SimPy 中的可选资源

Question

我正在使用 Simpy 进行离散事件模拟，但对于我最能描述为可选资源的内容存在问题。上下文是我正在模拟将要执行的任务，并且需要一些资产（资源）来执行该任务。与大多数 SimPy 实现不同，任务需要在分配的时间开始，否则会失败，并且可能会接受较少的资源来启动任务。

例如，一个任务在时间 = t 需要四辆车。在时间 t 只有三辆车可用，因此任务以三辆车开始，但结果较少。如果只有两辆或更少的车辆可用，任务将不会继续，将被视为失败。

抱歉，此示例中缺少代码，我很难理解它。任何帮助将不胜感激。

Answer 1

我使用容器来跟踪资源，但请看一下，让我知道这是否是您想要的

"""
Demos checking the availability of resources befor taking resouces

Each agent has a first choice and second choice for seizing resources
if neither choice is avalale will not wait and skip getting resources

programmer: Michael R. Gibbs
"""

import simpy
import random

def eval(choice, containers):
    """
    steps through each requirement and checks if there is enough resouces is available

    returns True if there is enough, else returns False
    """
    for k,v in choice.items():
        if containers[k].level < v:
            # not enough
            return False
    return True

def mission(env, id, firstChoice, secondChoice, startDelay, missingTime, containers):
    """
    Sims agent checking if first or second choice of resouces is available
    If so, will seize the resources, hold for a time, and then release the resouces
    """

    choice = None   # which choice to use

    # wait for mission start
    yield env.timeout(startDelay)

    print()
    print(f'{env.now} agent {id} is starting a mission')
    x = [(k, v.level) for k, v in containers.items()]
    print(f'available: {x}')
    print(f'first choice {firstChoice}')
    print(f'second choice {secondChoice}')

    # evaluate the choices and see if either will work
    if eval(firstChoice, containers):
        choice = firstChoice
        print(f'{env.now} agent {id} is going with first Choice')
    else:
        if eval(secondChoice,containers):
            choice = secondChoice
            print(f'{env.now} agent {id} is going with second Choice')
        else:
             print(f'{env.now} agent {id} is abort because not enough resources')

    if choice is not None:
        # found a choice that works
        # seize the resouces, we checked that they are availale, so there should not be any queue time
        for k,v in choice.items():
            containers[k].get(v)
        
        yield env.timeout(missingTime)

        # return resouces
        for k,v in choice.items():
            containers[k].put(v)

        print(f'{env.now} agent {id} has finished the mission ')

def build_choice():
    """
    quick helper to generate random first and second choice requirements
    (may not use all three resouces)

    returns a dictionary of the requirements where key=type, value=how much
    """
    choice = {}
    while len(choice) == 0:
        for k in ['A','B','C']:
            v = random.randint(0,3)
            if v > 0:
                choice[k] = v
    return choice


def gen_missions(env, containers):
    """
    Generate a series of agents to seize the resources
    """

    i = 0
    while True:
        i += 1 # id generator
        firstChoice = build_choice()
        secondChoice = build_choice()

        env.process(mission(env,i,firstChoice,secondChoice,random.randint(1,4),random.randint(2,8),containers))

        # put some time between agents
        yield env.timeout(1)

env = simpy.Environment()

# load the resouces
containers = {}
containers['A'] = simpy.Container(env, init=10, capacity=10)
containers['B'] = simpy.Container(env, init=10, capacity=10)
containers['C'] = simpy.Container(env, init=10, capacity=10)

# start generating agents
env.process(gen_missions(env,containers))

env.run(100)