JP2004212225A - Abnormality diagnostic apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an abnormality diagnostic apparatus capable of reducing maintenance and management costs by reducing the frequency of disassembly and assembly work, which requires time and labor. <P>SOLUTION: The abnormality diagnostic apparatus for a bearing device for an axle of a rolling stock for diagnosing the presence or absence of abnormalities of the bearing device for the axle of the rolling stock is provided with a detection processing part incorporated in the bering device for outputting a signal, as an electric signal, generated from the bearing deice; an arithmetic processing part for performing arithmetic processing for determining the presence or absence of the abnormalities in machinery facilities on the basis of the output of the detection processing part; a result outputting part for outputting the result of determination by the arithmetic processing part. The detection processing part comprises a vibration detector for detecting vibrations of the bearing device on at least two directions or more. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

【００６２】
図１８は、本方法の処理フローを示すフローチャートである。本方法における周波数スペクトル算出までの処理は、図６のフローチャートのステップＳ１０１〜ステップＳ１０８に至るまでの処理と同一である。図１８には、ステップＳ１０８以降の処理を示す。
【００６３】
まず、図４に示す式を参照して、軸受の異常に起因する異常周波数を軸受の各部位（外輪、内輪、転動体、及び保持器）毎に算出し（ステップＳ５０１）、その後、算出された周波数を含む周波数帯域について、自乗平均（Ｖｉ）又はパーシャルオーバオール（Ｓｉ）、及び、周波数スペクトルの帯域全体の自乗平均（Ｖ_ＲＭＳ）又はオーバーオール（Ｓ_ＯＡ）である正規化値を算出する（ステップＳ５０２）。そして、前述の一つの次数成分帯域の自乗平均（Ｖｉ）又はパーシャルオーバオール（Ｓｉ）を前記正規化値（Ｖ_ＲＭＳ又はＳ_ＯＡ）で除した値又は差分の値を算出する（ステップＳ５０３）。
【００６４】
次に、除した値又は差分の値を保存されている参照データと比較照合し、除した値又は差分の値が正常な範囲であるかどうか、具体的には所定の基準値以上かどうかを判断する（ステップＳ５０４）。除した値又は差分の値が所定の基準値以上または以下であれば、異常有りと判断し、周波数帯に基づき、異常発生箇所を特定する（ステップＳ５０５）。ここで、基準値以上であるか以下である場合に異常有りとするかは、実際の測定によって定めればよい。それ以外の場合は、異常なしと判断する（ステップＳ５０６）。
【００６５】
以上の方法を、実際の測定結果を引用して説明する。図１９は、外輪に異常がある場合の周波数スペクトルを、図２０は、外輪に異常が無い場合のスペクトルを示すグラフである。図１９の左端近傍（１０〜２０Ｈｚあたり）には、異常ピーク周波数帯が存在する。このスペクトル全体の自乗平均値Ｖａは、０．０１６である。一方、図２０の対応するスペクトル全体の自乗平均値Ｖｎは、０．００８である。ここで、外輪きずに起因する異常周波数帯（基本周波数）に対して抽出する周波数帯域幅を２Ｈｚとすると、その帯域における自乗平均値をＶで正規化した値は、図１９の場合９０．７８であり、図２０の場合３８．４７となる。異常を有する場合は、正常時に比べて約２．４倍正規化した値が大きいことがわかる。従って、９０．７８と３８．４７の間又は正常時と異常時の比に所定のしきい値を設け、しきい値より大きい場合には、外輪に異常が発生していると判断することができる。
【００６６】
一方、図２１及び図２２は、複数の帯域を用いる場合の例を示す。図２１は、外輪に損傷を有するころ軸受及び正常な歯車（歯数；３１）を有する機械設備のエンベロープ周波数スペクトルを示すグラフである。この図では、５つの周波数ピークが観測されており、基本周波数から、その整数倍毎に２次成分から５次成分までが観測されている。一方、図２２は、図２１に対応する正常時の観測データであり、特異周波数は見あたらない。
【００６７】
以下、図２１及び図２２のデータに関し、上記手法を適用してみる。外輪きずに起因する基本周波数およびその５次までの成分の各帯域における自乗平均値の和をスペクトル全体の自乗平均値で正規化した値は、図２１の場合は１１．６４であり、図２２の場合は５．１９となる。ここで、５次の高調波とは、基本周波数から数えて５番目のピークを意味する。異常を有する場合は、正常時に比べて約２．２倍正規化した値が大きいことがわかる。従って、１１．６４と５．１９の間又は正常時と異常時の比に所定のしきい値を設け、しきい値より大きい場合には、外輪に異常が発生していると判断することができる。
【００６８】
以上が、比較判定部３６により異常の有無の判断を行う場合の具体的な処理パターンである。比較判定部３６は、これらの方法のうち複数の判断方法を用いて異常診断を行うように構成してもよい。複数の判断方法により、異常と判断することは、異常診断の正確さが向上するため好ましい。
【００６９】
データ蓄積出力部３８は、比較判定部３６の判定結果を保存する保存部であり、ハードディスクやメモリ媒体等で構成される。データ蓄積出力部３８は、比較判定部３８の判定結果を制御部４１及び結果出力部４２に出力する。データ蓄積出力部３８は、制御部４１へは以下に説明する制御部４１の動作に必要な場合にのみ、結果を出力するように構成しても良い。
【００７０】
制御処理部４０は、演算処理部３０の分析結果や判定結果を所定の表示形態で表示する表示手段としての結果出力部４２と、軸受２１が組み込まれている車両の駆動機構の動作を制御する制御系に前記比較判定部３６の判定結果に応じた制御信号をフィードバックする制御部４１とを備えている。
【００７１】
結果出力部４２は、具体的には、モニターや画像表示やプリンタへ印刷出力によって、演算処理部３０の分析結果や判定結果を通知する他、演算処理部３０の判定結果が異常有りの場合には、警告灯の点滅や警報機の作動による通知を行う。
【００７２】
制御部４１は、例えば、演算処理部３０の判定結果が異常有りの場合に、異常の程度に応じて、車両の走行停止や、速度の減速等を示す制御信号を車両の走行制御器に送る。本実施形態では、複数のセンサ２２は、軸受装置の軸受の状態を連続で測定し、演算処理部３０は、測定されたデータに基づき順次異常診断するようにしている。従って、制御処理部４０は、異常が発生したら即座に異常を通知し、車両の制御を行う。即ち、検出、分析、判定及び結果出力の流れがリアルタイムに行われている。
【００７３】
なお、センサ２２は、常に測定を行うように構成してもよいし、または所定時間毎に測定を行うように構成しても構わない。また、リアルタイムに異常診断を行う代わりに、車両運行時には測定及び測定データの蓄積のみを行い、後に別の場所で解析を行うように構成してもよい。例えば、日中は測定のみを行い、夜間にまとめて分析、判定及び結果出力を行うように構成しても構わない。
【００７４】
図２３は、本実施形態の異常診断装置１の多方向同時検出素子２２ｄによって検出された振動信号のうちの一成分の時間波形を示すグラフであり、そして、図２４及び図２５は、それぞれ図２３の時間波形の周波数スペクトルおよびエンベロープ処理後の周波数スペクトルを示すグラフである。図２３から、本実施形態の異常診断装置を用いることにより、周期的で感度の良い信号が検出されていることがわかる。また、図２５より、強いピークを持つ周期的な周波数ピークが観測されていることがわかり、ノイズと信号とが非常にはっきりと区分されていることがわかる。
【００７５】
一方、図２６は、一方向のみ検出可能な振動検出素子を用いて検出された振動信号の時間波形を示すグラフであり、そして、図２７及び図２８は、それぞれ図２６の時間波形の周波数スペクトルおよびエンベロープ処理後の周波数スペクトルを示すグラフである。この振動検出素子と、本実施形態の多方向同時検出素子２２ｄの取り付け位置は同じ位置であり、図２３のグラフを検出した場合と同様の欠陥を有する軸受を同じ条件で稼動させ測定を行っている。しかし、図２３とは異なり、図２６の時間波形には周期的な振動は観察されておらず、図２７及び図２８に示される周波数スペクトルを見ても周期的な強いピークは現れていない。これは、ノイズ成分が大きいため、傷に起因する信号成分がノイズに埋もれてしまったものと推察される。これは、一方向のみ検出可能な振動検出素子の最も感度の良い検出方向がと振動大の方向とが一致していないため、正しく検出されていないと考えられる。本実施形態の多方向同時振動検出素子２２ｄによれば、複数の検出方向から検出された複数の振動成分が得られる。したがって、一方向のみ検出可能な振動検出素子では、振動大の方向と素子の検出可能方向が構造所一致させられないような場合であっても、本実施形態の多方向同時振動検出素子２２ｄは、何れかの検出方向を振動大の方向と一致させることが可能であるため、ＳＮ比の良好な振動成分を的確に得ることが可能である。
【００７６】
以上、説明した本実施形態の車軸用軸受装置の異常診断装置１は、鉄道車両の車軸用軸受装置の軸受装置の異常の有無を診断する異常診断装置であって、軸受装置から発生する信号を電気信号として出力する検出器としての複数のセンサ２２を有する検出処理部２０と、検出処理部２０の出力を基に軸受装置の異常診断を行う演算処理を行う演算処理部３０と、演算処理部３０による判定結果を出力する結果出力部４２と、判定結果に基づき鉄道車両の制御系に制御信号をフィードバックする制御処理部４１と、を備えている。この異常診断装置１は、センサ付き軸受２１の構成部品の摩耗や破損に起因した異常の有無を判定するため、センサ付き軸受２１自体や軸受２１を含む鉄道車両自体を分解をせずに通常の使用状態のままで異常の有無を判定することができる。
【００７７】
従って、手間のかかる分解・組み立て作業の頻度を減少させて保守・管理コストを低減させることができる。また、規定の演算処理による分析や比較で機械的に判定を行うため、従来の目視検査と比較すると、検査担当者の熟練度や個人差によって判定がばらつく虞がなく、異常の有無の診断の信頼性を向上させることができる。
【００７８】
また、転がり軸受２１を構成する機構部品である外輪等に直にセンサ２２が組み込まれて、センサ２２が高感度で転がり軸受２１の発生する物理量を検出するため、転がり軸受２１の周囲の他の器物が発生する音や振動の周波数成分のピークが、センサ２２の検出する信号のＳＮ比に悪影響を及ぼす危険が低減し、センサユニット５の出力信号のＳＮ比の改善によって、分析・判定の精度の向上を図ることができる。特に、振動検出素子は、多方向の振動を検出する多方向同時振動検出素子で構成されているため、複数の振動情報を得ることが可能であり、振動の解析における信頼性を向上することが可能である。
【００７９】
なお、本実施形態の異常診断装置１は、軸受装置の軸受の異常の有無及び異常発生箇所を診断するとしたがこれに限られず、車輪のフラット部を診断するように構成してもよいし、また軸受装置の歯車の異常の有無及び異常発生箇所を診断するように構成してもよい。
【００８０】
また、図２９に示すように検出処理部２０の代わりに、図２（ｃ）に示すように、演算処理部３０と無線通信を行うセンサ５２を有する検出部５１ａ、５１ｂ、５１ｃからなる検出処理部５１を用いてもよい。各検出部５１ａ、５１ｂ、５１ｃは、それぞれ軸受２１の外輪２３上にセンサ５２が取り付けられて構成されている。このセンサ５２は、センサケース５２ａの内部に温度検出素子５２ｂと、回転検出素子５２ｃと、多方向同時振動検出素子５２ｄと無線通信用の送信部５２ｈが取り付けられている。各検出素子には、各検出素子５２ｂ〜５２ｄが検出した信号を所定の増幅率で増幅する増幅器を設けても良い。送信部５２ｈは、演算処理部３０に設けられた受信部６０に無線を介して信号を送信する。
【００８１】
このように構成することにより、検出処理部５１と演算処理部３０間の配線等を考慮することなく、センサを軸受装置に取り付けることが可能となるため、センサの配置自由度が増し、精度が高くなる位置にセンサを取り付けることが容易になる。同様の送信部及び受信部を設けることにより、演算処理部３０と制御処理部４０との間を無線通信で行うように構成しても良い。
【００８２】
さらに、図３０に示すように、演算処理部３０をマイクロコンピュータ６０を用いて構成しても良い。マイクロコンピュータ６０は、本実施形態のシステム用に開発された１チップマイクロコンピュータ、又は１ボードマイクロコンピュータであり、ＩＣチップ、メモリ等から構成される。マイクロコンピュータ６０が行う処理は、演算処理部３０で行われる処理と同一である。このようなマイクロコンピュータ６０を演算処理部３０として用いることにより、演算処理部の小型化を図ることが可能になり、鉄道車両への組み込みが容易となる。
【００８３】
あわせて、図３１に示すように、マイクロコンピュータ６０に通信用の送受信機７０を併設、あるいはマイクロコンピュータ６０に組み込み、車両上ではなく、地上に配置された制御処理部４０と無線を介して通信を行うように構成しても良い。例えば、地上側の信号送受信装置８０を線路際、駅、踏切等に設置し、鉄道車両が通過するときに、送受信機７０と信号送受信装置８０との間で信号の送受信を行い、電話回線等を用いた有線またはワイヤレスを用いた無線等、既存の通信インフラを利用して信号送受信装置８０と制御処理部４０間で送受信するように構成してもよい。この場合、制御処理部４０は、地上に設置された情報処理センタに配置されることとなる。
【００８４】
上記構成によれば、地上に配置された制御処理部により、異常診断装置が組み込まれた複数の車両の状態を同時に観測する一括集中管理が可能となる。従って、情報センター等の中央局に、制御処理部を配置することにより、異常診断を通常の業務内で効率的に行うことが可能となる。
【００８５】
【発明の効果】
以上、本発明の異常診断装置は、鉄道車両の軸受装置に組み込まれているセンサ付き軸受のセンサの出力を情報処理装置としての演算処理部によって分析すると共に、基準データと比較することで、摺動部材の構成部品の摩耗や破損に起因した異常の有無を判定するため、鉄道車両およびその軸受装置を分解をせずに通常の使用状態のままで異常判定することができる。
従って、手間のかかる分解・組み立て作業の頻度を減少させて保守・管理コストを低減させることができる。
また、規定の演算処理による分析や比較で機械的に判定を行うため、従来の目視検査と比較すると、検査担当者の熟練度や個人差によって判定がばらつく虞がなく、異常の有無の診断の信頼性を向上させることができる。
また、多方向の振動を同時に検出可能な多方向同時振動検出素子を用いているため、軸受へのセンサの取り付けが容易であり、感度良く異常に起因する振動を検出することができる。
【図面の簡単な説明】
【図１】本発明に係る鉄道車両用の異常診断装置の実施形態を示す図である。
【図２】センサの内部構造を示すブロック図である。
【図３】データ蓄積分配部を示す図である。
【図４】軸受の各部材の欠陥と各部材で発生する異常振動周波数の関係を示す関係式である。
【図５】軸受における負荷圏と非負荷圏の関係を示す図である。
【図６】第１の方法における処理のフローを示すフローチャートである。
【図７】異常が発生していない場合の周波数スペクトルを示すグラフである。
【図８】外輪に異常が発生している場合の周波数スペクトルを示すグラフである。
【図９】
保持器に傷がある場合の周波数スペクトルと基準値の関係を示すグラフである
。
【図１０】
第２の方法における処理のフローを示すフローチャートである。
【図１１】
周波数スペクトルを示す図である。
【図１２】
第３の方法における処理のフローを示すフローチャートである。
【図１３】
外輪に傷がある場合の周波数スペクトルを示す図である。
【図１４】
エンベロープ処理後の周波数スペクトルを示す図である。
【図１５】
剥離の大きさと、実測周波数スペクトルデータに現れるピークと基準レベルと
の間のレベル差の関係を示したものである。
【図１６】
第５の方法における処理のフローを示すフローチャートである。
【図１７】
周波数スペクトルのレベルと基準線との関係を示す図である。
【図１８】
第６の方法における処理のフローを示すフローチャートである。
【図１９】
外輪に異常がある場合の周波数スペクトルを示すグラフである。
【図２０】
外輪に異常が無い場合の周波数スペクトルを示すグラフである。
【図２１】
外輪に異常がある場合の周波数スペクトルを示すグラフである。
【図２２】
外輪に異常が無い場合の周波数スペクトルを示すグラフである。
【図２３】本実施形態の多方向同時検出素子によって検出された振動信号のうちの一成分の時間波形を示すグラフである。
【図２４】図２３の時間波形の周波数スペクトルを示すグラフである。
【図２５】図２３の時間波形のエンベロープ処理後の周波数スペクトルを示すグラフである。
【図２６】一方向のみ検出可能な振動検出素子によって検出された振動信号の時間波形を示すグラフである。
【図２７】図２６の時間波形の周波数スペクトルを示すグラフである。
【図２８】図２６の時間波形のエンベロープ処理後の周波数スペクトルを示すグラフである。
【図２９】本実施形態の変形例を示す図である。
【図３０】本実施形態の別の変形例を示す図である。
【図３１】図３０の変形例の応用例を示す図である。
【符号の説明】
１，２ 異常診断装置
２０，５１ 検出処理部
２１ 転がり軸受
２２，５２ センサ
２３ 外輪
２４ 内輪
２５ 転動体
３０ 演算処理部
３１ データ蓄積分配部
３２ 温度分析部
３３ 回転分析部
３４ フィルタ処理部
３５ 振動分析部
３６ 比較判定部
３７ 内部データ保存部
３８ データ蓄積出力部
４０ 制御処理部
４１ 制御部
４２ 結果出力部
５０ 増幅器
６０ マイクロコンピュータ（演算処理部）
７０ 送受信機
８０ 信号送受信機
１００ 入力部[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a device for diagnosing an abnormality of a bearing device for an axle of a railway vehicle, and more particularly to detecting a defect of a bearing without disassembling a bearing of a transmission of a bogie, a bearing in a gearbox, or a bearing device supporting an axle. The present invention relates to an abnormality diagnosis device that can be used.
[0002]
[Prior art]
Conventionally, in a bearing device that rotatably supports an axle of a railway vehicle, in order to prevent occurrence of inconvenience due to wear or damage of a bearing component that is a sliding member, a periodic visual inspection is performed. I have.
In this disassembly visual inspection, after using the vehicle for a certain period of time, the bearing is removed from the vehicle and disassembled, and a skilled professional inspector visually checks the disassembled components for the degree of wear and whether there are scratches . If an abnormality such as unevenness or wear not found in the new part is detected by this check, the part is replaced with a new part and assembly is performed again.
[0003]
[Problems to be solved by the invention]
However, this disassembly inspection requires a great deal of labor to disassemble the bearing from the vehicle and to reassemble the inspected bearing components, resulting in a significant increase in vehicle maintenance and management costs. there were.
[0004]
Further, for example, when reassembly is performed, a dent not formed before the inspection may be made on the bearing component, and the inspection itself may cause a defect of the bearing.
Further, since a large number of bearings are visually inspected within a limited time, there is a disadvantage that a defect may be overlooked.
Furthermore, in the visual inspection, there is an individual difference in the judgment of the degree of the defect, and even if there is substantially no defect, it may be considered that there is a defect and the parts may be replaced, which results in wasteful cost. Also.
[0005]
In order to reduce the time required for disassembly, a method of arranging a sensor for detecting vibration information or the like on a vehicle and performing abnormality diagnosis based on the detected signal may be considered. However, due to the inability to secure a mounting plane, insufficient rigidity, and the sensor being easily dropped, there is a problem in that the mounting position of the sensor is not flexible. There are also problems such as a deterioration in the ratio, and further improvement is desired.
[0006]
The present invention has been made in view of the above-described problems, and has as its object to provide an abnormality diagnosis device that can reduce maintenance and management costs by reducing the frequency of laborious disassembly and assembly operations. I do.
Another object of the present invention is to provide a bearing abnormality diagnosis device that can detect a defect that may be overlooked by visual inspection.
[0007]
[Means for Solving the Problems]
The object of the present invention is achieved by the following means.
(1) An abnormality diagnosis device for an axle bearing device of a railway vehicle, which diagnoses whether or not there is an abnormality in the axle bearing device of the railway vehicle,
A detection processing unit that is incorporated in the bearing device and outputs a signal generated from the bearing device as an electric signal.
An arithmetic processing unit that performs arithmetic processing to determine whether there is an abnormality in the mechanical equipment based on the output of the detection processing unit,
A result output unit that outputs a determination result by the arithmetic processing unit,
The abnormality diagnosis device according to claim 1, wherein the detection processing unit includes a vibration detector that detects vibration of the bearing device in at least two or more directions.
(2) The abnormality diagnosis device according to (1), wherein the detection processing unit is incorporated in a bearing of the bearing device.
(3) The abnormality diagnosis device according to (1) or (2), wherein the bearing with the sensor outputs the determination result to the result output unit via wireless communication.
(4) The abnormality diagnosis device according to any one of (1) to (3), wherein the arithmetic processing unit is configured by a microcomputer.
[0008]
As described above, according to the present invention, it is possible to make an abnormality determination in a normal use state without disassembling the railway vehicle and its bearing device. Therefore, it is possible to reduce the frequency of laborious disassembly / assembly work and reduce maintenance / management costs. In addition, since the judgment is made mechanically by analysis and comparison by prescribed arithmetic processing, there is no possibility that the judgment will vary due to the skill of the inspector and individual differences, and the diagnosis of the presence or absence of abnormality is not compared with the conventional visual inspection. Reliability can be improved.
[0009]
In addition, since the detection processing unit has a vibration detector that detects vibration information in at least two or more directions, even if a large vibration is not detected in one direction, the vibration is detected in another direction. By clearly detecting the vibration caused by the abnormality, a highly accurate abnormality diagnosis can be performed.
[0010]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, an embodiment of an abnormality diagnosis device for mechanical equipment according to the present invention will be described in detail with reference to the accompanying drawings.
[0011]
FIG. 1 shows an embodiment of an abnormality diagnosis device for a railway vehicle according to the present invention. The abnormality diagnosis device 1 includes a detection processing unit 20, an arithmetic processing unit 30, and a control processing unit 40.
[0012]
The abnormality diagnosis device 1 detects the occurrence of abnormality due to wear and breakage of a plurality of rolling bearings 21 in a bearing device that supports an axle of a railway vehicle. Each of the rolling bearings 21 includes an outer ring 23 internally fitted on the vehicle body side, an inner ring 24 externally fitted on the axle and rotating with the axle, and an outer ring raceway formed on the inner peripheral surface side of the outer ring 23 and an outer peripheral surface of the inner ring 24. A rolling element 25 such as a ball, a roller, etc., which is held by a retainer (not shown) between the inner raceway formed on the side and is rotatably arranged between the two raceways, and a detector attached to the outer race 21 Sensor 22. In FIG. 1, the detection processing unit 20 includes detection units 20a, 20b, and 20c each including a rolling bearing 21 and a sensor 22.
[0013]
The sensor 22 includes a plurality of detection elements that detect information such as a sound J1, a vibration J2, a rotation speed J3 of the bearing, a bearing temperature J4, a distortion J5 generated on the bearing outer ring, and the like as physical quantities that change according to the rotation state of the bearing 21. Is a detector having: Each detection element outputs each detected physical quantity to the arithmetic processing unit 30 as an electric signal. Each of the detection elements may be independently arranged in a different place, or instead of independently arranging a plurality of detection elements, a plurality of detection elements may be housed in the housing and a plurality of types of signals may be provided. May be configured to use as the sensor 22 a composite sensor unit that simultaneously detects In the present embodiment, as shown in FIG. 2A, the sensor 22 has a configuration in which various sensors 22b, 22c, and 22d are housed in a unit case 22a. Hereinafter, the sensor 22 includes a temperature detection element 22b for detecting the temperature of the bearing 21 inside the unit case 22a, a rotation detection element 22c for detecting the rotation speed of the inner ring (axle) of the bearing 21, and vibration generated in the bearing 21. The description will be made assuming that the multi-directional simultaneous vibration detecting element 22d to be detected is provided.
[0014]
The multi-directional simultaneous vibration detecting element 22d is configured by a pickup or a detector capable of detecting vibrations in a plurality of directions. As an example, it is possible to use a two-axis acceleration detector that detects acceleration in two directions of x and y axes, or a three-axis acceleration detector that detects acceleration in three directions of x, y and z axes. is there. Further, a configuration may be adopted in which a plurality of acceleration (vibration) detectors are arranged to measure vibrations in multiple directions. The multi-directional simultaneous vibration detecting element 22d is preferably mounted so as to detect at least vibration in the radial direction when the bearing 21 is a radial bearing, and at least in the axial direction when the bearing 21 is a thrust bearing. As the type of the detecting element, a piezoelectric vibration acceleration detector may be used, and a vibration acceleration detector other than a piezoelectric vibration detector, a speed detection vibration detector, a displacement detection vibration detector, or the like is used. It is also possible.
[0015]
Each of the detection elements 22b to 22d amplifies an electric signal corresponding to the detected temperature, rotation speed, and vibration via the amplifier 50, and outputs the amplified electric signal to the arithmetic processing unit 30. The amplifiers 50 may be provided inside the sensor unit case 22a as shown in FIG. 2B, or may be provided between the sensor 22 and the arithmetic processing unit 30 as shown in FIG. , May be provided inside the arithmetic processing unit 30. Preferably, amplifiers 22e, 22f, and 22g are attached to the detection elements 22b, 22c, and 22d, respectively, in the unit case 22a. Noise may enter while the signal output from the sensor 22 is transmitted to the arithmetic processing unit 30 via the cable, and the reliability of the measurement may be reduced. By doing so, it is less susceptible to noise and reliability is improved.
[0016]
A signal is transmitted between the sensor 22 and the arithmetic processing unit 30 by wire (cable). The wire (cable) preferably has waterproof, oil-proof, dust-proof, rust-proof, moisture-proof, heat-resistant, and electromagnetic noise-resistant in order to improve measurement accuracy such as noise reduction. Similarly, each of the detection elements 22b to 22d of the sensor 22 has waterproof, oil-proof, dust-proof, rust-proof, moisture-proof, heat-resistant, and electromagnetic noise-resistant in order to improve measurement accuracy such as noise reduction. Is preferred. For example, all the detection elements are housed in the sensor unit, and the housing 22a of the sensor 22 is provided with waterproof, oil-proof, dust-proof, rust-proof, moisture-proof, heat-resistant, and electromagnetic noise-resistant properties. is there.
[0017]
The arithmetic processing unit 30 is a unit that performs an arithmetic process on an electric signal, which is an output received from each of the detection elements 22b to 22d, and specifies whether or not the bearing is abnormal and the location where the abnormality has occurred. The arithmetic processing unit 30 may be configured using, for example, a personal computer in which an existing operation system and a software application for executing abnormality diagnosis are installed, or an arithmetic unit including independent processing and storage circuits for each unit. It may be constituted as.
[0018]
The arithmetic processing unit 30 includes a data accumulation / distribution unit 31, a temperature analysis unit 32, a rotation analysis unit 33, a filter processing unit 34, a vibration analysis unit 35, a comparison determination unit 36, an internal data storage unit 37, And a data accumulation output unit 38. Hereinafter, the configuration and function of each unit of the arithmetic processing unit 30 will be described in detail.
[0019]
FIG. 3 is a diagram showing a data accumulation / distribution unit 31 which is a first data accumulation unit. The data storage / distribution unit 31 includes a data storage unit 31a, a sampling unit 31b, and a sampling reference setting unit 31c. The data storage unit 31a is a data storage medium for storing output signals from the detection elements 22b to 22d for each signal, and can be configured by various memories, a hard disk, or the like.
[0020]
The data storage unit 31a receives and temporarily stores the signals sent from the detection elements 22b to 22d, and distributes the signals to any of the analysis units 32, 33, and 34 according to the type of the signal. Various signals are converted into digital signals by an A / D converter (not shown) before being sent to the data accumulation / distribution unit 31.
[0021]
The sampling criterion setting unit 31b sets a criterion value based on information obtained from the external input unit 100 for excluding a region where the influence of noise is large from the analog signal output by the multidirectional simultaneous vibration detection element 22d. Things. Here, the input unit 100 is input means such as a mouse and a keyboard, and a user can arbitrarily set a reference value via the input unit 100.
[0022]
The sampling unit 31c cuts out the vibration, temperature, and rotation speed data, which are time data, into a predetermined length, and performs sampling for outputting a signal to a subsequent analysis unit. When the output signal from the multi-directional simultaneous vibration detecting element 22d includes noise larger than the reference value set by the sampling reference setting unit 31b, the sampling unit 31c performs signal sampling for a time period in which the noise is included. Sampling is not performed, and no signal is output to the subsequent filter processing unit 34. Specifically, two points A and B where the signal level is equal to or more than a certain value are detected, and data is output to the filter processing unit 34 and the vibration analysis unit 35 during the time from A to B. Control not to. As a result, it is possible to prevent the frequency analysis in the time domain including large noise data from being performed, and to reduce the possibility of erroneous abnormality diagnosis. The sampling reference setting unit 31b and the sampling unit 31c are not necessarily required to be provided. If a similar effect can be obtained, the sampling reference setting unit 31b and the sampling unit 31c are arranged in different places, for example, in front of the data storage unit 31a. May be.
[0023]
The temperature analysis unit 32 calculates the temperature of the bearing based on the output signal from the temperature detection element 22b, and sends the calculated temperature to the comparison determination unit 36. The analysis unit 32 has, for example, a temperature conversion table corresponding to the characteristics of the detection element, and calculates temperature data based on the level of the detection signal.
[0024]
The rotation analysis unit 33 calculates the rotation speed of the inner ring 24, that is, the axle, based on the output signal from the rotation detection element 22c, and sends the calculated rotation speed to the comparison determination unit 36. For example, when the rotation detection element 22c is composed of an encoder attached to the inner ring 24 and a magnet and a magnetic detection element attached to the outer ring 23, the signal output by the rotation detection element 22c is It becomes a pulse signal according to the rotation speed. The rotation analysis unit 33 has a predetermined conversion function or conversion table corresponding to the shape of the encoder, and calculates the rotation speed of the inner wheel 24 and the axle from the pulse signal according to the function or table.
[0025]
The vibration analysis unit 35 analyzes the frequency of the vibration generated in the bearing 21 based on the output signal from the multidirectional simultaneous vibration detection element 22d. Specifically, the vibration analysis unit 35 is an FFT calculation unit that calculates the frequency spectrum of the vibration signal, and calculates the frequency spectrum of the vibration based on the FFT algorithm. The calculated frequency spectrum is sent to the comparison / determination unit 36. Further, the vibration analysis unit 35 may be configured to perform an envelope process for obtaining an envelope of the vibration signal as a pre-process for performing the FFT to reduce noise. The vibration analysis unit 35 also outputs the envelope data after the envelope processing to the comparison determination unit 36 as necessary.
[0026]
Generally, an abnormal frequency band of vibration caused by rotation of a bearing is determined depending on the size of the bearing, the number of rolling elements, and the like. The relationship between the defect of each member of the bearing and the abnormal vibration frequency generated in each member is as shown in FIG. In the frequency analysis, the maximum frequency (Nyquist frequency) at which the Fourier transform can be performed is determined according to the sampling time. Therefore, it is preferable that frequencies higher than the Nyquist frequency are not included in the vibration signal. Therefore, in the present embodiment, a filter processing unit 34 is provided between the data accumulation / distribution unit 31 and the vibration analysis unit 35, a predetermined frequency band is cut out by the filter processing unit 34, and a vibration signal including only the cut out frequency band. Is transmitted to the vibration analysis unit 35. When the axle is rotating at a low speed in a railway vehicle, for example, it may be configured to extract only a frequency component of 1 kHz or less.
[0027]
Further, the filter processing unit 34 does not perform the filter processing at first, but first obtains the frequency spectrum in the vibration analysis unit 35, estimates in advance the frequency band in which the peak is observed, and then performs the filter processing corresponding to the frequency band. After that, the frequency analysis may be performed again. With this configuration, unnecessary noise can be effectively eliminated, and highly accurate frequency analysis can be performed.
[0028]
When the sensor 22, particularly the multidirectional simultaneous vibration detecting element 22d, is attached to a portion where a load is applied (load zone), a signal can be detected with high sensitivity, so that more accurate measurement can be performed. Here, the load zone indicates a region where a load is applied to the rolling elements as shown in FIG.
[0029]
Also, when there is no choice but to install the sensor in the non-load zone when there is no space to mount the sensor in the load zone or when a high voltage cable that generates noise is installed, increase the signal detection sensitivity by filtering etc. By doing so, measurement becomes possible.
[0030]
The comparing and judging unit 36 compares the frequency spectrum of the vibration obtained by the vibration analyzing unit 35 with a reference value stored in the internal data storage unit 37 or a reference value calculated from the frequency spectrum. It is determined whether or not. Here, the reference value is data of frequency components caused by wear or breakage of a specific portion of the bearing, or a predetermined value included in a spectrum obtained for each frequency spectrum. The comparing and judging unit 36 simultaneously judges with the frequency components, analyzes the temperature and rotation speed obtained from the temperature analyzing unit 32 and the rotation analyzing unit 33, and stores various types of bearings stored in the internal data storage unit 37. Refer to specification data such as data to ensure the accuracy of judgment.
[0031]
Specifically, when it is determined that there is an abnormality based on the frequency spectrum of the vibration, the comparison determination unit 36 checks the temperature of the bearing, and if the temperature exceeds a predetermined value, a serious abnormality has occurred. Judge. If only one of them indicates an abnormality, it is determined that some abnormality has occurred. If both results are normal, it is determined that there is no abnormality. If only one of them indicates an abnormality, it may be determined that there is an abnormality if the result does not change even if the determination is made a plurality of times. The comparison / determination unit 36 outputs the result of the abnormality diagnosis to the data accumulation output unit 38.
[0032]
As a specific process of the abnormality diagnosis based on the vibration information performed by the comparison determination unit 36, the following method can be used.
[0033]
(1) Method of Using Effective Value of Envelope Data as Reference Value In this method, a frequency component occurring at the time of abnormality is obtained based on the equation in FIG. Then, an effective value of the envelope data is calculated, and a reference value for comparison is obtained from the effective value. Then, a frequency equal to or higher than the reference value is calculated and compared with a frequency component generated at the time of abnormality. Hereinafter, description will be made with reference to FIG.
[0034]
First, the vibration of the bearing is detected via the multi-directional simultaneous vibration detecting element 22d housed in the sensor unit 22a (Step S101). The detected signal is amplified at a predetermined amplification rate, and is converted into a digital signal by the A / D converter (Step S102). The vibration signal converted into the digital signal is stored in the data storage / distribution unit 31 in a predetermined format (step S103).
[0035]
Next, the frequency spectrum of the digital signal is obtained (Step 104), and based on the obtained frequency spectrum, the filter processing unit 34 selects a filter band applied to the digital signal (Step S105). The filter processing unit 34 performs a filter process for removing a frequency component other than the selected filter band (step S106), and outputs the digital signal after the filter process to the vibration analysis unit 35. Then, the vibration analysis unit 35 performs an envelope process on the digital signal after the filter process (step S107), and obtains a frequency spectrum of the digital signal after the envelope process (step S108).
[0036]
At the same time, an effective value of the digital signal after the envelope processing is calculated (step S109), and a reference value used for abnormality diagnosis is calculated based on the effective value (step S112). Here, the effective value is obtained as an average value of the absolute value of the amplitude of the digital signal after the envelope processing. The reference value is calculated based on the following formula (1) or (2) based on the effective value.
(Reference value) = (effective value) + α (1)
(Reference value) = (effective value) × β (2)
α, β: predetermined values variable depending on the type of data
Next, based on the table shown in FIG. 4, the frequency generated due to the abnormality of the bearing is obtained (step S110), and the level of the abnormal frequency component of each member corresponding to the obtained frequency, that is, the inner ring damage component Si ( Zfi), the outer wheel injury component So (Zfc), the rolling element component Sb (2fb), and the cage component Sc (fc) are extracted (step S111), and are compared with the reference values calculated in step 112 (step S113). ). If the values of all components are smaller than the reference value, it is determined that no abnormality has occurred in the bearing (step S114). It is determined that an abnormality has occurred (step S115).
[0038]
FIG. 7 is a graph illustrating a frequency spectrum when no abnormality occurs, and FIG. 8 is a graph illustrating a frequency spectrum when an abnormality occurs in the outer ring. In the example of FIG. 7, the reference value was obtained as −29.3 dB from the envelope data. A line of a reference value is drawn in the graph in FIG. 7, and a comparison is made with the inner wheel injury component Si (Zfi), the outer wheel injury component So (Zfc), the rolling element component Sb (2fb), and the cage component Sc (fc). , The level of each component was smaller than the reference value. Therefore, this bearing can be determined to be normal. On the other hand, in the case of FIG. 8, the outer ring damage component So (Zfc) protrudes more than the reference value, and it can be determined that an abnormality has occurred in the outer ring of the bearing.
[0039]
FIG. 9 is a graph showing the relationship between the frequency spectrum and the reference value when the cage is damaged. In FIG. 9, a peak larger than the reference value is observed at the frequency fc corresponding to the cage flaw. As described above, since the presence or absence of the peak of the generated frequency is determined by comparing the level at the frequency due to the bearing with the reference value, it is necessary to appropriately diagnose even a small peak as shown in FIG. Is possible.
[0040]
(2) Method of Finding Peak of Spectrum and Comparing Peak Frequency and Abnormal Frequency In this method, a frequency component occurring at the time of abnormality is found based on the equation in FIG. Then, the frequency analysis unit 36 checks whether a predetermined number or more than a reference value of the frequency spectrum in the frequency spectrum obtained corresponds to a frequency component in which an abnormality occurs. Hereinafter, the details will be described based on the flowchart shown in FIG.
[0041]
The flow up to step S108 is the same as that described in the method (1), and a description thereof will be omitted. In this method, first, the peak of the obtained frequency spectrum is calculated (step S109). Here, in order to obtain the frequency peak, first, difference data indicating the difference between the level of the data point of each frequency component and the level of the data point of the immediately preceding frequency component is obtained. Then, an inflection point at which the sign of the difference data changes from positive to negative is searched for, and it is determined that the peak is at a frequency value related to the difference data on which the positive or negative is based. However, since the peak value necessary for the diagnosis is only for a peak having a sharp peak (slope), when the slope is larger or smaller than a predetermined reference value (for example, 1 or −1). Only at this point, it is determined that the peak is reached.
[0042]
FIG. 11 is a diagram showing a frequency spectrum. In FIG. 11, for three consecutive points A (X ₀ , Y ₀ ), B (X ₁ , Y ₁ ) and C (X ₂ , Y ₂ ), the point B is a peak. In this case, the difference data δ ₁ = Y ₁ −Y ₀ > 0 between A and B is satisfied, and the difference data δ ₂ = Y ₂ −Y ₁ <0 between B and C is satisfied. Has changed from positive to negative. If the slope (Y ₁ −Y ₀ ) / (X ₁ −X ₀ )> 1 or (Y ₂ −Y ₁ ) / (X ₂ −X ₁ ) <− 1, the point B is Judge as peak.
[0043]
Then, based on the specifications of the bearing, the abnormal frequency is calculated based on FIG. 4 (step S202), and the level of the abnormal frequency component of each member corresponding to the obtained frequency, that is, the inner wheel damage component Si (Zfi) and the outer wheel damage component So (Zfc), rolling element component Sb (2fb) and cage component Sc (fc) are extracted (step S203). Then, the peak frequency is compared with the frequency generated at the time of abnormality, and it is determined whether the peak frequency matches the calculated abnormal frequency (step S204). If a certain peak matches the abnormal frequency, it is determined that an abnormality has occurred in a member corresponding to the abnormal frequency (step S206). If not coincident with any frequency, it is determined that there is no abnormality (step S205).
[0044]
(3) Method using a fundamental frequency and a specific harmonic The method uses a first order value which is a fundamental frequency of an abnormal frequency component, a second order value having a frequency twice as much as the fundamental frequency, and a value four times the fundamental frequency. For the fourth order value having a frequency, it is compared whether the peak frequency and the frequency generated at the time of abnormality match, and if it is determined that there is an abnormality at at least two frequencies, it is finally determined that there is an abnormality. If the number of frequencies determined to be abnormal is one or less, it is determined that there is no abnormality. Hereinafter, this will be described in detail with reference to FIG.
[0045]
The process from calculating the frequency spectrum to calculating the frequency generated due to the abnormality is the same as the flow of the method (1). In the present method, in the comparison, first, as shown in FIG. 12, it is determined whether or not the value of the spectrum is equal to or higher than the reference value at the frequency of the basic component (primary component) generated at the time of abnormality (step S301). If the spectrum value is equal to or greater than the reference value, it is determined that the primary components match, and the process proceeds to step S302. On the other hand, if they do not match, the process proceeds to step S311.
[0046]
In step S302, it is determined whether or not the value of the spectrum is equal to or greater than the reference value at the frequency of the secondary component having a frequency twice as high as the fundamental component generated at the time of abnormality. If the value of the spectrum is equal to or greater than the reference value, it is determined that the secondary components match, and it is finally determined in step S321 that an abnormality has occurred at the corresponding location. On the other hand, if they do not match, the process proceeds to step S312.
[0047]
Also in step S311, it is determined whether or not the value of the spectrum is equal to or larger than the reference value at the frequency of the secondary component having a frequency twice as high as the fundamental component generated at the time of abnormality. If the value of the spectrum is equal to or larger than the reference value, it is determined that the secondary components match, and the process proceeds to step S312. On the other hand, if they do not match, the process proceeds to step S321, and it is finally determined that no abnormality has occurred in the corresponding location.
[0048]
In step S312, it is determined whether or not the value of the spectrum is equal to or greater than the reference value at the frequency of the fourth-order component having a frequency four times that of the fundamental component generated at the time of abnormality. If the value of the spectrum is equal to or greater than the reference value, it is determined that the fourth-order component has matched, and it is finally determined in step S321 that an abnormality has occurred at the corresponding location. On the other hand, if they do not match, it is finally determined that no abnormality has occurred in the corresponding location.
[0049]
FIG. 13 is a diagram showing a frequency spectrum when the outer ring has a flaw. It can be seen that a harmonic that is a natural number multiple of the fundamental frequency Zfc is observed. When the reference value in this case is -10 dB, it can be seen that the spectrum value is equal to or higher than the reference value for all of the first, second, and fourth order components. Therefore, it is determined by the processing of the present method that an abnormality has occurred in the outer ring.
[0050]
Generally, a case may be considered in which a large peak is accidentally observed due to the influence of noise or the like at the frequency corresponding to the abnormality. However, according to the present method, at least two of the first, second and fourth order components are used. Unless the peak value is equal to or higher than the reference value, the frequency is not determined to be abnormal, so that the possibility of erroneous determination can be reduced.
[0051]
In the flowchart of FIG. 12, the comparison is performed in the order of first, second, and fourth. However, the comparison may be performed in order from the one having the highest peak level. In this case, if the peak of the frequency having the largest peak is equal to or less than the reference value, it can be determined that there is no abnormality at that point, and the calculation time can be reduced.
Further, the frequency component to be combined may be a combination of a first-order value, a second-order value, and a third-order value, or a combination of a second-order value, a fourth-order value, and a sixth-order value.
[0052]
(4) Method of Estimating the Size of Damage Along with Diagnosis of Abnormality In the methods (1) to (3), the presence or absence of an abnormality is diagnosed. However, it is also possible to estimate the size of the damage as follows. . FIG. 14 is a diagram illustrating a frequency spectrum after the envelope processing. In the figure, a large peak is observed at the frequency Zfc, which indicates that the outer ring is damaged. Comparing the reference level L ₀ is an average value of the entire value Ln and the frequency spectrum of the peak at this ZFC, it is possible to estimate the magnitude of the damage in the outer ring which abnormally.
[0053]
FIG. 15 shows the relationship between the magnitude of the peeling and the level difference between the peak appearing on the actually measured frequency spectrum data d1 and the reference level when the bearing ring is peeled, which is damage to the bearing ring. It is. As described above, since the level difference generally increases in proportion to the magnitude of the damage, the magnitude of the damage is estimated by calculating the level difference at the peak on the actually measured frequency spectrum data d1. It is possible. The increase in the peak level on the actually measured frequency spectrum data d1 is most noticeable at the peak corresponding to the primary value of the frequency component. Therefore, when an abnormality is detected by calculating the level difference l between the primary value Ln and the reference level L ₀ of the frequency component, estimating the degree of damage, depending on the degree of damage, damaged parts Exchange time can be determined. As a result, it is possible to reduce the maintenance cost by performing replacement at an appropriate time instead of performing excessive component replacement or maintenance.
[0054]
(5) A method using a level difference from a harmonic component that is a natural number multiple of the fundamental frequency as a reference value This method uses the fundamental frequency 2, 3, 3, 4,... Counting the number of n-th, second, third, fourth,... N-th levels having n times the frequency is equal to or more than the reference value. Is determined to have occurred. Specifically, counting is performed when the value of the n-th order is equal to or more than {(first-order level) − (n−1) · a} (dB) with respect to the first-order level. Here, “a” is an arbitrary value. Hereinafter, description will be made with reference to the flowchart shown in FIG.
[0055]
FIG. 16 is a flowchart showing a processing flow in the present method. The processing up to the calculation of the frequency spectrum in this method is the same as the processing up to steps S101 to S108 in the flowchart of FIG. FIG. 16 shows the processing after step S108.
[0056]
First, referring to the equation shown in FIG. 4, an abnormal frequency caused by an abnormality in the bearing is calculated for each part (outer ring, inner ring, rolling element, and cage) of the bearing (step S401), and the abnormal frequency is calculated. The level of the frequency spectrum to be extracted is extracted (step S402). Then, the level of a frequency spectrum corresponding to a natural number times (2, 3,..., N times) of the abnormal frequency is calculated (step S403). Here, it is assumed that the second, third, fourth, and fifth order components having frequencies two, three, four, and five times the fundamental abnormal frequency are extracted.
[0057]
Next, the level of each of the second, third, fourth and fifth order components is confirmed based on the basic first order value (step S404). Here, when the level of each component is equal to or more than {(first-order level) −3 (n · 1)} (dB), the count of the presence of an abnormality is performed. Specifically, in the following cases, the count of abnormalities is performed for each component.
(Level of secondary component)> (Level of primary component) -3
(Level of third-order component)> (Level of first-order component) -6
(Level of fourth-order component)> (Level of first-order component) -9
(5th-order component level)> (1st-order component level) -12
[0058]
Then, a final abnormality determination is made by confirming whether or not the number of abnormal occurrence counts is equal to or greater than a predetermined number (step S405). Here, if there is a count of two or more abnormalities, it is finally determined that there is an abnormality, and if it is one or less, it is determined that there is no abnormality.
[0059]
FIG. 17 is a diagram showing the relationship between the level of the frequency spectrum and the reference line when the inner ring of a cylindrical roller bearing (outside diameter 215 mm, inside diameter 100 mm, width 47 mm, number of rollers 14) is rotated at about 300 min ⁻¹ . . The straight line in the figure is a determination reference line connecting the above-described reference values with a line. When there is a flaw in the bearing, the values of the second and higher order components are above the judgment reference line, but the peak levels corresponding to the rolling noise generated even in a normal state are the second and fourth order components. Is below this determination reference line. In general, the rolling noise (rolling member falling sound) has a higher-order component lower than that in the case where the outer ring has a defect. Therefore, as shown in FIG. 17, most of the values fall below the determination reference line. Thereby, even in the case where a peak appears at the same frequency as the case where the outer ring has a defect, such as a roller dropping sound, by comparing the levels of the higher-order components, it is possible to more accurately determine whether the abnormality is normal or abnormal. It is possible.
[0060]
(6) Method using effective value for each frequency band In this method, abnormality diagnosis is performed using the effective value of the frequency band including the frequency caused by the abnormality, not the value of the peak level of the frequency caused by the abnormality. . Specifically, the effective value of the frequency band including the frequency due to the abnormality is the root mean square or partial overall of the frequency band level. Here, the square average V _i and the partial over-ol S _i is given by the following equation. Here, _VRMS and _SOA are the root mean square and overall in all frequency bands. Overall means the sum of a specific designated section.
[0061]
(Equation 1)

[0062]
FIG. 18 is a flowchart showing the processing flow of the present method. The processing up to the calculation of the frequency spectrum in this method is the same as the processing up to steps S101 to S108 in the flowchart of FIG. FIG. 18 shows the processing after step S108.
[0063]
First, referring to the equation shown in FIG. 4, an abnormal frequency caused by an abnormality in the bearing is calculated for each part (outer ring, inner ring, rolling element, and cage) of the bearing (step S501), and thereafter, the calculated frequency is calculated. and the frequency band including the frequency, root-mean-square (Vi) or partial over-ol (Si), and calculates the normalized value is a mean square of the entire band of the frequency spectrum _{(V RMS)} or overalls _{(S OA)} ( Step S502). Then, a value obtained by dividing the root mean square (Vi) or partial overall (Si) of the one order component band by the normalization value ( _VRMS or _SOA ) or a difference value is calculated (step S503).
[0064]
Next, the divided value or the difference value is compared with the stored reference data, and it is determined whether the divided value or the difference value is within a normal range, specifically, whether the divided value or the difference value is equal to or more than a predetermined reference value. A determination is made (step S504). If the divided value or the difference value is equal to or greater than or equal to a predetermined reference value, it is determined that there is an abnormality, and an abnormality occurrence location is specified based on the frequency band (step S505). Here, whether or not there is an abnormality when the value is equal to or more than the reference value may be determined by actual measurement. Otherwise, it is determined that there is no abnormality (step S506).
[0065]
The above method will be described with reference to actual measurement results. FIG. 19 is a graph showing a frequency spectrum when there is an abnormality in the outer ring, and FIG. 20 is a graph showing a spectrum when there is no abnormality in the outer ring. An abnormal peak frequency band exists near the left end of FIG. 19 (around 10 to 20 Hz). The root mean square value Va of the entire spectrum is 0.016. On the other hand, the root mean square value Vn of the entire corresponding spectrum in FIG. 20 is 0.008. Here, assuming that a frequency bandwidth to be extracted for an abnormal frequency band (fundamental frequency) caused by an outer ring flaw is 2 Hz, a value obtained by normalizing a root mean square value in the band to V is 90.78 in the case of FIG. This is 38.47 in the case of FIG. It can be seen that, when there is an abnormality, the normalized value is about 2.4 times larger than in the normal state. Therefore, a predetermined threshold value is provided between 90.78 and 38.47 or the ratio between the normal state and the abnormal state, and when the threshold value is larger than the threshold value, it can be determined that an abnormality has occurred in the outer ring. it can.
[0066]
21 and 22 show examples in which a plurality of bands are used. FIG. 21 is a graph showing an envelope frequency spectrum of a machine having a roller bearing having a damaged outer ring and a normal gear (number of teeth: 31). In this figure, five frequency peaks are observed, and from the fundamental frequency, the second to fifth order components are observed at every integral multiple thereof. On the other hand, FIG. 22 shows normal observation data corresponding to FIG. 21, and no singular frequency is found.
[0067]
Hereinafter, the above method will be applied to the data in FIGS. 21 and 22. The value obtained by normalizing the sum of the root-mean-square values of the fundamental frequency caused by the outer ring flaw and the components up to the fifth order in each band with the root-mean-square value of the entire spectrum is 11.64 in FIG. 21 and FIG. In the case of, it is 5.19. Here, the fifth harmonic means the fifth peak counted from the fundamental frequency. In the case where there is an abnormality, it is understood that the normalized value is about 2.2 times larger than that in the normal state. Therefore, a predetermined threshold value is provided between 11.64 and 5.19 or the ratio between the normal time and the abnormal time, and if it is larger than the threshold value, it can be determined that an abnormality has occurred in the outer ring. it can.
[0068]
The above is a specific processing pattern when the comparison determination unit 36 determines whether there is an abnormality. The comparison determination unit 36 may be configured to perform the abnormality diagnosis using a plurality of determination methods among these methods. It is preferable to determine an abnormality by a plurality of determination methods because accuracy of abnormality diagnosis is improved.
[0069]
The data accumulation output unit 38 is a storage unit that stores the determination result of the comparison determination unit 36, and is configured by a hard disk, a memory medium, or the like. The data accumulation output unit 38 outputs the determination result of the comparison determination unit 38 to the control unit 41 and the result output unit 42. The data accumulation output unit 38 may be configured to output the result to the control unit 41 only when necessary for the operation of the control unit 41 described below.
[0070]
The control processing unit 40 controls the operation of a result output unit 42 as a display unit for displaying the analysis result and the determination result of the arithmetic processing unit 30 in a predetermined display form, and the driving mechanism of the vehicle in which the bearing 21 is incorporated. The control system includes a control unit 41 that feeds back a control signal according to the determination result of the comparison determination unit 36.
[0071]
Specifically, the result output unit 42 notifies the analysis result and the determination result of the arithmetic processing unit 30 by a monitor, an image display, and a printout to a printer, and also outputs the result of the arithmetic processing unit 30 when there is an abnormality. Provides a notification by flashing a warning light or activating an alarm.
[0072]
For example, when the determination result of the arithmetic processing unit 30 indicates that there is an abnormality, the control unit 41 sends a control signal indicating a stop of traveling of the vehicle, a reduction in speed, or the like to the traveling controller of the vehicle according to the degree of abnormality. . In the present embodiment, the plurality of sensors 22 continuously measure the state of the bearing of the bearing device, and the arithmetic processing unit 30 sequentially diagnoses abnormalities based on the measured data. Therefore, when an abnormality occurs, the control processing unit 40 immediately notifies the abnormality and controls the vehicle. That is, the flow of detection, analysis, determination, and result output is performed in real time.
[0073]
Note that the sensor 22 may be configured to always perform the measurement, or may be configured to perform the measurement every predetermined time. Further, instead of performing the abnormality diagnosis in real time, only the measurement and the accumulation of the measurement data may be performed during the operation of the vehicle, and the analysis may be performed later in another place. For example, only the measurement may be performed during the day, and the analysis, the determination, and the result output may be collectively performed at night.
[0074]
FIG. 23 is a graph showing a time waveform of one component of the vibration signal detected by the multidirectional simultaneous detection element 22d of the abnormality diagnosis device 1 of the present embodiment, and FIGS. 24 and 25 are diagrams respectively. 23 is a graph showing a frequency spectrum of a time waveform of No. 23 and a frequency spectrum after envelope processing. From FIG. 23, it can be seen that a periodic and highly sensitive signal is detected by using the abnormality diagnosis apparatus of the present embodiment. FIG. 25 also shows that a periodic frequency peak having a strong peak is observed, and that the noise and the signal are very clearly separated.
[0075]
On the other hand, FIG. 26 is a graph showing a time waveform of a vibration signal detected using a vibration detecting element capable of detecting only one direction, and FIGS. 27 and 28 are frequency spectra of the time waveform of FIG. 26, respectively. 7 is a graph showing a frequency spectrum after envelope processing. The mounting position of the vibration detecting element and the multi-directional simultaneous detecting element 22d of the present embodiment are the same position, and the bearing having the same defect as that detected in the case of detecting the graph of FIG. I have. However, unlike FIG. 23, no periodic oscillation is observed in the time waveform of FIG. 26, and a strong periodic peak does not appear in the frequency spectra shown in FIGS. 27 and 28. This is presumed to be because the noise component is large, and the signal component resulting from the flaw has been buried in the noise. This is presumably because the most sensitive detection direction of the vibration detecting element capable of detecting only one direction does not match the direction of the large vibration, and thus the vibration is not correctly detected. According to the multidirectional simultaneous vibration detection element 22d of the present embodiment, a plurality of vibration components detected from a plurality of detection directions can be obtained. Therefore, in the case of a vibration detecting element that can detect only one direction, even if the direction of the large vibration and the detectable direction of the element cannot be matched with each other in the structure, the multidirectional simultaneous vibration detecting element 22d of the present embodiment is Since it is possible to make any of the detection directions coincide with the direction of the large vibration, it is possible to accurately obtain a vibration component having a good SN ratio.
[0076]
The abnormality diagnosis device 1 for the axle bearing device according to the present embodiment described above is an abnormality diagnosis device for diagnosing the presence or absence of an abnormality in the bearing device of the axle bearing device of the railway vehicle, and outputs a signal generated from the bearing device. A detection processing unit 20 having a plurality of sensors 22 as detectors that output electric signals, a calculation processing unit 30 performing calculation processing for diagnosing abnormality of the bearing device based on the output of the detection processing unit 20, and a calculation processing unit The system includes a result output unit 42 that outputs a result of the determination by the control unit 30 and a control processing unit 41 that feeds back a control signal to a control system of the railway vehicle based on the result of the determination. This abnormality diagnosis device 1 determines the presence or absence of abnormality due to wear or breakage of the components of the sensor-equipped bearing 21, so that the sensor-equipped bearing 21 itself or the railway vehicle itself including the bearing 21 is not disassembled and used in a normal manner. It is possible to determine the presence or absence of an abnormality in the state of use.
[0077]
Therefore, it is possible to reduce the frequency of laborious disassembly / assembly work and reduce maintenance / management costs. In addition, since the judgment is made mechanically by analysis and comparison by prescribed arithmetic processing, there is no possibility that the judgment will vary due to the skill of the inspector and individual differences, and the diagnosis of the presence or absence of abnormality is not compared with the conventional visual inspection. Reliability can be improved.
[0078]
Further, the sensor 22 is directly incorporated into an outer ring or the like which is a mechanical component constituting the rolling bearing 21, and the sensor 22 detects the physical quantity generated by the rolling bearing 21 with high sensitivity. The risk that the peak of the frequency component of the sound or vibration generated by the object adversely affects the S / N ratio of the signal detected by the sensor 22 is reduced, and the improvement of the S / N ratio of the output signal of the sensor unit 5 improves the accuracy of analysis and determination. Can be improved. In particular, since the vibration detecting element is configured by a multi-directional simultaneous vibration detecting element that detects multi-directional vibrations, it is possible to obtain a plurality of pieces of vibration information, thereby improving reliability in vibration analysis. It is possible.
[0079]
Note that the abnormality diagnosis device 1 of the present embodiment diagnoses the presence / absence of an abnormality in the bearing of the bearing device and the location where the abnormality has occurred. However, the present invention is not limited thereto, and the abnormality diagnosis device 1 may be configured to diagnose a flat portion of a wheel. Further, a configuration may be adopted in which the presence / absence of an abnormality in the gear of the bearing device and the location of the abnormality are diagnosed.
[0080]
In addition, as shown in FIG. 2C, a detection process including detection units 51a, 51b, and 51c having a sensor 52 that performs wireless communication with the arithmetic processing unit 30 as shown in FIG. The unit 51 may be used. Each of the detection units 51a, 51b, and 51c is configured such that a sensor 52 is mounted on the outer ring 23 of the bearing 21. In the sensor 52, a temperature detecting element 52b, a rotation detecting element 52c, a multi-directional simultaneous vibration detecting element 52d, and a wireless communication transmitting section 52h are mounted inside a sensor case 52a. Each detection element may be provided with an amplifier that amplifies the signal detected by each of the detection elements 52b to 52d at a predetermined amplification factor. The transmitting unit 52h transmits a signal to the receiving unit 60 provided in the arithmetic processing unit 30 via radio.
[0081]
With such a configuration, the sensor can be attached to the bearing device without considering wiring between the detection processing unit 51 and the arithmetic processing unit 30 and the like. It becomes easy to mount the sensor at a higher position. By providing the same transmission unit and reception unit, the configuration may be such that the arithmetic processing unit 30 and the control processing unit 40 are wirelessly communicated.
[0082]
Further, as shown in FIG. 30, the arithmetic processing unit 30 may be configured using a microcomputer 60. The microcomputer 60 is a one-chip microcomputer or a one-board microcomputer developed for the system of the present embodiment, and includes an IC chip, a memory, and the like. The processing performed by the microcomputer 60 is the same as the processing performed by the arithmetic processing unit 30. By using such a microcomputer 60 as the arithmetic processing unit 30, the size of the arithmetic processing unit can be reduced, and it is easy to incorporate the arithmetic processing unit into a railway vehicle.
[0083]
In addition, as shown in FIG. 31, a transceiver 70 for communication is provided in the microcomputer 60 or incorporated in the microcomputer 60, and communicates with the control processing unit 40 arranged on the ground, not on the vehicle, via radio. May be performed. For example, the signal transmitting / receiving device 80 on the ground side is installed at a track, at a station, at a railroad crossing, or the like, and when a railroad vehicle passes, signals are transmitted and received between the transmitter / receiver 70 and the signal transmitting / receiving device 80, and a telephone line, The transmission and reception may be performed between the signal transmitting / receiving device 80 and the control processing unit 40 using an existing communication infrastructure such as wired using wireless or wireless using wireless. In this case, the control processing unit 40 is arranged at an information processing center installed on the ground.
[0084]
According to the above configuration, the central processing unit that simultaneously observes the states of a plurality of vehicles incorporating the abnormality diagnosis device can be performed by the control processing unit disposed on the ground. Therefore, by arranging the control processing unit in a central office such as an information center, it is possible to efficiently perform abnormality diagnosis in a normal business.
[0085]
【The invention's effect】
As described above, the abnormality diagnosis device of the present invention analyzes the output of the sensor of the bearing with the sensor incorporated in the bearing device of the railway vehicle by the arithmetic processing unit as the information processing device, and compares the output with the reference data to thereby make the sliding diagnosis possible. In order to determine the presence or absence of an abnormality due to wear or breakage of a component of a moving member, it is possible to determine an abnormality in a normal use state without disassembling the railway vehicle and its bearing device.
Therefore, it is possible to reduce the frequency of laborious disassembly / assembly work and reduce maintenance / management costs.
In addition, since the judgment is made mechanically by analysis and comparison by prescribed arithmetic processing, there is no possibility that the judgment will vary due to the skill of the inspector and individual differences, and the diagnosis of the presence or absence of abnormality is not compared with the conventional visual inspection. Reliability can be improved.
Further, since the multi-directional simultaneous vibration detecting element capable of simultaneously detecting the multi-directional vibrations is used, it is easy to attach the sensor to the bearing, and the vibration caused by the abnormality can be detected with high sensitivity.
[Brief description of the drawings]
FIG. 1 is a diagram showing an embodiment of an abnormality diagnosis device for a railway vehicle according to the present invention.
FIG. 2 is a block diagram showing an internal structure of a sensor.
FIG. 3 is a diagram illustrating a data accumulation / distribution unit.
FIG. 4 is a relational expression showing a relationship between a defect of each member of the bearing and an abnormal vibration frequency generated in each member.
FIG. 5 is a diagram showing a relationship between a load zone and a non-load zone in a bearing.
FIG. 6 is a flowchart showing a flow of processing in the first method.
FIG. 7 is a graph showing a frequency spectrum when no abnormality occurs.
FIG. 8 is a graph showing a frequency spectrum when an abnormality occurs in the outer ring.
FIG. 9
It is a graph which shows the relationship between a frequency spectrum and a reference value when a retainer has a flaw.
FIG. 10
It is a flowchart which shows the flow of a process in the 2nd method.
FIG. 11
It is a figure showing a frequency spectrum.
FIG.
It is a flowchart which shows the flow of a process in the 3rd method.
FIG. 13
It is a figure showing a frequency spectrum when there is a flaw in an outer ring.
FIG. 14
It is a figure showing a frequency spectrum after envelope processing.
FIG.
It shows the relationship between the magnitude of peeling and the level difference between a peak appearing in measured frequency spectrum data and a reference level.
FIG.
It is a flowchart which shows the flow of a process in the 5th method.
FIG.
FIG. 4 is a diagram illustrating a relationship between a level of a frequency spectrum and a reference line.
FIG.
It is a flowchart which shows the flow of a process in the 6th method.
FIG.
It is a graph which shows the frequency spectrum at the time of abnormality in an outer ring.
FIG.
It is a graph which shows the frequency spectrum when there is no abnormality in an outer ring.
FIG. 21
It is a graph which shows the frequency spectrum at the time of abnormality in an outer ring.
FIG.
It is a graph which shows the frequency spectrum when there is no abnormality in an outer ring.
FIG. 23 is a graph showing a time waveform of one component of a vibration signal detected by the multidirectional simultaneous detection element of the present embodiment.
FIG. 24 is a graph showing a frequency spectrum of the time waveform of FIG. 23;
FIG. 25 is a graph showing a frequency spectrum after the envelope processing of the time waveform of FIG. 23;
FIG. 26 is a graph showing a time waveform of a vibration signal detected by a vibration detection element that can detect only one direction.
FIG. 27 is a graph showing a frequency spectrum of the time waveform of FIG. 26;
FIG. 28 is a graph showing a frequency spectrum after the envelope processing of the time waveform of FIG. 26;
FIG. 29 is a diagram showing a modification of the present embodiment.
FIG. 30 is a diagram showing another modification of the present embodiment.
FIG. 31 is a diagram illustrating an application example of the modification example of FIG. 30;
[Explanation of symbols]
1, 2 abnormality diagnosis device 20, 51 detection processing unit 21 rolling bearing 22, 52 sensor 23 outer ring 24 inner ring 25 rolling element 30 arithmetic processing unit 31 data storage and distribution unit 32 temperature analysis unit 33 rotation analysis unit 34 filter processing unit 35 vibration analysis Unit 36 comparison and determination unit 37 internal data storage unit 38 data accumulation and output unit 40 control processing unit 41 control unit 42 result output unit 50 amplifier 60 microcomputer (arithmetic processing unit)
70 transceiver 80 signal transceiver 100 input unit

Claims (4)

An abnormality diagnosis device for an axle bearing device of a railway vehicle for diagnosing presence or absence of an abnormality of an axle bearing device for a railway vehicle,
A detection processing unit that is incorporated in the bearing device and outputs a signal generated from the bearing device as an electric signal.
An arithmetic processing unit that performs arithmetic processing to determine whether there is an abnormality in the mechanical equipment based on the output of the detection processing unit,
A result output unit that outputs a determination result by the arithmetic processing unit,
The abnormality diagnosis device according to claim 1, wherein the detection processing unit includes a vibration detector that detects vibration of the bearing device in at least two or more directions. The abnormality diagnosis device according to claim 1, wherein the detection processing unit is incorporated in a bearing of the bearing device. The abnormality diagnosis device according to claim 1, wherein the sensor-equipped bearing outputs the determination result to the result output unit via wireless communication. The abnormality diagnosis device according to claim 1, wherein the arithmetic processing unit includes a microcomputer.

Images