AgentSkillsCN

experiment-planner-doe

精通实验设计方法,可系统性优化纳米材料的合成与加工工艺

SKILL.md
--- frontmatter
name: experiment-planner-doe
description: Design of Experiments skill for systematic optimization of nanomaterial synthesis and processing
allowed-tools:
  - Read
  - Write
  - Glob
  - Grep
  - Bash
metadata:
  specialization: nanotechnology
  domain: science
  category: infrastructure-quality
  priority: high
  phase: 6
  tools-libraries:
    - JMP
    - Design-Expert
    - Minitab
    - scipy.stats

Experiment Planner DOE

Purpose

The Experiment Planner DOE skill provides systematic experimental design for nanomaterial synthesis and processing optimization, enabling efficient exploration of parameter space and robust process development.

Capabilities

  • Factorial design generation
  • Response surface methodology
  • Taguchi method implementation
  • ANOVA analysis
  • Optimization predictions
  • Robustness testing

Usage Guidelines

DOE Workflow

  1. Design Selection

    • Identify factors and levels
    • Choose appropriate design
    • Calculate required runs
  2. Execution Planning

    • Randomize run order
    • Include replicates
    • Plan blocking if needed
  3. Analysis

    • Perform ANOVA
    • Build response models
    • Optimize parameters

Process Integration

  • Nanoparticle Synthesis Protocol Development
  • Thin Film Deposition Process Optimization
  • Nanolithography Process Development

Input Schema

json
{
  "factors": [{
    "name": "string",
    "low": "number",
    "high": "number",
    "type": "continuous|categorical"
  }],
  "responses": ["string"],
  "design_type": "factorial|fractional|rsm|taguchi",
  "constraints": {
    "max_runs": "number",
    "blocking": "boolean"
  }
}

Output Schema

json
{
  "design": {
    "type": "string",
    "runs": "number",
    "run_table": [{
      "run": "number",
      "factors": {},
      "block": "number"
    }]
  },
  "analysis": {
    "anova_table": {},
    "significant_factors": ["string"],
    "r_squared": "number"
  },
  "optimization": {
    "optimal_settings": {},
    "predicted_response": "number",
    "confidence_interval": {"lower": "number", "upper": "number"}
  }
}