高德实时路况数据获取接口说明：

link

高德数据获取流程

数据库设计

CREATE TABLE "public"."gaode_date" (

"name" varchar(50) COLLATE "default",

"status" int4,

"direction" varchar(50) COLLATE "default",

"angle" int4,

"lcodes" varchar(255) COLLATE "default",

"polyline" text COLLATE "default",

"id" int4 DEFAULT nextval('gaode_id_seq'::regclass) NOT NULL,

"geom" "public"."geometry",

"date" date,

"time" time(6),

"speed" float8,

CONSTRAINT "gaode_date_pkey" PRIMARY KEY ("id")

)

WITH (OIDS=FALSE)

;

ALTER TABLE "public"."gaode_date" OWNER TO "postgres";

COMMENT ON COLUMN "public"."gaode_date"."name" IS '道路名称';

COMMENT ON COLUMN "public"."gaode_date"."status" IS '路况 0：未知 1：畅通 2：缓行 3：拥堵';

COMMENT ON COLUMN "public"."gaode_date"."direction" IS '方向描述';

COMMENT ON COLUMN "public"."gaode_date"."angle" IS '车行角度，判断道路正反向使用。以正东方向为0度，逆时针方向为正，取值范围：[0,360]';

COMMENT ON COLUMN "public"."gaode_date"."lcodes" IS '方向';

COMMENT ON COLUMN "public"."gaode_date"."polyline" IS '道路坐标集，坐标集合';

COMMENT ON COLUMN "public"."gaode_date"."date" IS '请求日期';

COMMENT ON COLUMN "public"."gaode_date"."time" IS '请求时间';

COMMENT ON COLUMN "public"."gaode_date"."speed" IS '速度';

CREATE TRIGGER "beforeinsertinsertgaodedata_trigger" BEFORE INSERT ON "public"."gaode_date"

FOR EACH ROW

EXECUTE PROCEDURE "beforeinsertgaodedata"();

请求高德数据Python脚本

import requests

import json

import time

from pandas import DataFrame

import psycopg2

#请求高德实时路况数据

def getGaodeTrafficStatus(key,rectangle,currentTime):

insert_list = []

TrafficStatusUrl = 'http://restapi.amap.com/v3/traffic/status/rectangle?key='+key+'&rectangle='+rectangle+'&extensions=all';

res = requests.get(url=TrafficStatusUrl).content;

total_json = json.loads(res);

print(total_json);

jsondata = total_json['trafficinfo']['roads'];

print(jsondata);

currentDate = time.strftime("%Y-%m-%d", time.localtime());

if any(jsondata):

for i in jsondata:

name = i['name']

status = i['status']

direction = i['direction']

angle = i['angle']

speed = i.get('speed');

# 若速度参数缺失，补为Null

if speed == None:

speed = None;

print(speed);

lcodes = i['lcodes']

polyline = i['polyline']

list = [name, status, direction, angle, lcodes, polyline, currentDate, currentTime, speed];

insert_list.append(list);

conn = psycopg2.connect(database="superpower", user="postgres", password="123456", host="localhost", port="5432");

cur = conn.cursor();

for i in insert_list:

if len(i):

cur.execute(

"insert into gaode_date(name, status, direction, angle, lcodes, polyline, date, time,speed) values (%s,%s,%s,%s,%s,%s,%s,%s,%s)",

i)

conn.commit()

conn.close()

else:

pass

#注册多个高德key,轮流使用，并记录请求的个数，满2000次换下一个key继续请求。

keyList=[{}];

#建立覆盖研究区的rectangle数组

rectangleList=[];

#创建请求时间，确保一次请求的时间都相同

currentTime = time.strftime("%H:%M:%S", time.localtime()); # 设置请求时间

key='yourKey';

rectangle='120.12924,30.30185;120.16018,30.27150';

getGaodeTrafficStatus(key,rectangle，currentTime);

空间信息解析(在此采用触发器，对数据进行解析，插入数据的同时完成空间信息的解析，减少人工参与。)，触发器Sql脚本如下：

--高德数据插入触发器

CREATE OR REPLACE FUNCTION BeforeInsertGaodeData() RETURNS TRIGGER AS $example_table$

BEGIN

--自动进行坐标计算，并赋值

NEW.geom=ST_LineFromText('LINESTRING('||replace(replace(New.polyline,',',' '),';',',')||')',4326);

return new;

End ;

$example_table$ LANGUAGE plpgsql;

CREATE TRIGGER BeforeInsertInsertGaodeData_trigger Before INSERT ON gaode_date FOR EACH ROW EXECUTE PROCEDURE BeforeInsertGaodeData ();

数据查看，可视化获取的高德数据，如下(未正儿八经的进行渲染：)

image

爬取整个研究区域的实时路况数据

构建整个研究区域的网格，如下图，构建网格过程略。

image

网格数据库脚本

CREATE TABLE "public"."hangzhou_grid" (

"gid" int4 DEFAULT nextval('hangzhou_grid_gid_seq'::regclass) NOT NULL,

"id" numeric(10),

"__xmin" numeric,

"__xmax" numeric,

"ymin" numeric,

"ymax" numeric,

"geom" "public"."geometry",

CONSTRAINT "hangzhou_grid_pkey" PRIMARY KEY ("gid")

)

WITH (OIDS=FALSE)

;

ALTER TABLE "public"."hangzhou_grid" OWNER TO "postgres";

COMMENT ON COLUMN "public"."hangzhou_grid"."__xmin" IS '最小x';

COMMENT ON COLUMN "public"."hangzhou_grid"."__xmax" IS '最大x';

COMMENT ON COLUMN "public"."hangzhou_grid"."ymin" IS '最小y';

COMMENT ON COLUMN "public"."hangzhou_grid"."ymax" IS '最大y';

CREATE INDEX "hangzhou_grid_geom_idx" ON "public"."hangzhou_grid" USING gist ("geom");

3.请求整个杭州高德实时路况数据Python脚本

#请求整个杭州的区域

conn = psycopg2.connect(database="superpower", user="postgres", password="123456", host="localhost", port="5432");

cur = conn.cursor();

#获取覆盖研究区域的网格

cur.execute("select t.__xmin||','||t.ymin||';'||t.__xmax||','||t.ymax rectangle from hangzhou_grid t");

rectangleData=cur.fetchall();

cur.close;

conn.close();

#创建请求时间，确保一次请求的时间都相同

currentTime = time.strftime("%H:%M:%S", time.localtime()); # 设置请求时间

for rec in rectangleData:

print(rec[0]);

getGaodeTrafficStatus(key,rec[0],currentTime);

查看请求结果，如下图

image

总结

该方法基本能够将请求区域里的数据，进行覆盖。

获取的数据只是原始数据，若需要进行使用可能需要进行进一步处理，如将某时刻重复数据去除等。

高德API接口有请求次数的限制，因此要覆盖大多数时间，需要申请多个key,大致是60个key,目前不知道高德有没有限制一个IP地址访问次数。(后面持续改进方向：建立key表记录每个key一天访问的次数，动态分配key)

该方法也可以构建研究区域的主要道路，解放道路采集的人力。