Tracks
/
Python
Python
/
Exercises
/
Meltdown Mitigation
Meltdown Mitigation

Meltdown Mitigation

Learning Exercise

Introduction

In Python, if, elif (a contraction of 'else and if') and else statements are used to control the flow of execution and make decisions in a program. Unlike many other programming languages, Python versions 3.9 and below do not offer a formal case-switch statement, instead using multiple elif statements to serve a similar purpose.

Python 3.10 introduces a variant case-switch statement called structural pattern matching, which will be covered separately in another concept.

Conditional statements use expressions that must resolve to True or False -- either by returning a bool type directly, or by evaluating as "truthy" or "falsy".

x = 10
y = 5

# The comparison '>' returns the bool 'True',
# so the statement is printed.
if x > y:
    print("x is greater than y")
...
>>> x is greater than y

When paired with if, an optional else code block will execute when the original if condition evaluates to False:

x = 5
y = 10

# The comparison '>' here returns the bool 'False',
# so the 'else' block is executed instead of the 'if' block.
if x > y:
    print("x is greater than y")
else:
    print("y is greater than x")
...
>>> y is greater than x

elif allows for multiple evaluations/branches.

x = 5
y = 10
z = 20

# The 'elif' statement allows for the checking of more conditions.
if x > y:
    print("x is greater than y and z")
elif y > z:
    print("y is greater than x and z")
else:
    print("z is greater than x and y")
...
>>> z is great than x and y

Boolean operations and comparisons can be combined with conditionals for more complex testing:


>>> def classic_fizzbuzz(number):
        if number % 3 == 0 and number % 5 == 0:
            return 'FizzBuzz!'
        elif number % 5 == 0:
            return 'Buzz!'
        elif number % 3 == 0:
            return 'Fizz!'
        else:
            return str(number)

>>> classic_fizzbuzz(15)
'FizzBuzz!'

>>> classic_fizzbuzz(13)
'13'

Instructions

In this exercise, we'll develop a simple control system for a nuclear reactor.

For a reactor to produce the power it must be in a state of criticality. If the reactor is in a state less than criticality, it can become damaged. If the reactor state goes beyond criticality, it can overload and result in a meltdown. We want to mitigate the chances of meltdown and correctly manage reactor state.

The following three tasks are all related to writing code for maintaining ideal reactor state.

1. Check for criticality

The first thing a control system has to do is check if the reactor is balanced in criticality. A reactor is said to be critical if it satisfies the following conditions:

  • The temperature is less than 800 K.
  • The number of neutrons emitted per second is greater than 500.
  • The product of temperature and neutrons emitted per second is less than 500000.

Implement the function is_criticality_balanced() that takes temperature measured in kelvin and neutrons_emitted as parameters, and returns True if the criticality conditions are met, False if not.

>>> is_criticality_balanced(750, 600)
True

2. Determine the Power output range

Once the reactor has started producing power its efficiency needs to be determined. Efficiency can be grouped into 4 bands:

  1. green -> efficiency of 80% or more,
  2. orange -> efficiency of less than 80% but at least 60%,
  3. red -> efficiency below 60%, but still 30% or more,
  4. black -> less than 30% efficient.

The percentage value can be calculated as (generated_power/theoretical_max_power)*100 where generated_power = voltage * current. Note that the percentage value is usually not an integer number, so make sure to consider the proper use of the < and <= comparisons.

Implement the function reactor_efficiency(<voltage>, <current>, <theoretical_max_power>), with three parameters: voltage, current, and theoretical_max_power. This function should return the efficiency band of the reactor : 'green', 'orange', 'red', or 'black'.

>>> reactor_efficiency(200,50,15000)
'orange'

3. Fail Safe Mechanism

Your final task involves creating a fail-safe mechanism to avoid overload and meltdown. This mechanism will determine if the reactor is below, at, or above the ideal criticality threshold. Criticality can then be increased, decreased, or stopped by inserting (or removing) control rods into the reactor.

Implement the function called fail_safe(), which takes 3 parameters: temperature measured in kelvin, neutrons_produced_per_second, and threshold, and outputs a status code for the reactor.

  • If temperature * neutrons_produced_per_second < 90% of threshold, output a status code of 'LOW' indicating that control rods must be removed to produce power.

  • If the value temperature * neutrons_produced_per_second is within 10% of the threshold (so either 0-10% less than the threshold, at the threshold, or 0-10% greater than the threshold), the reactor is in criticality and the status code of 'NORMAL' should be output, indicating that the reactor is in optimum condition and control rods are in an ideal position.

  • If temperature * neutrons_produced_per_second is not in the above-stated ranges, the reactor is going into meltdown and a status code of 'DANGER' must be passed to immediately shut down the reactor.

>>> fail_safe(temperature=1000, neutrons_produced_per_second=30, threshold=5000)
'DANGER'
Edit via GitHub The link opens in a new window or tab
Python Exercism

Ready to start Meltdown Mitigation?

Sign up to Exercism to learn and master Python with 17 concepts, 140 exercises, and real human mentoring, all for free.