Cloud Based Fleet Dispatching System
As the cloud brain of autonomous vehicle fleet, the fleet dispatching management system undertakes the task of connecting logic application layer and autonomous vehicle fleet. The fleet dispatching system of Hirain has six functional modules, including ToS interconnection system, vehicle dispatching, task planning, route planning, traffic management and remote driving interaction. Driven by the above six functional modules, the autonomous vehicle fleet can be applied in the scenarios of horizontal container transportation in the port.
Business Interconnection System
By connecting with the business system, the fleet dispatch system receives the task information to be performed by the autonomous fleet and the operation equipment information to cooperate with the autonomous fleet providing the necessary input for the task planning module. The fleet dispatch system has successfully realized the interconnection with Navis container production operation system and Yantai East China Electronic Container production operation system. In addition, the interconnection system of quayside container crane (quayside bridge) and rail container gantry crane (rail crane) has been realized with Shanghai Zhenhua and ABB respectively, which provides a bridge of information for driverless fleets serving ports.
As the core of the Fleet dispatch system of Hirain, the quality of vehicle dispatching module directly affects the operational efficiency of autonomous fleet. Based on operations research and reinforcement learning, Hirain developed the vehicle dispatching module for autonomous container horizontal transport trucks combining on-site business scenarios, which has been fully verified in port production.
The task planning module is responsible for analyzing the task instructions sent by the business system. After the instruction analysis is completed, the path planning module plans a reasonable driving path according to the road conditions in the operation area. The task planning module sends the planned task and path demand to the autonomous vehicle. In the process of executing tasks, the task planning module monitors the vehicle status and task execution status of vehicle. During the task execution process of the driverless fleet, the task planning module needs to maintain the state machine of the vehicle according to the vehicle state and makes decisions on the tasks to be executed in the next stage of the autonomous vehicle according to the rule engine.
According to the path planning request sent by the task planning module, the path planning module plans the starting point and ending point of the autonomous vehicle by reading the high-precision map information and real-time road condition information in combination with the current position, target position, task type and other information of the autonomous vehicle. Aiming at the port operation scenario, the path planning module also develops a multi-entry multi-exit for route management subsystem on the wharf area. By docking with quay bridge location and the bollard for ship berthing, the intersections of quay bridge and quay surface are dynamically selected.
The traffic management system controls the autonomous vehicles by combining the roadside perception information provided by V2X, the vehicle task information and the path planning information provided by the path planning module. The control area is mainly concentrated on intersection and operation area. Through control means such as barrier gate and the virtual traffic lights, the traffic management system makes sure that driverless vehicles and human-driven vehicles pass through the intersection area one by one at different times, ensuring a smooth traffic flow at the operation site of autonomous vehicle fleet.
Remote driving interaction
As a supplement of autonomous driving, remote driving plays a very important role in the operation of driverless fleet. How to integrate the business - vehicle - cloud control platform - remote driving system has become the key to the successful landing of the remote driving system. Remote driving interactive system fully integrates vehicle alarm information and business information. After a long time of practice, a perfect remote driving takeover scenario library has been established and the priority of remote driving takeover is clearly defined. Combined with the intelligent dispatching of the remote driving scheduling module, the remote driving system is perfectly integrated with the fleet dispatch system.
The fleet dispatch system developed by Hirain provides strong technical support for the landing of autonomous fleets in specific scenes like ports. The combination of fleet dispatch system, remote driving system and autonomous driving system makes it possible for autonomous fleet to operate normally in fixed scenes.