congestion and deadlocks. Our research results have shown that the main problem in AV control systems is the lack of methods to consider all influential factors simultaneously in a unified mathematical model. Historically researchers have failed to combine energy consumption, collision avoidance and path planning objectives in the same linear programming model due to the nonlinear features in fuel consumption and collision avoidance problems. Hence, during next five years, we plan to investigate the linearization of various physical conditions which are inherently nonlinear phenomenon in various transportation systems so that the development of unified linear programming models to tackle AV control systems. In our case, we will focus on developing methodologies for the Air Traffic Management (ATM) problem which is a special class of the vehicle routing problem. Collision and Conflict avoidance and fuel consumption issues are crucial for the air transportation industry due to safety and cost concerns as well as their impacts on the environment. While our immediate focus will be the collision and conflict avoidance and fuel consumption problems in ATM problems; it is important to note that all other types of AV operations planning problem requires explicit consideration of collision avoidance and energy management at the planning phase in order to lay the grounds for successful implementations. Therefore, it is safe to claim that the outcome of the proposed research program has potentials to revolutionize the next generation AV control systems such as ATM, AGV, UAV, drone and driverless vehicle systems.