如何将指数曲线转换为Java代码？AnyLogic沙林扩散建模求助

Hey there! Let's walk through this step by step—since you're new to programming, I'll keep things clear and practical, focused on what you need for your AnyLogic model.

First: Define the Exponential Diffusion Formula #

Sarin's concentration decay with distance typically follows an exponential curve. A standard form for this is:

C(d) = C₀ * e^(-k*d)

Where:

• C(d) = Sarin concentration at distance d from the release point
• C₀ = Initial concentration at the release point (you'll need to calibrate this with real data from the 1995 attack)
• k = Decay coefficient (controls how fast concentration drops off—use values from toxicology studies or attack reports)
• d = Distance from the release point (keep units consistent, e.g., meters)

Step 1: Set Up Your AnyLogic Parameters #

First, add these parameters to your AnyLogic model (via the "Parameters" panel):

• double initialSarinConcentration = [your calibrated value, e.g., 100.0]
• double decayCoefficient = [your calibrated value, e.g., 0.05]
• MapPoint releasePoint = The exact location of the sarin release on your Tokyo map (drag a MapPoint onto the map and set its coordinates)

Step 2: Write the Java Loop to Calculate Concentration for All Positions #

Now, we'll loop through all the positions you care about (e.g., grid points, or zone agents representing city blocks) and assign concentration values.

Example 1: Looping Through a Collection of Map Points #

If you've created a list of key positions (e.g., every 50m on a grid), use this code (put it in an Event or Button action):

// Assume you have a Collection<MapPoint> monitoringPoints (define this in your model)
for (MapPoint point : monitoringPoints) {
    // Calculate distance from release point to current point
    double distance = releasePoint.getDistance(point);
    
    // Compute exponential decay concentration
    double currentConcentration = initialSarinConcentration * Math.exp(-decayCoefficient * distance);
    
    // Optional: Round to 2 decimal places for readability
    currentConcentration = Math.round(currentConcentration * 100.0) / 100.0;
    
    // Assign the concentration to the linked agent (e.g., a CityBlock agent)
    CityBlock block = findAgent(CityBlock.class, c -> c.getLocation().equals(point));
    if (block != null) {
        block.sarinLevel = currentConcentration;
    }
    
    // Optional: Add a label to the map showing concentration
    map.addLabel(String.valueOf(currentConcentration), point);
}

Example 2: Looping Through All Zone Agents #

If you're using AnyLogic's Zone Agents to represent Tokyo city blocks, loop through them directly:

// Loop through every CityBlock agent in your model
for (CityBlock block : cityBlocks) {
    // Get the block's center reference point
    MapPoint blockLocation = block.getLocation();
    
    // Calculate distance to release point
    double distance = releasePoint.getDistance(blockLocation);
    
    // Compute concentration
    double currentConcentration = initialSarinConcentration * Math.exp(-decayCoefficient * distance);
    
    // Assign to the block's concentration parameter
    block.sarinLevel = currentConcentration;
    
    // Optional: Update block color to create a heatmap
    block.setFillColor(ColorUtil.interpolate(Color.GREEN, Color.RED, currentConcentration / initialSarinConcentration));
}

Key Tips for Your Model #

• Unit Consistency: Ensure your distance units (meters/kilometers) match the decay coefficient you're using—mixing units will break your results.
• Calibration: Look up peer-reviewed studies on the 1995 Tokyo attack to get realistic values for initialSarinConcentration and decayCoefficient—this makes your model accurate.
• Performance: If you have thousands of points/agents, run this loop in a periodic Event (e.g., every 1 minute in model time) instead of non-stop to avoid slowing down your model.
• Visualization: Use color interpolation (like in Example 2) to create a heatmap—it makes the diffusion pattern easy to interpret at a glance.

If you get stuck on specific parts (like creating the monitoringPoints collection, or linking agents to map positions), just clarify what you're working on and we can tweak this further!

内容的提问来源于stack exchange，提问作者T.Nosek