基于OpenClaw的砂型铸造仿真程序自动化调用与参数辅助设置方法OA

With the continuous penetration of artificial intelligence technology into the field of casting,and in view of the current issues in numerical simulation software for sand casting—such as large amounts of input required for casting simulation programs,complex inter-stage dependencies,and parameter settings relying heavily on manual experience—this study proposes a method for automated program invocation and parameter-assisted configuration based on OpenClaw.This method uses the existing simulation program as the computational core,with OpenClaw serving as the interaction interface and task scheduling layer.Through the Model Context Protocol(MCP)tool interface,it enables 3D model recognition,mold generation,parameter organization,rationality checking,computational task invocation,and result aggregation.Taking stepped test blocks and bearing housings as case studies,the proposed workflow was validated in terms of automated invocation processes,abnormal parameter detection,and batch computations based on orthogonal experiments.The results show that,under the optimized model(DeepSeek-V4-Flash)-driven approach,task configuration time is reduced from 20 min in manual operation to 4.5 min.The system can also identify and prompt various abnormal parameter settings,and is capable of executing task calls under multiple parameter combinations and performing shrinkage cavity and shrinkage porosity volume statistics.