Crowd simulation

It seems that humans are really right COVID-19 There's no way , It's really annoying .

So we might as well change our mind , Think about how to minimize its harm .

---

Computational simulation

I have a front-end tool at hand , be called ATOMIC AGENTS

Atomic Agents Spatial agent-based modeling in JavaScriptDocsExamples This module was written for the Visualising Contact Networks in Response to COVID-19 UKRI-funded project (University of Warwick and Swansea University). It is still under active development — contributions are welcome.

Atomic Agents

Untitled

Its advantages are high efficiency and simplicity , It is especially suitable for repeatable and interactive epidemiological analysis simulation . I am based on the official sample , Constructed its own analysis program . The code is visible in my front-end repository

Atomic Agents: Simulitis

This is a diagram of the main program

Untitled

The lower part represents a simulation scenario , There's... In the scene 1000 Members , Red represents the infected person （Infect）, Green represents rehabilitation （Recover） personnel , Blue represents uninfected persons . Some of these people are free to move （Move）, Others need to stay at home （Still）.

The upper part is the real-time statistics of various indicators , That means we can know in real time

• How many people are infected ;
• How many of them are still , How many people are athletic ;
• In the whole communication process , What is the peak value of each value .

In this way, we can get a dynamic simulation of epidemiology , That is, how infection spreads in society . As shown in the video .

【】

The results of the analysis

Next , We make simple statistics and analysis of the results .

First, let's see how effective the home isolation policy is , We adjust Move and Still The ratio between

• When the proportion of home isolation is 0.8 when , The proportion of people infected at the same time in the total population is 0.04 and 0.12;
• When the proportion of home isolation is 0.6 when , The proportion of people infected at the same time in the total population is 0.21 and 0.27;
• When the proportion of home isolation is 0.2 when , The proportion of people infected at the same time in the total population is 0.34 and 0.35.

This shows that even if the home isolation policy can not stop the virus infection forever , But it has at least one effect , That is to reduce the number of simultaneous infections to a very low level , From a utilitarian point of view , It can avoid the run on medical resources caused by the simultaneous infection of too many people to the greatest extent .

StillProb0.8

StillProb0.8

StillProb0.6

StillProb0.6

StillProb0.2

StillProb0.2

Next, we will examine the way of regional isolation , The parameters of previous area isolation are large . Shown on the graph is left 、 The passage between the two areas on the right is wide . We'll narrow it down next ,

• The proportion of home isolation is still set at 0.2, It means that there is little need for people to stay still , To the left 、 After the passage between the right areas narrows , The peak proportion of people infected at the same time increased to 0.48 and 0.49;
• In addition, we set the proportion of home isolation at 0.8, Represents the need for personnel to remain stationary , After narrowing , At the same time, the peak proportion of infected persons is still 0.06 and 0.12.

This may be due to increased intra regional action , It is easy to cause cross infection ; And when most people are still , Zone isolation has little additional effect .

StillProb0.2Narrow

StillProb0.2Narrow

StillProb0.8Narrow

StillProb0.8Narrow