双机械臂操控模式下船体外板氢氧焰线加热成形机理及智能作业决策支持系统研究

Shipbuilding industry adopts line heating technology (in short LH technology) to achieve the surface of ship hull plate forming process. However, the process is still completed by skilled workers using oxyacetylene flame now manually. This leads to forming quality hard to guarantee, large energy consumption and environmental pollution. It becomes the ship a "bottleneck" problem of construction cycle and quality. Hydrogen flame is applied in LH technology in the project, aiming to the better forming accuracy, lower energy consumption and pollution. (1)studying forming mechanism of hydrogen flame, and deeply grasping relationships between technology parameters and the deformation controlled by double manipulators; (2) applying large processing data to revision and optimization of forming mechanism model and founding close to practical production; (3)founding model based on multi-source heterogeneous information, and constructing knowledge base with intelligent hybrid model, and applying artificial intelligence algorithms to build the inference engine, and constantly optimizing the performance of the knowledge base and inference engine, and building intelligent processing decision system based on double manipulators to provide a reasonable processing decision to the operator. Above desired achievements provides beneficial theoretical support to enhance automation, information and greening degree of LH technology.

船舶工业普遍采用线状水火弯板方法（简称“LH成形方法”）实现“船体外板曲面成形工序”，然该工序目前仍靠工人使用氧乙炔焰手工作业完成，导致质量难于保证、能耗大和污染大，成为制约船舶建造周期和质量的“瓶颈”问题。本项目将节能、高效和环保能源——氢氧焰用于“LH成形方法”，以提高成形效率和精度、降低能耗和减少污染为目标开展研究：（1）研究氢氧焰线加热成形机理建模，深层次掌握多种双机械臂工作模式下加工参数与变形之间的关系规律；（2）将大量过程数据用于机理模型修正和优化，构建智能混合机理模型，逼近生产实际；（3）构建基于多源异构信息的产品信息模型，与（2）中模型一同组建智能作业支持系统知识库，同时采用人工智能算法实现推理机，并不断优化知识库和推理机的性能，实现双机械臂操控下的智能作业决策支持系统，向操作员提供合理的加工策略。以上成果为提升成形工序实现信息化、自动化、智能化和绿色化提供有利理论支撑。

我国已成为造船大国，三项综合指标已多年位列全世界第一。船舶建造过程中，需要对大量的钢板进行处理和加工，所需费用占整个造船成本的20%以上。而船体外板曲面成形工艺目前自动化、智能化和绿色化水平总体偏弱，大部分工作都由熟练工人手工完成，其成形精度和效率难以保证，已成为船舶建造的瓶颈环节。本项目创新采用节能、高效和环保的能源——氢氧焰用于船体外板成形加工中，提出一种新的“线加热成形方法”，显著提高了船体外板曲面成形的效率和精度，降低生产能耗和减少环境污染。本项目通过研究氢氧焰线加热成形机理，构建双椭球体热源和背面水冷模型，深层次掌握加工参数、过程数据与变形之间的关系规律，并研究在不同工况和加工生产要求下基于复杂轨迹的新加工模式对钢板变形的效率和精度的影响；采用人工智能算法实现作业决策支持，向操作员提供合理的加工策略，并不断优化知识库和推理机的性能，使得整个加工过程在效率和精度上有大幅度的提升，为实现船舶智能制造提供有利支撑和保障。