simulation_handler.run_simulation

  1"""
  2This module runs the port simulation, generating ships, trucks, and trains,
  3creating resources, and executing the simulation until the specified time.
  4"""
  5
  6import gc
  7import psutil
  8import time
  9import os
 10import shutil
 11
 12import simpy
 13import pandas as pd
 14
 15from simulation_handler.helpers import clear_logs, clear_env
 16from simulation_handler.preprocess import generate_ships, generate_trucks, generate_trains, get_piplines_import
 17from simulation_handler.generators import ship_generator, truck_generator, train_generator, data_logger
 18from simulation_handler.helpers import clean_data, create_terminal_data_cache, create_terminal_tuple_cache
 19from simulation_classes.port import create_resources
 20from simulation_classes.channel import Channel
 21from simulation_analysis.resource_utilization import bottleneckAnalysis
 22from simulation_analysis.whatif_scenarios import *
 23from simulation_analysis.results import plot_channel, gen_logs_and_plots
 24import constants
 25
 26
 27def print_memory_usage():
 28    """Print the current memory usage of the process."""
 29    process = psutil.Process(os.getpid())
 30    memory_info = process.memory_info()
 31    print(f"Memory usage: {memory_info.rss / 1024**2:.2f} MB (RSS), {memory_info.vms / 1024**2:.2f} MB (VMS)")
 32
 33def run_simulation(seed):
 34    """
 35    Run the port simulation with the given seed.
 36    This function initializes the simulation environment, generates ships, trucks, and trains,
 37    creates resources, and runs the simulation until the specified time.
 38    Args:
 39        seed (int): Random seed for reproducibility.
 40    """
 41    print("Processing seed: ", seed)
 42    start_time = time.time()
 43    run_id = f"Results_{seed}_{int(constants.NUM_MONTHS)}_months_{constants.ARRIVAL_INCREASE_FACTOR}"
 44    clear_logs(run_id)
 45    ship_logs = []
 46
 47    print("Preprocessing data...")
 48    terminal_data_df = clean_data(constants.directory)
 49    terminal_data = create_terminal_data_cache(terminal_data_df, run_id, seed)
 50    terminal_tuple_cache = create_terminal_tuple_cache(terminal_data_df, run_id, seed)
 51
 52    df = pd.DataFrame(terminal_data.items(), columns=['Key', 'Amount Allocated'])
 53    df[['Terminal type', 'Terminal Number', 'Resource Allocated']] = pd.DataFrame(df['Key'].tolist(), index=df.index)
 54    df = df.drop(columns=['Key'])
 55    df = df[['Terminal type', 'Terminal Number', 'Resource Allocated', 'Amount Allocated']]
 56    df.to_csv(f'.{run_id}/logs/terminal_data_cache.csv', index=False)
 57
 58    num_container_terminals = terminal_data_df[(terminal_data_df['Cargo'] == 'Container')]['Terminal'].nunique()
 59    num_liquid_terminals = terminal_data_df[(terminal_data_df['Cargo'] == 'Liquid')]['Terminal'].nunique()
 60    num_drybulk_terminals = terminal_data_df[(terminal_data_df['Cargo'] == 'DryBulk')]['Terminal'].nunique()
 61    num_terminals_list = [num_container_terminals, num_liquid_terminals, num_drybulk_terminals]
 62
 63    print("Generating ships...")
 64    generate_ships(run_id, num_terminals_list, seed)
 65    print("Generating trucks...")
 66    generate_trucks(run_id, num_terminals_list, terminal_data_df, terminal_tuple_cache, seed)
 67    print("Generating trains...")
 68    generate_trains(run_id, num_terminals_list, terminal_data, terminal_data_df, terminal_tuple_cache, seed)
 69
 70    liq_terminals_with_pipeline_source, liq_terminals_with_pipeline_sink = get_piplines_import(num_terminals_list, terminal_data)
 71
 72    plot_channel(run_id)
 73
 74    print("Creating resources...")
 75    env = simpy.Environment()
 76    events = []
 77    train_events = {}
 78    channel_events = []
 79    channel_logs = [] 
 80    chassis_bays_utilization = {}
 81    for terminal_type in ["Container", "Liquid", "DryBulk"]:
 82        chassis_bays_utilization[terminal_type] = {}
 83        for terminal_id in range(1, num_terminals_list[["Container", "Liquid", "DryBulk"].index(terminal_type)] + 1):
 84            chassis_bays_utilization[terminal_type][terminal_id] = []
 85
 86    terminal_resouces = create_resources(terminal_data, run_id, terminal_data_df, num_terminals_list, env, seed)
 87
 88    if constants.MODEL_HURRICANE:
 89         model_hurricane(env, terminal_resouces, num_terminals_list, terminal_data, run_id, seed)
 90
 91    ship_data_df = pd.read_csv(f".{run_id}/logs/ship_data.csv")
 92    ship_data = ship_data_df.to_dict(orient="index")
 93    SIMULATION_TIME = constants.SIMULATION_TIME
 94    NUM_CHANNEL_SECTIONS = constants.NUM_CHANNEL_SECTIONS
 95    CHANNEL_SAFETWOWAY = constants.CHANNEL_SAFETWOWAY
 96    port_berths_container_terminals, port_yard_container_terminals, port_berth_liquid_terminals, port_tanks_liquid_terminals, \
 97    port_berth_drybulk_terminals, port_silos_drybulk_terminals, port_loading_bays_liquid_terminals, port_drybulk_bays_drybulk_terminals, \
 98    port_chassis_container_terminals, truck_gates_ctr, truck_gates_liquid, truck_gates_dk, train_loading_racks_ctr, train_loading_racks_liquid, \
 99                         train_loading_racks_dk, day_pilots, night_pilots, tugboats, channel_scheduer = terminal_resouces
100
101    # Creating channel...
102    SHIPS_IN_ANCHORAGE = [0,0,0] 
103    SHIPS_IN_CHANNEL = []
104    SHIPS_IN_CHANNEL.append(0) 
105    if constants.MODEL_FOG:
106        turnoffTime = {"switch": "channel_closed", "closed_between": constants.FOG_CLOSURES}
107    else:
108        turnoffTime = {"switch": "channel_open"}
109
110    channel = Channel(ship_logs, env, NUM_CHANNEL_SECTIONS, SIMULATION_TIME, CHANNEL_SAFETWOWAY, channel_events, channel_logs, day_pilots, night_pilots, tugboats, turnoffTime, channel_scheduer, seed)
111
112    # Starting simulation...
113    ship_proc = env.process(ship_generator(run_id, env, chassis_bays_utilization, port_berths_container_terminals, port_yard_container_terminals, port_berth_liquid_terminals, port_tanks_liquid_terminals, 
114                            port_berth_drybulk_terminals, port_silos_drybulk_terminals, channel, day_pilots, night_pilots, tugboats, events, ship_logs, channel_events, channel_logs, SHIPS_IN_ANCHORAGE, SHIPS_IN_CHANNEL, ship_data, terminal_data, liq_terminals_with_pipeline_source, liq_terminals_with_pipeline_sink))
115    truck_proc = env.process(truck_generator(run_id, liq_terminals_with_pipeline_source, liq_terminals_with_pipeline_sink, chassis_bays_utilization, env, terminal_tuple_cache, port_tanks_liquid_terminals, port_yard_container_terminals, port_silos_drybulk_terminals, port_loading_bays_liquid_terminals, port_drybulk_bays_drybulk_terminals, port_chassis_container_terminals, truck_gates_ctr, truck_gates_liquid, truck_gates_dk, events, seed, terminal_data))
116    train_proc = env.process(train_generator(run_id, liq_terminals_with_pipeline_source, liq_terminals_with_pipeline_sink, env, terminal_tuple_cache, train_loading_racks_ctr, train_loading_racks_liquid, train_loading_racks_dk, train_events, port_tanks_liquid_terminals, port_yard_container_terminals, port_silos_drybulk_terminals, seed))
117
118    pilots_tugs_data = pd.DataFrame(columns=['Time', 'Day Pilots', 'Night Pilots', 'Tugboats'])
119    data_taker_proc = env.process(data_logger(run_id, env, pilots_tugs_data, day_pilots, night_pilots, tugboats))
120
121    env.run(until=SIMULATION_TIME)
122
123    clear_env(env, ship_proc, truck_proc, train_proc, data_taker_proc)
124    gen_logs_and_plots(run_id, ship_logs, events, chassis_bays_utilization, num_terminals_list, train_events, channel_logs, channel_events, channel, animate=False)
125    bottleneckAnalysis(run_id)
126    
127    # Free memory
128    # print_memory_usage()
129
130    del env, ship_proc, truck_proc, terminal_resouces, terminal_data, events, channel_events, channel_logs
131    del port_berths_container_terminals, port_yard_container_terminals, port_berth_liquid_terminals, port_tanks_liquid_terminals
132    del port_berth_drybulk_terminals, port_silos_drybulk_terminals, port_loading_bays_liquid_terminals, port_drybulk_bays_drybulk_terminals
133    del port_chassis_container_terminals, truck_gates_ctr, truck_gates_liquid, truck_gates_dk, day_pilots, night_pilots, tugboats
134    gc.collect()
135
136    # remove truck pickle file from "/.Results*" output folder (saves hard drive space)
137    if os.path.exists(f".{run_id}/logs/truck_data.pkl"):
138        os.remove(f".{run_id}/logs/truck_data.pkl")
139    else:
140        print("Truck pkl file does not exist")
141        pass 
142
143    # revome the availablity folder 
144    dir_path = f".{run_id}/logs/availability"
145
146    if os.path.exists(dir_path):
147        shutil.rmtree(dir_path) 
148    else:
149        print("Availability folder does not exist")
150
151    # close all open files and plots
152    plt.close('all')
153