1359943 九、發明說明 【發明所屬之技術領域】 本發明係關於一種對訊號的數量進行計數的計數裝 置'以及使用計數裝置測定干涉波形的數量而求取與測定 對象之距離的干涉型距離計^ 【先前技術】 利用因雷射所産生的光的干涉的距離測量因屬於非接 觸測定而不會干擾測定對象，自古以來一直被作爲高精度 的測定方法加以使用。近來，爲了實現裝置的小型化，半 導體雷射係作爲光測量用光源而被加以利用。以其代表例 而言’存在一種採用 FM外差式干涉儀（heterodyne interferometry)者。該者可進行較長距離的測量且精度良 好’但是由於在半導體雷射的外部採用干涉儀，故具有光 學系統較爲複雜的缺點。 相對於此，一種利用雷射的輸出光與來自測定對象的 返回光在半導體雷射內部的干涉（自耦合效應）的測量器 已被提出（參照例如非專利文獻1、非專利文獻2、非專 利文獻3 )。根據如上所示之自耦合型的雷射測量器，由 於內建光電二極體的半導體雷射_具有發光、干涉、受光 的各功能’故可大幅簡化外部干涉光學系統。因此，感測 器部僅爲半導體雷射與透鏡，與習知技術相比較，較爲小 型。另外’具有距離測定範圍大於三角測量法的特徵。 在第23圖中顯示ρρ型（Fabry-Perot型）半導體雷 1359943 射的複合共振器模型。在第23圖中，101係半導體雷射， 1 02係半導體結晶的壁開面，1 〇3係光電二極體，1 04係測 定對象。來自測定對象104的反射光的一部分容易返回到 振盪區域內。返回來的微弱的光與共振器101內的雷射光 耦合，動作變得不穩定而産生噪音（複合共振器雜訊或返 回光雜訊）。即使相對輸出光的相對返回光量極其微小， 因返回光引起的半導體雷射的特性的變化仍顯著呈現。如 上所示的現象並不限於法布里-伯羅（Fabry-Perot )型 (以下稱爲FP型）半導體雷射，在垂直共振腔面射雷射 (Vertical Cavity Surface Emitting Laser)型（以下稱爲 VCSEL 型）、分佈反饋雷射（Distributed Feedback Laser)型（以下稱爲DFB雷射型）等其他種類的半導體 雷射中也同樣呈現。 若將雷射的振盪波長設爲λ，由接近測定對象104的 壁開面1 02到測定對象1 04爲止的距離設爲L，則在滿足 以下的共振條件時，返回光和共振器1 0 1內的雷射相互加 強，雷射輸出稍稍增加。 L = ηλ y 2 ··· (1) 在式（1)中，η爲整數。該現象係即便在來自測定對 象104的散射光極其微弱的情況下，亦可因半導體雷射的 共振器101內的表觀反射率增加，而産生放大作用，而足 以充分觀測。 -5- 1359943 在半導體雷射中，由於按照注入電流的大小而放射頻 率不同的雷射光，因此在將振盪頻率進行調變時，不需要 外部調變器，即可藉由注入電流直接進行調變。第24圖 係顯示在按照某個恒定的比例使半導體雷射的振盪波長改 變時的振盪波長和光電二極體103的輸出波形的關係圖。 在滿足式（1)所示的！>=ηλ/2時，返回光和共振器101 內的雷射光的相位差變成〇° (同相位），返回光和共振器 101內的雷射光最爲相互加強，當ηλ/ 2+ λ/ 4時， 相位差爲180° (逆相位），返回光和共振器101內的雷射 光相互最爲相互減弱。由此，如果使半導體雷射的振盪波 長改變，則雷射輸出增強的情形和減弱的情形交替反覆地 出現，若利用設置於共振器1 〇 1的光電二極體1 〇3檢測此 時的雷射輸出時，如第24圖所示，獲得恒定周期的階梯 狀波形。這樣的波形一般稱爲干涉條紋。 將該階梯狀的波形，亦即干涉條紋中的一個個稱爲模 式跳躍脈衝（mode hop pulse)(以下稱之爲MHP)。 MHP是與模式跳躍（mode hopping)現象不同的現象。例 如，在距測定對象104的距離爲L1時，如果MHP的數量 爲10個，則在一半的距離L2時，MHP的數量爲5個。亦 即，在某一定時間使半導體雷射的振盪波長改變時，Μ HP 的數量與測定距離成正比地改變。因此，如果利用光電二 極體103檢測MHP，並測定MHP的頻率，則可容易進行 距離測量。 (非專利文獻1 )上田正、山田諄、紫藤進，“利用 1359943 半導體雷射的自耦合效應的距離計” ，1994年度電氣關係 學會東海支部聯合大會演講論文集，1994年 (非專利文獻2)山田諄、紫藤進、津田紀生、上田 正’ “利用半導體雷射的自耦合效應的小型距離計的相關 硏究”，愛知工業大學硏究報告，第31號B，p.35至 42 ， 1996 年 (非專利文獻 3) Guido Giuliani，Michele Norgia， Silvano Donati and Thierry Bosch， Laser diode se 1 f-mixing technique for sensing applications” ’ JOURNAL OF OPTICS A ： PURE AND APPLIED OPITCS，p.283 至 294 ， 2002 年 【發明內容】 (發明所欲解決之課題） 在包含自耦合型之習知的干涉型距離計中，係藉由使 用計數裝置來測定MHP的數量，或者使用 FFT ( Fast Fourier Transform，快速傅立葉變換）來測定MHP的頻 率，以求取與測定對象的距離。 但是，在使用FFT的距離計中，當雷射的振盪波長 變化相對於時間爲非線性，會有在以F F Τ所計算出的峰値 頻率與原本應求出的ΜΗΡ的平均頻率中產生差異，而使 所測定出的距離產生誤差的問題。 此外’在使用計數裝置的距離計中，例如將擾亂光等 雜訊作爲ΜΗΡ而進行計數或者因欠缺訊號而有無法計數 1359943 的的MHP，而會有在計數裝置所計數的MHP數目 誤差，而使所測定出的距離產生誤差的問題。 其中，如上所示之計數誤差並不限於距離計， 計數裝置中亦同樣會發生。 本發明係爲了解決上述課題而硏發者，其目的 可補正計數誤差的計數裝置及計數方法，且可補] 之計數誤差而使距離的測定精度提升的距離計及距 方法。 (解決課題之手段） 本發明係特定的物理量與訊號數量具有線性關 述物理量爲一定時，係對成爲大致單一頻率的前述 行計數的計數裝置，其係具有：計數手段，對一定 間中的輸入訊號數量進行計數；周期測定手段，在 輸入訊號時，即測定前述計數期間中之前述輸入訊 期；頻率分佈作成手段，由該周期測定手段的測定 作成前述計數期間中之訊號周期的頻率分佈；代表 手段，根據前述頻率分佈，計算前述輸入訊號之周 的代表値；以及補正値計算手段，根據前述頻率分 出爲前述代表値之第1預定數倍以下之等級的頻 Ns、及爲前述代表値之第2預定數倍以上之等級的 和Nw，根據該等頻率Ns及Nw，補正前述計數手 數結果。 此外，本發明之計數裝置之一構成例中，前述 中發生 在其他 在提供 三 MHP 離測量 係，前 訊號進 計數期 每次被 號的周 結果， 値計算 期分佈 佈，求 率總和 頻率總 段的計 代表値 -8- 1359943 係中位數、眾數、平均値中的任一者。 此外，本發明之計數裝置之一構成例中，前述補正値 計算手段係當將前述計數手段的計數結果設爲N時，藉由 Ν’ = N + Nw-Ns，求出補正後之計數結果N，。 此外，本發明之計數裝置之一構成例中，前述第1預 定數爲0.5，前述第2預定數爲1.5。 此外，本發明之距離計係具有：半導體雷射，對測定 對象放射雷射光；雷射驅動器，以使至少包含振盪波長連 續單調遞增的期間的第1振盪期間與至少包含振盪波長連 續單調遞減的期間的第2振盪期間交替存在的方式使前述 半導體雷射進行動作；受光器，將由前述半導體雷射所放 射的雷射光與來自前述測定對象的返回光的干涉光轉換成 電訊號；計數手段，對該受光器的輸出訊號所包含之因由 前述半導體雷射所放射的雷射光與來自前述測定對象的返 回光所產生的干涉波形的數量進行計數；周期測定手段， 在每次被輸入干涉波形時，即測定對前述干涉波形數量進 行計數的前述計數期間中之前述干涉波形的周期；頻率分 佈作成手段，由該周期測定手段的測定結果，作成前述計 數期間中之干涉波形之周期的頻率分佈；代表値計算手 段，根據前述頻率分佈，計算前述干涉波形之周期分佈的 代表値；補正値計算手段，根據前述頻率分佈，求出爲前 述代表値之第1預定數倍以下之等級的頻率總和Ns、及 爲前述代表値之第2預定數倍以上之等級的頻率總和 Nw，根據該等頻率Ns及Nw，補正前述計數手段的計數 -9- 1359943 結果；以及運算手段’根據由該補正値計算手段所補正的 計數結果，求出與前述測定對象的距離。 此外’本發明之距離計係具有：半導體雷射，對測定 對象放射雷射光；雷射驅動器，以使至少包含振盪波長連 續單調遞增的期間的第1振盪期間與至少包含振盪波長連 續單調遞減的期間的第2振盪期間交替存在的方式使前述 半導體雷射進行動作；檢測手段，檢測包含由前述半導體 雷射所放射的雷射光與來自前述測定對象的返回光因自耦 合效應所產生的干涉波形的電訊號；計數手段，對該檢測 手段的輸出訊號所包含之前述干涉波形的數量進行計數； 周期測定手段’在每次被輸入干涉波形時，即測定對前述 干涉波形數量進行計數的計數期間中之前述干涉波形的周 期；頻率分佈作成手段，由該周期測定手段的測定結果， 作成前述計數期間中之干涉波形之周期的頻率分佈；代表 値計算手段，根據前述頻率分佈，計算前述干涉波形之周 期分佈的代表値；補正値計算手段，根據前述頻率分佈， 求出爲前述代表値之第1預定數倍以下之等級的頻率總和 N s、及爲前述代表値之第2預定數倍以上之等級的頻率總 和Nw ’根據該等頻率Ns及Nw，補正前述計數手段的計 數結果；以及運算手段，根據由該補正値計算手段所補正 的計數結果，求出與前述測定對象的距離。 此外，本發明之距離計之一構成例中，前述檢測手段 係將前述半導體雷射的光輸出轉換成電訊號的受光器。 此外’本發明之距離計之一構成例中，前述檢測手段 -10- 1359943 係用以檢測前述半導體雷射之端子間電壓的電壓檢測手 段。 此外，本發明之距離計之一構成例中，前述代表値係 中位數、眾數、平均値中的任一者。 此外，本發明之距離計之一構成例中，前述補正値計 算手段係當將前述計數手段的計數結果設爲N時，藉由 N，= N + Nw-Ns，求出補正後之計數結果N，。 此外，本發明之距離計之一構成例中，前述第1預定 數爲0.5，前述第2預定數爲1.5。 此外，本發明之計數方法係具有：計數步驟，對一定 計數期間中的輸入訊號數量進行計數；周期測定步驟，在 每次被輸入訊號時，即測定前述計數期間中之前述輸入訊 號的周期；頻率分佈作成步驟，由該周期測定步驟的測定 結果，作成前述計數期間中之訊號周期的頻率分佈；代表 値計算步驟，根據前述頻率分佈，計算前述輸入訊號之周 期分佈的代表値；以及補正値計算步驟，根據前述頻率分 佈，求出爲前述代表値之第1預定數倍以下之等級的頻率 總和Ns、及爲前述代表値之第2預定數倍以上之等級的 頻率總和Nw，根據該等頻率Ns及Nw，補正前述計數步 驟的計數結果。 此外，本發明之距離測量方法係具備有：振盪步驟， 以使至少包含振盪波長連續單調遞增的期間的第1振盪期 間與至少包含振盪波長連續單調遞減的期間的第2振盪期 間交替存在的方式使前述半導體雷射進行動作；檢測步 -11 - 1359943 與 數 波 波 中 測 周 計 9 以 數 該 象 期 期 步 前 電 包 每 計 驟，藉由受光器，將由前述半導體雷射所放射的雷射光 來自前述測定對象的返回光的干涉光轉換成電訊號；計 步驟，對在該檢測步驟所獲得的輸出訊號所包含的干涉 形的數量進行計數；周期測定步驟，在每次被輸入干涉 形時，即測定對前述干涉波形數量進行計數的計數期間 之前述干涉波形的周期；頻率分佈作成步驟，由該周期 定步驟的測定結果，作成前述計數期間中之干涉波形之 期的頻率分佈；代表値計算步驟，根據前述頻率分佈， 算前述干涉波形之周期分佈的代表値；補正値計算步驟 根據前述頻率分佈，求出爲前述代表値之第1預定數倍 下之等級的頻率總和Ns、及爲前述代表値之第2預定 倍以上之等級的頻率總和Nw，根據該等頻率Ns及Nw 補正前述計數步驟的計數結果；以及運算步驟，根據由 補正値計算步驟所補正的計數結果，求出與前述測定對 的距離。 此外，本發明之距離測量方法係具備有：振盪步驟 以使至少包含振盪波長連續單調遞增的期間的第1振盪 間與至少包含振盪波長連續單調遞減的期間的第2振盪 間交替存在的方式使前述半導體雷射進行動作；檢測 驟，檢測包含由前述半導體雷射所放射的雷射光與來自 述測定對象的返回光因自耦合效應所產生的干涉波形的 訊號：計數步驟，對在該檢測步驟所獲得的輸出訊號所 含的前述干涉波形的數量進行計數；周期測定步驟，在 次被輸入干涉波形時，即測定對前述干涉波形數量進行 -12- 1359943 數的計數期間中之前述干涉波形的周期；頻率分佈作成步 驟，由該周期測定步驟的測定結果’作成前述計數期間中 之干涉波形之周期的頻率分佈；代表値計算步驟，根據前 述頻率分佈，計算前述干涉波形之周期分佈的代表値；補 正値計算步驟，根據前述頻率分佈，求出爲前述代表値之 第1預定數倍以下之等級的頻率總和Ns、及爲前述代表 値之第2預定數倍以上之等級的頻率總和Nw，根據該等 頻率Ns及Nw，補正前述計數步驟的計數結果；以及運算 步驟，根據由該補正値計算步驟所補正的計數結果，求出 與前述測定對象的距離。 (發明之效果） 藉由本發明，測定計數期間中之輸入訊號的周期，根 據該測定結果作成計數期間中的訊號周期的頻率分佈，根 據該頻率分佈來計算輸入訊號之周期的代表値，根據頻率 分佈’求取爲代表値之第1預定數倍以下之等級的頻率總 和Ns、及爲代表値之第2預定數倍以上之等級的頻率總 和Nw’根據該等頻率Ns和Nw來補正在計數手段的計數 結果’藉此可去除計數時的缺漏或過剩計數的影響，而補 正計數裝置的計數誤差。 此外’在本發明中，測定計數期間中的干涉波形的周 期’根據該測定結果作成計數期間中之干涉波形之周期的 頻率分佈’根據頻率分佈來計算干涉波形之周期的代表 値’根據頻率分佈，求取爲代表値之第丨預定數倍以下之 -13- 1359943 等級的頻率總和Ns、及爲代表値之第2預定數倍以上之 等級的頻率總和Nw，根據該等頻率Ns和Nw來補正計數 手段的計數結果，藉此可去除計數時的缺漏或過剩計數的 影響’而補正計數裝置的計數誤差，因此可在使用計數手 段來測定干涉波形的數量而求取與測定對象的距離的距離 計中，使距離的測定精度提升。 【實施方式】 (第1實施形態） 本發明係一種根據在採用波長調變的感測（sensing ) 時射出的波與由對象物反射的波的干涉訊號，來測量距離 的手法。因此，亦可適用於自耦合型以外的光學式干涉 計 '光以外的干涉計。若針對採用半導體雷射的自耦合的 情形更加具體說明，當一面由半導體雷射對測定對象照射 雷射光，一面使雷射的振盪波長改變時，在振盪波長從最 小振盪波長變化爲最大振盪波長的期間（或從最大振盪波 長至最小振盪波長變化的期間）中之測定對象的位移係反 映在MHP的數量。因此，可藉由調查使振盪波長改變時 的MHP的數量，來檢測測定對象的狀態。以上爲干涉計 的基本原理。 以下參照圖示，說明本發明的第1實施形態。第1圖 係顯示本發明第1實施形態的距離計的構成的方塊圖。第 1圖的距離計係具有：對測定對象1 2放射雷射光的半導體 雷射1;將半導體雷射1的光輸出轉換成電訊號的光電二 -14- 1359943 極體2;分別將來自半導體雷射1的光聚光而照射 對象12，並且將來自測定對象12的返回光聚光而 入半導體雷射1的透鏡3;使半導體雷射1交替地 盪波長連續地增加的第1振盪期間和振盪波長連續 的第2振盪期間的雷射驅動器4;將光電二極體2 電流轉換成電壓並進行放大的電流-電壓轉換放大器 電流-電壓轉換放大器5的輸出電壓去除載波的濾 路11;對濾波器電路11的輸出電壓所包含的MHP 進行計數的計數裝置8 ;根據MHP的數量，計算與 象12的距離的運算裝置9;以及顯示運算裝置9的 果的顯示裝置10。 以下爲了容易說明，假定在半導體雷射1採用 模式跳躍（mode hopping)現象的類型（VCSEL型 雷射型）。 例如’雷射驅動器4係將關於時間按照恒定的 而反覆增減的三角波驅動電流作爲注入電流而供給 體雷射1。由此，半導體雷射1係以交替反覆第1 間和第2振盪期間的方式被驅動，該第1振盪期間 波長與注入電流的大小成正比以恒定的變化率連 加，該第2振盪期間係振盪波長以恒定的變化率連 少。 第2圖係顯示半導體雷射丨的振盪波長之時 圖。在第2圖中，t-Ι表示第（t_;i)個振盪期間； 第t個振盪期間；t+1表示第（t+丨）個振盪期間； 在測定 使其射 反覆振 地減少 的輸出 5 :由 波器電 的數量 測定對 計算結 不具有 、DFB 變化率 至半導 振盪期 係振盪 續地增 續地減 間變化 t表示 t + 2表 -15- 1359943 示第（t + 2 )個振盪期間；t + 3表不第（t + 3 )個振盪期 間；t + 4表示第（t + 4 )個振盪期間；A a表示各期間中的 振盪波長的最小値；Ab表示各期間中的振盪波長的最大 値：T表示三角波的周期。在本實施形態中，振盪波長的 最大値λ b及振盪波長的最小値A a係分別恒爲一定，該 等的差値Xb-Xa亦恒爲一疋。 由半導體雷射1射出的雷射光係藉由透鏡3予以聚光 而射入測定對象1 2。以測定對象1 2予以反射的光係藉由 透鏡3予以聚光，並射入半導體雷射1。其中，透鏡3的 聚光並非爲必須。光電二極體2係將半導體雷射1的光輸 出轉換成電流。電流-電壓轉換放大器5係將光電二極體2 的輸出電流轉換成電壓並進行放大。 濾波器電路Π具有從調變波抽出重疊訊號的功能。 第3圖（A)係以模式顯示電流-電壓轉換放大器5的輸出 電壓波形圖，第3圖（B )係以模式顯示濾波器電路Π的 輸出電壓波形圖》該等圖係表示由相當於光電二極體2之 輸出的第3圖（A)的波形（調變波），去除第2圖的半 導體雷射1的振盪波形（載波），以抽出第3圖（B)的 MHP波形（重疊波）的過程。 計數裝置8係針對濾波器電路11的輸出電壓所包含 之MHP的數量，就第1振盪期間t-1、t+1' t + 3與第2振 盪期間t、t + 2、t + 4的各個期間進行計數。第4圖係顯示 計數裝置8之構成之一例的方塊圖。計數裝置8係由：判 定部 81、邏輯與運算部（AND) 82、計數器83、計數結 -16- 1359943 果補正部84 '及記億部85所構成。電流-電壓轉換放大器 5、滬波器電路11、及計數裝置8的判定部81、AND82與 計數器83係構成計數手段。 第5圖係顯示計數結果補正部84之構成之1例的方 塊圖。計數結果補正部84係由周期測定部840、頻率分佈 作成部841、代表値計算部842、以及補正値計算部843 _ 所構成。 φ 第6圖（A)至第6圖（F)係用以說明計數裝置8的 動作圖’第ό圖（A)係以模式顯示濾波器電路π之輸出 電壓的波形，亦即ΜΗΡ的波形的圖，第6圖（Β)係顯示 與第ό圖（Α)相對應的判定部81的輸出的圖，第6圖 (C)係顯示被輸入至計數裝置8的閘極訊號GS的圖， 第6圖（D)係顯示與第6圖（Β)相對應之計數器83的 計數結果的圖，第6圖（Ε)係顯示被輸入至計數裝置8 的時鐘訊號CLK的圖，第6圖（F)係顯示與第6圖 φ ( Β )相對應之周期測定部840之測定結果的圖。 • 首先’計數裝置8的判定部81係判定第6圖（A )所 . 示的濾波器電路11的輸出電壓爲高位準（H)或低位準 (L )，輸出如第6圖（B )所示的判定結果。此時，判定 部81係在濾波器電路11的輸出電壓上升至臨限値TH1以 上時，判定爲高位準；濾波器電路11的輸出電壓下降至 臨限値TH2 ( TH2 < TH1 )以下時，判定爲低位準，藉此 將濾波器11的輸出進行二値化。 AND82係輸出判定部81的輸出與第6圖（C)所示 -17- 1359943 的閘極訊號GS的邏輯與運算的結果，計數器83係對 AND82的輸出的上升進行計數（第6圖（D))。在此， 閘極訊號G S係在計數期間（在本實施形態中係第〗振盪 期間或第2振盪期間）的起始時上升，在計數期間的結束 時下降的訊號。因此，計數器83係對計數期間中之 AND 8 2的輸出的上升邊緣的數量（亦即MHP之上升邊緣 的數量）進行計數。 另一方面’計數結果補正部84的周期測定部840係 在每次發生上升邊緣時，即對計數期間中之AND82的輸 出的上升邊緣的周期（亦即MHP的周期）進行測定。此 時’周期測定部840係以第6圖（E)所示之時鐘訊號 CLK的周期爲1個單位，測定MHP的周期。在第6圖 (F )之例中’周期測定部8 4 0係依序測定τ <»、Τ θ及T r 作爲MHP的周期。由第6圖（E)、第6圖（F)可知，. 周期Τα、Τθ及Tr的大小係分別爲5時鐘、4時鐘、2時 鐘。時鐘訊號CLK的孽率遠大於MHP所可取得的最高頻 率。 記憶部8 5係記憶計數器8 3的計數結果及周期測定部 840的測定結果。 在閘極訊號G S下降、計數期間結束之後，計數結果 補正部8 4的頻率分佈作成部8 4 1係根據記憶在記億部8 5 的測定結果’作生計數期間中之MHP的周期的頻率分 佈。 接著’計數結果補正部84的代表値計算部842係根 -18- 1359943 據由頻率分佈作成部841所作成的頻率分佈，計算 之周期的中位數（median) TO。 計數結果補正部84的補正値計算部84 3係根據 分佈作成部841所作成的頻率分佈，求取爲周期中 TO之0.5倍以下的等級的頻率總和Ns、及爲周期中 T0之1.5倍以上的等級的頻率總和Nw，並如下補 器8 3的計數結果。 N’=N + Nw-Ns --(7 在式（2)中，N係作爲計數器83之計數結果的 的數量，Ν’係補正後的計數結果^ 在第7圖中顯示頻率分佈之一例。在第7圖中， 周期中位數Τ0的0.5倍的等級値，Tw係周期中位| 的1.5倍的等級値。當然第7圖中的等級乃爲MHP的 的代表値。其中，在第7圖中爲了簡化記載，省略圖 位數T0與Ts之間、及中位數T0與Tw之間的頻 佈。 第8圖係用以說明計數器83之計數結果之補正 的圖，第8圖（A)係以模式顯示濾波器電路11之輸 壓的波形’亦即MHP的波形的圖，第8圖（B)係顯 第8圖（A)相對應之計數器83的計數結果的圖。 原本MHP的周期係依與測定對象12的距離而異 是若與測定對象1 2的距離不變，則MHP係以相同的 MHP 頻率 位數 位數 計數 MHP Ts係 設T0 周期 示中 率分 原理 出電 示與 ，但 周期 -19- 1359943 不 誤 的 周 缺 訊 果 Ts 爲 因 訊 此 倍 正 第 的 出現。但是由於雜訊而會在MHP波形發生缺漏或産生 應作爲訊號進行計數的波形，而在ΜΗΡ的數量產生 差。 當發生訊號缺漏時，在已發生缺漏的部位的ΜΗΡ 周期Tw係成爲原本周期的大約2倍。亦即，當ΜΗΡ的 期約爲中位數的2倍以上時，係可判斷在訊號中已發生 漏。因此’將周期Tw以上之等級的頻率總和Nw視爲 號缺漏的次數，並將該Nw加算在計數器83的計數結 Ν’藉此可補正訊號的缺漏。 此外，在將雜訊進行計數後的部位的ΜΗΡ的周期 係成爲原本周期的大約0.5倍。亦即，當ΜΗΡ的周期約 中位數的〇 . 5倍以下時，係可判斷已過剩計數訊號。 此，將周期T s以下之等級的頻率總和N s視爲過剩計數 號的次數，並由計數器83的計數結果Ν減算該Ns，藉 可補正誤數的雜訊。 以上爲式（2)所示之計數結果的補正原理。其中 在本實施形態中，係將Ts設爲周期中位數T0的0.5倍 將Tw設爲中位數T0的1 .5倍的値而非中位數T0的2 的値，設爲1. 5倍的理由容後陳述。 補正値計算部8 4 3係將藉由式（2 )計算所得之補 後計數結果Ν’輸出至運算裝置9。計數裝置8係按每個 1振盪期間t-1、t+1、t + 3及每個第2振盪期間t、t + 2 t + 4進行如以上所示之處理。 接著’運算裝置9係根據藉由計數裝置8測量而得 -20-1359943 IX. Description of the Invention [Technical Field] The present invention relates to a counting device for counting the number of signals and an interference type distance meter for determining the distance from an object to be measured by measuring the number of interference waveforms using a counting device. [Prior Art] The distance measurement using the interference of the light generated by the laser does not interfere with the measurement target because it belongs to the non-contact measurement, and has been used as a highly accurate measurement method since ancient times. Recently, in order to achieve miniaturization of a device, a semiconductor laser system is utilized as a light source for light measurement. In its representative case, there is a person who uses a heterodyne interferometry. This person can perform long-distance measurement with good precision. However, since an interferometer is used outside the semiconductor laser, it has the disadvantage that the optical system is complicated. On the other hand, a measuring device that uses the laser output light and the return light from the measuring object to interfere with the inside of the semiconductor laser (self-coupling effect) has been proposed (see, for example, Non-Patent Document 1, Non-Patent Document 2, Non-Patent Document) Patent Document 3). According to the self-coupling type laser measuring device as described above, since the semiconductor laser having the built-in photodiode has various functions of light emission, interference, and light reception, the external interference optical system can be greatly simplified. Therefore, the sensor portion is only a semiconductor laser and a lens, and is relatively small compared with the conventional technology. In addition, there is a feature that the distance measurement range is larger than the triangulation method. In Fig. 23, a composite resonator model of a ρp type (Fabry-Perot type) semiconductor mine 1359943 is shown. In Fig. 23, a 101-semiconductor laser, a wall-opening surface of a 012-series semiconductor crystal, a 1 〇3-series photodiode, and a 04-series measurement object are shown. A part of the reflected light from the measurement object 104 is easily returned to the oscillation area. The weak light returned is coupled with the laser light in the resonator 101, and the operation becomes unstable to generate noise (composite resonator noise or return light noise). Even if the amount of relative return light relative to the output light is extremely small, the change in the characteristics of the semiconductor laser due to the return light is remarkably exhibited. The phenomenon shown above is not limited to the Fabry-Perot type (hereinafter referred to as FP type) semiconductor laser, and is a vertical cavity cavity type laser (Vertical Cavity Surface Emitting Laser) type (hereinafter referred to as The same is true for other types of semiconductor lasers such as the VCSEL type and the Distributed Feedback Laser type (hereinafter referred to as DFB laser type). When the oscillation wavelength of the laser is λ and the distance from the wall opening surface 102 of the measurement target 104 to the measurement target 104 is L, the return light and the resonator 10 are returned when the following resonance conditions are satisfied. The lasers in 1 are mutually reinforcing and the laser output is slightly increased. L = ηλ y 2 (1) In the formula (1), η is an integer. In this case, even when the scattered light from the measurement target 104 is extremely weak, the apparent reflectance in the resonator 101 of the semiconductor laser is increased to cause amplification, and sufficient observation is sufficient. -5- 1359943 In semiconductor lasers, since laser light of different frequencies is radiated according to the magnitude of the injection current, when the oscillation frequency is modulated, an external modulator is not required, and the current can be directly adjusted by the injection current. change. Fig. 24 is a graph showing the relationship between the oscillation wavelength and the output waveform of the photodiode 103 when the oscillation wavelength of the semiconductor laser is changed at a constant ratio. In the satisfaction of the formula (1)! >=ηλ/2, the phase difference between the return light and the laser light in the resonator 101 becomes 〇° (in-phase), and the return light and the laser light in the resonator 101 are mutually mutually enhanced, when ηλ/ 2+ λ At /4, the phase difference is 180° (reverse phase), and the return light and the laser light in the resonator 101 mutually weaken each other most. Therefore, if the oscillation wavelength of the semiconductor laser is changed, the case where the laser output is enhanced and the case where the laser is amplified alternately appear repeatedly, if the photodiode 1 〇 3 provided in the resonator 1 〇 1 is detected at this time. When the laser output is output, as shown in Fig. 24, a stepped waveform of a constant period is obtained. Such waveforms are generally referred to as interference fringes. One of the staircase waveforms, i.e., the interference fringes, is referred to as a mode hop pulse (hereinafter referred to as MHP). MHP is a phenomenon different from the mode hopping phenomenon. For example, when the distance from the measurement target 104 is L1, if the number of MHPs is 10, the number of MHPs is five at half the distance L2. That is, when the oscillation wavelength of the semiconductor laser is changed for a certain period of time, the number of ΜHP changes in proportion to the measured distance. Therefore, if the MHP is detected by the photodiode 103 and the frequency of the MHP is measured, the distance measurement can be easily performed. (Non-Patent Document 1) Ueda Masahiro, Yamada Aya, and Wisteria, "Distance Meter Using Self-Coupling Effect of 1359944 Semiconductor Laser", Proceedings of the 1994 Joint Meeting of the East Asia Branch of the Institute of Electrical Relations, 1994 (Non-Patent Document 2) ) Yamada Aya, Wisteria Jin, Tsuda Kisho, and Ueda Masa's "Research on Small Distance Meters Using Self-Coupling Effects of Semiconductor Lasers", Aichi University of Technology Research Report, No. 31 B, p. 35 to 42, 1996 (Non-Patent Document 3) Guido Giuliani, Michele Norgia, Silvano Donati and Thierry Bosch, Laser diode se 1 f-mixing technique for sensing applications" ' JOURNAL OF OPTICS A : PURE AND APPLIED OPITCS, p.283 to 294, 2002 [Problem to be Solved by the Invention] In an interference type distance meter including a self-coupling type, the number of MHPs is measured by using a counting device, or FFT (Fast Fourier Transform) is used. Transform) to determine the frequency of the MHP to determine the distance from the measured object. However, in the distance meter using FFT, when the laser The variation of the oscillation wavelength is nonlinear with respect to time, and there is a problem that the peak 値 frequency calculated by FF 与 differs from the average frequency of ΜΗΡ which should be originally obtained, and the measured distance is erroneous. In addition, in the distance meter using the counting device, for example, the noise such as disturbing light is counted as ΜΗΡ or the MHP which cannot count 1359943 due to the lack of signal, and there is an error in the number of MHP counted by the counting device. The problem of causing an error in the measured distance is not limited to the distance meter, and the same is true in the counting device. The present invention is intended to solve the above problems, and the purpose thereof is to correct the counting. The distance counting device and the distance method for improving the measurement accuracy of the error by the counting device and the counting method of the error. (Means for Solving the Problem) The present invention has a linear relationship between a specific physical quantity and a signal quantity. In a certain period of time, it is a counting device that counts the aforementioned rows that become a substantially single frequency, and has a counting means, a pair The number of input signals in the interval is counted; the period measuring means measures the input period in the counting period when the signal is input; the frequency distribution forming means determines the signal period in the counting period by the measurement of the period measuring means a frequency distribution; a representative means for calculating a representative 値 of the circumference of the input signal according to the frequency distribution; and a correction 値 calculation means for dividing the frequency Ns of the first predetermined number of times or less of the representative 値 according to the frequency And the sum Nw of the level of the second predetermined number or more of the representative 値, and the result of the counting hand is corrected based on the frequencies Ns and Nw. In addition, in one of the configuration examples of the counting device of the present invention, the foregoing occurs in the other three in the provision of the three MHP off-set measurement system, the weekly result of the number of the pre-signal count period, the distribution period of the calculation period, and the total sum frequency. The segment count 値-8- 1359943 is any of the median, mode, and average. Further, in a configuration example of the counting device of the present invention, the correction correction means calculates the result of the correction by Ν' = N + Nw-Ns when the counting result of the counting means is N. N,. Further, in a configuration example of the counting device of the present invention, the first predetermined number is 0.5, and the second predetermined number is 1.5. Further, the distance meter of the present invention includes: a semiconductor laser that radiates laser light to a measurement target; and a laser driver that continuously monotonically decreases the first oscillation period including at least the oscillation wavelength continuously including the oscillation wavelength. The semiconductor laser is operated alternately in the second oscillation period of the period; the light receiver converts the laser light emitted by the semiconductor laser and the interference light from the measurement target into electrical signals; Counting the number of interference waveforms generated by the laser light emitted by the semiconductor laser and the return light from the measurement target included in the output signal of the photoreceiver; the period measuring means, each time the interference waveform is input And measuring a period of the interference waveform in the counting period in which the number of the interference waveforms is counted; and a frequency distribution forming means for generating a frequency distribution of a period of the interference waveform in the counting period by the measurement result of the period measuring means; Representing 値 calculation means, according to the aforementioned frequency The cloth calculates a representative 値 of the periodic distribution of the interference waveform; and the correction 値 calculating means obtains, based on the frequency distribution, a frequency sum Ns of a level equal to or less than a first predetermined number of times of the representative 値, and a 2 a frequency sum Nw of a predetermined number of times or more, correcting the result of the counting means -9 - 1359943 based on the frequencies Ns and Nw; and calculating means "based on the counting result corrected by the correcting means" The distance from the aforementioned measurement target. Further, the distance meter of the present invention includes: a semiconductor laser that radiates laser light to a measurement target; and a laser driver that continuously monotonically decreases the first oscillation period including at least the oscillation wavelength continuously including the oscillation wavelength. The semiconductor oscillation is performed in such a manner that the second oscillation period of the period alternates, and the detection means detects an interference waveform generated by the self-coupling effect of the laser light emitted from the semiconductor laser and the return light from the measurement target. a counting means for counting the number of the interference waveforms included in the output signal of the detecting means; the period measuring means 'measuring the counting period of counting the number of the interference waveforms each time the interference waveform is input a period of the interference waveform; a frequency distribution forming means for generating a frequency distribution of a period of the interference waveform in the counting period by the measurement result of the period measuring means; and representing the 値 calculating means, calculating the interference waveform based on the frequency distribution Representative of the periodic distribution値The correction 値 calculation means obtains, based on the frequency distribution, a frequency sum N s of a level equal to or less than a first predetermined number of times of the representative 値, and a frequency sum Nw ' of a level of a second predetermined number or more of the representative 値The counting result of the counting means is corrected based on the frequencies Ns and Nw, and the calculating means obtains the distance from the measurement target based on the counting result corrected by the correction calculating means. Further, in a configuration example of the distance meter of the present invention, the detecting means is a light receiver for converting the light output of the semiconductor laser into an electric signal. Further, in the configuration example of the distance meter of the present invention, the detecting means -10- 1359943 is a voltage detecting means for detecting the voltage between the terminals of the semiconductor laser. Further, in one configuration example of the distance meter of the present invention, the aforementioned representative system is any one of a median, a mode, and an average value. Further, in a configuration example of the distance meter according to the present invention, the correction 値 calculation means obtains the count result after correction by N, = N + Nw - Ns when the count result of the counting means is N. N,. Further, in a configuration example of the distance meter of the present invention, the first predetermined number is 0.5, and the second predetermined number is 1.5. In addition, the counting method of the present invention has a counting step of counting the number of input signals in a certain counting period, and a period measuring step of determining the period of the input signal in the counting period each time the signal is input; a frequency distribution forming step of generating a frequency distribution of the signal period in the counting period by the measurement result of the period measuring step; representing a 値 calculating step of calculating a representative 値 of the periodic distribution of the input signal according to the frequency distribution; and correcting 値a calculation step of obtaining a frequency sum Ns of a level equal to or less than a first predetermined number of times of the representative 値 and a frequency sum Nw of a level of a second predetermined number or more of the representative 値 based on the frequency distribution, according to the frequency distribution The frequencies Ns and Nw correct the count result of the aforementioned counting step. Further, the distance measuring method of the present invention includes an oscillating step of alternately presenting a first oscillation period including at least a period in which the oscillation wavelength continuously monotonically increases, and a second oscillation period including a period in which the oscillation wavelength continuously monotonically decreases. Performing the foregoing semiconductor laser to operate; detecting steps -11 - 1359943 and the number of waves in the measurement week 9 to count the number of steps before the image period, and by the receiver, the semiconductor laser is radiated The laser beam is converted into an electrical signal by the interference light of the return light from the measuring object; the counting step counts the number of the interference shapes included in the output signal obtained in the detecting step; the periodic measuring step is performed every time the input is interfered a shape, that is, a period of the interference waveform during a counting period in which the number of the interference waveforms is counted; a frequency distribution forming step of generating a frequency distribution of the period of the interference waveform in the counting period from the measurement result of the periodic determining step; Representing the 値 calculation step, calculating the circumference of the aforementioned interference waveform according to the aforementioned frequency distribution The representative of the distribution 値; the calculation step of the correction 求出 is based on the frequency distribution, and obtains the sum of the frequencies Ns of the level of the first predetermined number of times of the representative 値 and the sum of the frequencies of the second predetermined times or more of the representative 値Nw corrects the counting result of the counting step based on the frequencies Ns and Nw, and an arithmetic step of obtaining the distance from the measurement pair based on the counting result corrected by the correction 値 calculating step. Further, the distance measuring method according to the present invention is characterized in that the oscillating step is such that the first oscillation between the first oscillation period including at least the oscillation wavelength continuously increasing monotonously and the second oscillation period including at least the oscillation wavelength continuous monotonously decreasing period are alternately present. The semiconductor laser is operated; and the detecting step detects a signal including an interference waveform generated by the self-coupling effect of the laser light emitted by the semiconductor laser and the return light from the measuring object: a counting step, in the detecting step Counting the number of the interference waveforms included in the obtained output signal; the period measuring step, when the interference waveform is input, that is, measuring the interference waveform in the counting period of the number of the interference waveforms of -12 - 1359943 a frequency distribution forming step of generating a frequency distribution of a period of the interference waveform in the counting period by the measurement result of the period measuring step; representing a 値 calculating step of calculating a representative of a periodic distribution of the interference waveform based on the frequency distribution ; correct 値 calculation steps, according to the aforementioned frequency The distribution is obtained as a frequency sum Ns of a level equal to or less than a first predetermined number of times of the representative 値, and a frequency sum Nw of a level which is a second predetermined number or more of the representative 値, and is corrected based on the frequencies Ns and Nw The counting result of the counting step; and the calculating step determine the distance from the measurement target based on the counting result corrected by the correction calculating step. (Effect of the Invention) According to the present invention, the period of the input signal in the counting period is measured, the frequency distribution of the signal period in the counting period is generated based on the measurement result, and the representative of the period of the input signal is calculated based on the frequency distribution, according to the frequency The distribution 'supplied as the frequency sum Ns representing the level of the first predetermined number of times or less and the frequency sum Nw' representing the second predetermined number of times or more of the 値 is complemented by the frequencies Ns and Nw The counting result of the means 'by this can remove the influence of the missing or excess count at the time of counting, and correct the counting error of the counting device. Further, in the present invention, the period of the interference waveform in the measurement count period is measured, and the frequency distribution of the period of the interference waveform in the counting period is calculated based on the measurement result 'the representative of the period of the interference waveform is calculated from the frequency distribution' according to the frequency distribution , which is a sum of frequencies Ns of the level of -13,359,943, which is a predetermined number of times less than a predetermined number of times, and a frequency sum Nw of a level that is more than a predetermined number of times of the second predetermined number of times, according to the frequencies Ns and Nw By correcting the counting result of the counting means, the counting error of the counting device can be corrected by removing the influence of the missing or excessive counting at the time of counting. Therefore, the number of interference waveforms can be measured using the counting means to determine the distance from the measuring target. In the distance meter, the measurement accuracy of the distance is improved. [Embodiment] (First Embodiment) The present invention is a method for measuring a distance based on an interference signal of a wave emitted by sensing when wavelength modulation is applied and a wave reflected by an object. Therefore, it can also be applied to an optical interferometer other than the self-coupling type. More specifically, for the case of self-coupling using a semiconductor laser, when the laser beam is irradiated to the measuring object by a semiconductor laser, and the oscillation wavelength of the laser is changed, the oscillation wavelength is changed from the minimum oscillation wavelength to the maximum oscillation wavelength. The displacement of the measurement object in the period (or the period from the maximum oscillation wavelength to the minimum oscillation wavelength) is reflected in the number of MHPs. Therefore, the state of the measurement target can be detected by investigating the number of MHPs at which the oscillation wavelength is changed. The above is the basic principle of the interferometer. Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. Fig. 1 is a block diagram showing the configuration of a distance meter according to a first embodiment of the present invention. The distance meter of Fig. 1 includes: a semiconductor laser 1 that emits laser light to the measurement target 12; a photodiode of the semiconductor laser 1 that converts the light output into an electric signal, a diode 2 - 1359943 polar body 2; The light of the laser 1 is condensed to illuminate the object 12, and the return light from the measurement target 12 is condensed into the lens 3 of the semiconductor laser 1; the first oscillation period in which the semiconductor laser 1 is alternately increased in wavelength continuously a laser driver 4 during a second oscillation period in which the oscillation wavelength is continuous; a current-voltage conversion amplifier that converts the photodiode 2 current into a voltage and amplifies the output voltage of the current-voltage conversion amplifier 5 to remove the carrier filter 11; The counting means 8 for counting the MHP included in the output voltage of the filter circuit 11, the arithmetic means 9 for calculating the distance from the image 12 based on the number of MHPs, and the display means 10 for displaying the fruit of the arithmetic means 9. Hereinafter, for the sake of easy explanation, it is assumed that a type of mode hopping phenomenon (VCSEL type laser type) is employed in the semiconductor laser 1. For example, the laser driver 4 supplies the body laser 1 as an injection current with a triangular wave drive current which is repeatedly increased or decreased with respect to time. Thereby, the semiconductor laser 1 is driven to alternately repeat the first and second oscillation periods, and the wavelength of the first oscillation period is proportional to the magnitude of the injection current and is continuously increased at a constant rate of change. The oscillation wavelength is continuously reduced at a constant rate of change. Fig. 2 is a timing chart showing the oscillation wavelength of the semiconductor laser beam. In Fig. 2, t-Ι represents the (t_;i)th oscillation period; the tth oscillation period; t+1 represents the (t+丨)th oscillation period; the output is measured to reduce the oscillation 5: The number of waves is determined by the number of waves, the DFB rate of change to the semi-inductive period is oscillating continuously, and the change of t is t + 2 Table -15 - 1359943 shows (t + 2) During the oscillation period; t + 3 represents the (t + 3)th oscillation period; t + 4 represents the (t + 4)th oscillation period; A a represents the minimum 値 of the oscillation wavelength in each period; Ab represents each period The maximum 値 of the oscillation wavelength in the middle: T represents the period of the triangular wave. In the present embodiment, the maximum 値λ b of the oscillation wavelength and the minimum 値A a of the oscillation wavelength are always constant, and the difference 値Xb-Xa is also constant. The laser light emitted from the semiconductor laser 1 is condensed by the lens 3 and incident on the measurement object 12. The light reflected by the measuring object 12 is condensed by the lens 3 and incident on the semiconductor laser 1. Among them, the condensing of the lens 3 is not essential. The photodiode 2 converts the light output of the semiconductor laser 1 into a current. The current-voltage conversion amplifier 5 converts the output current of the photodiode 2 into a voltage and amplifies it. The filter circuit Π has a function of extracting overlapping signals from the modulated wave. Fig. 3(A) shows the output voltage waveform diagram of the current-voltage conversion amplifier 5 in a mode, and Fig. 3(B) shows the output voltage waveform diagram of the mode display filter circuit 》. The waveform (modulated wave) of Fig. 3 (A) of the output of the photodiode 2 is removed, and the oscillation waveform (carrier) of the semiconductor laser 1 of Fig. 2 is removed to extract the MHP waveform of Fig. 3 (B) ( The process of overlapping waves). The counting means 8 is for the number of MHPs included in the output voltage of the filter circuit 11, for the first oscillation period t-1, t+1't + 3 and the second oscillation period t, t + 2, t + 4 Counting is performed for each period. Fig. 4 is a block diagram showing an example of the configuration of the counting device 8. The counting device 8 is composed of a determination unit 81, a logical AND operation unit (AND) 82, a counter 83, a counter knot -16-1359943, a correction unit 84', and a commemorative unit 85. The current-voltage conversion amplifier 5, the Shanghai circuit circuit 11, and the determination unit 81, the AND 82 of the counter device 8, and the counter 83 constitute a counting means. Fig. 5 is a block diagram showing an example of the configuration of the count result correcting unit 84. The count result correcting unit 84 is composed of a period measuring unit 840, a frequency distribution creating unit 841, a representative unit calculating unit 842, and a correction unit calculating unit 843_. φ Fig. 6(A) to Fig. 6(F) are diagrams for explaining the operation diagram of the counting device 8. The figure (A) shows the waveform of the output voltage of the filter circuit π in a mode, that is, the waveform of the ΜΗΡ In the figure, Fig. 6 (Β) shows the output of the determination unit 81 corresponding to the second diagram (Α), and Fig. 6(C) shows the diagram of the gate signal GS input to the counting device 8. Fig. 6(D) is a view showing the count result of the counter 83 corresponding to Fig. 6 (Β), and Fig. 6 (Ε) is a view showing the clock signal CLK input to the counter device 8, 6th. Fig. (F) is a view showing the measurement results of the period measuring unit 840 corresponding to Fig. 6 φ ( Β ). • First, the determination unit 81 of the counting device 8 determines that the output voltage of the filter circuit 11 shown in Fig. 6(A) is at a high level (H) or a low level (L), and the output is as shown in Fig. 6(B). The result of the judgment shown. At this time, the determination unit 81 determines that the output voltage of the filter circuit 11 is higher than the threshold 値TH1, and determines that the output voltage is lower than the threshold 値TH2 (TH2 < TH1 ). It is determined to be a low level, whereby the output of the filter 11 is binarized. The AND82 is the result of the logical AND operation of the output of the determination unit 81 and the gate signal GS of -17-1359943 shown in Fig. 6(C), and the counter 83 counts the rise of the output of the AND82 (Fig. 6 (D) )). Here, the gate signal G S is a signal that rises at the start of the counting period (in the present embodiment, the oscillation period or the second oscillation period), and falls at the end of the counting period. Therefore, the counter 83 counts the number of rising edges of the output of the AND 8 2 in the counting period (i.e., the number of rising edges of the MHP). On the other hand, the period measuring unit 840 of the counting result correcting unit 84 measures the period of the rising edge of the output of the AND 82 (i.e., the period of the MHP) every time the rising edge occurs. At this time, the period measuring unit 840 measures the period of the MHP by using the period of the clock signal CLK shown in Fig. 6(E) as one unit. In the example of Fig. 6(F), the period measuring unit 804 sequentially measures τ <», Τ θ, and T r as the period of the MHP. As can be seen from Fig. 6(E) and Fig. 6(F), the periods Τα, Τθ, and Tr are 5 clocks, 4 clocks, and 2 clocks, respectively. The rate of the clock signal CLK is much higher than the highest frequency that the MHP can achieve. The memory unit 805 is the result of counting the memory counter 8.3 and the measurement result of the period measuring unit 840. After the gate signal GS falls and the counting period is completed, the frequency distribution creating unit 8 4 1 of the counting result correcting unit 84 is based on the frequency of the period of the MHP in the measurement result period of the memory unit counted in the recording unit 8 5 distributed. Next, the representative 値 calculation unit 842 of the count result correcting unit 84 bases -18 - 1359943 based on the frequency distribution made by the frequency distribution creating unit 841, and calculates the median TO of the period. The correction/calculation unit 84 3 of the count result correcting unit 84 calculates the frequency sum Ns of the level of 0.5 times or less of the TO in the cycle, and 1.5 times or more of the T0 in the cycle, based on the frequency distribution made by the distribution creating unit 841. The frequency of the rank is summed by Nw, and the result of the count of the accumulator 8 3 is as follows. N'=N + Nw-Ns --(7 In the equation (2), the number of N is used as the count result of the counter 83, and the result of the count after the correction is ^^ shows an example of the frequency distribution in Fig. 7 In Fig. 7, the level of the period median Τ0 is 0.5 times, and the level of the Tw period is 1.5 times the level 値. Of course, the level in Fig. 7 is the representative of MHP. In Fig. 7, in order to simplify the description, the frequency between the picture bits T0 and Ts and the median T0 and Tw is omitted. Fig. 8 is a diagram for explaining the correction of the count result of the counter 83, 8th Fig. (A) is a diagram showing the waveform of the voltage of the filter circuit 11 in the mode display, that is, the waveform of the MHP, and Fig. 8(B) is a diagram showing the count result of the counter 83 corresponding to the eighth diagram (A). The original MHP cycle depends on the distance from the measurement object 12, and if the distance from the measurement object 12 is constant, the MHP is counted by the same MHP frequency digits. The MHP Ts is set to T0. The power supply shows, but the period -19- 1359943 is not wrong, the missing news Ts is due to the occurrence of this double positive. But because The MHP waveform will be missing or generate a waveform that should be counted as a signal, and the number of defects will be poor. When a signal leak occurs, the 周期 period Tw at the portion where the leak has occurred is about twice the original period. That is, when the period of the sputum is about twice the median, it is judged that a leak has occurred in the signal. Therefore, 'the sum of the frequencies Nw of the level above the period Tw is regarded as the number of misses, and The Nw is added to the count of the counter 83. This can correct the missing signal. Further, the period of the 后 of the portion where the noise is counted is about 0.5 times the original period. That is, when the period of the ΜΗΡ is When the median is less than 5 times, the excess count signal can be judged. This is the frequency sum N s of the level below the period T s is regarded as the number of excess count numbers, and the count result by the counter 83 The Ns is subtracted to correct the erroneous noise. The above is the correction principle of the counting result shown in the formula (2). In the present embodiment, Ts is set to 0.5 times the period median T0. Set to The 1 of 1.5 times the number of bits T0, and the 値 of 2 of the median T0, is set to 1.5 times the reason for the statement. The correction calculation unit 8 4 3 will be calculated by the formula (2) The complemented count result Ν' is output to the arithmetic unit 9. The counting means 8 performs for each of the 1 oscillation periods t-1, t+1, t + 3 and each of the second oscillation periods t, t + 2 t + 4 The processing as shown above is followed. The 'computing device 9 is based on the measurement by the counting device 8 -20-
1359943 MHP的數量Ν’，求取與測定對象12的距離。 的Μ HP的數量係與測定距離成正比。因此， 一定計數期間（在本實施形態中的第1振盪期 盪期間）中的MHP的數量與距離的關係並預 算裝置9的資料庫（未圖示），運算裝置9即 取得與藉由計數裝置8測量而得的MHP的數 的距離的値，藉此可求取與測定對象12的距离 或者，預先求取表示計數期間中之MHP 離的關係的數學式並預先設定，運算裝置9係 裝置8測量而得的MHP的數量Ν’代入數學式 算出與測定對象12的距離。運算裝置9係按 盪期間t-1、t+1、t + 3及每個第2振盪期間t、 行如以上所示之處理。 顯示裝置10係即時（real time)顯示藉白 計算所得之與測定對象1 2的距離（位移）。 如以上所示，在本實施形態中，測定計 MHP的周期，根據該測定結果作成計數期間中 周期的頻率分佈，根據頻率分佈計算MHP的 數，根據頻率分佈，求取爲中位數之〇·5倍以 頻率總和Ns、及爲中位數的1 .5倍以上的等級 Nw，根據該等頻率Ns與和Nw ’補正計數 果，藉此可補正MHP的計數誤差’因此可提 定精度。 其中，本實施形態中的計數裝置8及運算 一定期間中 若預先求取 間或第2振 :先登記在運 I可由資料庫 量Ν’相對應 I ° 的數量與距 丨將藉由計數 ，，藉此可計 每個第1振 t+2 、 t+4 進 3運算裝置9 數期間中的 1之 MHP的 周期的中位 下的等級的 的頻率總和 器的計數結 升距離的測 裝置9係可 -21 - 1359943 藉由例如具備有CPU、記億裝置、介面的電腦及控制該等 硬體資源的程式來實現。用以使如上所示之電腦動作的程 式係在被記錄在軟性磁碟、CD_R0M、DVD_R0M、記億卡 等記錄媒體的狀態下予以提供。CPU係將所讀取到的程式 寫入記憶裝置，按照該程式’執行在本實施形態中所說明 的處理。 接著，就使用周期之頻率分佈的中位數作爲MHP的 基準周期的理由、及將求取頻率Nw時之周期的臨限値設 爲中位數的1. 5倍的理由加以說明。 首先，針對由於已誤數雜訊而將MHP的周期分割爲2 的情形的計數結果的補正加以說明。當半導體雷射的振盪 波長變化呈線性時，MHP的周期T係將測量期間Tc除以 MHP的數量N所得的 TO爲中心進行常態分佈（第 9 圖）。 接著，考慮因雜訊而分割爲2的MHP的周期。過剩 計數雜訊的結果而分割爲2的MHP的周期係以隨機的比 例分割爲2，但是分割前的周期爲以TO爲中心的常態分 佈，因此成爲相對於〇·5 TO呈對稱的頻率分佈（第1〇圖 的a)。 針對包含該雜訊的 MHP的周期的頻率分佈，假設 MHP的η%因雜訊而將周期分割爲2時，計算MHP的周期 的平均値及中位數。 所有周期的和恒爲測量期間 Tc，並不會改變，但是 當MHP的η%因雜訊而將周期分割爲2時，頻率的積分値 -22- 1359943 會成爲（l+n〔 ％〕）N，因此MHP的周期的平均値成爲 (1/ ( 1+η〔 ％〕））TO。 另一方面，當忽略以雜訊的分佈而與常態分佈相重疊 的部分時，分割爲2的雜訊累積頻率係成爲中位數與T0 之間的等級所包含的頻率的兩倍，因此，M HP的周期的中 位數係位於第11圖之b的面積爲a的面積的2倍的位 置。 在屬於Microsoft公司之軟體的Excel (註冊商標）中 有可利用由常態分佈的平均値與α σ間之兩側値的內部比 例爲「（l-(bNORMSDIST(a) ) x2 ) x 10〇 ( % )」來 表現的NORMSDIST ()的函數，若利用該函數，可以如 下數式，表示MHP的周期的中位數。 (l-(l-NORMSDIST((中位數-Τ0)/σ))χ2)χ(100-η)/2=η〔 ％〕 …（3) 根據如以上所示，若將標準偏差σ設爲〇·〇2ΤΟ，而計 算出ΜΗΡ的10%因雜訊而將周期分割爲2時之ΜΗΡ之周 期的平均値Τ0’及中位數Τ0’，如以下所示。 TO’ =(1/(1 + 0.1))T0 = 0.91T0 …（4) TO，= 0.99 5T0 ··· (5) 其中，在此係將平均値、中位數均以TO’表示。計數器値 (頻率的積分値）係成爲1 · 1 N，計數誤差成爲1 ° -23- 1359943 在此’考慮在某周期Ta的MHP被分割爲2之後之2 個周期T1、T2(設爲T12T2)所可取得的期間的機率。 假設雜訊是隨機産生，如第12圖所示，Τ2係可以相同的 機率取得0 < T2S Ta/2的値。同樣地，Τ1亦可以相同的 機率取得T/2客Tl<Ta的値。第12圖中的T1所可取得 的機率分佈的面積與T2所可取得的機率分佈的面積均爲 1 ° 周期Ta係呈以TO爲中心的常態分佈，因此，若將 Ta看作集合，則T2所可取得的機率的頻率分佈係形成爲 與平均値爲0.5T0、標準偏差爲0·5σ的常態分佈的累積頻 率分佈相同的形狀。 此外，如第13圖所示，T1所可取得的機率的頻率分 佈係形成爲將平均値爲0.5T0、標準偏差爲〇·5 σ的常態分 佈的累積頻率分佈、與平均値爲Τ0、標準偏差爲σ的常 態分佈的累積頻率分佈相重疊的形狀。在此’ T1、Τ2的 各數量係與周期被分割爲2的ΜΗΡ的數量η〔 ％〕· Ν相 等。 若可對於因雜訊而使周期被分割爲2的ΜΗΡ的數量η 〔％〕. Ν進行計數，即可使用以下數式’導出ΜΗΡ的數 量Ν。 Ν = Ν，- η〔 ％〕· Ν …（6) 如第14圖所示，若可以使具有Tb以下之周期的ΜΗΡ -24- 1359943 的數量Ns與被分割爲2之MHP的數量Μ%〕.Ν成爲 相等的方式來設定Tb ’即可藉由對於具有Tb以下之周期 的MHP的數量Ns進行計數，而間接地對於周期被分割爲 2的MHP的數量n〔％〕. N進行計數。 在第14圖中，當具有Tb以上之周期的MHP的周期 T2的頻率（第14圖的c)與具有未達Tb之周期的MHP 的周期T1的頻率（第14圖的d)爲相同時，具有Tb以 下之周期的MHP的數量係與T2的數量，亦即周期被分割 爲2的MHP的數量Ns ( =n〔 ％〕· Ν )成爲相等。亦即， MHP的數量N係可以如下數式表示。 N = N，- η〔 〇/〇〕 · Ν = Ν，- Ns ··· (7) T1及T2的頻率形狀係在0.5 Ta呈對稱的形狀，因此 將0.5Ta作爲臨限値而進行判斷時，可正確地對周期被分 割爲2的MHP的頻率Ns ( = η〔 ％〕· Ν )進行計數。 接著，藉由對於具有0.5T0之下之周期的MHP的數 量進行計數，可對周期被分割爲2的MHP的數量η 〔％〕· Ν的數間接地進行計數，但是並無法根據包含雜 訊的ΜΗΡ的周期的頻率分佈（第10圖）來計算出Τ0。若 MHP的母群體如第10圖的頻率分佈所示眾數（m〇de)愈 與 T0相等愈爲理想而且母體參數（popuiation parameter)愈大，即可使用眾數作爲TO’。 在此記載使用平均値或中位數TO’所得之MHP的數量 -25- 1359943 n〔 ％〕. N的計數。若以TO’= y· TO表示，代入TO,取代 TO以求出Ns時，比作爲周期被分割爲2之MHP的數量 所進行判斷的0.5Τ0’爲小的周期的頻率Ns’係成爲y. η 〔％〕 · Ν (第 1 5 圖）。 若使用平均値或中位數TO ’，補正後的計數値Nt係以 下所示。1359943 The number of MHPs Ν', and the distance from the measurement object 12 is obtained. The number of HP is proportional to the measured distance. Therefore, the relationship between the number of MHPs and the distance in the predetermined counting period (the first oscillation period in the present embodiment) and the distance are calculated in the database of the device 9 (not shown), and the arithmetic unit 9 obtains and counts by The distance 数 of the number of MHPs measured by the device 8 can be obtained by calculating the distance from the measurement target 12 or obtaining a mathematical expression indicating the relationship of the MHP in the counting period in advance, and setting the calculation device 9 in advance. The number of MHPs measured by the device 8 is substituted into the mathematical formula to calculate the distance from the measurement target 12. The arithmetic unit 9 performs processing as shown above in the swing periods t-1, t+1, t + 3 and each of the second oscillation periods t and lines. The display device 10 displays the distance (displacement) from the measurement target 12 calculated by the white space in real time. As described above, in the present embodiment, the period of the measurement meter MHP is calculated, the frequency distribution of the period in the counting period is generated based on the measurement result, the number of MHPs is calculated from the frequency distribution, and the median is obtained from the frequency distribution. · 5 times the frequency sum Ns, and the level Nw of the median of 1.5 times or more, according to the frequencies Ns and Nw 'correct the count, thereby correcting the MHP count error 'so the accuracy can be determined . In the counting device 8 and the calculation period in the present embodiment, if the inter-time or the second vibration is obtained in advance, the number and the distance corresponding to the I° can be counted by the amount of the database. According to this, it is possible to measure the counting rise distance of the frequency summator of the level of the MHP of one of the first oscillations t+2 and t+4. The 9 Series can be implemented by, for example, a computer having a CPU, a device, an interface, and a program for controlling such hardware resources. The program for causing the computer to operate as described above is provided in a state of being recorded on a recording medium such as a flexible disk, a CD_ROM, a DVD_R0M, or a card. The CPU writes the read program to the memory device, and executes the processing described in the present embodiment in accordance with the program. Next, the reason for using the median of the frequency distribution of the period as the reference period of the MHP and the threshold of the period when the frequency Nw is obtained is set to 1.5 times the median. First, the correction of the counting result in the case where the period of the MHP is divided into two due to the erroneous noise is explained. When the oscillation wavelength of the semiconductor laser changes linearly, the period T of the MHP is a normal distribution centering on the TO obtained by dividing the measurement period Tc by the number N of MHPs (Fig. 9). Next, consider the period of the MHP divided into 2 by noise. The period of the MHP divided into 2 by the result of the excess count noise is divided into 2 by a random ratio, but the period before the division is a normal distribution centered on TO, and thus becomes a symmetric frequency distribution with respect to 〇·5 TO. (a in the first picture). For the frequency distribution of the period of the MHP containing the noise, the average 値 and median of the period of the MHP are calculated assuming that η% of the MHP divides the period into two due to noise. The sum of all periods is constant for the measurement period Tc, and does not change, but when the η% of the MHP divides the period into 2 due to noise, the integral of the frequency 値-22- 1359943 becomes (l+n[%]) N, so the average 値 of the period of the MHP becomes (1/(1+η[%])) TO. On the other hand, when the portion overlapping with the normal distribution is ignored by the distribution of the noise, the noise accumulation frequency divided into 2 becomes twice the frequency included in the level between the median and T0, and therefore, The median of the period of M HP is located at a position twice the area of area a of b in Fig. 11 . In Excel (registered trademark) belonging to Microsoft Corporation software, there is an internal ratio of "(l-(bNORMSDIST(a)) x2) x 10〇) that can be used by the average 値 and α σ between the normal distributions. The function of NORMSDIST() expressed by %) can use the function to represent the median of the period of the MHP as follows. (l-(l-NORMSDIST((median-Τ0)/σ))χ2)χ(100-η)/2=η[%] (3) According to the above, if the standard deviation σ is set For 〇·〇2ΤΟ, the average 値Τ0' and the median Τ0' of the period after dividing the period into 2 by the noise of 10% of the ΜΗΡ are calculated as shown below. TO' = (1/(1 + 0.1)) T0 = 0.91T0 (4) TO, = 0.99 5T0 (5) where, in this case, the average 値 and median are expressed by TO'. The counter 値 (integral 频率 of the frequency) becomes 1 · 1 N, and the counting error becomes 1 ° -23 - 1359943. Here, consider the two periods T1 and T2 after the MHP of a certain period Ta is divided into 2 (set to T12T2). The probability of the period that can be obtained. Assuming that the noise is randomly generated, as shown in Fig. 12, the Τ2 system can obtain the & of T < T2S Ta/2 with the same probability. Similarly, Τ1 can also achieve the same T 2 passenger Tl < Ta 値. The area of the probability distribution that can be obtained by T1 in Fig. 12 and the area of the probability distribution that can be obtained by T2 are both 1 °. The period Ta is a normal distribution centered on TO. Therefore, if Ta is regarded as a set, then The frequency distribution of the probability that T2 can obtain is formed into the same shape as the cumulative frequency distribution of the normal distribution having an average 値 of 0.5 T0 and a standard deviation of 0.5 σ. Further, as shown in Fig. 13, the frequency distribution of the probability that T1 can obtain is formed as a cumulative frequency distribution with a mean 値 of 0.5T0 and a standard deviation of 〇·5 σ, and an average 値0, standard. The deviation is a shape in which the cumulative frequency distribution of the normal distribution of σ overlaps. Here, the number of each of 'T1 and Τ2' is equal to the number η [%]· Ν of the ΜΗΡ which is divided into two by the period. If the number η [%]. 使 of the period in which the period is divided into 2 by the noise is counted, the number Ν of the ΜΗΡ can be derived using the following equation ’. Ν = Ν, - η [ %]· Ν (6) As shown in Fig. 14, if the number of ΜΗΡ -24 - 1359943 having a period of Tb or less and the number of MHPs divided into 2 can be made Μ% Ν.Ν is equal to set Tb' by counting the number Ns of MHPs having a period of Tb or less, and indirectly counting the number of MHPs whose period is divided into 2 by n[%]. . In Fig. 14, when the frequency of the period T2 of the MHP having the period of Tb or more (c of Fig. 14) is the same as the frequency of the period T1 of the MHP having the period of not reaching the period of Tb (d of Fig. 14) The number of MHPs having a period of Tb or less is equal to the number of T2, that is, the number Ns (=n[%]·Ν) of the MHP whose period is divided into two. That is, the number N of MHPs can be expressed by the following equation. N = N, - η [ 〇 / 〇 ] · Ν = Ν, - Ns · (·) The frequency shape of T1 and T2 is symmetrical in 0.5 Ta, so the judgment is made by using 0.5Ta as a threshold. At this time, the frequency Ns (= η [ % ]· Ν ) of the MHP whose period is divided into 2 can be correctly counted. Then, by counting the number of MHPs having a period below 0.5T0, the number of η [%]· M of the MHP whose period is divided into 2 can be indirectly counted, but cannot be included according to the inclusion of noise. The frequency distribution of the chirp period (Fig. 10) is used to calculate Τ0. If the mother group of the MHP is as shown in the frequency distribution of Fig. 10, the more the majority (m〇de) is equal to T0, and the larger the parent parameter (popuiation parameter), the more the number can be used as TO'. Here, the count of the number of MHPs obtained using the average 値 or the median TO' -25 - 1359943 n [%]. N is described. When TO is represented by TO'= y· TO, and substitutes TO to replace the TO to obtain Ns, the frequency Ns' which is smaller than the number of MHPs which are divided into 2 by the period is determined to be y. η [%] · Ν (Fig. 15). If the average 中 or median TO ’ is used, the corrected 値Nt is shown below.
Nt = N’ - Ns’ =(1 + n〔％〕）N-yn〔％〕N -(1+(1 - y)n [ %) )N = N +(1 - y)n ( %] N …（8) 其中，作爲補正後誤差的（1-y) n〔％〕N係第16圖的e 的部分的頻率。 在此，就使用平均値或中位數TO’的計數器83的計數 結果的補正例進行說明。 若將標準偏差設爲σ=0·02Τ0，且MHP的10%因雜訊 而使周期分割爲2時（計數結果爲10%的誤差），由於 ΜΗΡ 的周期的平均値 Τ0’爲 0.91Τ0，中位數 Τ0’爲 0.9949T0，因此使用平均値T0’時的y爲〇.91，使用中位 數T0’的y爲0.9949，補正後的計數結果Ν’係如以下所示 予以計算。 Ν， =(1 + 0.1(1 -0.9 1 ))N = 1 .009Ν …（9) Ν， =(1 + 0.1(1 -0.995))N = 1.0005 Ν •••(10) -26- 1359943 數式（9)係表示使用平均値TO’時的補正後的計數結 果Ν’，數式（10)係表示使用中位數TO’時的補正後的計 數結果N’。使用平均値TO’時的計數結果N.’的誤差爲 0.9%，使用中位數T0’時的計數結果Ν’的誤差爲0·05 %。 接著，若將標準偏差設爲σ=0.05Τ0，且ΜΗΡ的20% 因雜訊而使周期分割爲2時（計數結果爲2 0%的誤差）， 由於ΜΗΡ的周期的平均値Τ0’爲0.83Τ0，中位數Τ0’爲 0.9682T0 >因此使用平均値T0’時的y爲0.83，使用中位 數T0’的y爲0.968，補正後的計數結果Ν’係如以下所示 予以計算。 N5 =(1 + 0.2(1 - 0.83))N = 1.034Ν ...(11) Ν，=( 1 + 0.2( 1 - 0.968))N = 1.0064N ··· (12) 數式（11)係表示使用平均値το’時的補正後的計數 結果Ν’，數式（12)係表示使用中位數T0’時的補正後的 計數結果Ν’。使用平均値T0’時的計數結果Ν’的誤差爲 3.4%，使用中位數Τ0’時的計數結果Ν’的誤差爲0.64%。 由以上可知，若使用ΜΗΡ的周期的中位數來補正計 數結果Ν，可減小補正後的計數結果Ν’的誤差。 接著說明在ΜΗΡ波形產生缺漏時的計數結果的補 正。由於ΜΗΡ的強度較小而在計數時產生缺漏時之ΜΗΡ 的周期係由於原本的ΜΗΡ的周期係以Τ0爲中心的常態分 佈，因此成爲平均値爲2T0，標準偏差2 σ的常態分佈 -27- 1359943 (第17圖中的f)。假設缺漏m〔％〕的MHP，因該缺漏 而成爲2倍的ΜΗΡ的周期的頻率爲Nw ( =m〔 ％〕· N) °此外’因計數時的的缺漏而減少後之大約τ〇之周期 的頻率爲第17圖所示的g，第17圖的h所示的頻率減小 份爲2Nw ( =2m〔 ％〕）。因此，在計數時沒有發生ΜΗΡ 缺漏時之原本ΜΗΡ的數量Ν，係可以下式表示。Nt = N' - Ns' = (1 + n [%]) N-yn [%] N - (1 + (1 - y) n [ %) ) N = N + (1 - y) n ( %] N (8) where (1-y) n [%] N is the frequency of the portion of e of Fig. 16 as the corrected error. Here, the counter 83 of the average 値 or the median TO' is used. The correction example of the counting result will be described. If the standard deviation is σ=0·02Τ0, and 10% of the MHP is divided into 2 due to noise (the counting result is 10% error), due to the period of ΜΗΡ The average 値Τ0' is 0.91Τ0, and the median Τ0' is 0.9949T0. Therefore, the y when using the average 値T0' is 〇.91, and the y using the median T0' is 0.9949, and the result of the correction is Ν' Calculated as shown below. Ν, =(1 + 0.1(1 -0.9 1 ))N = 1 .009Ν ...(9) Ν, =(1 + 0.1(1 -0.995))N = 1.0005 Ν ••• (10) -26- 1359943 The equation (9) indicates the result of the correction after the correction using the average 値TO', and the equation (10) indicates the result of the correction after the median TO' is used. '. The error of the count result N.' when using the average 値TO' is 0.9%, using the median The error of the counting result Ν' at T0' is 0.05 %. Next, if the standard deviation is σ = 0.05 Τ 0, and 20% of ΜΗΡ is divided into 2 due to noise (counting result is 2 0 % error), since the average 値Τ0' of the period of ΜΗΡ is 0.83Τ0, the median Τ0' is 0.9682T0 > therefore, the y using the average 値T0' is 0.83, and the y using the median T0' is 0.968 The count result after correction is calculated as follows. N5 = (1 + 0.2 (1 - 0.83)) N = 1.034 Ν ... (11) Ν, = ( 1 + 0.2 ( 1 - 0.968) N = 1.0064N · (12) The equation (11) indicates the result of the correction after the correction using the average 値το', and the equation (12) indicates the correction when the median T0' is used. The result of the counting is Ν'. The error of the counting result Ν' when using the average 値T0' is 3.4%, and the error of the counting result Ν' when using the median Τ0' is 0.64%. From the above, if the period of ΜΗΡ is used, The median is used to correct the counting result Ν, and the error of the counting result 补' after correction can be reduced. Next, the correction of the counting result when the ΜΗΡ waveform is missing is explained. Since the intensity of enthalpy is small, the period of 缺 at the time of counting is due to the normal distribution of ΜΗΡ0 centered on the original ΜΗΡ, so the average 値 is 2T0, and the standard deviation 2 σ is normal distribution -27- 1359943 (f in Figure 17). It is assumed that the MHP of the missing m [%] has a frequency of Nw (=m[%]·N) in the period of the ΜΗΡ which is twice as large as the defect, and is reduced by approximately τ〇 due to the omission at the time of counting. The frequency of the period is g shown in Fig. 17, and the frequency-reduced portion shown by h in Fig. 17 is 2Nw (= 2m [%]). Therefore, the number of original flaws at the time of counting without missing faults can be expressed by the following formula.
N’=N + m〔％〕 = N + Nw …（13) 接著，考慮在對用以補正計數結果的Nw進行計數時 之周期的臨限値。在此，假設爲因計數時的缺漏而使周期 成爲2倍的MHP的周期的頻率Nw中因雜訊而分割爲2的 情形。缺漏的MHP中被分割爲2的MHP的周期的頻率爲 Nw’（=m.p〔％〕· N )。再次分割爲2的MHP的周期 的頻率分佈係如第18圖所示。當將視爲Nw的周期的臨 限値設爲1 ·5Τ0時，周期爲0.5T0以下之MHP的周期的頻 率爲 0.5Nw，（=0.5p〔％〕 . Nw )，周期爲 0.5T0 至 1.5T0之MHP的周期的頻率爲Nw，（=p〔％〕· Nw)， 周期爲1.5T0以上之MHP的周期的頻率爲0.5Nw，（= 0.5 p〔 ％〕 · N w )。 因此，所有MHP的周期的頻率分佈成爲如第19圖所 示，若將Ns的臨限値設爲0.5T0，將Nw的臨限値設爲 1 _ 5 TO，計數結果N係可以下式表示。 -28- 1359943 N =(N，- 2Nw) + (Nw - Nw，）+ 2Nw，= N，一 Nw + Nw， ··· ( 14 ) 由數式（14)予以補正的結果如以下所示’可知計算 出計數時未發生ΜΗ P之缺漏之情形下之原本的MHP的數 量Ν，〇N' = N + m [%] = N + Nw (13) Next, the threshold 周期 of the period when Nw for counting the count result is counted is considered. Here, it is assumed that the frequency Nw of the period of the MHP in which the period is doubled due to the omission at the time of counting is divided into two by noise. The frequency of the period of the MHP divided into 2 in the missing MHP is Nw' (= m.p [%]· N ). The frequency distribution of the period of the MHP divided again to 2 is as shown in Fig. 18. When the threshold 周期 of the period regarded as Nw is set to 1·5Τ0, the frequency of the period of the MHP having a period of 0.5T0 or less is 0.5Nw, (=0.5p[%]. Nw), and the period is 0.5T0 to 1.5. The frequency of the period of the MHP of T0 is Nw, (=p[%]·Nw), and the frequency of the period of the MHP having a period of 1.5 T0 or more is 0.5 Nw, (= 0.5 p [%] · N w ). Therefore, the frequency distribution of all MHP cycles is as shown in Fig. 19. If the threshold N of Ns is set to 0.5T0, the threshold N of Nw is set to 1 _ 5 TO, and the counting result N can be expressed by the following formula. . -28- 1359943 N = (N, - 2Nw) + (Nw - Nw,) + 2Nw, = N, a Nw + Nw, ··· ( 14 ) The result of correction by the formula (14) is as follows 'It can be seen that the number of original MHPs in the case where the ΜΗ P is not missing at the time of counting is calculated, 〇
Ν 0 _ 5 N w ’ + (0 _ 5 N w ’ + (N w - N w ’）） =(N - Nw + Nw’）+ (0.5Nw’ +(Nw - Nw’）） =N， ·_· (15) 由以上可知，若將求取頻率Nw之周期的臨限値設爲 中位數的1.5倍，即可補正計數結果N。其中，與因雜訊 而將MHP的周期分割爲2的情形相同地，由於取代το而 使用中位數來進行補正，因此發生同樣的誤差。 在以上說明中，係分別說明過剩計數雜訊的結果使 MHP的周期分割爲2的情形、及因計數時的缺漏而使 MHP的周期成爲2倍的情形，惟該等情形獨立發生，因此 若將該等情形表現爲1個頻率分佈，即如第2 0圖所示。 若將Ns的臨限値設爲0.5T0，將Nw的臨限値設爲 1.5T0，計數結果ν即可以下式表示。 N =(N’ - 2Nw - Ns) + (Nw _ Nw’）+ 2Nw’ + 2Ns ··，（16) -29- 1359943 由數式（16)予以補正的結果如以下所示，可知計算 出計數時未發生缺漏或過剩計數之情形下之原本的MHP 的數量Ν’。 Ν - {0.5Nw5 + Ns} + {0.5Nw5 +(Nw - Nw5)} ={N - Nw + Nw5 + Ns} - {0.5Nw5 + Ns} + {0.5Nw，+(Nw - Nw’）}Ν 0 _ 5 N w ' + (0 _ 5 N w ' + (N w - N w ')) = (N - Nw + Nw') + (0.5Nw' + (Nw - Nw')) = N, _· (15) From the above, it can be seen that the count result N can be corrected by setting the threshold 周期 of the period for obtaining the frequency Nw to 1.5 times the median. However, similarly to the case where the period of the MHP is divided into two due to noise, the median is used instead of τ, and the same error occurs. In the above description, the case where the result of the excess count noise is divided into two is the case where the period of the MHP is divided into two, and the period of the MHP is doubled due to the leak at the time of counting, but these cases occur independently. These cases are expressed as one frequency distribution, as shown in Fig. 20. If the threshold N of Ns is set to 0.5T0, the threshold N of Nw is set to 1.5T0, and the count result ν can be expressed by the following equation. N = (N' - 2Nw - Ns) + (Nw _ Nw') + 2Nw' + 2Ns ··, (16) -29- 1359943 The result of the correction by the formula (16) is as shown below, and it is known that The number of original MHPs in the case where no missing or excessive counts occurred during counting Ν'. Ν - {0.5Nw5 + Ns} + {0.5Nw5 +(Nw - Nw5)} ={N - Nw + Nw5 + Ns} - {0.5Nw5 + Ns} + {0.5Nw, +(Nw - Nw’)}
(第2實施形態） 接著說明本發明之第2實施形態。在第1實施形態 中，係使用中位數作爲MHP的周期的代表値來補正計數 結果，但是亦可使用眾數作爲周期的代表値。 例如因低頻之雜訊成分的影響，使得MHP的周期的 頻率分佈由第21圖所示之原本的分佈j偏移而變成如第 21圖之分佈k所示時，MHP周期的中位數亦由原本的値 TO偏移成爲Td。若在如上所示之情形下使用中位數Td來 補正計數結果，補正後的計數結果的誤差會變大。 因此’若考慮到如上所示因雜訊成分所造成的頻率分 佈偏移的影響時，係使用眾數作爲周期的代表値。具體而 言，計數結果補正部84的代表値計算部842若根據頻率 分佈作成部84 1所作成的頻率分佈，來計算MHP的周期 的眾數即可。計數結果補正部84的補正値計算部843若 使用眾數來取代中位數T 0，進行與第1實施形態相同的 -30- 1359943 處理即可。 此外’如使用數式（8)至數式（12)加以說明所 示，與使用中位數TO的情形相比較，補正後的計數結果 的誤差會變大，但是亦可使用平均値來作爲周期的代表 値。此時，代表値計算部842若計算MHP的周期的平均 値即可。 其中，在第1、第2實施形態中係就本發明的計數裝 置適用於距離計的情形加以說明，惟並非侷限於此，本發 明之計數裝置亦可適用於其他領域。本發明的計數裝置爲 有效時’作爲計數對象的訊號數量係與特定的物理量（在 第1、第2實施形態中爲距離）具有線性關係，若該物理 量爲一定時，訊號即成爲大致單一頻率。 此外，即使訊號非爲單一頻率，如同與計數期間相比 較，作爲計數對象的特定的物理量以十分低的頻率，例如 1/10以下的頻率振動的對象物的速度所示周期分佈的擴 展爲較小時，亦可作爲大致單一頻率，本發明的計數裝置 即爲有效。 此外’在第1、第2實施形態中，係針對MHP的缺漏 的補正’藉由1個缺漏，使MHP的周期成爲原本周期大 約2倍的情形加以說明，但是在連續發生2個以上之缺漏 的情形亦適用本發明。MHP連續缺漏2個時，係視爲中位 數的3倍的周期的MHP係3個MHP成爲1個者。此時， 求取爲周期的中位數的大約3倍以上之等級的頻率，若該 頻率設爲2倍’即可補正MHP的缺漏。若將如上所示的 -31 - 1359943 思考方式一般化’即可使用下式來取代數式（2)。 N ’ = N + N w 1 + Nw2 + Nw3 + …-Ns …（1 8 )(Second embodiment) Next, a second embodiment of the present invention will be described. In the first embodiment, the count result is corrected using the median as a representative of the period of the MHP, but the mode can also be used as the representative of the period. For example, due to the influence of the low-frequency noise component, the frequency distribution of the MHP period is shifted from the original distribution j shown in FIG. 21 to the distribution k as shown in FIG. 21, and the median of the MHP period is also The offset from the original 値TO becomes Td. If the median Td is used to correct the count result in the above situation, the error of the count result after correction becomes large. Therefore, when considering the influence of the frequency distribution offset caused by the noise component as described above, the mode is used as the representative of the period. Specifically, the representative 値 calculating unit 842 of the counting result correcting unit 84 may calculate the mode of the period of the MHP based on the frequency distribution created by the frequency distribution creating unit 84 1 . The correction correction unit 843 of the count result correcting unit 84 may perform the same processing as -30-1359943 in the first embodiment, instead of using the mode, instead of the median T 0 . In addition, as explained using the equations (8) to (12), the error of the count result after correction becomes larger than that of the case where the median TO is used, but the average value can also be used as the The representative of the cycle. At this time, the representative 値 calculation unit 842 may calculate the average 値 of the period of the MHP. In the first and second embodiments, the counting device of the present invention is applied to a distance meter. However, the present invention is not limited thereto, and the counting device of the present invention can be applied to other fields. When the counting device of the present invention is effective, the number of signals to be counted has a linear relationship with a specific physical quantity (distance in the first and second embodiments), and when the physical quantity is constant, the signal becomes a substantially single frequency. . Further, even if the signal is not a single frequency, as compared with the counting period, the specific physical quantity as the counting object is expanded at a very low frequency, for example, a velocity of an object vibrating at a frequency of 1/10 or less. The counting device of the present invention is effective as an approximate single frequency in hours. In addition, in the first and second embodiments, the correction of the missing defect of the MHP is described as a case where the period of the MHP is approximately twice as large as the original period by one missing, but two or more leaks occur continuously. The present invention also applies to the present invention. When there are two consecutive missing MHPs, one MHP system with three cycles of three times the median is one. At this time, the frequency of the level of about 3 times or more of the median of the period is obtained, and if the frequency is set to 2 times, the missing of the MHP can be corrected. If the above-mentioned -31 - 1359943 thinking mode is generalized, the following formula can be used instead of the formula (2). N ′ = N + N w 1 + Nw2 + Nw3 + ... - Ns ... (1 8 )
Nwl係爲周期之中位數之1.5倍以上之等級的頻率總 和，Nw2係爲周期之中位數之2.5倍以上之等級的頻率總 和，Nw3係爲周期之中位數之大約3.5倍以上之等級的頻 率總和。 其中，在將第1、第2實施形態應用於除了自耦合型 之外的距離計時，係將以測定對象1 2所反射的半導體雷 射1的光，藉由例如分束器（beam splitter)等與入射至 測定對象1 2的入射光分離，而以光電二極體2予以檢 測。如此一來，在自耦合型以外的距離計中，亦可獲得與 第1、第2實施形態相同的效果。 (第3實施形態） 在第1、第2實施形態中，從作爲受光器的光電二極 體的輸出訊號中抽取MHP波形，但是亦可在不使用光電 二極體的情形下抽取MHP波形。第22圖係顯示作爲本發 明之第3實施例之距離計之構成的方塊圖，關於與第1圖 相同的構成係標註相同的元件符號。本實施形態的距離計 係使用電壓檢測電路1 3來代替第1實施形態之光電二極 體2和電流-電壓轉換放大器5。 電壓檢測電路1 3係檢測並放大半導體雷射1的端子 -32- 1359943 間的電壓，亦即陽極-陰極間電壓。因由半導體雷射1所 發射的雷射光與來自測定對象12的返回光而產生干涉 時，在半導體雷射1的端子間電壓係呈現MHP波形。因 此，可由半導體雷射1的端子間電壓抽出MHP波形。 與第1實施形態相同地’濾波器電路1 1係具有由調 變波抽出重疊訊號的功能，由電壓檢測電路13的輸出電 壓抽出MHP波形。 半導體雷射1、雷射驅動器4、計數裝置8、運算裝置 9及顯示裝置10的動作係與第1實施形態相同。 如此一來，在本實施形態中，係可在不使用光電二極 體的情形下抽出MHP波形，與第1實施形態相比，可刪 減距離計的零件，並可減少距離計的成本。 其中，在第1、第2實施形態的情形下，不僅自耦合 型，亦可適用於自耦合型以外的距離計，而本實施形態之 適用對象僅爲自耦合型❶ (產業上利用可能性） 本發明係可適用於對於訊號數量進行計數的計數裝 置、使用計數裝置來測定干涉波形的數量而求取與測定對 象之距離的平涉型距離計。 【圖式簡單說明】 第1圖係顯示作爲本發明第丨實施形態之距離計之構 成的方塊圖。 -33- 1359943 第2圖係顯示本發明第〗實施形態中之半導體雷射的 振盪波長的時間變化的一例圖β 第3圖係以模式顯示本發明第1實施形態之電流-電 壓轉換放大器的輸出電壓波形及濾波器電路的輸出電壓波 形的圖。 第4圖係顯示本發明第】實施形態中之計數裝置之構 成之一例的方塊圖。 第5圖係顯示第4圖的計數裝置中之計數結果補正部 之構成之一例的方塊圖。 第6圖係用以說明第4圖之計數裝置的動作圖。 第7圖係顯示本發明第1實施形態中之周期之頻率分 佈之一例圖。 第8圖係用以說明本發明第1實施形態中之計數器之 計數結果之補正原理的說明圖。 第9圖係顯示模式跳躍脈衝之周期的頻率分佈圖。 第10圖係顯示包含雜訊的模式跳躍脈衝之周期的頻 率分佈圖。 第11圖係顯示包含雜訊的模式跳躍脈衝之周期之中 位數的圖。 第1 2圖係顯示周期經分割爲2之模式跳躍脈衝之周 期的頻率分佈圖。 第1 3圖係顯示周期經分割爲2之模式跳躍脈衝之周 期的頻率分佈圖。 第1 4圖係顯示周期經分割爲2之模式跳躍脈衝之周 -34- 1359943 期的頻率分佈圖。 第15圖係顯示周期經分割爲2之模式跳躍脈衝之周 期的頻率分佈圖。 第16圖係顯示計數器値補正後的誤差的圖。 第17圖係顯示形成爲2倍之周期的模式跳躍脈衝的 周期的頻率分佈圖。 第1 8圖係顯示在計數時所缺漏的模式跳躍脈衝中經 分割爲2之模式跳躍脈衝之周期的頻率分佈圖。 第1 9圖係顯示在計數時所缺漏的模式跳躍脈衝中經 分割爲2之模式跳躍脈衝之周期的頻率分佈圖。 第20圖係顯示在計數時同時發生缺漏與過剩計數時 之模式跳躍脈衝之周期的頻率分佈圖。 第2 1圖係用以說明本發明第1實施形態中在補正後 的計數結果發生誤差之例圖。 第22圖係顯示作爲本發明第3實施形態之距離計之 構成的方塊圖。 第23圖係顯示習知之雷射測量器中之半導體雷射之 複合共振模型圖。 第24圖係顯示半導體雷射的振盪波長與內建光電二 極體的輸出波形的關係圖。 【主要元件符號說明】 1 :半導體雷射 • 2 :光電二極體 -35- 1359943 3 :透鏡 4 =雷射驅動器 5 :電流-電壓轉換放大器 8 :計數裝置 9 :運算裝置 1 〇 :顯示裝置 1 1 :濾波器電路 1 2 :測定對象 1 3 :電壓檢測電路 8 1 :判定部 82 :邏輯與運算部（AND) 8 3 :計數器 84 :計數結果補正部 8 5 :記憶部 1 〇 1 :半導體雷射 102 :半導體結晶的壁開面 1 03 :光電二極體 104 :測定對象 840 :周期測定部 841 :頻率分佈作成部 842 :代表値計算部 8 4 3 :補正値計算部 -36Nwl is the sum of the frequencies of the order of 1.5 times or more of the median period, Nw2 is the sum of the frequencies of the order of 2.5 times or more of the median period, and Nw3 is about 3.5 times the median of the period. The sum of the frequencies of the ranks. In the first embodiment, the first and second embodiments are applied to the distance measurement other than the self-coupling type, and the light of the semiconductor laser 1 reflected by the measurement target 12 is used, for example, by a beam splitter. The light is incident on the incident light of the measurement object 12, and is detected by the photodiode 2. As a result, in the distance meter other than the self-coupling type, the same effects as those of the first and second embodiments can be obtained. (Third Embodiment) In the first and second embodiments, the MHP waveform is extracted from the output signal of the photodiode as the photoreceiver, but the MHP waveform can be extracted without using the photodiode. Fig. 22 is a block diagram showing the configuration of a distance meter as a third embodiment of the present invention, and the same components as those in Fig. 1 are denoted by the same reference numerals. In the distance meter of the present embodiment, the voltage detecting circuit 13 is used instead of the photodiode 2 and the current-voltage converting amplifier 5 of the first embodiment. The voltage detecting circuit 13 detects and amplifies the voltage between the terminals -32-1359943 of the semiconductor laser 1, that is, the voltage between the anode and the cathode. When the laser light emitted from the semiconductor laser 1 and the return light from the measurement target 12 interfere with each other, the voltage between the terminals of the semiconductor laser 1 exhibits an MHP waveform. Therefore, the MHP waveform can be extracted from the voltage between the terminals of the semiconductor laser 1. Similarly to the first embodiment, the filter circuit 11 has a function of extracting a superimposed signal from a modulated wave, and extracts an MHP waveform from the output voltage of the voltage detecting circuit 13. The operation of the semiconductor laser 1, the laser driver 4, the counter device 8, the arithmetic unit 9, and the display device 10 is the same as that of the first embodiment. As described above, in the present embodiment, the MHP waveform can be extracted without using the photodiode, and the parts of the distance meter can be deleted as compared with the first embodiment, and the cost of the distance meter can be reduced. In the case of the first and second embodiments, it is not only a self-coupling type but also a distance meter other than the self-coupling type, and the application of the present embodiment is only a self-coupling type (industrial use possibility) The present invention is applicable to a counting device that counts the number of signals, and a flat type distance meter that measures the distance from the measurement target by using the counting device to measure the number of interference waveforms. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a block diagram showing the construction of a distance meter as an embodiment of the present invention. -33- 1359943 FIG. 2 is a view showing an example of temporal change of an oscillation wavelength of a semiconductor laser in the embodiment of the present invention. FIG. 3 is a mode showing a current-voltage conversion amplifier according to the first embodiment of the present invention. A diagram of the output voltage waveform and the output voltage waveform of the filter circuit. Fig. 4 is a block diagram showing an example of the configuration of the counting device in the embodiment of the present invention. Fig. 5 is a block diagram showing an example of a configuration of a counting result correcting unit in the counting device of Fig. 4. Fig. 6 is a view for explaining the operation of the counting device of Fig. 4. Fig. 7 is a view showing an example of the frequency distribution of the period in the first embodiment of the present invention. Fig. 8 is an explanatory view for explaining the principle of correction of the counting result of the counter in the first embodiment of the present invention. Figure 9 is a graph showing the frequency distribution of the period of the mode skip pulse. Figure 10 is a graph showing the frequency distribution of the period of the mode skip pulse containing noise. Figure 11 is a graph showing the median period of the pattern skip pulse containing noise. Fig. 12 is a graph showing the frequency distribution of the period of the mode skip pulse divided into two by the period. Fig. 13 shows a frequency distribution diagram of the period of the mode skip pulse divided into two by the period. Figure 14 shows the frequency distribution of the period -34 - 1359943 period of the mode skip pulse divided into 2 cycles. Fig. 15 is a graph showing the frequency distribution of the period of the mode skip pulse divided into two by the period. Figure 16 is a graph showing the error after the counter 値 is corrected. Fig. 17 is a view showing a frequency distribution of a period of a mode skip pulse formed in a period of twice the period. Fig. 18 is a frequency distribution diagram showing the period of the mode skip pulse divided into 2 in the mode skip pulse which is missing at the time of counting. Fig. 19 is a frequency distribution diagram showing the period of the mode skip pulse divided into 2 in the mode skip pulse which is missing at the time of counting. Fig. 20 is a graph showing the frequency distribution of the period of the mode skip pulse when both the missing and the excess count occur at the time of counting. Fig. 2 is a view for explaining an example of an error in the counting result after correction in the first embodiment of the present invention. Fig. 22 is a block diagram showing the configuration of a distance meter according to a third embodiment of the present invention. Figure 23 is a diagram showing a composite resonance model of a semiconductor laser in a conventional laser measuring device. Fig. 24 is a graph showing the relationship between the oscillation wavelength of the semiconductor laser and the output waveform of the built-in photodiode. [Main component symbol description] 1 : Semiconductor laser • 2 : Photodiode - 35 - 1359943 3 : Lens 4 = Laser driver 5 : Current-voltage conversion amplifier 8 : Counting device 9 : Operation device 1 〇 : Display device 1 1 : Filter circuit 1 2 : Measurement target 1 3 : Voltage detection circuit 8 1 : Determination unit 82 : Logic AND operation unit (AND) 8 3 : Counter 84 : Counting result correction unit 8 5 : Memory unit 1 〇 1 : Semiconductor laser 102: wall opening surface of semiconductor crystal 103 : Photodiode 104 : Measurement target 840 : Period measuring unit 841 : Frequency distribution forming unit 842 : Representative 値 calculating unit 8 4 3 : Correcting 値 calculating unit - 36