【溫馨提示】 dwg后綴的文件為CAD圖,可編輯,無水印,高清圖,,壓縮包內文檔可直接點開預覽,需要原稿請自助充值下載,請見壓縮包內的文件及預覽,所見才能所得,請細心查看有疑問可以咨詢QQ:414951605或1304139763
為農業(yè)機械提供位置數據測量
Herman Speckmann
原文來源:Federal Agricultural Research Centre Braunschweig (FAL), Institute for Biosystems Engineering, Bundesallee 50, D-38116 Braunschweig, Germany
摘要
農業(yè)機械、車輛需要位置數據來指導和控制執(zhí)行最佳工作位置。位置數據也被需要用在像精細農作這樣的應用上。位置數據的必要的準確性、分辨率和頻率依照不同的應用而變化。只有一個系統,安裝在中央車輛(例句、拖拉機),應該提供對每項任務的位置數據。
提出的關于中央系統的基本概念是位置數據按照特定應用程序計算并且直接被傳送到它需要被應用到的那個點上。這片論文闡述了測量的基本原理和位置數據的計算,還對現有的傳送數據的農業(yè)網絡進行了簡要介紹。它集中建議了一個提供和轉移位置數據的網絡服務。被討論的解決方案是以農業(yè)BUS(總線)系統為基礎(DIN 9684, ISO 11783). ? 2000 Elsevier Science B.V. 版權所有.
1.前言
位置指導的目的是給生長在農田里一個固定的區(qū)域上的莊稼帶來增產的方法。莊稼或者它們在農田里所處的位置是指導的重要參照。
位置數據被用來指導農用車、實現控制和支持精耕農業(yè)。準確性、分辨率和頻率取決于他們的具體應用。
必須強調的是本文沒有合適的解決這個問題的傳感器來產生數據。更確切的說,這里研究的問題是參照移動單位的一定的位置進行了一個位置信號產生,但是這個位置和需要的位置數據并不是完全一致的。此外,位置信息有可能在同一時間被需要用于幾種目的, 車輛和工具組合的結構可能會經常改變。
正如 Freyberger 和 Jahns (1999), Wilson (1999)所提到的, 測量系統可以是一個絕對定位系統,比如Bell(1999)描述的衛(wèi)星系統,或者是一個相對的系統,比如Debain et al. (1999), Hague et al. (1999)描述的機器視覺系統。它可能也包括輔助傳感器。
傳感器只有在參考具體位置情況下測量位置,比如相機的安裝點、天線的底部。在接下來的描述中,這個位置被稱為測量點。由于各種原因,這位置測點的是預先設定好的,意味著衛(wèi)星天線將盡可能安裝在拖拉機的車頂上以便減少測量不到的區(qū)域。攝像機將會安裝在有保障最佳視覺的位置。粗糙或傾斜的表面引起的運動可能導致測量位置和運動表面的位置不同。例如,一輛車頂上裝有衛(wèi)星天線的車輛,大約3.5m,駕駛在10°的斜坡表面,傾斜方向造成的區(qū)別相差60cm。圖1闡述了這個情形。在這個例子中,計算一個參考點的位置可能更適當一些。貝爾(1999)提出把拖拉機的后方軸的中點作為參考點。表面上的一個點,例如,后方軸中間的下垂直面似乎顯得更適合與某些應用。像一些應用,比如控制實現,工具的一定點的位置可能最終重要。這個點將被稱作目標點。
在某些情況下位置數據需要用于不同的目的,分別為每個目的以一種獨立的測量系統測量位置不是很有效。當位置測量只有一次時多個硬件可以避免,同時工具上其他點的位置或者工具也被計算。假如位置和方法被測量,實驗測量和空間向量之間的地點測點的計算是眾所周知的,那么這種情況是可能的。如果兩個點嚴格耦合,這意味著兩點都在拖拉機上、兩點之間的向量是常數,一個簡單的矩陣運算就能產生結果。如果這些點沒被嚴格耦合,這意味著,例如,一處拖拉機,另一個是在附加工具上,矢量是可變的。額外的測量成為必要用來建立兩點之間的向量或必須應用其他原理計算目標點的位置。
2.數據處理和數據轉移
通過計量點上的測量位置和方法,在車輛或工具上任何點的位置數據可以被計算出來。計算結果可以被測量系統(中央數據處理)或由請求目標位置數據的各個系統(分布式數據處理)計算出來。
2.1 分布式數據處理
在分布式數據的情況下,測量系統僅作為智能傳感器服務。它測量需要的位置和計算,和提供這些未經處理的數據。頻率和精度等特點取決于請求的單位。這個單位執(zhí)行所有處理來計算位置。單位必須知道測點的位置和各有關參數。這樣處理的好處是測量裝置可以相對簡單。另一方面,每個請求的單位需要的充分的能力來履行這一運算。
2.2 中央數據處理
測量單位被擴展包括計算目標位置的各個組件。這個測量和處理系統形成了一個所謂的位置和導航服務的單元,這個單元提供任何目標點的最終位置數據。在這種情況下,只有一個測量與處理系統是必要的,即使位置數據必須被更多的用戶要求。這樣做,只有PNS必須知道所有相關的參數來進行計算。
2.3 數據傳送
無論數據在哪里處理,一個數據傳輸是必要的。對于這樣一個數據傳輸,一個標準的網絡是適當的。為了用于農業(yè)領域,存在一個在移動單位和固定農場電腦之間傳輸數據的汽車。農業(yè)總線系統(LBS)也已被標準化以便能在網路的各個電子單元(LBS節(jié)點或BUS節(jié)點)之間進行信息交換。這個標準定義了物理層網絡,網絡協議,系統管理,數據對象和常見任務的服務程序(Speckmann andJahns, 1999)。
LBS以DIN9684(DIN,1989–1998)作為標準。目前,正在努力建立一個國際標準(Nienhaus,1993),ISO 11783,為了這個目的,像LBS,ISO 11783也將定義一個農業(yè)BUS作為一個農業(yè)機械交換數據的開放系統,特別是在拖拉機-執(zhí)行工具的組合和從移動單位到靜止不動的農場計算機。這個標準是基于控制器區(qū)域網絡數據協議(CAN; BOSCH, 1991)。市場上有相應的硬件設備。
在LBS中,為一般位置數據(地理位置:經度、緯度、高度,或軌道位置)的傳輸定義了數據對象。這標準允許定義的額外的數據對象,例如多維的距離,方向和速度。沒有幾何實施參數的數據對象目前存在在LBS中。ISO 11783提供,在第7部分(信息實現應用層),實施航行偏移的第一個定義?,F行標準沒有定義數據在哪里進行處理。因此,關于BUS中哪個單元計算目標點的數據,哪個或那些單元測量數據不具體。
LBS提供所謂的LBS服務來執(zhí)行常見任務。LBS服務是為LBS的參與者頻繁地執(zhí)行復發(fā)的任務的功能單元。LBS用戶站就是這樣的一項服務。這是一個為用戶提供輸入和輸出BUS上節(jié)點(BUS參與者)處置的數據中央接口。另一項服務提供在移動單位和固定的電腦,農場的電腦之間的數據交換。一些服務在LBS中被定義但尚未有詳細的標準,例如診服務斷或“Ortung und Navigation”(位置和導航),將在下面作為PNS被討論。在圖2中,一個典型的農業(yè)網絡的簡化方案展示了一個拖拉機-噴霧器的組合。這個圖表包括物理BUS線路,即骨干網絡。在這個BUS上,參與單元如拖拉機的電子控制單元(ECUs)、霧化器被連接協作起來。另外,兩項LBS服務也被連接到BUS上。一項服務代表LBS用戶站。另外一項是位置和導航服務,即位置數據的測量和處理系統。
2.4 分布式和中央數據處理的比較
一個分布式數據處理,農業(yè)BUS,根據DIN 9684 或者ISO 11783, 定義了在測量系統和任何參賽者之間必要的數據交換;獨自地,任何一個ECU。每一個ECU怎樣得到計算機位置數據計算必要的幾何和運動參數的問題保持開放。每一個ECU知道從各自的結合點到目標點的參數,但它不知道從結合點到測量點的參數。這些參數必須由其他ECU提供。沒有標準定義相應的數據對象或請求數據的程序。對于分布式數據處理,這些定義必須補充。
另外,對于中央數據處理,一定要知道測量點和目標點之間所有的運動參數。此外,方法必需被定義以便使用中央服務計算目標點的位置數據。一個位置和導航服務需要擴展標準,但以下的優(yōu)點在實際使用中是至關重要的。
● 為了確定目標點的位置數據,相應的控制單元(ECU)只有一個對話伙伴網絡。它獨立工作于各自的網絡配置,僅僅發(fā)送自己的參數和只接受它特定位置數據。
● PNS從所有的ECU上接受參數。它知道所有一切幾何條件和車輛-工具組合的運動參數。因此,任何目標點位置的確定是可能的。
● 這個標準的定義了計算程序和明確的提出了目標點的位置數據。
● 計算位置數據的計算性能完全由PNS提供。沒有計算能力需要用于這個目的。
在前一節(jié)提到,提供位置和導航數據的服務已經在LBS的計劃中。在下文中,將提到PNS的一個試例解決方案。
3.一項定位和導航服務的提議
此時,應當指出,下面的PNS的介紹是一項建議。它提供了一個平臺進行討論,這可能導致這個服務標準化。
3.1 PNS的主要特征
PNS的特征首先依賴于它的使用目的。從前面所講的,很明顯的是,測量的位置數據在一個地點,用在不同的地點。為了提供需要的數據來指導車輛,控制工具的位置和協助任何一種精耕農業(yè),下面的條件必須滿足:
PNS提供有關測量點的數據。
PNS提供有關參考點的數據。
PNS提供有關目標點的數據。
這項服務的特點如下:
1.數據的請求和傳播的方式已經標準化,數據被LBS (DIN 9684)定義和將被ISO 11783標準化。因此,它將不會在此討論。在下面,LBS將作為一種標準化的農業(yè)BUS系統被使用。
2.數據的容量、準確性、頻率和范圍是由數據的目的決定的。
3.滿足這些要求的硬件和軟件不應被規(guī)范,應該取決于生產廠家。
3.2關于位置數據測量和計算方法標準的影響
各種測量系統和PNS中用于決定位置數據的方法不再標準的范圍之內?;谛l(wèi)星,機器視覺、慣性導航、地磁或這些情況的組合可能被應用。作為一種結果,生產企業(yè)可以決定如何產生位置數據,只要他滿足了規(guī)定的要求和準確性。
3.3 PNS在農業(yè)BUS系統中的整合
在LBS中整合定位和導航服務存在一些好處,因為許多特性已經被定義。LBS已經包括在PNS的選項作為試驗的標準。它允許實現服務作為一個獨立的物理單位或者為另外一個物理單位的邏輯單位。BUS接口和BUS協議的物理性能(DIN 9684, part 2)已經被標準定義。為LBS中服務的集成,系統的功能的定義是果斷的(DIN 9684,part3)。他們在LBS中定義節(jié)點的性能。第三部分也給了LBS服務一般的定義。
一項LBS服務形成與LBS參與者點對點的連接。LBS參與者使用服務時不會被其它使用者影響,一個LBS參與者也不能影響其他參與者對服務的使用。所有進一步PNS的定義還不規(guī)范。
3.4 PNS操作的一般模式
PNS設計應用以下的基本假設:
1.每一個ECU的只知道它自己的參數,包括參考點、目標點、結合點位置、車輛類型或軸距的坐標和數量。
2.只有ECU根據工作條件可以定義必要的時間間隔,準確度和位置數據的分辨。
3.每一個ECU的可以選擇不同的任意時刻的位置數據。
4.參數和計算和提供的位置數據的方法將會在田野機械開始運作過程之前被定義。
5.PNS提供了一些程序為實施標準和車輛類型計算位置數據。
6.位置數據自動(周期性)地或根據需求被提供。
為了滿足這些要求,服務窗口提供適當的工具,同時 ECUs 決定如何使用及使用哪個工具。這意味這它們定義一個或者多個任務。這樣一項任務基本上代表了一個命令表,包括激活具體工具使用的命令。這些任務被送到PNS,隨后PNS執(zhí)行這些任務。一個ECU的不同的任務相互獨立的被執(zhí)行。
圖3闡明了PNS與一個ECU之間的數據傳遞。同時,也顯示了PNS的主要部分。PNS的這些工具包括位置測量系統和測量點的數據,以及一系列處理這些數據的程序方法。程序如下:
1. 計算位置數據(位置程序);
2. 計算位置數據值的平均值,最大值、最小值和積分的方法(算術程序);
3. 輸入和輸出數據(傳輸程序);
4. 傳遞數據到ECU(傳遞程序);
5. 控制數據處理(數據控制程序);
為了這些方法的執(zhí)行,ECU必須定義相應的參數。它同時也定義位置數據的數據對象。
PNS的主要工具是一項執(zhí)行ECU定義的任務的程序系統。簡而言之,程序系統解釋任務指令,調動相應的方法,計算要求的位置以及把數據送到ECU(電子控制單元)。
為了一項任務的定義,ECU生成一個任務庫。一個務庫主要是一系列調動PNS的程序法或者調動內嵌的任務庫的指令。各種參數被定義并且放置在參數庫里。為了存儲被計算的位置數據,ECU必需定義數據庫。數據庫必需在激發(fā)相應任務程序之前通過BUS從ECU傳送到達PNS。
3.5 PNS預定義的程序
PNS預定義的程序是一些處理位置數據或者控制數據處理的程序。不同的程序執(zhí)行不同的功能。不同的程序被一些獨特的標識符區(qū)別。這些程序被稱為“內部任務”(任務庫)。他將會成為標準的一部分用來定義標識符,功能規(guī)格和調用程序規(guī)格。
3.5.1 位置程序
位置程序(計算位置數據的程序)是計算目標點位置數據的數據。這些方法計算從最初的位置(輸入位置數據、資料的參考點的數據或以前計算的數據)到一種新的點的位置(輸出的位置數據、數據的目標點或作為中間結果)。位置程序能夠滿足不同結構位置的計算(考慮一、二或三維模型,嚴格耦合點,幾個基本類型車輛的不嚴格耦合點,工具和車輛-工具的結合)。這些程序從有關ECU執(zhí)行定義的參數庫得到他們的實際參數(目標點的坐標,車輛的長度、寬度、高度、類型或軸距)這是確定的有關實施ECU的。
圖4顯示了使用一個位置程序的一段任務庫。PNS的程序系統執(zhí)行這個程序庫。在任務庫的某一點上,它發(fā)現調用位置程序的指令。這個調用指令包括特定程序的標識符和有關參數庫的引用。這時,程序系統擁有由以上的操作產生的實際位置數據?,F在它使用這些實際數據作為輸入數據,和引用參數庫用于位置程序。然后,它執(zhí)行特定的程序。該程序使用指定的參數計算輸出的位置數據。然后,它返回到程序系統。位置程序的輸出數據成為新的實際位置數據。程序系統繼續(xù)執(zhí)行下面的指令。
3.5.2 算術程序
算術方法被用來計算位置數據的平均值,最大值、最小值或者積分值。一個算術程序從程序系統的實際位置數據或從特定數據庫得到位置輸入數據。它使用在調用指令里決定的參數庫中的參數計算輸出位置數據。然后,計算結果數據被存儲在一個被定義的數據庫里。
圖5展示了一個算術程序使用的例子。在任務庫的某一點上,它發(fā)現調用算術程序的指令。這個調用包括具體程序的標識符,一個有關參數庫的引用,一個目的數據庫的引用和源數據庫選擇性的引用。這個程序系統采用實際數據和參考數據用于程序計算。根據調用規(guī)格,算術程序從程序系統(沒有定義的數據庫參考)或一種數據資源(數據資源I)得到輸入數據。它計算被要求的值并把計算結果存儲在一個數據庫里(數據庫II)。計算參數是從定義的參數庫中得到的。程序發(fā)揮到程序系統并繼續(xù)執(zhí)行。實際的位置數據沒有被改變。
3.5.3 傳輸程序
PNS定義了三種類型的傳輸程序。輸入程序是用來裝載作為實際位置數據的確定的數據庫位置數據到PNS的程序系統。輸出程序存儲實際位置數據到一個在調用指令里預先定義了的數據庫。輸入/輸出程序被用來從一個源數據庫到目的數據庫之間傳輸數據。
圖6顯示了一個使用輸入和輸出程序的例子。輸入程序的調用指令包括具體程序的標識符和源數據庫的引用。在執(zhí)行輸入程序之前,程序系統為程序提供源數據庫的引用。然后,程序執(zhí)行和得到位置數據,并將它作為實際位置數據返回給程序系統。以前的實際位置數據被損壞。系統繼續(xù)進行。對于輸出程序的使用,實際位置數據與目的位置數據庫提供參考。輸出程序將實際數據放到目的數據庫并返回到程序系統。實際位置數據仍然有效。
3.5.4 傳遞程序
傳遞程序發(fā)送具體的位置數據到ECU。源數據在調用指令(或一個數據庫或程序系統的實際數據)里被定義。當執(zhí)行一個傳遞程序時,它得到具體的位置數據并傳送到ECU。
3.5.5 數據控制程序
數據控制程序控制一個任務庫的執(zhí)行。程序流程是控制時間或距離。PNS的程序系統調查任務庫。假如確定的時間間隔已過期或已超出距離限制,程序將執(zhí)行下列指令。否則,程序系統跳到數據庫的結尾。
Computers and Electronics in Agriculture 25 (2000) 87–106 Providing measured position data for agricultural machinery Hermann Speckmann Federal Agricultural Research Centre Braunschweig (FAL), Institute for Biosystems Engineering, Bundesallee 50, D-38116 Braunschweig, Germany Abstract Agricultural machinery and vehicles require position data for guidance and to control implements for optimal working positions. Position data are also needed for such applica- tions as precision farming. The necessary accuracy, resolution and frequency of position data vary according to the specific application. Only one system, installed at a central vehicle (e.g. the tractor), should provide position data for each task. The basic concept for the proposed central system is that position data are calculated in accordance with the application and transferred directly to the point at which they will be used. The paper describes the fundamentals of measurement and calculation of position data, and gives a short introduc- tion to the existing agricultural networks to transfer these data. It concentrates on a proposal for a network service to provide and transfer position data. The solution discussed is based on the agricultural BUS (DIN 9684, ISO 11783). ? 2000 Elsevier Science B.V. All rights reserved. Keywords: Local area network; Controller area network; Agricultural BUS system; LBS; Calculation of position; Calculation of direction; LBS service :locate:compag 1. Introduction The purpose of position guidance is to bring the means of production to the plants, which grow at a fixed location on the field. The plants, or rather their location on the field surface, are the reference for guidance. Position data are needed to guide agricultural vehicles, to control implements and to support precision farming. Accuracy, resolution and frequency depend on their application. E-mail address: hermann.speckmann@fal.de (H. Speckmann) 0168-1699:00:$ - see front matter ? 2000 Elsevier Science B.V. All rights reserved. PII: S0168-1699(99)00057-5 H. Speckmann : Computers and Electronics in Agriculture 25 (2000) 87–10688 It must be emphasized that this paper does not address the problem of suitable sensors to generate the data. Rather, the problem studied here is that a position signal is generated with reference to a certain location on the mobile unit, but this position is not identical with the location where the position data are needed. Moreover, position information may be needed for several purposes at the same time, and the configuration of the vehicle–implement combination may change frequently. As mentioned by Freyberger and Jahns (1999), Wilson (1999), the measuring system can either be an absolute position system, such as the satellite system described by Bell (1999), or a relative system, such as the machine vision systems described by Debain et al. (1999), Hague et al. (1999). It may also include auxiliary sensors. Sensor systems measure position only in reference to a specific location, such as the mounting point of the camera or the foot of the aerial. In the following presentation, this location is called the measuring point. For various reasons, the location of this measuring point is predetermined, meaning the satellite antenna will be mounted as high as possible on the roof of the tractor cab to minimize shading. A camera will be mounted where optimal view is guaranteed. Movement caused by rough or sloping field surfaces may cause the measured position and the position on the field surface to differ widely. For example, for a vehicle with a satellite antenna mounted on top of the cab, at about 3.5 m, driving on a sloping surface of 10°, the difference in direction of the inclination will be about 60 cm. Fig. 1 illustrates this scenario for one dimension. In this example, it may be appropriate to calculate the position of a reference point. Bell (1999) proposes the middle rear axes of the Fig. 1. Difference in position for two locations due to sloping terrain. H. Speckmann : Computers and Electronics in Agriculture 25 (2000) 87–106 89 tractor as a reference point. A point in the field surface, for example, vertical under the middle of the rear axis seems more appropriate for some applications. For certain applications, such as the control of implements, the position of a certain point of the implement may be of final importance. This point will be called the target point. In cases where position data are needed for different purposes, it is not very efficient to measure the position for each purpose separately with an independent measuring system. Multiple hardware can be avoided when the position is measured only once, and the positions of the other points on the vehicle or implements are calculated. This is possible if position and attitude are measured, and the spatial vector between the measuring point and the point to be calculated is known. If both points are rigidly coupled, meaning that both points are on the tractor, the vector between these points is constant, and a simple matrix calculation yields the result. If these points are not rigidly coupled, meaning, for example, that one point is on the tractor and the other is on an attached implement, the vector is variable. Additional measurements become necessary to establish the vector between these two points or other principles to calculate the position of the target point must be applied. 2. Data processing and data transfer Position data of any point on the vehicle or implement can be calculated from the position and attitude measured at a measuring point. This calculation can be made by the measuring system (central data processing) or by each system requesting target position data (distributed data processing). 2.1. Distributed data processing The measuring system serves only as an intelligent sensor in the case of distributed data. It measures position and attitude on request, and provides these data without any processing. Characteristics such as frequency and accuracy are determined by the requesting unit. This unit performs all processing to calculate the position. The unit must know the position of the measuring point and all relevant parameters to do this. The advantage of this procedure is that the measuring device can be relatively simple. On the other hand, each requesting unit needs the full capacity to perform this calculation. 2.2. Central data processing The measuring unit is extended by components to calculate the position of target points for any user. This measuring and processing system forms one unit of a so-called position and navigation service (PNS), which provides final position data of any target point. In this case, only one measuring and processing system is necessary even when position data are requested by more than one user. To do so, only the PNS must know all of the relevant parameters for the calculation. H. Speckmann : Computers and Electronics in Agriculture 25 (2000) 87–10690 2.3. Data transfer A data transfer is necessary no matter where the data are processed. For such a data transfer, a standardized network is appropriate. For agricultural purposes, a BUS for data transfer between mobile units and stationary farm computers exists. The agricultural BUS system (LBS) has been standardized to exchange information between the electronic units (LBS participants or BUS nodes) in a network. The standard defines the physical layer of the network, network protocol, system management, data objects and central services for common tasks (Speckmann and Jahns, 1999). The LBS has been standardized as DIN 9684 (DIN, 1989–1998). Currently, efforts are being made to establish an international standard (Nienhaus, 1993), ISO 11783, for such purposes. Like LBS, ISO 11783 will also define an agri- cultural BUS as an open system to exchange data on agricultural machinery, particularly on tractor–implement combinations and from the mobile units to the stationary farm computer. The standards are based on the controller area network data protocol (CAN; BOSCH, 1991). Corresponding hardware is on the market. In the LBS, data objects are defined for the transmission of general position data (geographical positions: longitude, latitude, altitude, or position in a tramline). The standard allows definition of additional data objects such as multidimensional distances, directions and speeds. No data objects exist presently in the LBS for geometric implement parameters. ISO 11783 provides, in Part 7 (Implement Mes- sages Application Layer), the first definitions of implement navigational offsets. Current standards do not define where which data are processed. Therefore, it is immaterial on which unit the BUS calculates the data for the target point, and which unit or units measure the data. The LBS provides so-called LBS services to execute common tasks. LBS services are functional units, which perform frequently recurring tasks for LBS participants. Such a service is the LBS user station. This is a central interface to the user (operator) for input and output of data which is at the disposal of any node (LBS participant) on the BUS. Another service provides the data exchange between the mobile unit and the stationary computer, the farm computer. Some more services are named in the LBS but not yet stan- dardized in detail, such as for diagnosis services or the service ‘Ortung und Navigation’ (position and navigation), which will be discussed in the following as PNS. In Fig. 2, an exemplary simplified scheme of an agricultural network is shown for a tractor–sprayer combination. This scheme includes the physical BUS line, which is the backbone of the network. At this BUS, participants such as electronic control units (ECUs) of the tractor and sprayer are coupled. Additionally, two LBS services are connected on the BUS. One of these services represents the LBS user station. The other is the LBS service ‘position and navigation’, with the measuring and processing system for position data. H. Speckmann : Computers and Electronics in Agriculture 25 (2000) 87–106 91 Fig. 2. Scheme of an agricultural network in a tractor–sprayer combination. 2.4. Comparison of distributed and central data processing For a distributed data processing, the agricultural BUS, according to DIN 9684 or ISO 11783, defines the necessary data exchange between the measuring system and any participant; respectively, any ECU. The question how each ECU gets geometric and kinematic parameters that are necessary to compute position data remains open. Each ECU knows its own parameter from its coupling point to the target point, but it does not know the parameter from the coupling point to the measuring point. These parameters must be provided from other ECUs. None of the standards define corresponding data objects or procedures requesting the data. For distributed data processing, these definitions have to be supplemented. Also, for central data processing, all kinematic parameters between the measur- ing point and the target point must be known. In addition, methods are to be defined for the use of the central service with regard to the calculation of position data of target points. A position and navigation service requires an extension of the standards, but the following advantages in practical use are essential: To determine the position data of a target point, the corresponding ECU has only one dialogue partner in the network. It works independently from the respective network configuration, delivers only its own parameters and receives only its specific position data. The PNS receives parameters from all ECUs. It knows all geometric conditions and kinematic parameters of the vehicle–implement combination. Thereby, an unambiguous determination of the position of any target point is possible. H. Speckmann : Computers and Electronics in Agriculture 25 (2000) 87–10692 The standard defines the procedures to calculate and present the position data of a target point unambiguously. The computing performance to calculate the position data is provided solely by the PNS. No computing capacity is needed for this purpose from the ECUs. As mentioned in the previous section, a service to provide position and naviga- tion data is already planned in the LBS. In the following, a sample solution of a PNS is presented. 3. Proposal for a positioning and navigation service At this time, it should be mentioned that the following description of a PNS is a proposal. It provides a platform for discussion, which may lead to the standard- ization of such a service. 3.1. Main features of a PNS The features of a PNS depend, first of all, on the purpose for which it will be used. From the foregoing, it is clear that position data are measured at one location and used at different locations. The following requirements must be fulfilled to provide the data needed to guide a vehicle, to control positions of implements and to assist any kind of precision farming: The PNS provides data related to the measurement point(s). The PNS provides data related to the reference point(s). The PNS provides data related to the target point(s). The characteristics of such a service are as follows: 1. The way the data are requested and transmitted is already standardized and defined by the LBS (DIN 9684) and will be standardized by ISO 11783. Therefore, it will not be discussed here. In the following, LBS will be used as a standardized agricultural BUS system. 2. The volume, accuracy, frequency and range of the data are determined by the purpose of the data. 3. The hardware and software to fulfil these demands should not be standardized, but be determined by the manufacturers. 3.2. Influence of the standard on measuring and calculation methods for position data The kinds of measuring systems and methods used to determine position data by the PNS is not in the scope of the standard. Systems based on satellites, machine vision, inertial navigation, geomagnetics or a combination of these may be applied. As a consequence, the manufacturer may determine how to generate the position data as long as he meets the stated requirements and accuracy. H. Speckmann : Computers and Electronics in Agriculture 25 (2000) 87–106 93 3.3. Integration of the PNS into an agricultural BUS system There are some benefits of integrating the positioning and navigation service into the LBS, because many characteristics are already defined. The LBS already includes the option of a PNS as part of the standard. It allows the realization of a service either as an independent physical unit or as a logical unit inside of another physical unit. The physical properties of the BUS interface and the BUS protocol (DIN 9684, part 2) are defined by the standard. For integration of the service into the LBS, the definitions of the system functions are decisive (DIN 9684, part 3). They define the performance of the nodes at the LBS. Part 3 also gives the general definitions of LBS services. An LBS service forms a point-to-point link with LBS participants. The use of a service by an LBS participant can neither be influenced by other users, nor can an LBS participant influence links between the service and other participants. All further definitions of the PNS are not yet standardized. 3.4. General mode of operation of the PNS For the design of the PNS, the following basic assumptions apply: 1. Each ECU knows only its parameters, meaning coordinates and numbers of reference points, target points, positions of couplings, vehicle types or wheelbases. 2. Only the ECU can define necessary time intervals, accuracy and resolution for position data, depending on the working conditions. 3. Each ECU can get different position data at arbitrary times. 4. Parameters and the way of calculating and providing position data will be defined before the working processes of the field machinery are started. 5. The PNS provides a library of procedures to calculate position data for standard implement and vehicle types. 6. Position data are provided automatically (cyclically) or on demand. To meet these requirements, the service provides the tools, and the ECUs determine how and which tools are used. This means they define one or several task(s). Such a task basically represents a list that includes commands to activate the specific tools. These tasks are sent to the PNS, which subsequently performs these tasks. Different tasks of one ECU are executed independently of each other. Fig. 3 illustrates the data transfer between the PNS and one ECU. It also shows the main parts of the PNS. The tools of the PNS include the system for measuring the position and attitude data of the measuring point, and a library of methods to process these data. Methods exist: to calculate position data (position methods); to calculate mean, maximum, minimum and integral values of position data (arithmetic methods); to export and import data (transport methods); to send data to the ECU (transmission methods); and to control the data processing (data control methods). H. Speckmann : Computers and Electronics in Agriculture 25 (2000) 87–10694 For some of these methods, the ECU has to define corresponding parameters. It also defines data objects for position data. The central tool of the PNS is the program system to execute the tasks defined by the ECU. Simplified, the program system interprets the instructions of the task, calls the corresponding methods, calculates the demanded position and sends the data to the ECU. For the definition of a task, the ECU generates a task resource. A task resource is mainly a list of instructions to call methods of the PNS or to call nested task resources. Parameters are defined by the ECU and placed in parameter resources. To store calculated position data, the ECU has to define data resources. The resources have to be transmitted from the ECU via the BUS to the PNS before activating corresponding tasks. Fig. 3. Strcture of a PNS and its data exchange with one ECU. H. Speckmann : Computers and Electronics in Agriculture 25 (2000) 87–106 95 Fig. 4. Example of the use of a position method in the course of a task resource. 3.5. Predefined methods of the PNS Predefined methods of the PNS are procedures to process position data or to control this data processing. Methods exist to perform different functions. The different methods are distinguished by a unique designator. They are called ‘within tasks’ (task resources). It will be a part of the standard to define the designators, function specifications and calling specifications of the methods. 3.5.1. Position methods Position methods (methods to calculate position data) are the basis for calculat- ing position data of target points. These methods calculate from an initial position (input position data, data of a reference point or previously computed data) the position of a new point (output position data, data of a target point or as an interim result). Position methods exist for different configurations (one-, two- or three-dimensional model considerations, rigidly coupled points, non-rigidly coupled points for several basic types of vehicles, implements and vehicle–implement combinations). These methods get their actual parameters (coordinates of the target point, vehicle length, width, height, type or wheelbases) from parameter resources which are defined by the concerned implement ECU. Fig. 4 shows a section of a task resource using a position method. The program system of the PNS executes this task resource. At a certain part of the task resource, it finds a calling instruction for a position method. This calling instruction includes the designator of the specific method and a reference to a relevant parameter resource. At this moment, the program system owns actual position data, which result from previous operations. Now it uses these actual data as input data, and the parameter resource reference for the position method. Then, it executes the specified m