201000934 九、發明說明： 【發明所屬之技術領域】 本發明係關於一種以時空環境資料探勘技術偵測交通瓶頸點的方去 特別是指利用裝置在於車輛之智慧型車上單元與後端監控中心之通訊資 料’利用時空環境資料探勘技術（spatiotemporal data mining)、次 貝料叢集 (clustering)、地理資訊系統（GIS)的技術，找出都會區路網中六 塞 不同啟發柄方法分析各 的模式（traffic congestion patterns)，並利用三種 交通壅塞模式之間的關連性，然後進一步找出都會區路網中各種不门時办 環境之下的交通瓶頸點。 【先前技術】 交通瓶頸點為交通路網中特別容易引發壅塞的原點，若能夠找出交通 瓶頸點並加以適當的控制與疏導措施，則對交通的順暢會有彳艮大的幫助. 傳統的交通瓶頸點是依照運輸管理上瓶頸點的定義而來，認為對路網物件 (路段或路口）的需求（流量）高於該路段所能允許的容量，則該物件則 為一瓶頸點；但是此種定義很難應用到實際的都會區路網上，通常是因為 有下列原因： 1. 都會區路網太過複雜，路網物件通常會交互影響，並非各自獨立。 2. 路網物件的流量與容量資訊很難取得，且流量資訊是隨時間 化。 3. 都會區的路網瓶頸點並非固定不變’而是跟整體的交通起訖需求 有相關（OD demand)，舉例而言，上班時段的瓶頸點與下班時段 201000934 的瓶頸點有所不同，上班日的交通瓶頸點又細於假日的交通 瓶頸點。 4.各交通路網物件之箱訊（流量）收集困難，需要針對整個路 網大量佈建_器方可取得；目前在收鱗時交通資訊的方法有 >種在各個重要地段或疋路口架設交通偵測感應器或是監控攝 影機（，歧採_探_各觀段_訊(簡 稱探針車），但是這_方法各有其缺點： 1) 涵盍性不足（定點式、探針車），定點式的偵測器通常只安 裝在路口或是重要職，核置侧㈣地點無法得知其交 通路況；探針車則受限於車她量，雖可以轉的收集行經 路段的交«訊，但整__涵蓋面不足，探針車也無法 全年無休的進行交通資訊收集，故其時間的涵蓋面也不足。 2) 樣本數科（探針車），路段的交通歧會縣時間、氣候 或者人為活絲林同，但因峨探測車數量有限，一部探 針車在同-B夺間也只能在一個路段進行資料冤集，所以整個 交通路網收集的樣本數不足。 3)成本高（定‘點式、探針車），佈建交通侧感應器建設需要 不少的經費’路口攝影機的及時監控與感應器的維護更是需 要大量人力；探針車需要精密設備並配置駕駛與操作人員， 其才呆作成本偏南。 由此可見’上述習用方式仍有諸多缺失，實非一良善之設計，而亟待 201000934 加以改良 本案發明人鑑於上述制方式所衍生的各項缺點，乃歪思加以改良創 新’並經多年苦心職潛心研究後，終於成功研發完成本件以時空環境資 料探勘技術偵測交通瓶頸點的方法。 【發明内容】 本發明之目_在於提供—種鱗❸料觸技術侧交通瓶頸 點的方法’用以辅助並取代現有定點式與探針錢集交通資訊的方法，藉 由為晴職獅㈣州轉_，咖謝心的通 赠料，^過資料探勘與電子地_技術，分析出免f且大量的即時交通 資訊取代昂貝且稀少的交通資訊探測車； ，本發明之次要目的即在於提供—種以時空環境㈣探勘技賴測交通 賴點的方法’本發日歧行分析車輛_後端馳k之，結合電子 地圖來找出時啦間維度（各時段、各地點）的交職型，進—步分析找 點’解決行動通訊探測車數量不足、涵蓋性不足及樣本數不足 的問題。 疋 目的即在於提供一種以時峨細勘技術偵測交通 ▲、、’本發明可簡用在都會㈣交通控财心、，提供預測結果 ’ 口予父控中心管'3®昌丨、，a 广M、g ㈣作為進—步交通管理調控，或是提供給用路人作為 旅仃路m是旅行運具選擇之參考。 法，係利㈣2目㈣_嫩臟輸舰軸點的方 、3 〃、交通資訊產生、交通壅塞模式資料探勘及交通瓶頸 201000934 麟勘等方法，射該資驗錢交通f訊赵針對由魅粒車隊派遣 官理系統單元（Location_basedService，LBS)所回傳的經緯度、車速、方向 等資料’進行資訊收集與交通資訊產生模組收集及整理，並配合電子地圖 單元將該等資㈣換鱗段錢資訊，並依據各路段的交通速限統計各路 段、各時段的交通狀態，劃分為五個等級，其中等級！代表正常不奎塞， 等級5代表非常壅塞； 3本發明除了產生各時空環境之下、都會區路網之交通資訊外，並將衛 星定位車隊派遣管理系統單元中喊料轉換為車她次資料，其中包含車 輛旅次起迄點、旅行時間、經過點座標、速度與時間等詳細旅次資訊，此 資訊在下一階段找出交通壅塞模式時使用； 其中該父通壅塞模式資料探勘係利用—交通麥塞模式資料探勘模組找 出在時级塞區域之_模式（pattems);本模財定義了兩個交通奎 塞模式，分狀 1塞翻試（CGngestic)n c_quent Pattem, ccp)與產 (congestion drop downstream pattern, CDP)，CCP 因為個時工父通壅塞區域導致另一個時空交通壅塞區域；奎塞流差模式 則為《物件與下游物件的慶塞降，若降幅過大絲通過該壅塞的物 件之後其下游物件都不會壅塞，_物件可以欺為瓶頸點； 另外該交通瓶頸點探勘係利用一交通瓶頸點探勘模組以找出交通瓶頸 點’本模組定義了三種找出交通㈣點的方法，分般壅塞擴散法、產塞 M：法、流差比法’並可針對交通擁魏域_產出的交馳顧域規則， 某又通擁塞H域丨贴規貞彳的條件科親服魏，則將此區域 201000934 視為瓶頸區域，且此區域内的所有交通擁塞點視為交通瓶頸點候選人，最 後以擁塞信心指數以及擁塞程度來找出真正的且可依照時段之交通瓶頸 點。 【實施方式】 請參閱圖-，為本發明以時空環境資料探勘技術偵測交通瓶頸點的方 法之應用系統架構圖，由圖中可知，該系統架構用來提供資料來源之主要 元件包括： -智慧型車上單70 1，該智慧型車上單元i設置於車上，並與後端系 統10之間的通訊記錄都會記錄在資料庫中，此資料庫資訊可以透過本發明 再加利用，找出都會區網路中的交通瓶頸點； 後％系統1G ’該後端系統1Q經由智慧型車上單元1回報給後端系 統10之資财魅定錄置、車減態、行進方向及速度等，喃的時機 則刀為疋時回報、事件回報及被動回報，而回報所發生的地點稱之為回報 占因此根據回報時的車輛狀態資訊及地理資訊系統的道路路網資訊，我 V回報點田時所在路段的交通資訊，用以判斷當下該路段是否處於 交通阻塞狀態。 ' •叫參_二’為本發明鱗空環境資料娜技術制交通瓶頸點的方 法之曰慧型車上早元元件示意圖，由财可知，該智慧型車上單元之主要 元件有： -——jit- » 广- 钌疋位杈組8，該GPS衛星定位模組8可以透過全球衛星定 位系統GPS得到車輛目前所在地點的位置座標，而在無法接收GPS訊號的 201000934 區域，也紐過行動通信基地台定财式，推估車輛所在地之座標； —GPR_TS行動通訊模組9 ’該Gprs_s行動通訊模組9則 為智慧型車上單元與後端系統通訊的界面，透過行動通訊祕與後端通訊 進行資料回報與車隊派遣運作。 請參閱圖三’為本發明以時空環境資料探勘技術制交通瓶頸點的方 法之系統模域巾可知，該魏模鸿顯巾主要包括： -資訊收触交«喊生·4，„訊轉敏通資減生模… 收集衛星故轉轉管理祕單元2讀料，迦（去除祕、異常資 枓）後，透過GIS電子地圖單元3將電子地圖的經緯度轉換為地址，並配 合車速、方轉轉換為路段交訊，且依交通速限進行麟各路段、 各時段的交通狀態’賴分為五轉級，其巾等級丨代表正常轉塞，等 ^ 5代表非巾t塞’另外本触除了產生不同時空觀下的都會區路網之 乂通貝訊外，並將衛星定位車隊派遣管理系統單元中的資料轉換為車輛旅 次資料’ XC些資料包含車輛旅次起迄點、旅行時間、經過點座標、速度與 時間等詳細旅絲訊’這錄訊可提供錢壅塞模式·獅模組5以找 出交通壅塞模式。 父通壅塞模式資料探勘模組5，該交通壅塞模式資料探勘模組5係 根據貝訊收集與交通資訊產生触*所收集整理的資料，而找出在時空環 兄奎塞區域之_模式（Patefns);其方法共分為兩個步驟： 第—個步驟是將奎塞的物件（交通狀態等級為4或5的物件），依照地 理相鄰的關鱗#絲做通MM (spatiotempoml _gesti()n _ 201000934 STCA)； 第一個步驟則是找出g塞區域之間的關係，利用時空關聯法則找出每 兩兩壅塞區域的規則，如果規則中的兩個皇塞區域重複旅次比例高於門挺 值’則將這1¾彳睹塞區域列為有關聯；在本模财定義了_交通查塞模 式刀另1疋壅塞遞延模式（c〇ngesti〇n c〇nsequent pattem. Cep )與壅塞流差 模式（congestion drop downstream pattern，CDP)，CCP 模式之主要成因為 _ 個時空父通壅塞區域導致另—個時空交通蓬塞區域；.喊塞流差模式則為 蜜塞物件與下游物件的壅塞降幅比，若降幅過絲示通過職塞的物件之 後其下游物件都不會壅塞，則該物件可以判定為瓶頸點； 一父通瓶頸點探勘模組6,該交通瓶頸點探勘模組6係根據交通壅塞 杈式責料探勘模組5的兩種交通麥塞模式定義了三觀出交通瓶頸點的方 去刀另J疋.奎塞擴散法、g塞收敛法、流差比法；其中蜜塞擴散法係根 據產塞遞延模式，找出一個交通奎塞區域（STCA)會引發多個其他交通蹇 塞區域（STCA)’貞彳推論齡謎可能落在此交sg塞區域0的路網物件； 而壅塞收傲法則是相反推論，多個STCA都指到了同一個stca，則觀頸點 可能落在存在該STCA的路_件上；斜流差比關是根顧塞流差模 式’找出壅塞流差比超過門檻值的物件（上游物件之奎絲度與下游物件 之壅塞程度差異過大）’根據其的信心、指數（該物件發生餘狀況的發 生比率）與蜜塞程度來判定是否為交通瓶頸點，· -父通_點7 ’該交職獅7係將交通瓶賴探勘模組6找出的 依照時段之交通瓶頸點顯示於電子地圖上。 11 201000934 請參閱圖四，為本發明以時空環境資料探勘技術偵測交通瓶頸點的方 法之探勘偵測交通瓶頸點演算流程圖，由圖中可知，其探勘镇測交通瓶頸 點之主要步驟流程為· 步驟1 :資料收集與資料整理，係從衛星定位車隊派遣管理系統單元 收集資料，並做該資料之整理401 ; 步驟2:交通資料轉換及交通旅次資料彙整，配合電子地圖服務系統 將步驟1之資料轉換為交通資料及交通旅次資料，並儲存於時空交通狀態 資料庫402 ; 步驟3 :交通壅塞點分析，利用時空交通狀態資料庫進行交通壅塞點 分析，並產生時空交通壅塞物件403 ; 步驟4·父通壅塞點叢集演算，將壅塞的物件依照地理相鄰的關係群 集成數個時空交通壅塞區域404 ; 步驟5 :交通麥塞模式資料探勘，根據時空交通區域及其關聯性 定義了兩個交通麥塞模式’分別是壅塞遞延模式與奎塞流差模式4〇5 ; 步驟6 :交通瓶驅域分析，根據步驟5所定義之模式進行奎塞擴散 法、壅塞收斂法、流差比法之分析4〇6 ; 步驟7:判斷交通物件的蓬塞程度及信心指數，若超過門播值，則將 父通壅塞區域或交通瓶頸點顯示在電子地圖上4〇7。 請參閱圖五，為本發明以時空環境資料探勘技術偵測交通綱點的方 法之交通擁紐域叢集演算法，該演算法實現了圖四巾第二階段的交通擁 塞點叢集演算法，此演算法之運算係將上__步财所找出的時空交通擁塞 12 201000934 物件以地理相鄰的關係找出附近同樣是擁 區域，此演算法之步驟流程為： 路網上所制到的擁塞物件點，依擁塞程 的物件，將同一時間平面的擁塞 塞的物件並叢集成同一時空擁塞 步驟1:針對每一段時間， 度排序成一集合：{〇j5〇i; 步驟2 :如嶽中尚有物件，取繼程度最高的（第-個）物件， 建立一個新的叢集STC，並將該物件加入該叢集5〇2 ; 步驟3:逐次讀取該叢集之物件，依⑽路網地圖搜尋該物件之鄰近 物 物件是否落在奎塞物件集合中叫，若有物件落在餘集合中，則從奎塞 件集合中取出該物件並加入叢集5〇3 ; 步驟4:新加入叢集的物件再次搜尋其鄰近物件是否需要加入叢集， -直重複職叢細所有物件都已輯瞄完畢5〇4 ; 步驟5 :回到步驟2，若奎塞集中尚有物件，繼續相同步驟。 、請參閱圖六，為本發明以時空環境資料探勘技觸測交通瓶頸點的方 法之_交通瓶賴示_上班日、上⑷，制本發明所 產生並顯示於電子地圖上之都會區交通瓶頸點示意圖，其中依不同形式之 箭頭來表示各時段之规擁塞輯歧通綱點。 以 *頸點的方法，與 本發明所提供之以時^環境龍探純術侧交通如 其他習用技術相互比較時，更具備下列優點： 】·本發财以找料„«境讀會區路_技瓶獅，如上班 日與假日之交通_點有所不同，上班尖峰_與下班尖峰時刻之 交通瓶頸點也會有所不同。 ^ 13 201000934 本Ιχ月透雜程之機制，可自動重新處理失敗之訂單，大幅降低人 工處理之人力與日麵，·本侧具有交職塞删魏，由找出來的 交通瓶賴可以依照該路段的歷史資料預測其發生皇塞的機率，並 了以依知及時妓通資訊，套㈣料探勘所找出的餘模式法則， 預測下一個時段可能壅塞的路段為何，例如：本發明中資料探勘找 出的壅塞模式法則中有—條法則為上班日的〇8:〇_0 Α路段 為瓶頭點其導致B路段在壅塞的信心、減達娜，則若配 置曰慧里車上早疋的車輛在該時段行經該路段回報的即時路況為 «，則可以預測在08:40時，B路段應該會產生奎塞，信心指數 85%。 3. 本發明之定位服務（衛星定位車隊派遣管理㈣為㈣不 停機的運作’故本發明方法可以Μ小時，全年無休的運作，資料 收集的成本遠低於交通狀態探針車與訊麵測器。 4. 本發明可以透過大量的定位應用服務之車輛通訊資料，侧出交通 蜜塞之區域’本方法所策集的通訊資料樣本資料量大，且涵蓋區域 比傳統定點之偵測方式廣大。 本發月無時間、空間限制，可對所有定位應㈣、統(商用車輛營運 系統)的車輛行經的區域進行監測其交通奎塞狀況。 、上列詳細說明係針對本發明之一可行實施例之具體說明 ，惟該實施例 、、，限制本發明之專利範圍，凡未脫離本發明技藝精神所為之等效實 施或變更，均應包含於本案之專利範圍中。 14 201000934 综上所述，本案不但在技術思想上確屬創新，並能較習用物品増進上 述夕項功效’應以充分符合卿性及進錄之法定發明專利要件，麦依法 提出申凊’懇請貴局核准本件發明專利申請案，以勵發明，至感德便。 【圖式簡單說明】 圖-為本發明以時空環境資料探勘技術偵測交通瓶頸點的方法之衛星 疋位車隊派遣管理系統架構示意圖； 圖一為本發明以時空環境資料探勘技術偵測交通瓶頸點的方法之智慧 型車上單元元件示意圖； 圖—為本發明以時空環境資料探勘技術债測交通瓶頸點的方法之系統 模組架構圖； 圖四為本發明以時空環境資料探勘技術偵測交通瓶頸點的方法之探勘 偵測交通瓶頸點演算流程圖； 圖五為本發明以時空環境資料探勘技術偵測交通瓶頸點的方法之叢集 演算法流程圖；以及 圖六為本發明以時空環境資料探勘技術偵測交通瓶頸點的方法之都會 區父通版頸點示意圖（上班日、上/下午尖峰時刻）。 【主要元件符號說明】 1 智慧型車上單元 2 衛星定位車隊派遣管理系統單元 3 GIS電子地圖單元 4 資訊收集與交通資訊產生模組 15 201000934 5 交通壅塞模式探勘模組 6 瓶頸點探勘模組 7 交通瓶頸點 8 GPS衛星定位模組 9 GPRS/UMTS行動通訊模組 10 後端系統 / r·· '*·· ^ 16201000934 IX. Description of the invention: [Technical field of invention] The present invention relates to a method for detecting traffic bottlenecks by using space-time environmental data exploration technology, in particular, a smart car-mounted unit and a back-end monitoring center using a device in a vehicle The communication data uses the techniques of spatiotemporal data mining, clustering, and geographic information system (GIS) to find out the different inspirational methods in the metropolitan area network. (traffic congestion patterns), and use the correlation between the three modes of traffic congestion, and then further identify the traffic bottlenecks under the various circumstances of the network in the metropolitan area. [Prior Art] Traffic bottlenecks are the origins of traffic road networks that are particularly prone to congestion. If traffic bottlenecks can be identified and appropriate control and grooming measures are taken, it will greatly help the smoothness of traffic. Traditional The traffic bottleneck point is based on the definition of the bottleneck point in the transportation management. It is considered that the demand (flow) of the road network object (road section or intersection) is higher than the capacity allowed by the road section, and the object is a bottleneck point; However, this definition is difficult to apply to the actual metropolitan area network, usually for the following reasons: 1. The road network in the metropolitan area is too complicated, and the road network objects usually interact and are not independent. 2. The traffic and capacity information of road network objects is difficult to obtain, and the traffic information is time-dependent. 3. The road network bottleneck in the metropolitan area is not fixed. It is related to the overall traffic demand (OD demand). For example, the bottleneck point during the working hours is different from the bottleneck point of the 201000934 in the off-hours. The traffic bottleneck point of the day is finer than the traffic bottleneck point of the holiday. 4. The box information (flow) of each traffic network object is difficult to collect, and it needs to be built for the entire road network. The current method of traffic information when collecting scales is > planted in various important sections or forks. Set up a traffic detection sensor or a surveillance camera. However, this method has its own shortcomings: 1) Insufficient coverage (fixed point, probe) Car), the fixed-point detector is usually installed only at the intersection or important position, the nuclear side (4) location can not know the traffic conditions; the probe car is limited by the amount of the car, although it can be transferred to collect the road section If you don't have enough coverage, the probe car can't collect traffic information all year round, so the time coverage is not enough. 2) The number of samples (probe car), the traffic of the road section is different from the county time, climate or artificial live silk forest, but because of the limited number of probe vehicles, a probe car can only be in the same-B A section of the road is used for data collection, so the number of samples collected by the entire traffic network is insufficient. 3) High cost (fixed 'point type, probe car), construction of traffic side sensor construction requires a lot of money 'The timely monitoring of the intersection camera and the maintenance of the sensor requires a lot of manpower; the probe car needs precision equipment And the driver and the operator are configured to stay in the south. It can be seen that there are still many shortcomings in the above-mentioned methods of use. It is not a good design, but it is urgent to be improved by 201000934. In view of the shortcomings derived from the above-mentioned system, the inventors of this case are thinking and improving and innovating. After painstaking research, I finally successfully developed and completed this method of detecting traffic bottlenecks with space-time environmental data exploration technology. SUMMARY OF THE INVENTION The object of the present invention is to provide a method for measuring the traffic bottleneck point of the technical side of the scale material to assist and replace the existing method of fixed point and probe money collection information, by using the lion (4) State turn _, coffee thank you for the gift, ^ data exploration and electronic _ technology, analysis of a free f and a large amount of real-time traffic information to replace the Amber and rare traffic information detection vehicle;, the secondary purpose of the present invention That is to provide a method for measuring the traffic point in time and space (4) exploration technology. The analysis of the vehicle is based on the analysis of the vehicle, and the electronic map is used to find the time dimension (times, locations). The type of assignment, the step-by-step analysis to find the point to solve the problem of insufficient number of mobile communication vehicles, insufficient coverage and insufficient sample size. The purpose of the project is to provide a means of detecting traffic with time and detail. ▲, 'The invention can be used simply in the city (4) traffic control, and provide predictive results' to the parent control center tube '3® Chang Hao, a Wide M, g (4) as a step-by-step traffic management regulation, or provided to passers-by as a tourist road m is a reference for travel gear selection. Law, Department of profit (4) 2 orders (4) _ tender dirty ship shaft point, 3 〃, traffic information generation, traffic congestion mode data exploration and traffic bottleneck 201000934 Lin survey and other methods, shooting the money test traffic f news Zhao against the charm The latitude, longitude, speed, direction and other information returned by the Granville team's dispatching system system unit (Location_basedService, LBS) collects and organizes the information collection and traffic information generation modules, and cooperates with the electronic map unit to change the scales (4) Money information, and according to the traffic speed limit of each road section, the traffic status of each road section and each time period is divided into five grades, among which grade! On behalf of normal non-Quaisie, level 5 represents very congestion; 3 in addition to generating traffic information in the time and space environment, the metropolitan area road network, and the satellite positioning fleet dispatch management system unit converts the material into the vehicle data It contains detailed travel information such as the starting and ending points of travel time, travel time, point coordinates, speed and time. This information is used in the next stage to find the traffic congestion mode; The traffic jamce mode data exploration module finds the pattems in the time-slot area; this model defines two traffic queer modes, CGngestic n c_quent Pattem, ccp) and Congestion drop downstream pattern (CDP), CCP because the time of the construction of the stagnation area caused another time-space traffic congestion area; the Quécy flow difference mode is "the object and the downstream object of the Qingsai drop, if the reduction is too large through the wire After the obstructed object, the downstream objects will not be blocked, and the object can be bullied as a bottleneck point. In addition, the traffic bottleneck point exploration system utilizes a traffic bottleneck point to explore the model. In order to find the traffic bottleneck point, this module defines three methods for finding the traffic (four) points, such as the general congestion diffusion method, the production plug M: method, the flow difference ratio method, and can be used for the traffic In the case of the domain rules, some of the conditions in the H-domain 丨 贞彳 亲 魏 魏 , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , 魏 , 魏Finally, we use the Confidence Confidence Index and the degree of congestion to find out the real traffic bottlenecks that can be used according to the time period. [Embodiment] Please refer to FIG. 3 , which is an application system architecture diagram of a method for detecting a traffic bottleneck point by using a spatiotemporal environment data exploration technology. As can be seen from the figure, the main components of the system architecture for providing data sources include: The smart car order 70 1 , the smart car upper unit i is set in the car, and the communication record with the back end system 10 is recorded in the database, and the database information can be further utilized by the present invention. Find out the traffic bottlenecks in the network of the metropolitan area; After the % system 1G, the back-end system 1Q reports to the back-end system 10 via the smart car-on-unit 1 that the vehicle is recorded, the vehicle is reduced, and the direction of travel is Speed, etc., the timing is the return of the knife, the return of the event and the passive return, and the place where the return occurs is called the return. Therefore, according to the vehicle status information and the road network information of the geographic information system, I Report the traffic information of the section where the time is located to determine whether the road section is in traffic jam. ' 叫 参 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ —jit- » 广 - 杈 杈 group 8, the GPS satellite positioning module 8 can obtain the position coordinates of the current location of the vehicle through the global satellite positioning system GPS, and in the 201000934 area where the GPS signal cannot be received, The communication base station fixed the financial formula and estimated the coordinates of the vehicle location; - GPR_TS mobile communication module 9 'The Gprs_s mobile communication module 9 is the interface for communication between the intelligent vehicle upper unit and the back-end system, through the mobile communication secret and after End communication for data reporting and fleet dispatch operations. Please refer to Figure 3, which is a system model of the method for making traffic bottlenecks in space-time environmental data exploration technology. The Wei Mo Hong towel mainly includes: - information collection and contact « shouting · 4, „ Capital reduction model... Collecting satellites and transferring management secrets unit 2 reading materials, after Jia (removing secrets, abnormal assets), the latitude and longitude of the electronic map is converted into addresses through the GIS electronic map unit 3, and combined with vehicle speed and square conversion For the road section, and according to the traffic speed limit, the traffic status of each section of Lin and the time of each period is divided into five grades, the towel grade 丨 represents the normal transfer, and the other 5 represents the non-clothing t plug. The data of the metropolitan area network under the different time and space is transmitted to Beixun, and the data in the satellite positioning fleet dispatch management system unit is converted into vehicle travel data. XC These data include the starting and ending points of the vehicle travel time, travel time, After the coordinates, speed and time, etc., the detailed travel information can provide money congestion mode and lion module 5 to find the traffic congestion mode. Parent communication congestion data exploration module 5, the traffic congestion mode data The exploration module 5 is based on the data collected by the Beixun collection and the traffic information generation, and finds the _ pattern (Patefns) in the space-time ring-quiet area. The method is divided into two steps: The first step is to find the object of Quaisai (the object with traffic status rating 4 or 5) according to the geographically adjacent Guanxian # silk (spatiotempoml _gesti()n _ 201000934 STCA); the first step is to find out The relationship between the g-sag areas, using the space-time association rule to find the rules for every two or two occlusion areas, if the ratio of the two empire areas in the rule is higher than the threshold value, then the 13⁄4 区域 area is listed. For correlation; in this model, _ traffic check mode knife another 1 递 deferred mode (c〇ngesti〇nc〇nsequent pattem. Cep) and congestion drop downstream pattern (CDP), CCP The main reason for the pattern is that _ a space-time father stagnation area leads to another time-space traffic clogging area; the slamming flow difference mode is the ratio of the congestion of the honey-supplied object to the downstream object, and if the drop is over the wire, it passes through the job. Object downstream If the object is not blocked, the object can be determined as a bottleneck point; a parent bottleneck point exploration module 6, the traffic bottleneck point exploration module 6 is based on two types of traffic jams of the traffic congestion type The model defines three ways to observe the traffic bottleneck point, and the other method is to use the J疋. Quesser diffusion method, the g-sag convergence method, and the flow difference ratio method. The honey-soil diffusion method is based on the production delay model to find a traffic ku The plugging area (STCA) will trigger a number of other traffic congestion areas (STCA), which may fall on the road network object of the sg plug area 0; and the stipulation of the arrogance rule is the opposite inference, multiple STCA When the same stca is reached, the neck point may fall on the road where the STCA exists; the oblique flow ratio is the root of the plug flow difference pattern to find the object whose clogging flow ratio exceeds the threshold value (upstream object) The difference in the degree of congestion between the queuing degree and the downstream object is too large. 'According to its confidence, the index (the occurrence ratio of the occurrence of the object) and the degree of honey plug to determine whether it is a traffic bottleneck point, - Father Pass_Point 7 ' The lion 7 system will find out the transportation bottle Lai exploration module 6 Traffic bottleneck point as the time displayed on an electronic map. 11 201000934 Please refer to Figure 4, which is a flow chart of the exploration and detection of traffic bottleneck points for the method of detecting traffic bottlenecks by the space-time environmental data exploration technology. It can be seen from the figure that the main steps of the exploration of the traffic bottlenecks Step 1 : Data collection and data collation, collect data from the satellite positioning fleet dispatch management system unit, and do the sorting of the data 401; Step 2: Traffic data conversion and traffic travel data collection, with the electronic map service system The data of step 1 is converted into traffic data and traffic travel data, and stored in the space-time traffic state database 402; Step 3: Traffic congestion point analysis, using the space-time traffic state database for traffic congestion analysis, and generating spatio-temporal traffic congestion objects 403; Step 4: Parental congestion point cluster calculus, clustering the objects of the congestion into several time-space traffic congestion areas 404 according to the geographical neighbor relationship; Step 5: Traffic Maicer model data exploration, according to the space-time traffic area and its relevance The two traffic Messer modes are defined as 'depression delay mode and quexe flow difference mode 4〇5 respectively; Step 6: Analysis of the traffic bottle drive domain, according to the mode defined in step 5, the Quebec diffusion method, the congestion convergence method, and the flow difference ratio method are analyzed. 4 Step 6: Determine the degree of congestion and confidence index of the traffic object. If the gatecast value is exceeded, the parent traffic congestion area or the traffic bottleneck point is displayed on the electronic map 4〇7. Please refer to FIG. 5 , which is a traffic congestion domain clustering algorithm for detecting a traffic class by using space-time environment data exploration technology, and the algorithm implements a traffic congestion point clustering algorithm in the second stage of FIG. The operation of the algorithm is to accumulate the space-time traffic congestion found on the __step finances. 2010 20109934 Objects are geographically adjacent to find the same nearby areas. The process flow of this algorithm is: The congested object points, according to the objects of the congestion, integrate the congested objects of the same time plane into the same space-time congestion. Step 1: For each period of time, the degrees are sorted into a set: {〇j5〇i; Step 2: Ru Yuezhong There are still objects, the most highly taken (first) object, a new cluster STC is created, and the object is added to the cluster 5〇2; Step 3: The objects of the cluster are read one by one, according to the (10) road map Searching for the neighboring object of the object to fall in the Quayce object set, if any object falls in the remaining set, the object is taken out from the Quéme set and added to the cluster 5〇3; Step 4: newly added to the cluster object Search times whether it near objects need to join a cluster, - straight repeat Cong fine job all the items have been completed 5〇4 aiming series; Step 5: Return to step 2, if there are objects Kuise focus, continue the same procedure. Please refer to FIG. 6 , which is a method for measuring the traffic bottleneck point of the space-time environment data by using the space-time environment data exploration technology _ traffic bottle ray display _ working day, upper (4), the metropolitan area traffic generated by the invention and displayed on the electronic map A schematic diagram of a bottleneck point, in which arrows of different forms are used to indicate the congestion points of each time period. With the method of *neck point, when compared with the other conventional techniques provided by the present invention, the following advantages are obtained: 】· This is a fortune to find materials „«境读会区Road _Technical bottle lion, such as the traffic between the day and the holiday _ point is different, the peak of the work _ and the traffic bottleneck point at the peak of the work will be different. ^ 13 201000934 This month's mechanism of the process, automatically Re-processing the failed orders, greatly reducing the manpower and the day of manual processing. · This side has the job of deleting the Wei, and the traffic bottle that is found out can predict the probability of the occurrence of the Huangsai according to the historical data of the section. In accordance with the information and the rules of the residual model found in the (4) materials exploration, it is predicted that the road section may be blocked in the next period of time. For example, in the method of congestion detection found in the data of the present invention, there is a rule of law for work. Day 8: 〇 Α Α Α Α 为 为 为 为 为 为 为 为 为 为 为 为 为 为 为 为 为 为 为 为 信心 信心 信心 信心 信心 信心 信心 信心 信心 信心 信心 信心 信心 信心 信心 若 若 若 若 若 若 若 若 若 若 若 若 若 若",then It can be predicted that at 08:40, the B section should generate Quaysie, and the confidence index is 85%. 3. The positioning service of the present invention (satellite positioning fleet dispatch management (4) is (4) non-stop operation', so the method of the present invention can be used for hours. Throughout the year, the cost of data collection is much lower than that of the traffic state probe vehicle and the signal detector. 4. The present invention can provide a traffic jamming area through a large number of vehicle communication materials for positioning application services. The sample of communication data collected is large, and the coverage area is wider than that of traditional fixed-point detection. There is no time and space limitation in this month, and all vehicles that are positioned (4) and unified (commercial vehicle operating system) can be used. The area is monitored for its traffic condition. The detailed description above is a detailed description of one of the possible embodiments of the present invention, but the embodiment and the scope of the invention are limited, without departing from the spirit of the invention. Equivalent implementation or change shall be included in the patent scope of this case. 14 201000934 In summary, this case is not only innovative in terms of technical thinking, but also more The effect of using the items into the above-mentioned eve items should be in accordance with the statutory invention patents that fully meet the requirements of the Qing Dynasty and the entry, and the stipulations of the stipulations of the stipulations of the stipulations of the stipulations of the stipulations of the stipulations of the stipulations of the inventions. Brief Description of the Scheme] Figure - is a schematic diagram of the satellite fleet dispatch management system architecture for the method of detecting traffic bottlenecks by using space-time environmental data exploration technology; Figure 1 is a schematic diagram of the present invention for detecting traffic bottlenecks by using space-time environmental data exploration technology The schematic diagram of the system component of the intelligent vehicle on the vehicle; Figure 4 is a system module architecture diagram of the method for measuring the traffic bottleneck point of the space-time environment data exploration technology debt in the present invention; FIG. 4 is a diagram of the invention detecting the traffic bottleneck by using the space-time environment data exploration technology Point method for detecting traffic bottleneck point calculus flow chart; Figure 5 is a flow chart of a cluster algorithm for detecting a traffic bottleneck point by using spatiotemporal environment data exploration technology; and Fig. 6 is a time and space environment data exploration of the present invention The method of detecting the bottleneck point of the technology in the metropolitan area of the city through the neck point map (working day, upper / afternoon tip Time). [Main component symbol description] 1 Smart car unit 2 Satellite positioning fleet dispatch management system unit 3 GIS electronic map unit 4 Information collection and traffic information generation module 15 201000934 5 Traffic congestion mode exploration module 6 Bottleneck point exploration module 7 Traffic bottleneck point 8 GPS satellite positioning module 9 GPRS/UMTS mobile communication module 10 back-end system / r·· '*·· ^ 16