出租车轨迹地图匹配实例
2024-05-16 09:48:00
好用的轨迹地图匹配框架可以参见：
使用graphhopper(map-matching)进行地图匹配_lei吼吼的博客-CSDN博客
在上面这个博客中介绍了一些轨迹地图匹配的框架，并且详细介绍了graphhoper中的子模块map-matching(轨迹地图匹配框架)的使用方法。下面介绍一个出租车轨迹地图匹配实例：
需要进行地图匹配的数据样式：
1,30.624806,104.136604,1,2014/8/3 21:18:46
1,30.624809,104.136612,1,2014/8/3 21:18:15
1,30.624811,104.136587,1,2014/8/3 21:20:17
1,30.624811,104.136596,1,2014/8/3 21:19:16
1,30.624811,104.136619,1,2014/8/3 21:17:44
1,30.624813,104.136589,1,2014/8/3 21:19:46
1,30.624815,104.136585,1,2014/8/3 21:21:18
1,30.624815,104.136587,1,2014/8/3 21:20:48
1,30.624815,104.136639,1,2014/8/3 21:17:14
1,30.624816,104.136569,1,2014/8/3 21:22:50
1,30.624816,104.136574,1,2014/8/3 21:22:19
1,30.624816,104.136577,1,2014/8/3 21:21:49
1,30.624818,104.136564,1,2014/8/3 21:23:20
1,30.624818,104.136621,1,2014/8/3 21:15:42数据说明：第一列：车辆ID 第二列：纬度 第三列：经度 第四列：是否载客 第五列：时间戳
使用的轨迹地图匹配框架：graphhoper.map-matching
具体是匹配方法参见：
使用graphhopper(map-matching)进行地图匹配_lei吼吼的博客-CSDN博客
但是使用那个地图匹配框架之前还需要做一些数据预处理，下面我将一一介绍：
一、数据排序
从数据样例就可以看出，数据不是按照时间先后排序的，但是根据常理来看轨迹数据应该是通过时间排序的，所以第一步将数据按照时间先后排序，排序代码如下：
# Author:lei吼吼
# -*- coding=utf-8 -*-
# @Time :2023/1/6 10:56
# @File: 排序.py
# @Software:PyCharm
# 这个是排序成都的那个数据集（那个数据集原来不是按照时间排序的）# 这个函数是利用选择排序对文件进行排序
# 选择排序就是每轮都选择最小的那个放在第一个位置，然后循环
def sort_txt():f = open('2.txt')lines = f.readlines()for i in range(len(lines)):min_idx = ifor j in range(i + 1, len(lines)):time_1 = lines[min_idx].split(',')[4]time_2 = lines[j].split(',')[4]t_1 = time_1.split(' ')[1].strip('\n')t_2 = time_2.split(' ')[1].strip('\n')if compare_big(t_1, t_2):min_idx = jlines[i], lines[min_idx] = lines[min_idx], lines[i]for i in range(len(lines)):with open("2_sort.txt", "a") as f:# 这里注意一下:不能写成了"w"，w每次都会覆盖前面写入的，a是每次都是追加f.write(lines[i])print("排序后的数组：")for i in range(len(lines)):print(lines[i])# 这个函数是比较下面这样的数据结构的
# t1 = '21:18:46'
# t2 = '21:18:15'
# print(compare_big(t1, t2))
# 如果t1>t2，那么就输出True
def compare_big(t1, t2):t1 = t1.split(':')t2 = t2.split(':')hour1 = int(t1[0])hour2 = int(t2[0])minute1 = int(t1[1])minute2 = int(t2[1])second1 = int(t1[2])second2 = int(t2[2])if hour1 > hour2:return Trueif hour1 == hour2:if minute1 > minute2:return Trueif minute1 == minute2:if second1 > second2:return Trueif second1 < second2:return Falseif minute1 < minute2:return Falseif hour1 < hour2:return Falsesort_txt()产生的文件样式如下：
2_sort.txt
1,30.654470,104.121588,0,2014/8/3 06:00:53
1,30.654470,104.121588,0,2014/8/3 06:01:53
1,30.654470,104.121588,0,2014/8/3 06:02:54
1,30.654470,104.121588,0,2014/8/3 06:03:54
1,30.654470,104.121588,0,2014/8/3 06:04:54
1,30.654470,104.121588,0,2014/8/3 06:05:55
1,30.654470,104.121588,0,2014/8/3 06:06:55
1,30.654470,104.121588,0,2014/8/3 06:07:55
1,30.654470,104.121588,0,2014/8/3 06:08:55
1,30.654470,104.121588,0,2014/8/3 06:09:56
1,30.654470,104.121588,0,2014/8/3 06:10:57
1,30.654470,104.121588,0,2014/8/3 06:11:57
1,30.654470,104.121588,0,2014/8/3 06:12:58二、数据简化
从数据样例中可以看出，车辆的经纬度数据是一分钟基本会有两个，这样的轨迹数据点太过于密集，数据量太大，使得后续的轨迹匹配难度加大。 轨迹数据可能在收集的时候会出现中断的情况，我们认为：前后数据时间相差30分钟的轨迹数据是不同的轨迹。意思就是这辆车如果他的这个轨迹点和下一个轨迹点的时间相差30分钟，那么这辆车的这两个轨迹点存在于这辆车的不同的轨迹路线中。现在进行轨迹数据简化。（注意：10分钟、30分钟都是可以自己设定的）
# Author:lei吼吼
# -*- coding=utf-8 -*-
# @Time :2023/1/7 15:37
# @File: 轨迹截断.py
# @Software:PyCharm
# 5分钟是一个点，但是30分钟就是另外一个轨迹了
from datetime import datetime
# 会生成两种文件，traj+num的是去除30分钟之外的数据的文件（并且已经分轨迹了）
# traj_simplify是简化的的轨迹文件（就是10分钟一个gpx点）# 函数含义：判断是否为新的轨迹段（超过30分钟）
# 输入:2014/8/3 10:56:00类似的时间格式
def time_out(t1, t2):time_1_struct = datetime.strptime(t1, "%Y/%m/%d %H:%M:%S")time_2_struct = datetime.strptime(t2, "%Y/%m/%d %H:%M:%S")seconds = (time_2_struct - time_1_struct).secondsif seconds > 60 * 30:return Falseelse:return True# 时间点太多了，进行时间点的筛选，10分钟为一个间隔
# 输入：文件的lines
def time_ten(lines):i = 0interval = []traj_lines = [lines[0]]while i + 31 < len(lines):# 在这个点的后面30个点中选择下一个点traj_line = traj_lines[-1]for item in range(i + 1, i + 31):interval.append(abs(time_interval(traj_line.split(',')[4].strip('\n'),lines[item].split(',')[4].strip('\n')) - 10 * 60))# print(interval)# index这里+1是序号是从0开始的，为了得到下一个点的序号就需要+1index = interval.index(min(interval)) + 1# print(index)interval.clear()i = i + indextraj_lines.append(lines[i])# print(traj_lines)return traj_lines# 计算两个时间点之间的时间差
# 输入两个时间，时间样式：2014/8/3 10:56:00
def time_interval(t1, t2):time_1_struct = datetime.strptime(t1, "%Y/%m/%d %H:%M:%S")time_2_struct = datetime.strptime(t2, "%Y/%m/%d %H:%M:%S")seconds = (time_2_struct - time_1_struct).secondsreturn secondsdef main():f = open('2_sort.txt')lines = f.readlines()# 进行是否是新轨迹的判断num = 1traj_lines = [] # 这个是存放每次在同一个轨迹的的lines# 解释一下这个for循环# 如果不超过30分钟就是一个轨迹的，所以放入记录同一个轨迹的列表(traj_lines)中,直到找到另一个轨迹# 创建文件，将上一个轨迹的所有点写入，然后将轨迹列表traj_lines清除，写入本轨迹的第一个点，然后进入循环找本轨迹的其他点for item in range(len(lines)):t1 = lines[item].split(',')[4].strip('\n')t2 = lines[item].split(',')[4].strip('\n')if time_out(t1, t2):traj_lines.append(lines[item])else:file_name = 'traj' + str(num) + '.txt'for i in range(len(traj_lines)):with open(file_name, "a") as f:# 这里注意一下:不能写成了"w"，w每次都会覆盖前面写入的，a是每次都是追加f.write(traj_lines[i])num = num + 1traj_lines.clear()traj_lines.append(lines[item])# 这里使用了traj_lines.clear()注意会不会出错# 写num==1这个if的原因是：如果只有一条轨迹那么就不会写入文件，所以要把这种情况单独列出来if num == 1:file_name = 'traj' + str(num) + '.txt'for i in range(len(traj_lines)):with open(file_name, "a") as f:# 这里注意一下:不能写成了"w"，w每次都会覆盖前面写入的，a是每次都是追加f.write(traj_lines[i])# 进行轨迹简化for j in range(1, num + 1):file_name = 'traj' + str(j) + '.txt'file = open(file_name)tr_lines = file.readlines()sim_lines = time_ten(tr_lines)file_name = 'traj_simplify' + str(j) + '.txt'for s in range(len(sim_lines)):with open(file_name, "a", encoding='utf-8') as f:# 这里注意一下:不能写成了"w"，w每次都会覆盖前面写入的，a是每次都是追加f.write(sim_lines[s])if __name__ == '__main__':main()这个代码会产生两种文件，如下：
traj1.txt
1,30.654470,104.121588,0,2014/8/3 06:00:53
1,30.654470,104.121588,0,2014/8/3 06:01:53
1,30.654470,104.121588,0,2014/8/3 06:02:54
1,30.654470,104.121588,0,2014/8/3 06:03:54
1,30.654470,104.121588,0,2014/8/3 06:04:54
1,30.654470,104.121588,0,2014/8/3 06:05:55
1,30.654470,104.121588,0,2014/8/3 06:06:55
1,30.654470,104.121588,0,2014/8/3 06:07:55
1,30.654470,104.121588,0,2014/8/3 06:08:55
1,30.654470,104.121588,0,2014/8/3 06:09:56
1,30.654470,104.121588,0,2014/8/3 06:10:57
1,30.654470,104.121588,0,2014/8/3 06:11:57
1,30.654470,104.121588,0,2014/8/3 06:12:58traj_simplify1.txt
1,30.654470,104.121588,0,2014/8/3 06:00:53
1,30.654470,104.121588,0,2014/8/3 06:10:57
1,30.654470,104.121588,0,2014/8/3 06:21:00
1,30.654470,104.121588,0,2014/8/3 06:31:02
1,30.654470,104.121588,0,2014/8/3 06:41:06
1,30.654470,104.121588,0,2014/8/3 06:51:09
1,30.654470,104.121588,0,2014/8/3 07:00:13
1,30.654673,104.121396,0,2014/8/3 07:10:17
1,30.654682,104.121483,0,2014/8/3 07:20:30
1,30.654632,104.121445,0,2014/8/3 07:30:43
1,30.654605,104.121440,0,2014/8/3 07:40:56
1,30.654580,104.121703,0,2014/8/3 07:51:10三、txt文件转化成gpx文件
因为我们会使用轨迹地图匹配框架：graphhoper.map-matching来进行轨迹地图匹配，但是graphhoper.map-matching的输入是gpx文件，我们现在文件是txt文件，所以我们将txt文件转化成gpx文件
# Author:lei吼吼
# -*- coding=utf-8 -*-
# @Time :2023/1/4 19:34
# @File: txt2gpx.py
# @Software:PyCharm
# 这个是将成都的数据集的txt文件转成gpx文件def csv2gpx(outfilepath):f = open('traj_simplify1.txt')lines = f.readlines()car_id = []time = []lat = [] # 纬度lon = [] # 经度for item in lines:line = item.split(',')car_id.append(line[0])t = line[4].split(' ')t_0 = t[0].split('/')t = t_0[0]+'-0'+t_0[1]+'-0'+t_0[2] + 'T' + t[1].strip('\n') + '+00:00'time.append(t)lon.append(line[2]) # 经度lat.append(line[1].strip('\n')) # 纬度outstring = '<?xml version="1.0" encoding="UTF-8" standalone="yes"?>\n'outstring += '<gpx xmlns="http://www.topografix.com/GPX/1/1" ' \'xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" ' \'xsi:schemaLocation="http://www.topografix.com/GPX/1/1 http://www.topografix.com/GPX/1/1/gpx.xsd">'outstring += '<trk>\n<trkseg>\n'for i in range(len(lines)):item = '<trkpt lat="' + str(lat[i]) + '" lon="' + str(lon[i]) + '"><time>' + str(time[i]) + '</time></trkpt>'outstring = outstring + item + "\n"outstring += '</trkseg>\n</trk>\n</gpx>'fw = open(outfilepath, 'w')fw.write(outstring)fw.close()# ------------------------------------------Test----------------------------------
def forfolders():csv2gpx('traj_sim_gpx.gpx')forfolders()转化后的文件：
<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<gpx xmlns="http://www.topografix.com/GPX/1/1" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.topografix.com/GPX/1/1 http://www.topografix.com/GPX/1/1/gpx.xsd"><trk>
<trkseg>
<trkpt lat="30.654470" lon="104.121588"><time>2014-08-03T06:00:53+00:00</time></trkpt>
<trkpt lat="30.654470" lon="104.121588"><time>2014-08-03T06:10:57+00:00</time></trkpt>
<trkpt lat="30.654470" lon="104.121588"><time>2014-08-03T06:21:00+00:00</time></trkpt>
<trkpt lat="30.654470" lon="104.121588"><time>2014-08-03T06:31:02+00:00</time></trkpt>
<trkpt lat="30.654470" lon="104.121588"><time>2014-08-03T06:41:06+00:00</time></trkpt>
<trkpt lat="30.654470" lon="104.121588"><time>2014-08-03T06:51:09+00:00</time></trkpt>
<trkpt lat="30.654470" lon="104.121588"><time>2014-08-03T07:00:13+00:00</time></trkpt>
<trkpt lat="30.654673" lon="104.121396"><time>2014-08-03T07:10:17+00:00</time></trkpt>
<trkpt lat="30.654682" lon="104.121483"><time>2014-08-03T07:20:30+00:00</time></trkpt>
<trkpt lat="30.654632" lon="104.121445"><time>2014-08-03T07:30:43+00:00</time></trkpt>
<trkpt lat="30.654605" lon="104.121440"><time>2014-08-03T07:40:56+00:00</time></trkpt>
<trkpt lat="30.654580" lon="104.121703"><time>2014-08-03T07:51:10+00:00</time></trkpt>
<trkpt lat="30.654332" lon="104.122103"><time>2014-08-03T08:01:23+00:00</time></trkpt>
</trkseg>
</trk>四、进行轨迹匹配
使用graphhopper(map-matching)进行地图匹配_lei吼吼的博客-CSDN博客
匹配之后的文件
<?xml version="1.0" encoding="UTF-8" standalone="no" ?><gpx xmlns="http://www.topografix.com/GPX/1/1" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" creator="Graphhopper version 1f5419977ed205d48d5507b65fae231592e52c5b" version="1.1" xmlns:gh="https://graphhopper.com/public/schema/gpx/1.1">
<metadata><copyright author="OpenStreetMap contributors"/><link href="http://graphhopper.com"><text>GraphHopper GPX</text></link><time>2014-08-03T06:00:53Z</time></metadata>
<trk><name></name><trkseg>
<trkpt lat="30.653076" lon="104.121415"><time>2014-08-03T06:00:53Z</time></trkpt>
<trkpt lat="30.652163" lon="104.122879"></trkpt>
<trkpt lat="30.651934" lon="104.123171"></trkpt>
<trkpt lat="30.651732" lon="104.123339"><time>2014-08-03T06:01:07Z</time></trkpt>
<trkpt lat="30.65445" lon="104.123867"><time>2014-08-03T06:01:29Z</time></trkpt>
<trkpt lat="30.651732" lon="104.123339"><time>2014-08-03T06:01:51Z</time></trkpt>
<trkpt lat="30.651418" lon="104.123557"></trkpt>
<trkpt lat="30.650851" lon="104.124216"></trkpt>
<trkpt lat="30.650517" lon="104.124884"></trkpt>
<trkpt lat="30.650316" lon="104.125551"></trkpt>
<trkpt lat="30.650209" lon="104.12604"><time>2014-08-03T06:02:10Z</time></trkpt>
<trkpt lat="30.650316" lon="104.125551"></trkpt>
</trkseg>
</trk>
</gpx>五、将匹配之后的估计经纬度点与地图道路ID匹配
因为项目需要的数据格式是需要轨迹点转成道路点的，项目需要的数据样式
所以进行这一步（下面的代码中使用了高德地图的API，具体方法请参见通过 经纬度 获取 地理位置（Python、高德地图）_Oriental_1024的博客-CSDN博客_根据经纬度坐标获取地理信息python）
# Author:lei吼吼
# -*- coding=utf-8 -*-
# @Time :2023/1/8 15:45
# @File: gps转换.py
# @Software:PyCharm
# 这个文件是将gpx文件中的经纬度和道路id匹配，并生成最终结果文件
import requests
import gpxpy
import pandas as pd# 调用高德地图API，获取道路ID
def geocode(location):# 参数内容 可以写成字典的格式parameters = {'output': 'json', 'key': '你自己的高德的key', 'location': location,'extensions': 'all'}# 问号以前的内容base = 'http://restapi.amap.com/v3/geocode/regeo'response = requests.get(base, parameters)print('HTTP 请求的状态: %s' % response.status_code)return response.json()if __name__ == '__main__':# 主函数：获取道路ID 并生成最终结果文件id_line = []with open('traj_sim_gpx.gpx.res.gpx') as fh:gpx_file = gpxpy.parse(fh)segment = gpx_file.tracks[0].segments[0]# print(segment)coords = pd.DataFrame([{'lat': p.latitude,'lon': p.longitude,'time': p.time} for p in segment.points])# 下次要转换新的数据的时候就需要把这部分的代码激活# print(coords)# for i in range(len(coords)):# loc = str(coords['lon'][i]) + ',' + str(coords['lat'][i])# data = geocode(loc) # 获取的数据类型为dict# formatted_address = data['regeocode']['roads']# id_line.append(formatted_address[0]['id'])# print(id_line)# 因为每天的次数有限，所以只能把ID打印出来id_line=['028H48F017017604156', '028H48F017017604140', '028H48F017017604140', '028H48F01701724217', '028H48F01701724217', '028H48F01701724217', '028H48F01701724217', '028H48F0170173773', '028H48F0170173773', '028H48F01701723284', '028H48F017017683271', '028H48F01701723284', '028H48F0170173773', '028H48F0170173773', '028H48F01701724217', '028H48F01701724217', '028H48F01701724217', '028H48F0170175400', '028H48F01701722789', '028H48F01701722789', '028H48F01701722789', '028H48F01701722789', '028H48F01701722789', '028H48F01701722789', '028H48F017017659205', '028H48F017017659205', '028H48F016017659713', '028H48F016017659713', '028H48F016017605665', '028H48F016017605655', '028H48F01601720164', '028H48F01601720164', '028H48F0160173202', '028H48F0160173202', '028H48F0160173202', '028H48F0160173202', '028H48F0160173202', '028H48F0160173202', '028H48F0160173202', '028H48F0160173202', '028H48F01601727622', '028H48F01601727622', '028H48F016017603368', '028H48F016017603372', '028H48F016017603372', '028H48F016017632442', '028H48F016017604663', '028H48F0160173570', '028H48F0160173570', '028H48F0160173570', '028H48F01601740938', '028H48F016017725117', '028H48F016017725117', '028H48F016017725117', '028H48F01601740938', '028H48F0160173193', '028H48F01601740938', '028H48F0160173570', '028H48F0160173570', '028H48F0160173570', '028H48F016017604663', '028H48F016017632442', '028H48F016017632442', '028H48F0160173818', '028H48F01601724728', '028H48F016017691288', '028H48F016017691288', '028H48F016017691288', '028H48F016017691288', '028H48F016017691288', '028H48F016017691288', '028H48F0160172992', '028H48F016017705045', '028H48F01601718188', '028H48F01601718188', '028H48F01601718188', '028H48F01601718188', '028H48F016017705045', '028H48F016017705045', '028H48F016017691288', '028H48F0160173570', '028H48F016017630970', '028H48F017017659850', '028H48F0170174018', '028H48F0170173149', '028H48F017017603944', '028H48F017017603944', '028H48F017017603944', '028H48F017017749262', '028H48F017017749262', '028H48F0170174018', '028H48F0170174018', '028H48F017017659850', '028H48F0160174170', '028H48F0160174170', '028H48F016017630970', '028H48F0170175388', '028H48F0170175388', '028H48F01601724231', '028H48F0160173570', '028H48F0160173570', '028H48F016017604663', '028H48F016017632442', '028H48F016017603372', '028H48F016017632442', '028H48F0160173818', '028H48F01601724728', '028H48F016017691288', '028H48F016017691288', '028H48F016017691288', '028H48F016017691288', '028H48F016017691288', '028H48F016017691288', '028H48F0160172992', '028H48F016017705045', '028H48F01601718188', '028H48F01601718188', '028H48F01601718188', '028H48F01601718188', '028H48F016017705045', '028H48F016017705045', '028H48F016017691288', '028H48F0160173570', '028H48F016017630970', '028H48F017017659850', '028H48F0170174018', '028H48F0170173149', '028H48F017017603944', '028H48F017017603944', '028H48F017017603944', '028H48F017017749262', '028H48F017017749262', '028H48F0170174018', '028H48F0170174018', '028H48F017017659850', '028H48F0160174170', '028H48F0160174170', '028H48F0160174170', '028H48F0170175388', '028H48F0170175388', '028H48F0170175388', '028H48F0160174170', '028H48F0160174170', '028H48F016017630970', '028H48F0170175388', '028H48F0170175388', '028H48F01601724231', '028H48F0160173570', '028H48F0160173570', '028H48F016017604663', '028H48F016017632442', '028H48F016017632442', '028H48F0160173818', '028H48F01601724728', '028H48F016017691288', '028H48F016017691288', '028H48F016017691288', '028H48F016017691288', '028H48F016017691288', '028H48F016017691288', '028H48F0160172992', '028H48F016017705045', '028H48F01601718188', '028H48F01601718188', '028H48F01601718188', '028H48F01601718188', '028H48F016017705045', '028H48F016017705045', '028H48F016017691288', '028H48F0160173570', '028H48F016017630970', '028H48F017017659850', '028H48F0170174018', '028H48F017017603951', '028H48F0170174062', '028H48F017017639450', '028H48F0170173951', '028H48F017017639450', '028H48F017017639450', '028H48F017017643823', '028H48F017017643823', '028H48F0170178609', '028H48F0170178609', '028H48F017017636405', '028H48F017017636405', '028H48F0170174734', '028H48F0170174734', '028H48F017017603975', '028H48F017017603975', '028H48F017017603975', '028H48F017017613960', '028H48F017017613960', '028H48F01701724221', '028H48F017017604156', '028H48F017017604140', '028H48F017017604140', '028H48F01701724217', '028H48F01701724217', '028H48F01701724217', '028H48F01701724217', '028H48F01701724217', '028H48F01701724217', '028H48F0170173773', '028H48F0170173773', '028H48F01701723284', '028H48F017017683271', '028H48F017018604509', '028H48F017018604509', '028H48F017018622514', '028H48F017018630208', '028H48F017018630208', '028H48F0170181447', '028H48F017018615240', '028H48F017018144674', '028H48F017018144674', '028H48F0170187258', '028H48F017018601859', '028H48F017018144668', '028H48F0170187298', '028H48F017018144668', '028H48F017018601859', '028H48F0170187258', '028H48F017018144674', '028H48F017018144674', '028H48F017018604561', '028H48F017018604561', '028H48F0170186955', '028H48F0170189475', '028H48F017018361', '028H48F0170189475', '028H48F0170186971', '028H48F0170189475', '028H48F0170186971', '028H48F0170189475', '028H48F017018144992', '028H48F017018601309', '028H48F0170189475', '028H48F017018612203', '028H48F017018612203', '028H48F017018612203', '028H48F0170187523', '028H48F017018612203', '028H48F017018612203', '028H48F017018612203', '028H48F0170181447', '028H48F017018612972', '028H48F017018612968', '028H48F0170184796', '028H48F017018615263', '028H48F017018686', '028H48F017018686', '028H48F0170184282', '028H48F0170184181', '028H48F01701811127', '028H48F017018630208', '028H48F0170181447', '028H48F017018615240', '028H48F017018144674', '028H48F017018604561', '028H48F0170187057', '028H48F017018604561', '028H48F017018604561', '028H48F017018604561', '028H48F017018144674', '028H48F017018144674', '028H48F017018615240', '028H48F0170181447', '028H48F017018630208', '028H48F017018630208', '028H48F017018622514', '028H48F017018604509', '028H48F017018604509', '028H48F017017683271', '028H48F01701723284', '028H48F0170173773', '028H48F0170173773', '028H48F01701724217', '028H48F01701724217', '028H48F017017604140', '028H48F017017604140', '028H48F017017604156', '028H48F01701724221', '028H48F017017613960', '028H48F017017613960', '028H48F017017749278', '028H48F017017603975', '028H48F017017603975', '028H48F017017603975', '028H48F017017603975', '028H48F017017603975', '028H48F017017603975', '028H48F017017603975', '028H48F017017603975', '028H48F017017603975', '028H48F0170173149', '028H48F017017749262', '028H48F017017749262', '028H48F0170174018', '028H48F0170174018', '028H48F017017659850', '028H48F0160174170', '028H48F0160174170', '028H48F016017630970', '028H48F0170175388', '028H48F0170175388', '028H48F01601724231', '028H48F0160173570', '028H48F0160173570', '028H48F016017604663', '028H48F016017632442', '028H48F016017632442', '028H48F0160173818', '028H48F01601724728', '028H48F016017691288', '028H48F016017691288', '028H48F016017691288', '028H48F016017691288', '028H48F016017691288', '028H48F016017691288', '028H48F0160172992', '028H48F016017705045', '028H48F01601718188', '028H48F01601718188', '028H48F0160172992', '028H48F0160172992', '028H48F0160172992', '028H48F0160172992', '028H48F01601724728', '028H48F01601724728', '028H48F0160172992', '028H48F0160172992', '028H48F01601744821', '028H48F0160174864', '028H48F0160174864', '028H48F01601723244', '028H48F01601723244', '028H48F0160172995', '028H48F01601713787', '028H48F01601713787', '028H48F01601720302', '028H48F016017646111', '028H48F016017260', '028H48F016017260', '028H48F016017260', '028H48F01601741240', '028H48F0160171665', '028H48F0160171665', '028H48F016017624172', '028H48F0160171665', '028H48F0160171665', '028H48F0160171665', '028H48F016017574', '028H48F016017983', '028H48F016017983', '028H48F01601743578', '028H48F01601743578', '028H48F01601743578', '028H48F016017590', '028H48F016017590', '028H48F01601722496', '028H48F01601722496', '028H48F01601743416']# file_line = []file_lines = []for i in range(len(coords)):file_line = '1'+','+'1'+','+str(coords['time'][i]).strip('+00:00')+','+str(id_line[i])# file_line 前面两个要根据具体的变化的# time中带有gpx文件time的样子，现在要变回来file_lines.append(file_line)# print(file_lines)with open('sample.txt','a') as f:f.write('vehicle_id,trajectory_id,time,road_id')# 车辆ID 轨迹ID 时间戳 道路IDf.write('\n')for i in range(len(file_lines)):with open('sample.txt','a') as f:f.write(file_lines[i])f.write('\n')结果展示：
vehicle_id,trajectory_id,time,road_id
1,1,2014-08-03 06:00:53,028H48F017017604156
1,1,NaT,028H48F017017604140
1,1,NaT,028H48F017017604140
1,1,2014-08-03 06:01:07,028H48F01701724217
1,1,2014-08-03 06:01:29,028H48F01701724217
1,1,2014-08-03 06:01:51,028H48F01701724217
1,1,NaT,028H48F01701724217
1,1,NaT,028H48F0170173773
1,1,NaT,028H48F0170173773
1,1,NaT,028H48F01701723284
1,1,2014-08-03 06:02:1,028H48F017017683271
1,1,NaT,028H48F01701723284
1,1,NaT,028H48F0170173773
1,1,NaT,028H48F0170173773
1,1,NaT,028H48F01701724217
1,1,2014-08-03 06:02:29,028H48F01701724217出租车轨迹地图匹配实例相关推荐
- 地图匹配实例-几何匹配
网上很多说隐马尔可夫模型的地图匹配,但是说来说去全是理论性质的,形象的贴一点代码,让你云里雾里,还有网上说的graphhopper-mapmatching项目,实际无法运行.我在写论文之初也是冥思苦想 ...
- Mysql出租车轨迹的分析_一种基于出租车轨迹数据的交通出行共现现象的可视化分析方法与流程...
本发明涉及交通数据可视化领域,尤其涉及一种基于出租车轨迹数据的交通出行共现现象的可视化分析方法. 背景技术: :城市中人类移动规律中的共现现象是指来自两个区域的人们在同一个时间段内访问同一个城市区域. ...
- 出租车轨迹点地图匹配研究
背景 历经一年时间,从一个对地理信息一无所知的小白到入门的过程,在今天晚上,终于把轨迹点初步匹配到地图上了. 关于地图匹配,已经是一门成熟的技术了,但是网上对地图匹配的开源还是比较少.Github上面 ...
- 【地图匹配（ST-matching）】GPS 轨迹数据预处理——T-Driver数据集【持续更新中】
一.数据介绍 1.数据链接 https://download.csdn.net/download/skyman30/87432266?spm=1001.2014.3001.5503https://do ...
- 经纬度画轨迹图_空气质量、实时路况、出租车轨迹，城市中海量数据如何高效管理？...
原创 JUST团队-何华均 JUST时空数据 一.问题背景 城市中超过80%的数据都与时空有关,如加油站点.出租车轨迹.交通路况等.这些数据多为半结构化和非结构化数据,并且需要管理的数据量巨大. 传统 ...
- spark如何进行聚类可视化_基于Spark的出租车轨迹处理与可视化平台
由于城市化进程加剧以及汽车数量增加, 城市交通问题日益严重[, 通过分析各种空间数据解决交通问题是当前研究的热点. 出租车提供广泛且灵活的交通运输服务, 是城市交通的重要组成部分. 出租车轨迹数据记录 ...
- 浙江农林大学python期末考试_基于Python的地图匹配最短路径法实现
117 科学论坛 在智能交通领域,众所周知,智能 交通系统在当今世界道路交通网络的管 理中发挥着重要的作用,而车辆导航系 统又是智能交通系统的重要组成部分. 在车辆导航系统中,定位的精确性和实 时性是 ...
- 地图距离算法_基于权重的地图匹配技术
文章目录 目录 地图匹配技术概览 文章目录 前言 一.地图匹配概述 二.轨迹预处理 1.降噪 1.1 中值滤波 1.2 极值滤波 1.3 分段 1.4 插值算法 三.基于权重的地图匹配算法 3.1.计 ...
- python地图匹配_基于隐马尔科夫模型(HMM)的地图匹配(Map-Matching)算法
1. 摘要 本篇博客简单介绍下用隐马尔科夫模型(Hidden Markov Model, HMM)来解决地图匹配(Map-Matching)问题.转载请注明网址. 2. Map-Matching(MM ...
最新文章
- cesium加载百度地图_四大Webgis地图框架的对比选择
- python取均匀不重复的随机数
- heartbeat+drbd+mysql:实现最廉价的高可用组合
- 这位电子工程师，你不能错过。
- ASP.NET Core Blazor Webassembly 之 数据绑定
- 前端学习（3062）：vue+element今日头条管理-总页码处理
- php中学院与专业的级联,SQLite中使用CTE巧解多级分类的级联查询
- 刘强东宣布向瑞士捐赠160万只口罩及其他大量急需医疗物资
- c语言如果调用的函数没有return,C语言中，整型函数若在调用时未获得return值，此时其值是如何确定的？...
- 取消UltraEdit提示“文件可能不是DOS格式”
- 公司面试自我介绍个人简历PPT模板
- java super extend 区别_Java泛型中extends和super的区别
- 词云可视化——四行代码轻松上手
- 黑马程序员中的简单网页制作
- J.P. Morgan：AI for Investing（脱水解读）
- 设计模式学习（十一）：Builder建造者模式
- 2022-2028年全球与中国零售会计软件行业发展趋势及竞争策略研究
- 我们需要自由——献给所有大学生的一封信
- 电子版寸照可以在手机上修改底色吗
- 赛迪全球科技产业动态周报（第二十五期）发布！