引导式测验 · 中等 等级

能源中的人工智能测验

测试操作限制、罕见事件、不确定性、测量边界、操作员权限、监控和可信的净节省声明。

相关引导路径人工智能能源
问题 1 的 8

Why evaluate an energy model under outages and weather extremes?

预览每个问题8 问题

当您参加测验时，就会出现答案和解释。

  1. Why evaluate an energy model under outages and weather extremes?
    • Rare conditions can affect whether predictions or actions remain safe and useful
    • They always appear equally in training data
    • Extreme cases are irrelevant to grid operations
    • A model can only use historical averages
  2. A storage controller lowers simulated cost but violates a reserve requirement. What is the issue?
    • The simulation proves the controller is safe
    • The optimization objective missed a real operating constraint
    • Reserve requirements apply only to forecasting
    • Lower cost always means better grid performance
  3. Which operating decision needs a different horizon than day-ahead forecasting?
    • A yearly equipment inventory
    • A static document search
    • Real-time balancing
    • A one-time model card review
  4. How should AI’s energy effect be reported?
    • Count only the energy savings
    • Assume model serving uses no energy
    • Compare different periods without stating them
    • Measure the computation’s electricity use and any downstream savings at a defined boundary
  5. What should an operator do when a forecast is uncertain or outside trained conditions?
    • Use the defined escalation or operating procedure rather than treating the forecast as certain
    • Follow it because optimization is automatic
    • Disable every safety rule
    • Assume the input is typical
  6. Why version weather data and operating policies with the model?
    • They improve numerical precision automatically
    • They affect inputs and decision behavior and help reconstruct an operational result
    • They replace monitoring
    • They have no influence after training
  7. A demand model performs well on average but misses a rare outage scenario. What does the average conceal?
    • A guaranteed reduction in electricity use
    • That the model has no input drift
    • An important low-frequency failure condition
    • The operating boundary is irrelevant
  8. Which comparison can support a claim that an AI optimization saves energy?
    • Compare the model’s training time with a utility bill
    • Report only the best day
    • Use a simulation without stating assumptions
    • Measure computation and operating effects over the same boundary and period