File size: 7,106 Bytes
295accb
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
import numpy as np
import pandas as pd
import random
import warnings
warnings.filterwarnings('ignore')


def fill_demand_data(data, count_data, mode):
    for i in range(len(count_data)):
        Date = count_data.loc[i, 'Date']
        hour = count_data.loc[i, 'start_hour']
        cluster = count_data.loc[i, 'cluster']
        num = count_data.loc[i, 'count']
        if mode == 'rent':
            data.loc[(data['Date'] == Date) & (data['start_hour'] == hour), '{}_rent'.format(cluster)] = num
        else:
            data.loc[(data['Date'] == Date) & (data['start_hour'] == hour), '{}_return'.format(cluster)] = num
    return data


def fill_inventory(target_data, bike_inventory, battery_inventory):
    data = target_data.copy()
    data['penalty_1'] = 0
    data['penalty_0'] = 0
    for i in range(len(data)):
        if data["1_I_0"][i] < 0:
            for s in range(station_num):
                pure_demand = data["{}_rent".format(s)][i - 1] - data["{}_return".format(s)][i - 1]
                high_surplus = data["{}_I_2".format(s)][i - 1] + data["{}_dispatch_2".format(s)][i - 1] + \
                               data["{}_exchange_0".format(s)][i - 1] + data["{}_exchange_1".format(s)][i - 1]
                data.loc[i, "{}_I_2".format(s)] = max(0, high_surplus - pure_demand)  # -data["{}_pick".format(s)][i-1]
                low_satisfy = pure_demand - min(pure_demand, high_surplus)
                low_surplus = data["{}_I_1".format(s)][i - 1] + data["{}_dispatch_1".format(s)][i - 1] - \
                              data["{}_exchange_1".format(s)][i - 1]
                data.loc[i, "{}_I_1".format(s)] = max(0, low_surplus - low_satisfy)
                no_satisfy = low_satisfy - min(low_surplus, low_satisfy)
                data.loc[i, "{}_I_0".format(s)] = max(0,
                                                      data["{}_I_0".format(s)][i - 1] + data["{}_dispatch_0".format(s)][
                                                          i - 1] - data["{}_exchange_0".format(s)][i - 1])
                data.loc[i, "penalty_1"] = data.loc[i, "penalty_1"] + low_satisfy - no_satisfy
                data.loc[i, "penalty_0"] = data.loc[i, "penalty_0"] + no_satisfy
        else:
            for s in range(station_num):
                for e in soc:
                    data.loc[i, "{}_I_{}".format(s, e)] = int(data["{}_I_{}".format(s, e)][i] * random.random())
            car_bike_inventory = [0, 0, bike_inventory]
            car_battery_inventory = battery_inventory
        if data['start_hour'][i] == 14:
            for s in range(station_num):
                high_low = int(random.random() * data["{}_I_2".format(s)][i] * 0.5)
                data.loc[i, "{}_I_2".format(s)] = data["{}_I_2".format(s)][i] - high_low
                low_no = int(random.random() * data["{}_I_1".format(s)][i] * 0.5)
                data.loc[i, "{}_I_1".format(s)] = data["{}_I_1".format(s)][i] + high_low - low_no
                data.loc[i, "{}_I_0".format(s)] = data["{}_I_0".format(s)][i] + low_no
        # dispatch ebike
        for s in range(station_num):
            dock = dock_dict[s]
            station_inventory = [data["{}_I_{}".format(s, e)][i] for e in soc]
            for e in soc:
                num = random.randint(-5, 5)
                if num > 0:
                    real_num = min(car_bike_inventory[e], num, max(0, dock - sum(station_inventory)))
                else:
                    real_num = -min(-num, bike_capacity - car_bike_inventory[e], station_inventory[e])
                    # data.loc[i,"{}_pick".format(s)]=real_num
                    # car_inventory=car_inventory+real_num
                data.loc[i, "{}_dispatch_{}".format(s, e)] = real_num
                car_bike_inventory[e] = car_bike_inventory[e] - real_num
        # exchange battery
        for s in range(station_num):
            station_inventory = [data["{}_I_{}".format(s, e)][i]+data["{}_dispatch_{}".format(s, e)][i] for e in soc]
            for e in range(2):
                num = random.randint(0, 5)
                real_num = min(car_battery_inventory, num, station_inventory[e])
                #     real_num=-min(-num,battery_capacity-car_bike_inventory[e],station_inventory[e])
                #     # data.loc[i,"{}_pick".format(s)]=real_num
                #     # car_inventory=car_inventory+real_num
                data.loc[i, "{}_exchange_{}".format(s, e)] = real_num
                car_battery_inventory = car_battery_inventory - real_num
    return data


def fill_label(target_data):
    data = target_data.copy()
    data['label'] = 0
    for i in range(len(data)):
        label = 0
        for s in range(station_num):
            cost = sum(abs(data["{}_dispatch_{}".format(s, e)][i]) for e in soc) * bike_cost + sum(
                data["{}_exchange_{}".format(s, e)][i] for e in range(2)) * battery_cost + data["penalty_1"][
                       i] * penalty_1 + data["penalty_0"][i] * penalty_0
            # index=data["{}_I".format(s)][i]+data["{}_return".format(s)][i]+-data["{}_rent".format(s)][i]
            # # -data["{}_pick".format(s)][i]
            # if index<0:
            #     label=label+abs(index)
            # else:
            #     docks=dock_dict[s]
            #     label=label+max(0,index-docks)
            label = label + cost
        data.loc[i, 'label'] = label
    return data


def generate_train_data(origin_data, time=10):
    for t in range(time):
        data = origin_data.copy()
        grouped = data.groupby(['Date']).head(1)
        for i in range(station_num):
            num = dock_dict[i]
            grouped['{}_I_2'.format(i)] = num * 0.7
            grouped['{}_I_1'.format(i)] = num * 0.1
            grouped['{}_I_0'.format(i)] = num * 0.1
        for i in range(station_num):
            for e in soc:
                data.loc[grouped.index, '{}_I_{}'.format(i, e)] = grouped.loc[:, '{}_I_{}'.format(i, e)]

        bike_inventory = random.randint(int(bike_capacity* 0.7) , bike_capacity)
        battery_inventory = random.randint(int(battery_capacity* 0.7) , battery_capacity)
        data_1 = fill_inventory(data, bike_inventory, battery_inventory)
        data_2 = fill_label(data_1)
        if t == 0:
            target_data = data_2
        else:
            target_data = pd.concat([target_data, data_2], ignore_index=True)
    target_data.to_csv(r"ebike_data/data04_cluster_5.csv", index=False)
    return


if __name__ == '__main__':
    station_num = 5  #55
    bike_capacity = 60
    battery_capacity = 200
    random.seed(10)
    soc = [0, 1, 2]
    bike_cost = 1
    battery_cost = 1
    penalty_1 = 0.4
    penalty_0 = 0.8
    dock_dict = {}

    for i in range(station_num):#各个站点总停车位数量
        dock_dict[i] = int(50 * np.random.triangular(0.6, 0.8, 1))
    print(dock_dict)
    data = pd.read_csv(r"ebike_data/origin_data_4_cluster_5.csv")
    generate_train_data(data, time=1)  #要改文件名