Research On Process Planning Method For Intelligent CNC Controller Based On Cloud Knowledge Base

Facing current variable market and the needs of increasing customized production,enterprises must adopt advanced production mode and manufacturing system to survive in such fierce global economic competition.As the main manufacturing equipment of manufacturing industry,CNC machine tools are the foundation and core of advanced manufacturing technology.CNC controller is considered as the brain of CNC machine tool and its performance determines the level of CNC machine tool.However,both foreign and domestic CNC controllers lack intelligent support,which has become the bottleneck in the implement of advanced manufacturing mode.Intelligent decision-making ability is one of the development goals for intelligent equipment,which involves intelligent process planning technology.However,current CNC controllers still lack the ability of autonomous process planning.To solve thisproblem,process planning method of intelligent CNC controller based on cloudknowledge base is studied in this paper to improve the intelligence of the controller from aspect of process planning.Based on the functional analysis of autonomous process planning of intelligent CNC controllers,a cloud-local two-layer intelligent controller framework is proposed to solvethe contradiction between the storage and processing requirements for massive process planning knowedge and the limited resource of a single machine.Through putting complex decision-making process on the cloud,storage and computing required for the openness of process planning system can be satisfied.Then through real-time interaction between cloud knowledge based and local controller,requirements of on-site process decision-making to adapt to unexpected situation of workshop to can be meet.The design flow of intelligent CNC controller is designed with design of process planning model,design of cloud knowledge base and design of local controller.The running mechanism of intelligent CNC controller is also introduced from two aspects of knowledge gathering and response to process planning requests.Aiming at the heterogeneous problem in process planning knowledge gathering and the problem that CNC controller can’t understand the upstream design information for autonomous decision-making,a process planning model is designed based on ontology to solve these problems.Through analysing the application purpose of process planning model,content of the model is classified and the relationship between each part is clarified.Then the modeling method using Web Ontology Language(OWL)and ontology editor protege is given.Subsequently,the sub-models of process planning model,including machining task model,machining resource model and machining process model,are analyzed and designed using ontology respectively to construct a unified process planning model.At last,a design method of instance-generating software is given based on the ontology model and Sematic Web Rule Language(SWRL),which provides support for the construction and running of intelligent CNC controller.To solve the problems existed in process planning system,such as inadequate openness and difficulty to integrate with CNC controller,which leads to rigid decision-making and obstacle on-site process planning,cloud knowledge base is designed as the brain of intelligent CNC controllers.Then the autonomous process planning ability of CNC controller can be realized through cooperation between cloud base with local controller.Cloud knowledge base is responsible for the gathering,storage and utilization of massive process planning knowledge.The communication interface between cloud and local is designed based on Client/Server architecture,which can achieve the cross-regional gathering of process planning knowledge and the multi-client real-time interaction in autonomous process planning.Process planning case set is designed based on the column-based HBase.its distributed architechture meet the requirements of open system for extensible storage space and high throughput.At the same time.to realize the management for process planning knowledge,a process evaluation mechanism is designed to select high quality knowledge,and a credibility evaluation mechanism is also constructed to realize the optimization of scoring group.On the basis of realizing the gathering and storage of massive process planningknowledge,workflow of autonomous process planning is designed based on the MapReduce distributed programing framework,which realizes the matching of machining tasks and the rapid acquisition of corresponding machining processing.In addition,if existed machining processing cannot be appied for submitted machining task,new process planning solution can be generated through reasoning on workpiece level or feature level.Then,autonomous process planning can be realized based on cloud knowledge base.The main role of local controller is to cooperate with cloud knowledge base,which includes functions of process planning requests submition and process planning solution interpretation,execution and feedback.In the pre-processing stage,through designing process buffer and pre-planning modules,the processing feasibility of machining task can be detected,and the local machining task match can be realized.In the post-processing stage,through designing post-processing module,parser and tool path generator,process planning solution pushed by cloud knowledge base can be adjusted,interpreted and the tool path required by NC kernel can be generated.In the user feedback stage,though designing user feedback module,quantitative scoring of used process planning solution and submission of new process planning knowledge can be realized,which enriches theway of knowledge gathering.Based on the above researches,Hadoop development kit is used to develop the cloud knowledge base using Java.Local controller is developed based on the CoDeSys platform.Servo system control is realized by using field bus EtherMAC,which is developed by our research group.Then the intelligent CNC controller prototype is built.Finally,through a turning case,autonomous process planning ability and corresponding implementation ability of intelligent CNC controller are verified.