JP2004133553A - Diagnostic device for equipment - Google Patents

Diagnostic device for equipment Download PDF

Info

Publication number
JP2004133553A
JP2004133553A JP2002295412A JP2002295412A JP2004133553A JP 2004133553 A JP2004133553 A JP 2004133553A JP 2002295412 A JP2002295412 A JP 2002295412A JP 2002295412 A JP2002295412 A JP 2002295412A JP 2004133553 A JP2004133553 A JP 2004133553A
Authority
JP
Japan
Prior art keywords
failure
equipment
distribution
data
type
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
JP2002295412A
Other languages
Japanese (ja)
Inventor
Kenzo Yonezawa
米沢 憲造
Tomio Yamada
山田 富美夫
Nobutaka Nishimura
西村 信孝
Yuichi Hanada
花田 雄一
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Toshiba Corp
Original Assignee
Toshiba Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Toshiba Corp filed Critical Toshiba Corp
Priority to JP2002295412A priority Critical patent/JP2004133553A/en
Publication of JP2004133553A publication Critical patent/JP2004133553A/en
Withdrawn legal-status Critical Current

Links

Images

Abstract

<P>PROBLEM TO BE SOLVED: To support an operator in operating equipment and an equipment manager in maintaining the equipment by predicting the failure of the equipment, and accurately and quickly tracking the cause of the failure to point out the location of the failure. <P>SOLUTION: This diagnostic device for equipment includes a history data analyzing means 43 which, should the equipment fail, analyzes the tendency of changes in history data leading to the failure, from measurement data retained in a process history database 42; and an equipment trouble progress monitoring means 44 for predicting any failure by monitoring changes in the history of the measurement data of equipment of the same kind from the analysis results. The device further includes a failure time distribution type specifying means 46 for specifying the type of distribution of each equipment failure data group of each building by analyzing a failure database 45; a distribution parameter estimating means 47 for determining the distribution parameter of each type of failure data group by combining the failure data of the equipment of the same kind from type-based failure databases 45N-45EX stored for each different type based on the identification result; and an equipment diagnosing means 48 for predicting any failure of the equipment in each building and detecting where trouble is progressing according to information coming from the monitoring means 44 and the estimating means 47. <P>COPYRIGHT: (C)2004,JPO

Description

【0001】
【発明の属する技術分野】
本発明は空調、エレベータ等の各種の設備と、これを監視制御する装置を備えた複数の建物を集中的に管理して、設備機器のフィールドデータを解析して、各ビルの設備機器の故障予知、異常進行個所発見等を行う設備診断装置に関するものである。
【0002】
【従来の技術】
ビルの設備にはいろいろなものがあるが、今日、暖房や冷房を行う空調システムは建物にとって欠かせないものになっている。空調システムはビルの階毎に数台必要で、大規模ビルでは台数が多くなり、さらにこの空調システムは複数の装置から成り立っているため、しばしば不具合が生じる。不具合が生じると空調システムが正常な動作をしなくなるので、居住者に快適な空調環境を提供できなくなる。さらに省エネ制御が導入されていても設備機器の不具合が発見されずそのまま運転していると、機器効率が大幅に低下しエネルギーが無駄に使われることになる。近年、建物不具合を迅速に見つけるために、多数の建物と通信回線で結んで集中的に監視するセンタが設けられるようになった。そのようなセンタにおいては、蓄積し解析するためのデータ量も膨大になり、人手の監視による不具合の検知には限界があり、解析処理の自動化と検知の支援が求められている。
【0003】
従来から設備の異常診断では、機器の劣化状況を定量的に把握するため、フィールドデータ統計的解析手法である、信頼性工学が重要である。
【0004】
信頼性工学でよく用いられる諸式を以下に示す。
【0005】
【数1】

Figure 2004133553
建築設備のシステムや機器を解析するうえで特に重要な概念であり、故障解析によく使われるのが上記指数分布とワイブル分布である。
【0006】
指数分布は、故障が偶発的に発生する場合を表現するのに適している。したがって、システムのように多くの部品で構成されているものほど指数分布で表現される。指数分布の特徴は、故障率λ(t)が時間の経過にかかわらず常に一定であるという点である。
【0007】
一方、ワイブル分布は、W.Weibullによって提案された分布で、指数分布を拡張したものと考えられる。ワイブル分布はきわめて融通性があり、故障データの解析においてもっとも重要な働きをする。ワイブル分布は尺度パラメータη、形状パラメータβをもち、とくに形状パラメータβによって指数分布(β=1.0)から正規分布(β≒3.2)まで表現可能であるという点がすぐれ、これが融通性を保有する理由になっている。
【0008】
β<1 故障率減少形:初期故障期で、事後保全が有利である。
【0009】
β=1 故障率一定形:偶発故障期で、事後保全が有利である。
【0010】
β>1 故障率増加形:摩耗故障期で、予防保全が有利である。
【0011】
また、上記3つの故障率パターンと、バスタブ曲線による概念図を図20に、さらにワイブル分布の形状を図21に示す。
【0012】
バスタブ曲線と各分布との関係では、偶発故障期で指数分布が、摩耗故障期では正規分布が用いられる。さらにワイブル分布では、上記のように推定される形状パラメータβが1より小さいか、1に等しいか、あるいは1よりも大きいかによって、それぞれ初期故障期、偶発故障期、摩耗故障期に分けることができる。
【0013】
【発明が解決しようとする課題】
しかし、設備異常診断のために信頼性工学の精度の良い、上述したような故障解析を行うためにはフィールドの故障データが多数必要であるが、個々のビル、特に新しいビルにおいてはそのようなデータが少ないのが現状である。
【0014】
機器あるいは制御システムに何らかの異常が発生すると、連続計測している設備の運転状態値は正常時の値から外れた値が計測される。すなわち突発的に起こる異常の発生は、計測値が域値を越える回数で診断することができる。しかし域値の幅をどのように設定するかは、診断予知の精度を決める重要な要因である。従来は、機器設計値等を参考に初期設定して、運転実績に基づいて試行錯誤で調整して行く方式であったため、実際に故障するよりだいぶ前に重故障警報を出すか、あるいは故障を予知できない場合が多々あった。
【0015】
本発明は上記事情に鑑みてなされたもので、複数の建物の設備を集中的に常時監視を行い、リアルタイムで各ビルの設備機器が不具合に至るまでの過程で発生する異常を検知して故障を予知し、また異常が進行している個所を指摘して、オペレータの運転やビル管理者の設備保全を支援する設備診断装置を提供することを目的としている。
【0016】
【課題を解決するための手段】
上記の目的を達成するために本発明は、監視対象側から入力されたプロセス値データや機器故障データを定期的にかつ、各監視対象別に分類・編集してプロセス履歴データベース及び故障データベースに蓄積するデータ収集・分類・記憶手段と、監視対象となる機器が故障した時、前記プロセス履歴データベースに蓄積されたその機器に関連する計測データに基づき、故障に至るまでの履歴データ変化の傾向を解析する履歴データ解析手段と、この履歴データ解析手段の解析結果を用いて各監視対象における同種の機器の計測データの履歴変化を監視することにより、当該監視対象となる機器の故障を予測する機器異常進行監視手段と、前記故障データベースを解析することにより、各監視対象における各機器故障データ集団の分布タイプを同定する故障時間分布タイプ同定手段と、この故障時間分布タイプ同定手段の同定結果に基づいて、異なる分布タイプ毎に分けて蓄えられたタイプ毎故障データベースから、同種の機器の故障データを組み合わせて故障データ集団の各タイプの分布パラメータを求める分布パラメータ推定手段と、前記機器異常進行監視手段と分布パラメータ推定手段からの情報に基づいて、各監視対象における設備機器の故障予知、異常進行個所発見を行う設備診断手段とを具備することを特徴としている。
【0017】
本発明は以上のような手段を講じたことにより、例えば、複数の建物の設備を集中的に常時監視を行い、各ビルの各設備機器が故障に至るまでの過程で発生する異常を予知したり、異常が発生した場合、正確で迅速な故障原因追跡を行い原因個所を指摘し、オペレータの運転やビル管理者の設備保全を支援することができる。
【0018】
【発明の実施の形態】
以下、図面を参照して本発明の実施の形態について説明する。図1は本発明によるの設備診断装置の一実施形態である建物設備診断装置を示す全体構成図である。
【0019】
多数のビル(図では5つのみ表示)を通信回線を結んで集中的に監視・診断するセンタとして(遠隔)建物設備診断装置1が設けられている。
【0020】
各ビルに設けられた制御システムが設備の省エネルギー制御を含む各種制御を行っており、かつ監視・診断対象となっている各建物、例えば事務所ビルA1やA2、デパートビルB1やB2、病院ビルC1等にはそれぞれ計測監視制御手段21〜25が設けられている。計測監視制御手段21〜25は、それぞれの建物に設置されている空調装置を含む各種設備機器関連のプロセス値を検出し、その検出データ及び、各機器の故障データをネットワーク30を介して、本発明に係る建物設備診断装置1に送出する。ここでプロセス値には、例えば外気温や、室温、室内湿度等も含まれる。各建物の図示していない制御システムは空調制御やエレベータ制御など、各設備の制御を行っている。
【0021】
建物設備診断装置1は、プロセス履歴データベース42と故障データベース(故障に関するデータベース)45を有するデータ収集・分類・記憶手段41、履歴データ解析手段43、機器異常進行監視手段44、故障時間分布(タイプ)同定手段46、分布パラメータ同定手段47および設備診断手段48を備えるとともに、マンマシンインターフェースのための端末50を備えている。
【0022】
データ収集・分類・記憶手段41は、各建物の計測監視制御手段21〜25から定期的に、例えば1回/日にその日のプロセスデータをまとめて収集し、各建物及び機器別毎に編集処理を行って光磁気記録媒体などの記憶装置にプロセス履歴データベース42として記憶する。また、各建物で機器故障が発生した場合、その機器が故障するまでの時間等の機器故障データを計測監視制御手段21〜25から随時送信してもらい、各建物及び機器別毎に編集処理を行って光磁気記録媒体などの記憶装置に故障データベース45として記憶する。
【0023】
次に履歴データ解析手段43について詳述する。
【0024】
履歴データ解析手段43は、あるビルの機器が故障した時、その機器に関連する計測データの故障に至るまでの履歴データの変化を、度数分布にして、正常時の値から外れるまでの傾向を解析する。その解析結果から、上記の域値、異常と判定するための域値を計測値が越える回数等のパラメータを機器異常進行監視手段44に渡す。
【0025】
具体的な例を用いて以下に説明する。
【0026】
例えば、あるビルの空調用ターボ冷凍機に不具合が発生したことを想定すると、履歴データ解析手段43は、不具合の発生時点で、プロセス履歴データベース42の中から、この機器関連の計測項目(電力量・冷水入口及び出口温度・蒸発器圧力及び温度・凝縮器圧力及び温度・その他)データの過去1ヶ月分以上の空調機運転中のデータを抽出する。なお、このプロセス履歴データベース43には計測監視制御手段21〜25が1分間隔で連続に計測したデータを設備診断用として、データ収集・分類・記憶手段41が30分間隔で間引きして収集したものが記憶されている。
【0027】
図2は、上記で抽出したデータを用いて、ターボ冷凍機の不具合発生から過去1ヶ月分の蒸発器圧力を度数分布グラフにしたものである。蒸発器圧力の設定値は−0.525(kg/cm)であり、−0.650から−0.400(kg/cm)が正常範囲である。図3は不具合発生から2日分前のデータを削除したものである。図4は不具合発生から3日分前のデータを削除したものである。さらに過去のデータを解析した結果、月別頻度分布では−0.375(kg/cm)の級のところに15回以下で現れることは何回かあったが、−0.325(kg/cm)の級のところは1回も出現していないことが判明した。この結果から−0.375(kg/cm)の級のところの頻度が15回以上になると、数日後に故障すると予想され、−0.325(kg/cm)の級のところに数回でも現れると、その日の内に故障することが予想される。
【0028】
この解析結果を機器異常進行監視手段44に渡す。機器異常進行監視手段44では、同じビルあるいは他のビルの、同種でほぼ同仕様機器の同じ計測データから履歴データ解析手段43で作られた度数分布を上記のパラメータを用いて監視することにより、異常が進行している場合、事前に故障を予知する。上記の具体的な例では、空調用ターボ冷凍機の過去1ヶ月の蒸発器圧力計測値の度数分布が、圧力設定値より+0.125(kg/cm)から+0.175(kg/cm)の級に15回以上現れたら、数日後にターボ冷凍機系統に不具合が生じる故障警報を設備診断手段48に送る。
【0029】
故障時間分布(タイプ)同定手段46では、各ビルの各機器の故障データ集団がどの分布に従うかを同定する。実施形態では確率紙法を用いてワイブル分布、正規分布、対数正規分布、指数分布の4つの中から、どの分布に一番良く従うかを判定し、故障データベース(各ビルの各機器別)45を図1に示すように上記4つの分布ごとの故障データベース45W,45N,45LN,45EXの4つに細分する。図中ではWはワイブル分布、Nは正規分布、LNは対数正規分布、EXは指数分布を示す。
【0030】
図5〜図8は、表1に示すAビルの機器Xの故障データ(データ個数26個)を分布タイプ同定手段46により上記4つの分布確率紙にプロットした結果である。図5は正規分布、図6は指数分布、図7は対数正規分布、図8はワイブル分布を示す。ここでは従来技術で述べたように、きわめて融通性があり、故障データの解析において最も重要な働きをするワイブル分布の確率紙について説明するが、他の確率紙も同様である。
【0031】
ワイブル確率紙は故障データがワイブル分布に従う場合は、データをこの上に打点した時に直線になることを利用して作られた確率紙である。左側の縦軸が不信頼度F(t)=1−R(t)の%目盛り、下側の横軸に時間が目盛られている。具体的には、y座標がln(−ln(1−F)で、x座標はln(時間)で目盛られている。したがって、故障時間tに対して不信頼度F(t)をペアにして打点する。なお、パラメータ推定は通常Fが30〜80%の点が直線にのることを重視して推定する。
【0032】
【表1】
Figure 2004133553
【表2】
Figure 2004133553
以下にAビルのX機器故障データ(表1)を用いて説明する。まず表1のデータを故障までの時間の小さい順に並べ直し、さらにi番目のデータt(故障までの時間)に対して(表2参照)、次式で計算される累積確率Fを対応させる。前者が横軸、後者が縦軸の値となる。
【0033】
【数2】
Figure 2004133553
ここで、nはデータ個数である。
【0034】
他の確率紙も、同様にデータをこの上にプロットした時に、その分布にデータが従うなら直線になるようにx座標、y座標が目盛られている。因みに指数分布確率紙はy座標が−ln(1−F)で、x座標は時間で目盛られている。
【0035】
図5〜図8を比較すると判るように、4つのグラフの中で図8に示すワイブル確率紙がいちばん良く直線にフィットしているので、このAビルの機器Xの故障データはワイブル分布に従うものとして同定する。
【0036】
次にデータ統合による分布パラメータ同定手段47について述べる。
【0037】
分布パラメータ同定手段47は、W分布パラメータ推定手段47Wと、N分布パラメータ推定手段47Nと、LN分布パラメータ推定手段47LNと、EX分布パラメータ推定手段47EXとを備え、個別の故障データベース45W,45N,45LN,45EXに保存されたデータ中のビル群の中から類似ビルの、同種機器の故障データ群をN組持ってきてデータを統合し(ワイブルの)確率紙の推定法により、精度の良い尺度パラメータηと形状パラメータβの推定を行う。
【0038】
まず、単独で尺度パラメータηと形状パラメータβの値を算出し、どちらの値も近い組(α%以内)のみデータを統合して、推定を行う。どの単独の誤差よりも統合したデータ群の誤差が小さい時精度の良いパラメータが推定できたとする。
【0039】
以下、具体的な実施形態を示す。故障データ群 4組(機器名 X:空調システム関連の機器)
Figure 2004133553
それぞれの解析結果をワイブル確率紙の上に示したのが図9〜図12である。プロットの傾きから尺度(Scale)パラメータηと形状(Shape)パラメータβの値を算出する。
【0040】
図9〜図12に示す4つのグラフを一つのグラフにプロットしたのが、図13である。また、パラメータ推定結果と誤差を表3に示した。
【0041】
【表3】
Figure 2004133553
この結果から、ビルDの機器Xの故障データの集団のみ、パラメータが異なるものとして除外する。残りのビルA,B,Cの同種の機器Xのデータ70個は、パラメータが同じ集団と見なして、データを統合して表1,表2と同様な手法でワイブル確率紙にプロット(この場合nは70)し、尺度パラメータηと形状パラメータβを推定する。
【0042】
結果を図14と表4に示した。
【0043】
【表4】
Figure 2004133553
表4より、尺度パラメータ、形状パラメータどちらもビルA,B,Cの機器X単独に求めたパラメータ誤差(表3)よりも小さい。よって、表4のパラメータをビルA,B,Cの機器Xのパラメータとする。
【0044】
ビルA,B,Cの同種の機器Xのデータ70個の集団を確認のため、分布タイプ同定手段46に入力した結果が図15〜図18である。図15は正規分布、図16は指数分布、図17は対数正規分布、図18はワイブル分布をそれぞれ示す。形状パラメータβが1に近いので図15に示す指数分布確率紙が、図18に示すワイブル分布確率紙の次に直線にフィットしている。なお、この確認は省略しても良い。
【0045】
この実施形態では、簡単のためデータ群として4組だけを取り上げたので、ビルDの機器Xと同じパラメータ値を持つ集団はないが、データ群の組を増やせばDの機器Xと同じパラメータ値を持つ集団が見つかる確率は増す。
【0046】
上記のようにして推定されたパラメータは、設備診断手段48に渡される。
【0047】
設備診断手段48は、従来のエキスパートシステムや信頼性工学のFTA(フォールツリーアナリシス)手法等を用いて、分布パラメータ同定手段47と機器異常進行監視手段44からの情報を元に、各ビルの各設備機器が故障に至るまでの過程で発生する異常を予知したり、異常が発生している個所を指摘する。
【0048】
FTAは、トップ事象である故障原因に対して、測定値などの異状現象として現れる基本事象を複数あらかじめ列挙し、結果として現れた基本事象から故障原因個所を推論しようとする手法である。このFTAを応用すれば、設備診断において、結果として現れた基本事象に関連するトップ事象の中で故障確率(これは実データに基づいて実証的に導ける客観的確率である)の大きい順に原因追跡を行うことができる。
【0049】
故障原因となるトップ事象はビルの空調衛生設備だけでも下記のように数多くある。
【0050】
空調熱源機器:ボイラ、ターボ冷凍機、冷温水発生機、ヒートポンプチラーなど
空調熱源補機:冷却塔、循環ポンプ、オイルタンク、熱交換器など
空調2次側機器:空調機、パッケージ型空調機、ファンコイル、ヒートポンプエアコンなど
空調配管設備:冷水、温水、冷却水、蒸気、還水、ドレン配管など
換気設備:送風機、全熱交換器、厨房排気設備、天井扇、圧力扇など
風道設備:給気ダクト、排気ダクト、換気ダクト、ダンパー類など
衛生機器設備:受水槽、高架水槽、揚水ポンプ、加圧給水ポンプ、給湯設備、ボイラ
衛生配管設備:給水、給湯、雑排水、汚水排水、通気配管など
上記のトップ事象に対して、測定値などの異状現象として現れる基本事象は相互に絡みあっている。図19はトップ事象の1つであるターボ冷凍機を、さらにその構成要素(蒸発器、凝縮器、圧縮機、電動機等)に基づいてトップ事象を分解(図の右側)した、実際のFTA例の一部分を示したものである。測定値の状態として、図19の左にある冷水出口温度HOTの基本事象が生じた場合、ここには図示しない全体のFTAから、トップ事象である故障原因はターボ冷凍機、冷却塔、送水ポンプ、蓄熱槽(効率低下)、空調機内冷却コイル低下、その他が挙げられる。
【0051】
ここで、この原因探索を設備診断手段48が行う場合、分布パラメータ同定手段47で求めた、精度の良い機器の故障確率を用いて、その大きい順に原因追跡を行うことができる。例えばターボ冷凍機の故障確率が一番大きかった場合は、まず、この機器の構成要素から原因追及を行う。この時、機器異常進行監視手段44で作られている蒸発器、凝縮器、圧縮機、電動機等のプロセス計測値履歴データ度数分布の傾向に変化が現れているか否かの情報を元に、エキスパートシステムを用いることにより正確で迅速な故障原因追跡ができる。得られた結果は該当するビルの計測監視制御手段21〜25にネットワークを通じて送信される。
【0052】
以上説明したような構成とアルゴリズムにより、本発明の目的を達成することができる。
【0053】
[他の実施形態]
なお、本願発明は上記実施の形態に限定されるものではなく、その要旨を逸脱しない範囲で種々変形して実施できるものである。
【0054】
以上では、本発明の実施形態を建物群を例に取って説明したが、他のシステムでも同様に適用できることは自明である。
【0055】
故障解析のパラメータ推定法として確率紙法を用いたが、最尤法、ベイズ法などを用いても良い。
【0056】
また、本発明の実施形態の故障解析では、故障した機器のデータのみ用いる手法を用いたが、動作中の機器のデータも含めて信頼度解析を行うハザード解析の手法を同様に用いることもできる。
【0057】
建物設備診断手段の中で設備の故障個所と原因を自動的に推定できるようにするため、エキスパートシステムを用いたが、他のAI(人工知能Artificial Intelligence)技術を用いても良い。
【0058】
各実施形態は可能な限り組み合わせて実施することが可能であり、その場合には組合せによる効果が得られる。
【0059】
【発明の効果】
以上、説明したように本発明によれば、複数の建物の設備を集中的に常時監視を行い、各ビルの各設備機器が故障に至るまでの過程で発生する異常を予知したり、異常が発生した場合、正確で迅速な故障原因追跡を行い原因個所を指摘し、オペレータの運転やビル管理者の設備保全を支援することができる設備診断装置を提供することができる。
【図面の簡単な説明】
【図1】本発明による建物設備診断装置の一実施形態を示す全体構成図である。
【図2】ターボ冷凍機の蒸発器圧力値の度数分布例を示す説明図である。
【図3】ターボ冷凍機の蒸発器圧力値の度数分布例を示す説明図である。
【図4】ターボ冷凍機の蒸発器圧力値の度数分布例を示す説明図である。
【図5】機器故障データを4つのタイプの分布確率紙のうち正規分布確率紙にプロットした説明図である。
【図6】機器故障データを4つのタイプの分布確率紙のうち指数分布確率紙にプロットした説明図である。
【図7】機器故障データを4つのタイプの分布確率紙のうち対数正規分布確率紙にプロットした説明図である。
【図8】機器故障データを4つのタイプの分布確率紙のうちワイブル分布確率紙にプロットした説明図である。
【図9】4組の故障データをそれぞれワイブル確率紙にプロットした説明図である。
【図10】4組の故障データをそれぞれワイブル確率紙にプロットした説明図である。
【図11】4組の故障データをそれぞれワイブル確率紙にプロットした説明図である。
【図12】4組の故障データをそれぞれワイブル確率紙にプロットした説明図である。
【図13】上記4組の故障データを1つのワイブル確率紙にプロットした説明図である。
【図14】上記4組の故障データの内、3組のデータを同じ集団として統合しワイブル確率紙にプロットした説明図である。
【図15】上記3組の統合した故障データを4つのタイプの分布確率紙のうち正規分布確率紙にプロットした説明図である。
【図16】上記3組の統合した故障データを4つのタイプの分布確率紙のうち指数分布確率紙にプロットした説明図である。
【図17】上記3組の統合した故障データを4つのタイプの分布確率紙のうち対数正規分布確率紙にプロットした説明図である。
【図18】上記3組の統合した故障データを4つのタイプの分布確率紙のうちワイブル分布確率紙にプロットした説明図である。
【図19】ターボ冷凍機のFTA例の一部分を示した説明図である。
【図20】バスタブ曲線による故障率パターンの概念説明図である。
【図21】形状パラメータβの値を変化させた場合のワイブル分布の形状を示す説明図である。
【符号の説明】
1 建物設備診断装置
21〜25 計測監視制御手段
30 ネットワーク
41 データ収集・分類・記憶手段
42 プロセス履歴データベース
43 履歴データ解析手段
44 機器異常進行監視手段
45 故障データベース
46 故障時間分布同定手段
47 分布パラメータ同定手段
48 設備診断手段
50 端末[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention centrally manages various facilities such as air conditioners and elevators, and a plurality of buildings equipped with devices for monitoring and controlling the same, analyzes the field data of the facilities, and detects the failure of the facilities in each building. The present invention relates to a facility diagnostic device for performing prediction, finding abnormally advanced points, and the like.
[0002]
[Prior art]
Although there are various types of building equipment, air conditioning systems for heating and cooling are indispensable for buildings today. Several air-conditioning systems are required for each floor of a building, and the number of air-conditioning systems is large in a large-scale building. Further, since the air-conditioning system is composed of a plurality of devices, problems often occur. If a malfunction occurs, the air conditioning system does not operate normally, and it is not possible to provide a comfortable air conditioning environment to the occupants. Further, even if energy saving control is introduced, if the equipment is operated as it is without detecting any malfunction of the equipment, the efficiency of the equipment is greatly reduced and energy is wasted. 2. Description of the Related Art In recent years, in order to quickly find a building defect, a center has been provided which is connected to a number of buildings via a communication line and performs intensive monitoring. In such a center, the amount of data to be accumulated and analyzed is enormous, and there is a limit in detecting a defect by manual monitoring, and automation of analysis processing and support for detection are required.
[0003]
Conventionally, in failure diagnosis of equipment, reliability engineering, which is a statistical analysis method of field data, is important in order to quantitatively grasp the degradation state of equipment.
[0004]
The formulas often used in reliability engineering are shown below.
[0005]
(Equation 1)
Figure 2004133553
The exponential distribution and the Weibull distribution are particularly important concepts for analyzing systems and equipment of building equipment, and are often used for failure analysis.
[0006]
The exponential distribution is suitable for expressing a case where a failure occurs accidentally. Therefore, a system composed of many components such as a system is represented by an exponential distribution. The characteristic of the exponential distribution is that the failure rate λ (t) is always constant over time.
[0007]
On the other hand, the Weibull distribution is based on W.I. It is considered to be an extension of the exponential distribution in the distribution proposed by Weibull. The Weibull distribution is very versatile and plays the most important role in analyzing failure data. The Weibull distribution has a scale parameter η and a shape parameter β, and it is excellent in that it can be expressed from an exponential distribution (β = 1.0) to a normal distribution (β ≒ 3.2) by the shape parameter β. Is the reason for holding.
[0008]
β <1 Failure rate reduction type: Early failure period, post-maintenance is advantageous.
[0009]
β = 1 Constant failure rate: In the event of accidental failure, post maintenance is advantageous.
[0010]
β> 1 Failure rate increase type: Preventive maintenance is advantageous during the wear failure period.
[0011]
FIG. 20 shows a conceptual diagram based on the three failure rate patterns and the bathtub curve, and FIG. 21 shows a Weibull distribution shape.
[0012]
In the relationship between the bathtub curve and each distribution, an exponential distribution is used in a random failure period, and a normal distribution is used in a wear failure period. Further, in the Weibull distribution, the shape parameter β estimated as described above can be divided into an initial failure period, a random failure period, and a wear failure period depending on whether the shape parameter β is smaller than 1, equal to 1, or larger than 1. it can.
[0013]
[Problems to be solved by the invention]
However, in order to perform the failure analysis as described above with high accuracy of reliability engineering for equipment abnormality diagnosis, a large number of field failure data is required. At present, there is little data.
[0014]
When any abnormality occurs in the device or the control system, the operation state value of the facility that is continuously measured deviates from the normal value. That is, the occurrence of a sudden abnormality can be diagnosed by the number of times the measured value exceeds the threshold value. However, how to set the range of the threshold value is an important factor that determines the accuracy of diagnosis prediction. Conventionally, initial setting was made with reference to equipment design values, etc., and adjustment was performed by trial and error based on operation results. In many cases, it was impossible to predict.
[0015]
The present invention has been made in view of the above circumstances, and constantly monitors the facilities of a plurality of buildings intensively, detects abnormalities that occur in the process until the equipment of each building reaches a failure in real time, and breaks down. It is an object of the present invention to provide an equipment diagnostic device that supports the operation of an operator and the maintenance of equipment by a building manager by predicting a location where an abnormality is progressing.
[0016]
[Means for Solving the Problems]
In order to achieve the above object, according to the present invention, process value data and equipment failure data input from a monitoring target are periodically and classified and edited for each monitoring target and stored in a process history database and a failure database. Data collection / classification / storage means, and when a device to be monitored fails, analyze the tendency of change in history data until the failure based on measurement data related to the device accumulated in the process history database. A device abnormality progress that predicts a failure of a device to be monitored by monitoring a history change of measurement data of the same type of device in each monitoring target by using a history data analysis unit and an analysis result of the history data analysis unit. By monitoring the failure database, the distribution type of each device failure data group in each monitoring target is analyzed. Failure time distribution type identification means to be determined, and failure data of the same type of equipment combined from failure data for each type stored separately for each different distribution type based on the identification result of the failure time distribution type identification means. Based on information from the distribution parameter estimating means for obtaining each type of distribution parameter of the data group, and the information from the equipment abnormality progress monitoring means and the distribution parameter estimating means, the failure prediction and the abnormal progress location of the equipment in each monitoring target are performed. And a facility diagnosis unit.
[0017]
By taking the above measures, the present invention, for example, constantly monitors the facilities of a plurality of buildings intensively and predicts abnormalities that occur in the course of failure of each facility equipment of each building. In the event of an abnormality or an abnormality, the cause of the failure can be accurately and promptly traced to point out the cause of the failure, and the operation of the operator and the facility maintenance of the building manager can be supported.
[0018]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is an overall configuration diagram showing a building equipment diagnosis apparatus which is an embodiment of the equipment diagnosis apparatus according to the present invention.
[0019]
A (remote) building equipment diagnostic device 1 is provided as a center for centrally monitoring and diagnosing a number of buildings (only five shown in the figure) by connecting communication lines.
[0020]
A control system provided in each building performs various controls including energy saving control of equipment, and is a building to be monitored and diagnosed, for example, office buildings A1 and A2, department stores B1 and B2, and hospital buildings. Measurement monitoring control means 21 to 25 are respectively provided in C1 and the like. The measurement / monitoring / controlling means 21 to 25 detect process values related to various types of equipment including the air conditioners installed in each building, and transmit the detected data and the failure data of each device to the The information is sent to the building equipment diagnosis apparatus 1 according to the invention. Here, the process values include, for example, the outside air temperature, the room temperature, the indoor humidity, and the like. A control system (not shown) of each building controls each facility such as air conditioning control and elevator control.
[0021]
The building equipment diagnosis apparatus 1 includes a data collection / classification / storage unit 41 having a process history database 42 and a failure database (database relating to failure) 45, a history data analysis unit 43, a device abnormality progress monitoring unit 44, and a failure time distribution (type). An identification unit 46, a distribution parameter identification unit 47 and a facility diagnosis unit 48 are provided, and a terminal 50 for a man-machine interface is provided.
[0022]
The data collection / classification / storage means 41 collects the process data of the building from the measurement monitoring control means 21 to 25 of each building periodically, for example, once / day, and edits the processing data for each building and each device. And store it as a process history database 42 in a storage device such as a magneto-optical recording medium. In addition, when a device failure occurs in each building, device failure data such as the time until the device fails is transmitted from the measurement monitoring control means 21 to 25 at any time, and editing processing is performed for each building and each device. Then, it is stored as a failure database 45 in a storage device such as a magneto-optical recording medium.
[0023]
Next, the history data analysis means 43 will be described in detail.
[0024]
When a device in a building fails, the history data analysis unit 43 converts the change in the history data up to the failure of the measurement data related to the device into a frequency distribution, and shows the tendency until the value deviates from a normal value. To analyze. From the analysis result, parameters such as the above-mentioned threshold value and the number of times the measured value exceeds the threshold value for determining an abnormality are passed to the device abnormality progress monitoring means 44.
[0025]
This will be described below using a specific example.
[0026]
For example, assuming that a malfunction has occurred in the air-conditioning centrifugal chiller of a certain building, the history data analysis unit 43, at the time of occurrence of the malfunction, searches the process history database 42 for this device-related measurement item (electric power amount).・ Chilled water inlet and outlet temperatures ・ Evaporator pressure and temperature ・ Condenser pressure and temperature ・ Others) Extract data during the air conditioner operation for more than one month in the past. In the process history database 43, the data collected, classified, and stored by the data collection / classification / storage means 41 is thinned out at 30-minute intervals for equipment diagnosis using data continuously measured at one-minute intervals by the measurement monitoring control means 21 to 25. Things are remembered.
[0027]
FIG. 2 is a graph showing a frequency distribution graph of evaporator pressures for the past month from the occurrence of a malfunction of the centrifugal chiller using the data extracted above. Set value of the evaporator pressure is -0.525 (kg / cm 2), from -0.650 -0.400 (kg / cm 2) it is in the normal range. FIG. 3 shows the data deleted two days before the occurrence of the malfunction. FIG. 4 shows the data deleted three days before the occurrence of the malfunction. Furthermore, as a result of analyzing the past data, it was found that in the monthly frequency distribution, the frequency was less than 15 times at the level of −0.375 (kg / cm 2 ), but it was −0.325 (kg / cm 2 ). It was found that the class of 2 ) did not appear even once. From this result, if the frequency at the class of −0.375 (kg / cm 2 ) becomes 15 times or more, it is expected that a failure will occur after several days, and the frequency at the class of −0.325 (kg / cm 2 ) If it appears several times, it is expected that it will break down within the day.
[0028]
This analysis result is passed to the device abnormality progress monitoring means 44. The device abnormality progress monitoring unit 44 monitors the frequency distribution created by the history data analysis unit 43 from the same measurement data of the same type and substantially the same specification of the same building or another building by using the above-mentioned parameters, If an abnormality is in progress, a failure is predicted in advance. In the above specific example, the frequency distribution of the evaporator pressure measurement values of the air conditioner centrifugal chiller for the past month is from +0.125 (kg / cm 2 ) to +0.175 (kg / cm 2 ) from the pressure set value. If it appears 15 times or more in the class, a failure alarm that causes a failure in the centrifugal chiller system is sent to the equipment diagnosis means 48 several days later.
[0029]
The failure time distribution (type) identification means 46 identifies which distribution the failure data group of each device in each building follows. In this embodiment, the probability paper method is used to determine which of the four distributions, Weibull distribution, normal distribution, lognormal distribution, and exponential distribution, is best followed, and a failure database (for each device in each building) 45 Is subdivided into four fault databases 45W, 45N, 45LN, and 45EX for each of the four distributions as shown in FIG. In the figure, W indicates Weibull distribution, N indicates normal distribution, LN indicates lognormal distribution, and EX indicates exponential distribution.
[0030]
FIGS. 5 to 8 show the results of plotting the failure data (26 pieces of data) of the device X of the building A shown in Table 1 on the four distribution probability sheets by the distribution type identification means 46. FIG. 5 shows a normal distribution, FIG. 6 shows an exponential distribution, FIG. 7 shows a lognormal distribution, and FIG. 8 shows a Weibull distribution. Here, as described in the related art, the probability paper of the Weibull distribution, which is extremely flexible and plays the most important role in analyzing failure data, will be described, but the same applies to other probability papers.
[0031]
The Weibull probability paper is a probability paper created by utilizing the fact that when the failure data follows the Weibull distribution, the data becomes a straight line when the data is hit on it. The vertical axis on the left side is a scale of% of unreliability F (t) = 1−R (t), and the horizontal axis on the lower side is time. Specifically, the y coordinate is scaled by ln (-ln (1-F), and the x coordinate is scaled by ln (time). Therefore, the unreliability F (t) is paired with the failure time t. Note that parameter estimation is usually performed with emphasis on the fact that a point where F is 30 to 80% is on a straight line.
[0032]
[Table 1]
Figure 2004133553
[Table 2]
Figure 2004133553
This will be described below using the X device failure data of Building A (Table 1). First, the data in Table 1 are rearranged in ascending order of the time to failure, and the cumulative probability F i calculated by the following equation corresponds to the i-th data t i (time to failure) (see Table 2). Let it. The former is the value on the horizontal axis, and the latter is the value on the vertical axis.
[0033]
(Equation 2)
Figure 2004133553
Here, n is the number of data.
[0034]
In other probability papers, similarly, when data is plotted thereon, the x-coordinate and the y-coordinate are graduated so as to form a straight line if the data follows the distribution. Incidentally, in the exponential distribution probability paper, the y coordinate is -ln (1-F), and the x coordinate is graduated with time.
[0035]
As can be seen by comparing FIGS. 5 to 8, the Weibull probability paper shown in FIG. 8 among the four graphs fits the straight line best, so the failure data of the device X in Building A follows the Weibull distribution. Identified as
[0036]
Next, the distribution parameter identification means 47 by data integration will be described.
[0037]
The distribution parameter identification unit 47 includes a W distribution parameter estimation unit 47W, an N distribution parameter estimation unit 47N, an LN distribution parameter estimation unit 47LN, and an EX distribution parameter estimation unit 47EX, and individual failure databases 45W, 45N, and 45LN. , 45EX, N sets of failure data groups of similar equipment of similar buildings from among the building groups in the data stored in 45EX are integrated, the data is integrated, and the (Weibull) probability paper estimation method is used to obtain accurate scale parameters. η and the shape parameter β are estimated.
[0038]
First, the values of the scale parameter η and the shape parameter β are calculated independently, and data is integrated only for a set (within α%) where both values are close to each other for estimation. It is assumed that a highly accurate parameter can be estimated when the error of the integrated data group is smaller than any single error.
[0039]
Hereinafter, specific embodiments will be described. Failure data group 4 sets (Equipment name X: Air conditioning system related equipment)
Figure 2004133553
9 to 12 show the respective analysis results on the Weibull probability paper. A value of a scale parameter η and a value of a shape parameter β are calculated from the slope of the plot.
[0040]
FIG. 13 is a plot of the four graphs shown in FIGS. 9 to 12 in one graph. Table 3 shows parameter estimation results and errors.
[0041]
[Table 3]
Figure 2004133553
From this result, only the group of the failure data of the device X of the building D is excluded as having different parameters. The remaining 70 data of devices X of the same type in buildings A, B and C are regarded as a group having the same parameters, and the data are integrated and plotted on Weibull probability paper in the same manner as in Tables 1 and 2 (in this case, n is 70), and the scale parameter η and the shape parameter β are estimated.
[0042]
The results are shown in FIG.
[0043]
[Table 4]
Figure 2004133553
From Table 4, both the scale parameter and the shape parameter are smaller than the parameter error (Table 3) obtained for the device X alone of the buildings A, B, and C. Therefore, the parameters in Table 4 are set as the parameters of the device X of the buildings A, B, and C.
[0044]
FIGS. 15 to 18 show the results of input to the distribution type identification means 46 to confirm a group of 70 data of the same type of equipment X of buildings A, B, and C. 15 shows a normal distribution, FIG. 16 shows an exponential distribution, FIG. 17 shows a lognormal distribution, and FIG. 18 shows a Weibull distribution. Since the shape parameter β is close to 1, the exponential distribution probability paper shown in FIG. 15 fits a straight line next to the Weibull distribution probability paper shown in FIG. This confirmation may be omitted.
[0045]
In this embodiment, only four sets are taken as a data group for simplicity. Therefore, there is no group having the same parameter value as the device X of the building D. However, if the number of data groups is increased, the same parameter value as that of the device X of the D will be obtained. The probability of finding a group with is increased.
[0046]
The parameters estimated as described above are passed to the equipment diagnosis means 48.
[0047]
The equipment diagnosis means 48 uses the conventional expert system or the FTA (Fall Tree Analysis) method of reliability engineering, etc., based on the information from the distribution parameter identification means 47 and the equipment abnormality progress monitoring means 44, for each building. Predict abnormalities that occur in the process of equipment equipment failure, and point out where abnormalities occur.
[0048]
The FTA is a method in which a plurality of basic events appearing as abnormal phenomena such as measured values are listed in advance with respect to a failure cause which is a top event, and an attempt is made to infer the cause of the failure from the basic events that appear as a result. If this FTA is applied, in equipment diagnosis, cause tracing is performed in the order of ascending failure probability (this is an objective probability that can be empirically derived based on actual data) among top events related to a basic event that has appeared as a result. It can be performed.
[0049]
There are a number of top events that can cause failures, even in air-conditioning and sanitation facilities in buildings, as described below.
[0050]
Air conditioning heat source equipment: Boiler, turbo refrigerator, cold / hot water generator, heat pump chiller, etc. Air conditioning heat source auxiliary equipment: Cooling tower, circulation pump, oil tank, heat exchanger, etc. Secondary air conditioning equipment: Air conditioner, packaged air conditioner, Air conditioning piping equipment such as fan coil and heat pump air conditioner: Ventilation equipment such as cold water, hot water, cooling water, steam, return water, drain piping: Blower, total heat exchanger, kitchen exhaust equipment, ceiling fan, pressure fan, etc. Sanitary equipment such as air ducts, exhaust ducts, ventilation ducts, dampers: receiving tanks, elevated water tanks, pumps, pressurized water supply pumps, hot water supply equipment, boiler sanitary plumbing equipment: water supply, hot water supply, miscellaneous drainage, sewage drainage, ventilation piping Basic events that appear as abnormal phenomena, such as measured values, are intertwined with each other for the above top events. FIG. 19 shows an example of an actual FTA in which a centrifugal chiller, which is one of the top events, is further decomposed (on the right side of the figure) based on its components (evaporator, condenser, compressor, electric motor, etc.). It shows a part of. When the basic event of the chilled water outlet temperature HOT on the left side of FIG. 19 occurs as the state of the measured values, the cause of the failure, which is the top event, is the turbo chiller, the cooling tower, the water pump , A heat storage tank (decrease in efficiency), a cooling coil in an air conditioner, and others.
[0051]
Here, when the equipment diagnosis unit 48 performs the cause search, the cause can be tracked in the descending order by using the failure probability of the device with high accuracy obtained by the distribution parameter identification unit 47. For example, when the failure probability of the centrifugal chiller is the highest, the cause is first investigated from the components of this device. At this time, based on information on whether or not the trend of the frequency distribution of process measurement value history data of the evaporator, the condenser, the compressor, the electric motor and the like made by the equipment abnormality progress monitoring means 44 is changed, the expert is used. The use of the system enables accurate and quick failure cause tracking. The obtained result is transmitted to the measurement monitoring control means 21 to 25 of the corresponding building via the network.
[0052]
With the configuration and algorithm described above, the object of the present invention can be achieved.
[0053]
[Other embodiments]
The present invention is not limited to the above-described embodiment, but can be implemented in various modifications without departing from the scope of the invention.
[0054]
In the above, the embodiment of the present invention has been described by taking a group of buildings as an example, but it is obvious that the present invention can be similarly applied to other systems.
[0055]
Although the probability paper method is used as a parameter estimation method for failure analysis, a maximum likelihood method, a Bayes method, or the like may be used.
[0056]
Further, in the failure analysis according to the embodiment of the present invention, a method using only data of a failed device is used, but a hazard analysis method of performing reliability analysis including data of an operating device can be used in the same manner. .
[0057]
Although an expert system is used to automatically estimate the failure location and cause of the equipment in the building equipment diagnosis means, other AI (Artificial Intelligence) technology may be used.
[0058]
The embodiments can be implemented in combination as much as possible, and in that case, the effect of the combination can be obtained.
[0059]
【The invention's effect】
As described above, according to the present invention, the equipment of a plurality of buildings is intensively monitored at all times, and it is possible to predict the abnormality that occurs in the course of the failure of each equipment in each building, When it occurs, it is possible to provide an equipment diagnosis apparatus that can accurately and promptly trace the cause of a failure, point out the cause of the failure, and support the operation of an operator and the maintenance of equipment by a building manager.
[Brief description of the drawings]
FIG. 1 is an overall configuration diagram showing an embodiment of a building facility diagnosis device according to the present invention.
FIG. 2 is an explanatory diagram showing an example of a frequency distribution of evaporator pressure values of a turbo refrigerator.
FIG. 3 is an explanatory diagram showing an example of a frequency distribution of evaporator pressure values of a turbo refrigerator.
FIG. 4 is an explanatory diagram showing an example of a frequency distribution of evaporator pressure values of a turbo refrigerator.
FIG. 5 is an explanatory diagram in which device failure data is plotted on a normal distribution probability sheet among four types of distribution probability sheets.
FIG. 6 is an explanatory diagram in which device failure data is plotted on an exponential distribution probability sheet among four types of distribution probability sheets.
FIG. 7 is an explanatory diagram in which device failure data is plotted on a lognormal distribution probability sheet among four types of distribution probability sheets.
FIG. 8 is an explanatory diagram in which device failure data is plotted on Weibull distribution probability paper among four types of distribution probability paper.
FIG. 9 is an explanatory diagram in which four sets of failure data are respectively plotted on Weibull probability paper.
FIG. 10 is an explanatory diagram in which four sets of failure data are respectively plotted on Weibull probability paper.
FIG. 11 is an explanatory diagram in which four sets of failure data are plotted on Weibull probability paper.
FIG. 12 is an explanatory diagram in which four sets of failure data are plotted on Weibull probability paper.
FIG. 13 is an explanatory diagram in which the four sets of failure data are plotted on one Weibull probability sheet.
FIG. 14 is an explanatory diagram in which three sets of data among the four sets of failure data are integrated as the same group and plotted on Weibull probability paper.
FIG. 15 is an explanatory diagram in which the three sets of integrated failure data are plotted on a normal distribution probability sheet among four types of distribution probability sheets.
FIG. 16 is an explanatory diagram in which the three sets of integrated failure data are plotted on an exponential distribution probability sheet among four types of distribution probability sheets.
FIG. 17 is an explanatory diagram in which the three sets of integrated failure data are plotted on a lognormal distribution probability sheet among four types of distribution probability sheets.
FIG. 18 is an explanatory diagram in which the three sets of integrated failure data are plotted on Weibull distribution probability paper among four types of distribution probability paper.
FIG. 19 is an explanatory view showing a part of an example of an FTA of a turbo refrigerator.
FIG. 20 is a conceptual explanatory diagram of a failure rate pattern based on a bathtub curve.
FIG. 21 is an explanatory diagram showing the shape of the Weibull distribution when the value of the shape parameter β is changed.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Building equipment diagnostic apparatus 21-25 Measurement monitoring control means 30 Network 41 Data collection / classification / storage means 42 Process history database 43 History data analysis means 44 Equipment abnormality progress monitoring means 45 Failure database 46 Failure time distribution identification means 47 Distribution parameter identification Means 48 Equipment diagnostic means 50 Terminal

Claims (8)

監視対象側から入力されたプロセス値データや機器故障データを定期的にかつ、各監視対象別に分類・編集してプロセス履歴データベース及び故障データベースに蓄積するデータ収集・分類・記憶手段と、
監視対象となる機器が故障した時、前記プロセス履歴データベースに蓄積されたその機器に関連する計測データに基づき、故障に至るまでの履歴データ変化の傾向を解析する履歴データ解析手段と、
この履歴データ解析手段の解析結果を用いて各監視対象における同種の機器の計測データの履歴変化を監視することにより、当該監視対象となる機器の故障を予測する機器異常進行監視手段と、
前記故障データベースを解析することにより、各監視対象における各機器故障データ集団の分布タイプを同定する故障時間分布タイプ同定手段と、
この故障時間分布タイプ同定手段の同定結果に基づいて、異なる分布タイプ毎に分けて蓄えられたタイプ毎故障データベースから、同種の機器の故障データを組み合わせて故障データ集団の各タイプの分布パラメータを求める分布パラメータ推定手段と、
前記機器異常進行監視手段と分布パラメータ推定手段からの情報に基づいて、各監視対象における設備機器の故障予知、異常進行個所発見を行う設備診断手段と、
を具備したことを特徴とする設備診断装置。
Data collection, classification, and storage means for periodically classifying and editing process value data and equipment failure data input from the monitoring target and for each monitoring target, and accumulating them in the process history database and the failure database;
When a device to be monitored fails, based on measurement data related to the device accumulated in the process history database, history data analysis means for analyzing a trend of change in history data until failure,
By monitoring the history change of the measurement data of the same type of device in each monitoring target using the analysis result of the history data analysis unit, a device abnormality progress monitoring unit for predicting a failure of the device to be monitored,
By analyzing the failure database, failure time distribution type identification means for identifying the distribution type of each device failure data group in each monitoring target,
Based on the identification result of the failure time distribution type identification means, a distribution parameter of each type of a failure data group is obtained by combining failure data of the same type of equipment from a failure database for each type stored separately for each different distribution type. Distribution parameter estimating means;
Equipment diagnosis means for performing failure prediction of equipment equipment in each monitoring target, finding an abnormal progress point, based on information from the equipment abnormality progress monitoring means and distribution parameter estimation means,
An equipment diagnostic device comprising:
前記履歴データ解析手段は、前記プロセス履歴データベースに蓄積された計測データから随時度数分布を作成し、機器が故障した時その機器に関連する計測データの度数分布から故障に至るまでの履歴データの変化として、正常時の値から外れるまでの傾向を解析し、域値及び異常と判定するための域値を計測値が越える回数を求めることを特徴とする請求項1記載の設備診断装置。The history data analysis means creates a frequency distribution from time to time from the measurement data accumulated in the process history database, and when a device fails, changes in the history data from the frequency distribution of the measurement data related to the device to the failure. 2. The equipment diagnosis apparatus according to claim 1, wherein a tendency until the value deviates from a normal value is analyzed, and the number of times the measured value exceeds a threshold value and a threshold value for determining an abnormality is obtained. 前記機器異常進行監視手段は、前記域値及び異常と判定するための域値を計測値が越える回数を求めた、故障した機器と同仕様機器の同じ計測データの前記履歴データ解析手段で作られた度数分布を監視し、その計測値が前記域値を越える回数が異常と判定するための回数になった時、その機器の異状が進行していると判断することを特徴とする請求項1または2記載の設備診断装置。The device abnormality progress monitoring means is obtained by the history data analyzing means of the same measurement data of the equipment having the same specification as the failed equipment, wherein the threshold value and the number of times the measured value exceeds the threshold value for determining an abnormality are obtained. The frequency distribution is monitored, and when the number of times the measured value exceeds the threshold value is the number of times for determining an abnormality, it is determined that the abnormality of the device is in progress. Or the equipment diagnostic device according to 2. 前記故障時間分布タイプ同定手段は、前記故障データベースに蓄積された、各監視対象の各機器の故障データ集団がどのタイプの分布に従うかを確率紙法により同定することを特徴とする請求項1乃至3のいずれか1項に記載の設備診断装置。The failure time distribution type identification means identifies, according to a probability paper method, which type of distribution the failure data group of each device to be monitored accumulates in the failure database follows. 3. The equipment diagnostic device according to any one of items 3. 前記故障時間分布タイプ同定手段は、前記故障データベースに蓄積された、各監視対象の各機器の故障データ集団がワイブル分布、正規分布、対数正規分布、指数分布の4つの中から、どの分布に一番良く従うかを確率紙法により判定することを特徴とする請求項1乃至4のいずれか1項に記載の設備診断装置。The failure time distribution type identification means determines which one of four distributions of a failure data group of each device to be monitored stored in the failure database from a Weibull distribution, a normal distribution, a lognormal distribution, and an exponential distribution. The equipment diagnosis apparatus according to any one of claims 1 to 4, wherein it is determined according to a probability paper method whether or not to follow best. 前記分布パラメータ同定手段は、前記故障時間分布タイプ同定手段により、異なる分布タイプ毎に分けて蓄えられたタイプ毎故障データベースから同種の機器の故障データを組み合わせて故障データ集団の各タイプの分布パラメータを求めるために確率紙法を用いることを特徴とする請求項1乃至5のいずれか1項に記載の設備診断装置。The distribution parameter identification means combines the failure data of the same type of equipment from the failure database for each type stored separately for each different distribution type by the failure time distribution type identification means to calculate the distribution parameter of each type of the failure data group. The equipment diagnosis apparatus according to any one of claims 1 to 5, wherein the probability paper method is used to obtain the equipment diagnosis. 前記分布パラメータ同定手段は、前記故障時間分布タイプ同定手段により、ワイブル分布、正規分布、対数正規分布、指数分布の4つのタイプ毎に分けて蓄えられたタイプ毎故障データベースから同種の機器の故障データを組み合わせて故障データ集団の前記4タイプの各分布パラメータを求めるために、確率紙法を用いることを特徴とする請求項1乃至6のいずれか1項に記載の設備診断装置。The distribution parameter identification unit is configured to store failure data of the same type of equipment from a failure database for each type stored by the failure time distribution type identification unit for each of four types of Weibull distribution, normal distribution, lognormal distribution, and exponential distribution. The equipment diagnosis apparatus according to any one of claims 1 to 6, wherein a probability paper method is used to obtain each of the four types of distribution parameters of the failure data group by combining. 前記設備診断手段は、前記機器異常進行監視手段からの情報を元にエキスパートシステムなどの手法を用いて機器故障予知を行い、FTA(フォールツリーアナリシス)手法を用いて前記分布パラメータ同定手段で求められた故障時間パラメータから、機器故障確率の大きい順に異常進行個所発見や異状発生原因追求を行うことを特徴とする請求項1乃至7のいずれか1項に記載の設備診断装置。The equipment diagnosis means performs equipment failure prediction using a method such as an expert system based on information from the equipment abnormality progress monitoring means, and is obtained by the distribution parameter identification means using an FTA (fall tree analysis) method. The equipment diagnosis apparatus according to any one of claims 1 to 7, wherein, based on the failure time parameter, an abnormally advanced location is found and a cause of occurrence of an abnormality is searched for in descending order of the equipment failure probability.
JP2002295412A 2002-10-08 2002-10-08 Diagnostic device for equipment Withdrawn JP2004133553A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2002295412A JP2004133553A (en) 2002-10-08 2002-10-08 Diagnostic device for equipment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP2002295412A JP2004133553A (en) 2002-10-08 2002-10-08 Diagnostic device for equipment

Publications (1)

Publication Number Publication Date
JP2004133553A true JP2004133553A (en) 2004-04-30

Family

ID=32285677

Family Applications (1)

Application Number Title Priority Date Filing Date
JP2002295412A Withdrawn JP2004133553A (en) 2002-10-08 2002-10-08 Diagnostic device for equipment

Country Status (1)

Country Link
JP (1) JP2004133553A (en)

Cited By (37)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2006040203A (en) * 2004-07-30 2006-02-09 Tokyo Gas Co Ltd Method and system for diagnosing equipment
JP2006143359A (en) * 2004-11-17 2006-06-08 Mitsubishi Electric Corp Building facility diagnosing device
JP2006199486A (en) * 2005-01-24 2006-08-03 Mitsubishi Electric Building Techno Service Co Ltd System for providing maintenance information for elevator
JP2007106525A (en) * 2005-10-12 2007-04-26 Mitsubishi Electric Building Techno Service Co Ltd System for making elevator maintenance work plan
JP2007199844A (en) * 2006-01-24 2007-08-09 Hitachi Ltd Component demand prediction program, component demand prediction method, and system executing this method
JP2007205614A (en) * 2006-01-31 2007-08-16 Mitsubishi Electric Corp Inspection service system for air conditioner
JP2007328522A (en) * 2006-06-07 2007-12-20 Tokyo Electric Power Co Inc:The Failure probability calculation device, method, and program
JP2008250716A (en) * 2007-03-30 2008-10-16 Railway Technical Res Inst Integrated risk management support system
KR100890028B1 (en) * 2008-04-16 2009-03-25 (주)크리티컬퍼실리티서비스 Consolidated system reliability test (csrt) method
JP2009155043A (en) * 2007-12-26 2009-07-16 Toshiba Elevator Co Ltd Elevator component improvement schedule system and elevator component improvement schedule method
JP2009215010A (en) * 2008-03-11 2009-09-24 Toshiba Corp Monitoring and diagnosing device and remote monitoring and diagnosing system
JP2010108108A (en) * 2008-10-29 2010-05-13 Sekisui Chem Co Ltd System for managing measurement relating to house
JP2011020758A (en) * 2009-07-13 2011-02-03 Toshiba Elevator Co Ltd Part improvement plan system of elevator and part improvement plan method thereof
KR101048727B1 (en) * 2009-08-21 2011-07-14 주식회사 서린바이오사이언스 Bio / medical equipment abnormal history management system using sensor tag and its operation method
JP2011170898A (en) * 2011-06-07 2011-09-01 Toshiba Corp User evaluation apparatus depending on hardware usage status
WO2012118550A1 (en) * 2011-03-02 2012-09-07 Carrier Corporation Spm fault detection and diagnostics algorithm
JP2013218550A (en) * 2012-04-10 2013-10-24 Toyota Home Kk Building facility management system
WO2016129775A1 (en) * 2015-02-15 2016-08-18 (주)부품디비 Device and method for managing failure types for predictive maintenance of marine resource production equipment
JP2016176611A (en) * 2015-03-18 2016-10-06 株式会社東芝 Air conditioning control device, control method, and program
WO2016174734A1 (en) * 2015-04-28 2016-11-03 三菱電機株式会社 Monitoring device and method for air conditioner
JP2017088314A (en) * 2015-11-10 2017-05-25 株式会社日立ビルシステム Equipment diagnostic apparatus, equipment diagnostic method, and equipment diagnostic system
JP2017215669A (en) * 2016-05-30 2017-12-07 日本電信電話株式会社 Probability density function estimation device, continuous value prediction device, method, and program
KR101830036B1 (en) * 2017-11-13 2018-02-19 (주)아이티공간 Preventive maintenance method of elevator driving part
JP2018085589A (en) * 2016-11-22 2018-05-31 パナソニックIpマネジメント株式会社 Diagnostic method, diagnostic device, and display device
WO2018124228A1 (en) * 2016-12-28 2018-07-05 ナブテスコ株式会社 Monitoring system for foreign substance removal device, foreign substance removal system, and monitoring method for foreign substance removal device
JP2019008354A (en) * 2017-06-20 2019-01-17 株式会社日立ビルシステム Monitoring apparatus, monitoring system, and abnormality detection method
KR101925357B1 (en) * 2017-12-12 2019-02-26 (주)위세아이텍 System and method for visualizing equipment health status and forecast maintenance requirements
JPWO2018123037A1 (en) * 2016-12-28 2019-04-04 三菱電機ビルテクノサービス株式会社 Elevator remote monitoring device
CN110114294A (en) * 2016-12-26 2019-08-09 三菱电机株式会社 Recovery system
CN110296504A (en) * 2019-06-19 2019-10-01 浙江大学 Air-conditioning O&M intelligent gateway and preservation & testing method
CN110371825A (en) * 2019-07-31 2019-10-25 中南大学 A kind of mine hoist tension method for diagnosing faults, system and control system
KR20200083702A (en) * 2018-12-28 2020-07-09 한라아이엠에스 주식회사 Fault Information Extraction Method for Industrial Tanks
CN111498634A (en) * 2016-04-26 2020-08-07 三菱电机株式会社 Elevator remote maintenance support system
CN111689328A (en) * 2020-06-24 2020-09-22 重庆电子工程职业学院 Intelligent elevator detection system
WO2022071189A1 (en) * 2020-09-30 2022-04-07 Ntn株式会社 State-monitoring device and state-monitoring method
CN115600130A (en) * 2022-11-15 2023-01-13 山东锦弘纺织股份有限公司(Cn) Plywood composite adhesive equipment operation management and control system based on data analysis
WO2023058190A1 (en) * 2021-10-07 2023-04-13 三菱電機株式会社 Elevator control inspection system and elevator control inspection method

Cited By (55)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2006040203A (en) * 2004-07-30 2006-02-09 Tokyo Gas Co Ltd Method and system for diagnosing equipment
JP2006143359A (en) * 2004-11-17 2006-06-08 Mitsubishi Electric Corp Building facility diagnosing device
JP2006199486A (en) * 2005-01-24 2006-08-03 Mitsubishi Electric Building Techno Service Co Ltd System for providing maintenance information for elevator
JP4668632B2 (en) * 2005-01-24 2011-04-13 三菱電機ビルテクノサービス株式会社 Elevator maintenance information provision system
JP2007106525A (en) * 2005-10-12 2007-04-26 Mitsubishi Electric Building Techno Service Co Ltd System for making elevator maintenance work plan
JP2007199844A (en) * 2006-01-24 2007-08-09 Hitachi Ltd Component demand prediction program, component demand prediction method, and system executing this method
JP4638359B2 (en) * 2006-01-31 2011-02-23 三菱電機株式会社 Air conditioner inspection service system
JP2007205614A (en) * 2006-01-31 2007-08-16 Mitsubishi Electric Corp Inspection service system for air conditioner
JP2007328522A (en) * 2006-06-07 2007-12-20 Tokyo Electric Power Co Inc:The Failure probability calculation device, method, and program
JP2008250716A (en) * 2007-03-30 2008-10-16 Railway Technical Res Inst Integrated risk management support system
JP2009155043A (en) * 2007-12-26 2009-07-16 Toshiba Elevator Co Ltd Elevator component improvement schedule system and elevator component improvement schedule method
JP4612699B2 (en) * 2008-03-11 2011-01-12 株式会社東芝 Monitoring / diagnosis device and remote monitoring / diagnosis system
US8306782B2 (en) 2008-03-11 2012-11-06 Kabushiki Kaisha Toshiba System for monitoring and diagnosing remote devices
JP2009215010A (en) * 2008-03-11 2009-09-24 Toshiba Corp Monitoring and diagnosing device and remote monitoring and diagnosing system
TWI400605B (en) * 2008-03-11 2013-07-01 Toshiba Kk Apparatus for monitoring and diagnosing an object device and system for monitoring and diagnosing grouped object devices
KR100890028B1 (en) * 2008-04-16 2009-03-25 (주)크리티컬퍼실리티서비스 Consolidated system reliability test (csrt) method
JP2010108108A (en) * 2008-10-29 2010-05-13 Sekisui Chem Co Ltd System for managing measurement relating to house
JP2011020758A (en) * 2009-07-13 2011-02-03 Toshiba Elevator Co Ltd Part improvement plan system of elevator and part improvement plan method thereof
KR101048727B1 (en) * 2009-08-21 2011-07-14 주식회사 서린바이오사이언스 Bio / medical equipment abnormal history management system using sensor tag and its operation method
WO2012118550A1 (en) * 2011-03-02 2012-09-07 Carrier Corporation Spm fault detection and diagnostics algorithm
US9977409B2 (en) 2011-03-02 2018-05-22 Carrier Corporation SPC fault detection and diagnostics algorithm
CN103403463A (en) * 2011-03-02 2013-11-20 开利公司 Spm fault detection and diagnostics algorithm
CN103403463B (en) * 2011-03-02 2016-06-01 开利公司 Fault detection and diagnosis algorithm
JP2011170898A (en) * 2011-06-07 2011-09-01 Toshiba Corp User evaluation apparatus depending on hardware usage status
JP2013218550A (en) * 2012-04-10 2013-10-24 Toyota Home Kk Building facility management system
WO2016129775A1 (en) * 2015-02-15 2016-08-18 (주)부품디비 Device and method for managing failure types for predictive maintenance of marine resource production equipment
JP2016176611A (en) * 2015-03-18 2016-10-06 株式会社東芝 Air conditioning control device, control method, and program
WO2016174734A1 (en) * 2015-04-28 2016-11-03 三菱電機株式会社 Monitoring device and method for air conditioner
JPWO2016174734A1 (en) * 2015-04-28 2017-11-30 三菱電機株式会社 Air conditioner monitoring device and method
JP2017088314A (en) * 2015-11-10 2017-05-25 株式会社日立ビルシステム Equipment diagnostic apparatus, equipment diagnostic method, and equipment diagnostic system
CN111498634B (en) * 2016-04-26 2021-07-06 三菱电机株式会社 Elevator remote maintenance support system
CN111498634A (en) * 2016-04-26 2020-08-07 三菱电机株式会社 Elevator remote maintenance support system
JP2017215669A (en) * 2016-05-30 2017-12-07 日本電信電話株式会社 Probability density function estimation device, continuous value prediction device, method, and program
JP2018085589A (en) * 2016-11-22 2018-05-31 パナソニックIpマネジメント株式会社 Diagnostic method, diagnostic device, and display device
CN110114294A (en) * 2016-12-26 2019-08-09 三菱电机株式会社 Recovery system
CN110114294B (en) * 2016-12-26 2020-11-03 三菱电机株式会社 Recovery system
WO2018124228A1 (en) * 2016-12-28 2018-07-05 ナブテスコ株式会社 Monitoring system for foreign substance removal device, foreign substance removal system, and monitoring method for foreign substance removal device
JPWO2018123037A1 (en) * 2016-12-28 2019-04-04 三菱電機ビルテクノサービス株式会社 Elevator remote monitoring device
KR20190083666A (en) 2016-12-28 2019-07-12 미쓰비시 덴키 빌딩 테크노 서비스 가부시키 가이샤 Elevator remote monitoring device
JP2019008354A (en) * 2017-06-20 2019-01-17 株式会社日立ビルシステム Monitoring apparatus, monitoring system, and abnormality detection method
WO2019093613A1 (en) * 2017-11-13 2019-05-16 주식회사 아이티공간 Preventive maintenance method for elevator driving unit
US11845632B2 (en) 2017-11-13 2023-12-19 Its Co., Ltd. Method of predictively maintaining elevator driving unit
KR101830036B1 (en) * 2017-11-13 2018-02-19 (주)아이티공간 Preventive maintenance method of elevator driving part
CN110167860A (en) * 2017-11-13 2019-08-23 It空间株式会社 The predictive maintenance method of elevator driving portion
KR101925357B1 (en) * 2017-12-12 2019-02-26 (주)위세아이텍 System and method for visualizing equipment health status and forecast maintenance requirements
KR20200083702A (en) * 2018-12-28 2020-07-09 한라아이엠에스 주식회사 Fault Information Extraction Method for Industrial Tanks
KR102176401B1 (en) * 2018-12-28 2020-11-10 한라아이엠에스 주식회사 Fault Information Extraction Method for Industrial Tanks
CN110296504A (en) * 2019-06-19 2019-10-01 浙江大学 Air-conditioning O&M intelligent gateway and preservation & testing method
CN110371825B (en) * 2019-07-31 2020-12-22 中南大学 Mine hoist tension fault diagnosis method, system and control system
CN110371825A (en) * 2019-07-31 2019-10-25 中南大学 A kind of mine hoist tension method for diagnosing faults, system and control system
CN111689328A (en) * 2020-06-24 2020-09-22 重庆电子工程职业学院 Intelligent elevator detection system
WO2022071189A1 (en) * 2020-09-30 2022-04-07 Ntn株式会社 State-monitoring device and state-monitoring method
WO2023058190A1 (en) * 2021-10-07 2023-04-13 三菱電機株式会社 Elevator control inspection system and elevator control inspection method
CN115600130A (en) * 2022-11-15 2023-01-13 山东锦弘纺织股份有限公司(Cn) Plywood composite adhesive equipment operation management and control system based on data analysis
CN115600130B (en) * 2022-11-15 2023-03-07 山东锦弘纺织股份有限公司 Plywood composite adhesive equipment operation control system based on data analysis

Similar Documents

Publication Publication Date Title
JP2004133553A (en) Diagnostic device for equipment
US11769118B2 (en) Systems and methods for automated diagnostics of HVAC systems
JP4334176B2 (en) Building energy conservation evaluation monitoring device
US10700942B2 (en) Building management system with predictive diagnostics
Wang et al. A robust fault detection and diagnosis strategy for multiple faults of VAV air handling units
US10663187B2 (en) Air conditioning system and control method thereof
US7031880B1 (en) Method and apparatus for assessing performance of an environmental control system
Yu et al. A data-driven approach to extract operational signatures of HVAC systems and analyze impact on electricity consumption
Wang et al. Sensor validation and reconstruction for building central chilling systems based on principal component analysis
US11703815B2 (en) Building control system with predictive maintenance based on time series analysis
Zhou et al. A model‐based fault detection and diagnosis strategy for HVAC systems
CN109065135B (en) Hospital logistics equipment full-life-cycle management cloud platform, method and system
Wang et al. A robust fault detection and diagnosis strategy for pressure-independent VAV terminals of real office buildings
CN101326415B (en) Remote diagnosis and estimation of refrigerant system
KR19990006755A (en) Expansion valve monitoring system
US20210080138A1 (en) Performance diagnosis device and performance diagnosis method for air conditioner
Yang et al. HVAC equipment, unitary: Fault detection and diagnosis
JP2005301582A (en) Process management device
CN108302739B (en) Temperature adjusting system and temperature adjusting method
CN107461881B (en) Refrigeration host energy efficiency diagnosis method and system for hospital air conditioner room
Wei et al. Fault detection and diagnosis for variable-air-volume systems using combined residual, qualitative and quantitative techniques
JP2004234302A (en) Process management device
CN111788598B (en) Energy efficiency analysis system using data of high-rise energy management system
Holcomb et al. Algorithms for green buildings: Learning-based techniques for energy prediction and fault diagnosis
JP4434905B2 (en) Process management device

Legal Events

Date Code Title Description
A621 Written request for application examination

Free format text: JAPANESE INTERMEDIATE CODE: A621

Effective date: 20050225

A977 Report on retrieval

Free format text: JAPANESE INTERMEDIATE CODE: A971007

Effective date: 20071122

A131 Notification of reasons for refusal

Free format text: JAPANESE INTERMEDIATE CODE: A131

Effective date: 20071130

A761 Written withdrawal of application

Free format text: JAPANESE INTERMEDIATE CODE: A761

Effective date: 20080124