TWI443591B - Inspection managing system and method - Google Patents
Inspection managing system and method Download PDFInfo
- Publication number
- TWI443591B TWI443591B TW97120647A TW97120647A TWI443591B TW I443591 B TWI443591 B TW I443591B TW 97120647 A TW97120647 A TW 97120647A TW 97120647 A TW97120647 A TW 97120647A TW I443591 B TWI443591 B TW I443591B
- Authority
- TW
- Taiwan
- Prior art keywords
- patrol
- object symbol
- inspection
- symbol group
- probability
- Prior art date
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Description
本案是關於一種巡檢管理系統及方法，特別是關於一種用於消防設備巡檢之巡檢管理系統及方法。The present invention relates to a patrol management system and method, and more particularly to a patrol management system and method for patrol inspection of fire equipment.
目前工廠或大樓中的消防設備幾乎都沒有中央監控，僅靠值班人員依排班方式至現場進行設備巡查與紀錄，紀錄完成後再靠人力完成資料整理與報表製作及主管簽核，對於查核異常缺失與重要監測數值也無統計分析，而無法達到設備重要監測項目的趨勢統計分析與預警功能。要能達到設備預警功能就必須在設備失效前提出警告，對於消防設備目前只有法規訂定某些參數的界限值，超過時，代表系統已經失效但無法達到預警功能。而利用統計方法可找出預警界限值，解決預警的問題。At present, there is almost no central monitoring of fire-fighting equipment in factories or buildings. Only on-duty personnel go to the site to conduct equipment inspections and records. After the completion of the records, they will rely on manpower to complete data sorting and report production and signing of supervisors. There is no statistical analysis of missing and important monitoring values, and it is unable to achieve trend statistical analysis and early warning functions for important monitoring items of equipment. In order to achieve the equipment warning function, it is necessary to provide a warning before the equipment fails. For the fire equipment, only the limits of certain parameters are set by the regulations. When it exceeds, the system has failed but the warning function cannot be achieved. The use of statistical methods can identify early warning thresholds and solve early warning problems.
此外，設備巡查目前多以外包為主，雇主希望外包巡檢人員，能確實到位進行檢查。然而，無論是使用光學條碼或無線電子標籤(RFID)符號設備，仍然有可能發生人員複製條碼或無線電子標籤而未到位的問題。另一問題是巡檢效率如何提升，巡檢人員常發生要花時間思考〝下一個查核設備是誰？查核地點在那？〞的問題，造成巡查時間與人力成本的浪費。若是能有一套系統能輔助主動告知巡查對象與位置，則可幫助巡檢人員在最短時間內完成巡查工作。In addition, equipment inspections are currently mostly outsourcing, and employers want to outsource inspection personnel and can actually check them in place. However, whether using optical bar codes or wireless electronic tag (RFID) symbol devices, there is still the possibility that a person copying a bar code or a wireless electronic tag may not be in place. Another problem is how to improve the efficiency of inspection. The inspection personnel often take time to think about who is the next inspection equipment. The problem of checking the location is that the inspection time and labor cost are wasted. If there is a system that can assist the initiative to inform the inspection object and location, it can help the inspection personnel to complete the inspection in the shortest time.
為了解決以上所述的問題，需要更有效的巡檢管理 系統及方法。職是之故，本案發明人鑑於以上的需求，經悉心之研究，並一本鍥而不捨的精神，終創作出本案『巡檢管理系統及方法』。In order to solve the above problems, more effective inspection management is needed. System and method. In view of the above needs, the inventor of this case, through careful research and a perseverance spirit, finally created the "patrol inspection management system and method".
本案之一目的為提出一種巡檢管理系統及方法，利用使用者巡檢時在區域空間中檢查設備順序具有一定習慣性的特徵，統計現有的巡檢資料，獲得複數巡檢時間順位與一對象符號組之間的一出現機率分佈，根據該出現機率分佈，確認一巡檢是否到位或提示一巡檢中一特定巡檢時間順位所對應的一特定對象符號，以節省人力與時間成本且達到設備之預知保養功效。One of the objectives of the present invention is to propose a patrol management system and method, which utilizes the habitual characteristics of checking the equipment sequence in the regional space when the user patrols, and counts the existing patrol data to obtain a plurality of patrol inspection time orders and an object. An occurrence probability distribution between the symbol groups, according to the probability distribution, confirming whether a patrol is in place or prompting a specific object symbol corresponding to a specific patrol time in a patrol, thereby saving manpower and time cost and achieving Predictive maintenance of the equipment.
本案係提出一種巡檢管理方法，該方法包括下列步驟：設置複數巡檢時間順位、複數第一對象符號組與一第二對象符號組，其中各第一對象符號組的複數份子分別對應於該等巡檢時間順位；利用統計地分析該等第一對象符號組的複數份子，決定該第二對象符號組的各份子於各巡檢時間順位的各出現機率；及，根據每一該出現機率，提示一特定巡檢時間順位所對應的一特定對象符號，其中該特定巡檢時間順位為該等巡檢時間順位的其中之一，且該特定對象符號出現於該等第一對象符號組的該等份子中與該第二對象符號組的複數份子中。The present invention provides a patrol management method, the method comprising the steps of: setting a plurality of patrol time units, a plurality of first object symbol groups and a second object symbol group, wherein the plurality of first object symbol groups respectively correspond to the Waiting for the patrol time order; using statistically analyzing the plural components of the first object symbol group to determine the probability of occurrence of each of the second object symbol groups in each patrol time; and, according to each occurrence probability a specific object symbol corresponding to a specific patrol time sequence, wherein the specific patrol time order is one of the patrol time order, and the specific object symbol appears in the first object symbol group The aliquot is in the plural of the second object symbol group.
較佳者，該巡檢管理方法，更包括下列步驟：經由複數第一巡檢，獲得該特定對象符號所對應的一特定被巡檢對象的一參數的量值資料，其中該量值資料按照該 等第一巡檢所對應的複數巡檢時間的順序排列。Preferably, the inspection management method further includes the following steps: obtaining, by the plurality of first inspections, quantity data of a parameter of a specific object to be inspected corresponding to the specific object symbol, wherein the quantity data is The The order of the plurality of inspection times corresponding to the first inspection is arranged.
較佳者，該巡檢管理方法，更包括下列步驟：從該量值資料中選取複數鄰近連續樣本；獲得該等鄰近連續樣本的一平均值與一標準差；根據該平均值與該標準差，制定一預警值上限與一預警值下限；及，根據該等鄰近連續樣本中的一樣本超過該預警值上限與該預警值下限之間的範圍，發出警告。Preferably, the inspection management method further comprises the steps of: selecting a plurality of adjacent continuous samples from the magnitude data; obtaining an average value and a standard deviation of the adjacent consecutive samples; and determining the standard deviation according to the average value And formulating an upper limit value of the warning value and a lower limit value of the warning value; and issuing a warning according to the range between the upper limit of the warning value and the lower limit of the warning value according to the same in the adjacent consecutive samples.
較佳者，該巡檢管理方法，更包括下列步驟：從該量值資料中選取複數連續樣本，其中該等連續樣本位於該等巡檢時間中的一第一巡檢時間與一第二巡檢時間之間；獲得該等連續樣本的一第一平均值；設定該第一平均值為該第二巡檢時間時該參數的一起始平均值；根據該起始平均值、該參數的一法規上限、該參數的一法規下限與該參數的一調整係數，制定一預警值上限與一預警值下限；及，根據在該第二巡檢時間之後的一巡檢所獲得該特定對象符號的該參數的一量值超過該預警值上限與該預警值下限之間的範圍，發出警告。Preferably, the patrol management method further comprises the steps of: selecting a plurality of consecutive samples from the statistic data, wherein the consecutive samples are located at a first patrol time and a second patrol in the patrol time Between the detection times; obtaining a first average value of the consecutive samples; setting the first average value as a starting average value of the parameter when the second inspection time; according to the initial average value, one of the parameters a legal upper limit, a lower limit of the parameter and an adjustment coefficient of the parameter, and an upper limit value of the warning value and a lower limit value of the early warning value; and, according to a inspection after the second inspection time, the specific object symbol is obtained A magnitude of the parameter exceeds the range between the upper limit of the warning value and the lower limit of the warning value, and a warning is issued.
較佳者，該特定被巡檢對象為一灑水系統，且該參數為一水壓、一出水流量與一幫浦運轉電流的其中之一。Preferably, the specific object to be inspected is a sprinkler system, and the parameter is one of a water pressure, a water flow rate and a pump running current.
本案又提出一種巡檢管理系統，該系統包括一處理單元。該處理單元設置複數巡檢時間順位、複數第一對象符號組與一第二對象符號組，利用統計地分析該等第一對象符號組的複數份子，決定該第二對象符號組的各份子於各巡檢時間順位的各出現機率，且根據每一該出現機率，提示一特定巡檢時間順位所對應的一特定對象 符號，其中各第一對象符號組的複數份子分別對應於該等巡檢時間順位，該特定巡檢時間順位為該等巡檢時間順位的其中之一，且該特定對象符號出現於該等第一對象符號組的該等份子中與該第二對象符號組的複數份子中。The present invention further proposes a patrol management system, which system includes a processing unit. The processing unit sets a plurality of patrol time units, a plurality of first object symbol groups and a second object symbol group, and statistically analyzes the plurality of elements of the first object symbol group to determine each of the second object symbol groups Each occurrence probability of each inspection time order, and according to each occurrence probability, prompting a specific object corresponding to a specific inspection time order a symbol, wherein the plurality of first object symbol groups respectively correspond to the patrol time order, the specific patrol time order is one of the patrol time orders, and the specific object symbol appears in the first The aliquot of an object symbol group is in the plural of the second object symbol group.
本案再提出一種巡檢管理方法，該方法包括下列步驟：設置複數巡檢時間順位、複數第一對象符號組與一第二對象符號組，其中各第一對象符號組的複數份子與該第二對象符號組的複數份子皆分別對應於該等巡檢時間順位；經由統計地分析該等第一對象符號組的複數份子與該第二對象符號組的該等份子，決定該第二對象符號組的各份子於各巡檢時間順位的各同輪出現機率；及，根據每一該同輪出現機率，確認一輪巡檢是否到位，其中該第二對象符號組由該輪巡檢產生。The present invention further provides a patrol management method, the method comprising the steps of: setting a plurality of patrol time units, a plurality of first object symbol groups and a second object symbol group, wherein the plurality of first object symbol groups and the second The plurality of elements of the object symbol group respectively correspond to the patrol time order; determining the second object symbol group by statistically analyzing the plurality of elements of the first object symbol group and the aliquot of the second object symbol group The probability that each of the pieces is in the same round of each inspection time; and, according to the probability of occurrence of each of the same wheels, confirm whether a round of inspection is in place, wherein the second object symbol group is generated by the round inspection.
本案更提出一種巡檢管理系統，該系統包括一處理單元。該處理單元設置複數巡檢時間順位、複數第一對象符號組與一第二對象符號組，經由統計地分析該等第一對象符號組的複數份子與該第二對象符號組的該等份子，決定該第二對象符號組的各份子於各巡檢時間順位的各同輪出現機率，且根據每一該同輪出現機率，確認一輪巡檢是否到位，其中各第一對象符號組的複數份子與該第二對象符號組的複數份子皆分別對應於該等巡檢時間順位，且該第二對象符號組由該輪巡檢產生。The present invention further proposes a patrol management system, which system includes a processing unit. The processing unit sets a plurality of patrol time order, a plurality of first object symbol groups and a second object symbol group, and statistically analyzes the plurality of the first object symbol group and the quarantine of the second object symbol group, Determining the probability of occurrence of each of the second object symbol group in each of the same time of the inspection time, and confirming whether a round of inspection is in place according to the probability of occurrence of each of the same rounds, wherein the plurality of first object symbol groups are plural The plurality of copies of the second object symbol group respectively correspond to the patrol time order, and the second object symbol group is generated by the round of patrol.
請參閱第一圖，其為本案第一實施例所提巡檢管理系統的示意圖。如圖所示，巡檢管理系統91包括一處理單元W1。處理單元W1設置複數巡檢時間順位A11、A12、A13、A14、…、A1M、複數對象符號組B1、B2、…、BN與一對象符號組V1。對象符號組B1包括複數份子S39、S40、S41、S42、…、S80，對象符號組B2包括複數份子S40、S39、S41、S42、…、S_NULL，且對象符號組BN包括複數份子S39、S40、S42、S_NULL、…、S80，其中份子S_NULL代表一空符號。Please refer to the first figure, which is a schematic diagram of the inspection management system proposed in the first embodiment of the present invention. As shown, the inspection management system 91 includes a processing unit W1. The processing unit W1 sets a plurality of patrol time units A11, A12, A13, A14, ..., A1M, a plurality of object symbol groups B1, B2, ..., BN and an object symbol group V1. The object symbol group B1 includes a plurality of components S39, S40, S41, S42, ..., S80, the object symbol group B2 includes a plurality of components S40, S39, S41, S42, ..., S_NULL, and the object symbol group BN includes a plurality of components S39, S40, S42, S_NULL, ..., S80, wherein the part S_NULL represents an empty symbol.
各對象符號組(如B2)的複數份子分別對應於該等巡檢時間順位A11、A12、A13、A14、…、A1M，例如，對象符號組B1的該等份子S39、S40、S41、S42、…、S80分別對應於該等巡檢時間順位A11、A12、A13、A14、…、A1M，對象符號組B2的該等份子S40、S39、S41、S42、…、S_NULL分別對應於該等巡檢時間順位A11、A12、A13、A14、…、A1M。對象符號組V1包括複數份子S1、…、S39、S40、S41、…、S80。The plural components of each object symbol group (such as B2) respectively correspond to the patrol time orders A11, A12, A13, A14, ..., A1M, for example, the equal parts S39, S40, S41, S42 of the object symbol group B1, ..., S80 respectively correspond to the patrol time orders A11, A12, A13, A14, ..., A1M, and the equal parts S40, S39, S41, S42, ..., S_NULL of the object symbol group B2 respectively correspond to the patrols The time is in order of A11, A12, A13, A14, ..., A1M. The object symbol group V1 includes a plurality of components S1, ..., S39, S40, S41, ..., S80.
處理單元W1利用統計地分析該等對象符號組B1、B2、…、BN的複數份子S39、S40、S41、S42、…、S80、S40、S39、S41、S42、…、S_NULL、…、S39、S40、S42、S_NULL、…、S80，決定對象符號組V1的各份子(如S39)出現於各巡檢時間順位(如A11)的各出現機率(如P61)，且根據每一該出現機率(如P61、P72、P83、…等)，提示一特定巡檢時間順位(如A12)所對應的一特定對象符號(如S40)。The processing unit W1 statistically analyzes the plural parts S39, S40, S41, S42, ..., S80, S40, S39, S41, S42, ..., S_NULL, ..., S39 of the object symbol groups B1, B2, ..., BN, S40, S42, S_NULL, ..., S80, determining the probability of each occurrence of the object symbol group V1 (such as S39) appearing in each inspection time order (such as A11) (such as P61), and according to each occurrence probability ( For example, P61, P72, P83, ..., etc.), a specific object symbol (such as S40) corresponding to a specific inspection time order (such as A12) is indicated.
該特定巡檢時間順位(如A12)為該等巡檢時間順位A11、A12、A13、A14、…、A1M的其中之一。該特定對象符號(如S40)出現於該等對象符號組B1、B2、…、BN的該等份子S39、S40、S41、S42、…、S80、S40、S39、S41、S42、…、S_NULL、…、S39、S40、S42、S_NULL、…、S80中，且出現於對象符號組V1的該等份子S1、…、S39、S40、S41、…、S80中。The specific inspection time order (such as A12) is one of the inspection time orders A11, A12, A13, A14, ..., A1M. The specific object symbols (such as S40) appear in the equal parts S39, S40, S41, S42, ..., S80, S40, S39, S41, S42, ..., S_NULL of the object symbol groups B1, B2, ..., BN, ..., S39, S40, S42, S_NULL, ..., S80, and appear in the equal parts S1, ..., S39, S40, S41, ..., S80 of the object symbol group V1.
該等巡檢時間順位A11、A12、A13、A14、…、A1M可代表該等對象符號組B1、B2、…、BN中的一預設對象符號組(如B2)的複數份子(如S40、S39、S41、S42、…、S_NULL)的複數排列順序位置或該等排列順序位置所對應的順序符號，其中該等排列順序位置對應於依照時間順序被巡檢的複數被巡檢對象(未顯示)。該特定巡檢時間順位(如A12)可為一輪巡檢中的一下一巡檢時間順位。對象符號組V1的該等份子S1、…、S39、S40、S41、…、S80可包括該等對象符號組B1、B2、…、BN的該等份子中的全部非重複份子或部分非重複份子。The patrol time units A11, A12, A13, A14, ..., A1M may represent plural elements of a preset object symbol group (such as B2) of the object symbol groups B1, B2, ..., BN (such as S40, S39, S41, S42, ..., S_NULL) a sequence of sequential positions corresponding to the positions of the arrangement order, wherein the arrangement order positions correspond to a plurality of inspected objects that are inspected in chronological order (not shown) ). The specific inspection time order (such as A12) may be the next inspection time in the round of inspection. The aliquots S1, ..., S39, S40, S41, ..., S80 of the object symbol group V1 may include all non-repetitive or partial non-repetitive elements in the aliquots of the object symbol groups B1, B2, ..., BN .
該特定對象符號(如S40)可對應於一特定被巡檢對象(未顯示)，例如，該特定對象符號(如S40)可對應於一消防設備、生物或裝置(未顯示)。該特定對象符號(如S40)可為該特定被巡檢對象的識別碼、名稱或敘述；對於該等對象符號組B1、B2、…、BN、V1，每一對象符號組(如B2)的複數分子(如S40、S39、S41、S42、…、S_NULL)所對應的複數被巡檢對象在趨勢上可按照空間次序排列。The particular object symbol (e.g., S40) may correspond to a particular object being inspected (not shown), for example, the particular object symbol (e.g., S40) may correspond to a firefighting device, creature, or device (not shown). The specific object symbol (such as S40) may be an identification code, a name or a description of the specific object to be inspected; for the object symbol groups B1, B2, ..., BN, V1, each object symbol group (such as B2) The plural quarantine objects corresponding to the complex numerator (such as S40, S39, S41, S42, ..., S_NULL) can be arranged in a spatial order according to the trend.
處理單元W1可為一電子化裝置，且可利用該特定對象符號(如S40)和一畫面或一語音，提示該特定被巡檢對象的位置與名稱。處理單元W1可安排一學習性的巡檢，可利用該學習性的巡檢，獲得分別對應於該等巡檢時間順位A11、A12、A13、A14、…、A1M的一第一對象符號組的複數份子，且可設置該第一對象符號組為該等對象符號組B1、B2、…、BN的其中之一。The processing unit W1 can be an electronic device, and can use the specific object symbol (such as S40) and a picture or a voice to prompt the location and name of the specific object to be inspected. The processing unit W1 can arrange a learning patrol, and the learning patrol can be used to obtain a first object symbol group corresponding to the patrol time units A11, A12, A13, A14, ..., A1M, respectively. A plurality of copies, and the first object symbol group can be set to be one of the object symbol groups B1, B2, ..., BN.
一第一輪巡檢可獲得該對象符號組B1，一第二輪巡檢可獲得該對象符號組B2，一第N輪巡檢可獲得該對象符號組BN。The object symbol group B1 can be obtained by a first round of inspection, and the object symbol group B2 can be obtained by a second round of inspection. The object symbol group BN can be obtained by an Nth round of inspection.
被巡檢對象可具有設備識別碼，被巡檢對象所對應的對象符號可代表一設備次序識別碼，例如，對象符號組B1所具有的該等份子S39、S40、S41、S42、…、S80可分別代表設備次序識別碼39、40、41、42、…、80，且該等巡檢時間順位A11、A12、A13、A14、…、A1M可分別代表該等份子S39、S40、S41、S42、…、S80的複數指示位置序號1、2、3、4、…、M，其中M為一自然數。對象符號組B2所具有的該等份子S40、S39、S41、S42、…、S_NULL可分別代表設備次序識別碼40、39、41、42、…、無識別碼，且該等巡檢時間順位A11、A12、A13、A14、…、A1M可分別代表該等份子S40、S39、S41、S42、…、S_NULL的複數指示位置序號1、2、3、4、…、M。The object to be inspected may have a device identification code, and the object symbol corresponding to the object to be inspected may represent a device order identification code, for example, the elements S39, S40, S41, S42, ..., S80 of the object symbol group B1 The device order identification codes 39, 40, 41, 42, ..., 80 may be respectively represented, and the patrol time orders A11, A12, A13, A14, ..., A1M may represent the equal parts S39, S40, S41, S42, respectively. The plural of S, ..., S80 indicates the position number 1, 2, 3, 4, ..., M, where M is a natural number. The squarings S40, S39, S41, S42, ..., S_NULL of the object symbol group B2 may respectively represent device order identification codes 40, 39, 41, 42, ..., no identification code, and the patrol time order A11 A12, A13, A14, ..., A1M may respectively represent position numbers 1, 2, 3, 4, ..., M of the equal parts S40, S39, S41, S42, ..., S_NULL.
第一輪巡檢時先訂出設備識別碼與設備次序識別碼，讓位於同一空間中的設備次序識別碼的數字盡量連 續。於一段時間內，依照該等巡檢時間順位A11、A12、A13、A14、…、A1M進行設備次序識別碼的紀錄，以作為系統學習之用。對於某一類設備，在巡檢時，隨該等巡檢時間順位A11、A12、A13、A14、…、A1M作紀錄，且紀錄所對應的設備次序識別碼。In the first round of inspection, the device identification code and the device order identification code are first set, so that the numbers of the device order identification codes located in the same space are connected as much as possible. Continued. For a period of time, the device order identification code is recorded in accordance with the inspection time ranks A11, A12, A13, A14, ..., A1M for system learning. For a certain type of equipment, during the inspection, the inspection time ranks A11, A12, A13, A14, ..., A1M are recorded, and the corresponding equipment order identification code is recorded.
當記錄所巡檢的資料一段時間後，處理單元W1中可具有該N個對象符號組B1、B2、…、BN，每一該巡檢時間順位(如A12)可對應於一對象符號組的複數份子(如S40、S39、…、S40)。After recording the patrol data for a period of time, the processing unit W1 may have the N object symbol groups B1, B2, ..., BN, and each of the patrol time units (such as A12) may correspond to an object symbol group. Multiple components (such as S40, S39, ..., S40).
請參閱第二圖，其為本案第一實施例所提巡檢管理系統中具有相同位置指標的複數設備次序識別碼的示意圖。如圖所示，巡檢時間順位A11(位置指標1)對應於一對象符號組H1的複數份子S39、S40、…、S39(具有相同巡檢時間順位A11的設備次序識別碼39、40、…、39)。Please refer to the second figure, which is a schematic diagram of a plurality of device order identification codes having the same position index in the inspection management system of the first embodiment of the present invention. As shown in the figure, the patrol time order A11 (position index 1) corresponds to the plural parts S39, S40, ..., S39 of an object symbol group H1 (device order identification codes 39, 40, ... having the same patrol time order A11). , 39).
請參閱第三圖(a)，其為本案第一實施例所提巡檢管理系統中各對象符號的出現個數的示意圖。如圖所示，橫座標代表對象符號組V1的該等份子S1、…、S39、S40、S41、S42、…、S80(本圖僅列出對象符號S10-S50，設備次序識別碼10、…、39、40、41、42、…、80)，對象符號組V1包括所感興趣的對象符號。縱座標代表『針對巡檢時間順位A11(位置指標1)所對應的對象符號組H1，該等份子S10、…、S39、S40、S41、S42、…、S80中的各份子(如S39)出現於對象符號組H1的該等份子S39、S40、…、S39中的出現個數』，曲線F1為利用一維高斯混合模型(Gaussian Mixed Model)近似實際出現 個數的一條出現個數分佈曲線。Please refer to the third figure (a), which is a schematic diagram of the number of occurrences of each object symbol in the inspection management system according to the first embodiment of the present invention. As shown, the abscissa represents the aliquots S1, ..., S39, S40, S41, S42, ..., S80 of the object symbol group V1 (this figure only lists the object symbols S10-S50, the device order identification code 10, ... , 39, 40, 41, 42, ..., 80), the object symbol group V1 includes the object symbol of interest. The ordinate represents "the object symbol group H1 corresponding to the inspection time order A11 (position index 1), and the elements (such as S39) appear in the equal parts S10, ..., S39, S40, S41, S42, ..., S80. The number of occurrences in the equal parts S39, S40, ..., S39 of the object symbol group H1, and the curve F1 is approximated by the one-dimensional Gaussian Mixed Model. A number distribution curve appears.
請參閱第三圖(b)，其為本案第一實施例所提巡檢管理系統中利用一維高斯混合模型分析近似各對象符號的出現機率的示意圖。如圖所示，橫座標代表對象符號組V1的該等份子。縱座標為利用第三圖(a)的資料與一維高斯混合模型，所獲得該等份子S1、…、S39、S40、S41、S42、…、S80中的各份子(如S39)出現於該對象符號組H1的該等份子S1、…、S40、S41、S42、…、S80中的出現機率。曲線g1 、g2 、g3 為高斯分佈曲線，曲線wi gi 為具有加權值wi 的高斯分佈曲線，曲線F2為將w1 g1 、w2 g2 、w3 g3 相加的結果，以近似第三圖(a)中的曲線F1，因此曲線F2為對應於曲線F1的出現機率分佈曲線。相同方法可獲得其他巡檢時間順位(如A14)所對應的出現機率的分佈曲線。Please refer to the third figure (b), which is a schematic diagram of analyzing the probability of occurrence of each object symbol by using the one-dimensional Gaussian mixture model in the inspection management system proposed in the first embodiment of the present invention. As shown, the abscissa represents the aliquot of the object symbol group V1. The ordinate is obtained by using the data of the third figure (a) and the one-dimensional Gaussian mixture model, and the obtained elements (such as S39) of the equal parts S1, ..., S39, S40, S41, S42, ..., S80 appear in the The probability of occurrence in the equal parts S1, ..., S40, S41, S42, ..., S80 of the object symbol group H1. The curves g 1 , g 2 , g 3 are Gaussian distribution curves, the curve w i g i is a Gaussian distribution curve with a weighting value w i , and the curve F2 is a sum of w 1 g 1 , w 2 g 2 , w 3 g 3 As a result, the curve F1 in the third graph (a) is approximated, so the curve F2 is an appearance probability distribution curve corresponding to the curve F1. The same method can obtain the distribution curve of the probability of occurrence of other inspection time order (such as A14).
請參閱第四圖，其為本案第一實施例所提巡檢管理系統中各對象符號出現於各巡檢時間順位的出現機率的示意圖。如圖所示，對象符號組V1可包括，例如，全部的設備次序識別碼1、…、39、40、41、…、80，且全部的設備次序識別碼的總量為N。該等巡檢時間順位A11、A12、A13、A14、…、A1M的總量為M，出現機率矩陣F(N,M)為經由高斯混合模型分析後所獲得的出現機率資料，其中F代表機率。對象符號S39出現於巡檢時間順位A11的出現機率P61的值為0.10，對象符號S40出現於巡檢時間順位A12的出現機率P72的值為0.12，對象符號S41出現於巡檢時間順位A13的出現機率P83的值 為0.09。Please refer to the fourth figure, which is a schematic diagram of the probability of occurrence of each object symbol appearing in each patrol time order in the patrol management system according to the first embodiment of the present invention. As shown, the object symbol group V1 may include, for example, all device order identification codes 1, ..., 39, 40, 41, ..., 80, and the total amount of all device order identification codes is N. The total amount of the inspection time ranks A11, A12, A13, A14, ..., A1M is M, and the probability matrix F(N, M) is the probability of occurrence data obtained by analyzing the Gaussian mixture model, where F represents the probability. . The probability that the object symbol S39 appears in the inspection time order A11 is 0.10, the probability that the object symbol S40 appears in the inspection time order A12 is 0.12, and the object symbol S41 appears in the inspection time order A13. Probability P83 Is 0.09.
隨時間的改變，巡檢的總輪數與該等對象符號組B1、B2、…、BN的數量跟著改變。在此情況下，利用自適應高斯混合模型(Adaptive GMM)，可使出現機率矩陣F(N,M)隨時間動態改變。As the time changes, the total number of rounds of the inspection changes with the number of the object symbol groups B1, B2, ..., BN. In this case, using the adaptive Gaussian Mixing Model (Adaptive GMM), the probability matrix F(N, M) can be dynamically changed with time.
請參閱第五圖，其為本案第一實施例所提巡檢管理系統中利用學習完成出現機率矩陣的流程示意圖。如圖所示，在步驟502中，第一輪巡檢時先訂出設備識別碼與設備次序識別碼，讓位在同一空間中的設備次序識別碼的數字盡量連續。在步驟504中，記錄每種設備每輪巡檢的時間順位與其對應的設備次序識別碼。Please refer to the fifth figure, which is a schematic flow chart of using the learning completion probability matrix in the inspection management system of the first embodiment of the present invention. As shown in the figure, in step 502, the device identification code and the device order identification code are first set in the first round of inspection, so that the numbers of the device order identification codes in the same space are as continuous as possible. In step 504, the time sequence of each round of inspections of each device is recorded with its corresponding device order identification code.
在步驟506中，記錄一段時間，取得多筆資料，讓每個時間順位上有多個設備次序識別碼。在步驟508中，對每個時間順位上的多個設備次序識別碼，利用一維高斯混合模型，找出每個次序識別碼的出現機率。在步驟510中，完成出現機率矩陣F(N,M)，其中N代表設備次序識別碼總量，M代表時間順位總量，F代表機率。In step 506, a period of time is recorded, and a plurality of pieces of data are acquired such that there are a plurality of device order identification codes in each time order. In step 508, a plurality of device order identification codes for each time order are used, and a one-dimensional Gaussian mixture model is used to find the probability of occurrence of each order identification code. In step 510, the occurrence probability matrix F(N, M) is completed, where N represents the total number of device order identification codes, M represents the total amount of time series, and F represents the probability.
請參閱第六圖與第四圖，第六圖為本案第一實施例所提巡檢管理系統中提示使用者下一個巡查設備的流程示意圖。如圖所示，在步驟602中，在查詢某一設備次序識別碼所對應設備的時候，該設備次序識別碼所對應的時間順位為m，例如為m=1。在步驟604中，設定一設備次序識別碼出現的一臨界出現機率f2，例如f2=0.1。在步驟606中，找出時間順位為m+1的狀態下，所有設備次序識別碼的複數同順位出現機率F(n,m+1)，例如出 現機率矩陣F(N,M)中的一橫行出現機率向量的複數份子，其中n代表對象符號組V1的每一該份子的一編號。Please refer to the sixth figure and the fourth figure. The sixth figure is a schematic flow chart of prompting the user to the next patrol device in the patrol management system of the first embodiment of the present invention. As shown in the figure, in step 602, when querying a device corresponding to a device order identifier, the time sequence corresponding to the device order identifier is m, for example, m=1. In step 604, a critical probability of occurrence f2 of a device order identification code is set, for example, f2 = 0.1. In step 606, in the state where the time order is m+1, the probability that all device order identification codes have the same order occurrence probability F(n, m+1), for example, A horizontal line in the current probability matrix F(N, M) appears as a complex number of probability vectors, where n represents a number of each of the object symbol groups V1.
在步驟608中，列出該等同順位出現機率F(n,m+1)中出現機率大於臨界出現機率f2的所有設備次序識別碼。例如，在m=1時，同屬對應於巡檢時間順位A12的該等同順位出現機率F(n,2)中的出現機率P72是0.12，且P72>f2=0.1，故符合的設備次序識別碼包括40。在步驟610中，以畫面或語音方式提示使用者該符合的設備次序識別碼所對應的設備名稱與設備位置。In step 608, all device order identification codes in which the probability of occurrence of the equivalent order occurrence rate F(n, m+1) is greater than the critical occurrence probability f2 is listed. For example, when m=1, the probability of occurrence P72 in the equivalent occurrence probability F(n, 2) corresponding to the patrol time order A12 is 0.12, and P72>f2=0.1, so the device order identification is met. The code includes 40. In step 610, the user name and device location corresponding to the device order identifier corresponding to the user are prompted in a screen or voice manner.
請參閱第七圖與第四圖，第七圖為本案第一實施例所提巡檢管理系統中處理巡檢到位確認的配置示意圖。如圖所示，配置D1包括該等巡檢時間順位A11、A12、A13、A14、…、A1M、所要確認是否到位巡檢的對象符號組(如B1)、所要確認對象符號組的同輪出現機率組Q(K,m)、一臨界出現機率組f1(K,m)與一比較結果值組r(K,m)。Please refer to the seventh figure and the fourth figure. The seventh figure is a schematic diagram of the configuration of processing the inspection in-position confirmation in the inspection management system of the first embodiment of the present invention. As shown in the figure, the configuration D1 includes the patrol time units A11, A12, A13, A14, ..., A1M, the object symbol group (such as B1) to be confirmed whether the patrol is in place, and the same round of the symbol group to be confirmed. The probability group Q (K, m), a critical probability probability group f1 (K, m) and a comparison result value group r (K, m).
根據第四圖的出現機率矩陣F(N,M)與所要確認的對象符號組(如B1)，可獲得同輪出現機率組Q(K,m)，例如，同輪出現機率組Q(K,m)中的一同輪出現機率Q(K,1)=P61=0.10(巡檢時間順位A11對應於對象符號組B1的份子S39，且對象符號組V1的份子S39出現於巡檢時間順位A11的出現機率P61為0.10)，一同輪出現機率Q(K,2)=P73=0.12，一同輪出現機率Q(K,3)=P83=0.09(巡檢時間順位A13對應於對象符號組B1的份子S41，且對象符號組V1的份子S41出現於巡檢時間順位A13的出現機率 P83為0.09)。利用一臨界出現機率fa，例如fa=0.05，可獲得臨界出現機率組f1(K,m)，其中臨界出現機率組f1(K,m)可為對應於同輪出現機率組Q(K,m)的一出現機率向量。利用關係式Q(K,m)<f1(K,m)，可獲得比較結果值組r(K,m)。根據比較結果值組r(K,m)，可確認第K輪巡檢是否到位，例如，Q(K,m)<f1(K,m)為真的總數R為10，若R/M>恕限值，則第K輪巡檢被判斷為未到位，其中M代表時間順位總量。According to the appearance probability matrix F(N, M) of the fourth figure and the object symbol group (such as B1) to be confirmed, the same-round probability group Q(K, m) can be obtained, for example, the same-round occurrence probability group Q (K) , the probability of occurrence of the same wheel in m), Q(K, 1) = P61 = 0.10 (the inspection time order A11 corresponds to the object S39 of the object symbol group B1, and the object S39 of the object symbol group V1 appears in the inspection time order A11 The probability of occurrence P61 is 0.10), the probability of occurrence of the same wheel Q (K, 2) = P73 = 0.12, the probability of occurrence of the same wheel Q (K, 3) = P83 = 0.09 (the inspection time order A13 corresponds to the object symbol group B1) Part S41, and the probability that the object S41 of the object symbol group V1 appears in the inspection time order A13 P83 is 0.09). Using a critical probability of occurrence fa, such as fa=0.05, a critical probability probability group f1(K,m) can be obtained, wherein the critical occurrence probability group f1(K,m) can correspond to the same-round probability group Q(K,m). An occurrence probability vector. Using the relation Q(K, m) < f1 (K, m), a comparison result value group r(K, m) can be obtained. According to the comparison result value group r(K, m), it can be confirmed whether the K-th rotation inspection is in place, for example, Q(K, m) < f1 (K, m) is true, the total number R is 10, if R/M> For the limit value, the Kth round inspection is judged as not being in place, where M represents the total amount of time.
請參閱第八圖，其為本案第一實施例所提巡檢管理系統中確認一輪巡檢是否到位的流程示意圖。如圖所示，在步驟702中，紀錄各設備該輪巡檢的各時間順位與其對應的各設備次序識別碼。在步驟704中，設定每一該設備次序識別碼出現的各臨界出現機率fa，開始時間順位m=1，總數R=0。在步驟706中，找出在m=每一該時間順位時，每一該設備次序識別碼的各同輪出現機率Q。在步驟708中，若Q<fa，則R++。在步驟710中，若m<M為真，則回到步驟706，若m<M為假，則進入步驟712。在步驟710中，計算R/M，若R/M>恕限值，則該輪巡檢被判斷為未到位，其中M代表時間順位總量。Please refer to the eighth figure, which is a schematic diagram of the process of confirming whether a round of inspections is in place in the inspection management system proposed in the first embodiment of the present invention. As shown in the figure, in step 702, each device sequence of the round of inspections of each device is recorded with its corresponding device order identification code. In step 704, each critical occurrence probability fa of each device order identification code is set, and the start time order is m=1, and the total number R=0. In step 706, it is found that each m of the device order identification code has a probability Q when m = each time of the time. In step 708, if Q < fa, then R++. In step 710, if m < M is true, then return to step 706. If m < M is false, proceed to step 712. In step 710, R/M is calculated. If R/M> the limit value, the round of inspection is judged as not being in place, where M represents the total amount of time.
在重要參數如消防設備之灑水系統水壓、出水流量或幫浦運轉電流的監測方面，針對監測的短期變動分析與長期趨勢分析兩種狀態，採用不同的統計方法找出短期與長期的預警上下限值，若超過該上下限則發出預警，告知有異常狀況發生，讓人員針對異常狀況能及早找出原因，達到預知保養目的。In the monitoring of important parameters such as water pressure, effluent flow or pump running current of fire-fighting equipment, for short-term fluctuation analysis and long-term trend analysis of monitoring, different statistical methods are used to identify short-term and long-term warnings. If the upper and lower limits are exceeded, an early warning is issued to inform the occurrence of an abnormal situation, so that the personnel can find out the cause early for the abnormal condition and achieve the purpose of predictive maintenance.
請參閱第九圖，其為本案第一實施例所提巡檢管理系統中短期變動預警上下限值分析的示意圖。如圖所示，經由複數第一巡檢，獲得該特定對象符號(如S34)所對應的一特定被巡檢對象的一參數的量值資料，其中該參數可為一物理參數或一化學參數，且該量值資料E2按照該等第一巡檢所對應的複數巡檢時間的順序排列。Please refer to the ninth figure, which is a schematic diagram of the analysis of the upper and lower limit values of short-term fluctuation early warning in the inspection management system proposed in the first embodiment of the present invention. As shown in the figure, the plurality of first patrols are used to obtain a quantity data of a parameter of a specific object to be inspected corresponding to the specific object symbol (such as S34), wherein the parameter may be a physical parameter or a chemical parameter. And the magnitude data E2 is arranged in the order of the plurality of inspection times corresponding to the first inspections.
取一段時間(如巡檢時間T5至巡檢時間T6)的監測值即時資料，例如，選取複數鄰近連續樣本，利用適應性一維高斯混合模式找出即時平均值與標準差σ，取95%為正常分佈，訂出+3σ為預警值上限SWU ，-3σ為預警值下限SWL ，若是監測值超過此預警值範圍則發出警告。由於是採適應性高斯混合模式，故平均值與標準差σ會隨著新的監測值不斷變動調整。Take the monitoring data for a period of time (such as the inspection time T5 to the inspection time T6), for example, select a plurality of adjacent continuous samples, and use the adaptive one-dimensional Gaussian mixture mode to find the instantaneous average With the standard deviation σ, take 95% as the normal distribution, set +3σ is the upper limit of the warning value SW U , -3σ is the lower limit of the warning value SW L , and if the monitored value exceeds the range of this warning value, a warning is issued. Average value due to adaptive Gaussian mixing mode The standard deviation σ will be adjusted as the new monitoring value changes.
請參閱第十圖，其為本案第一實施例所提巡檢管理系統中長期變動預警上下限值分析的示意圖。如圖所示，量值資料E1隨巡檢時間變動，取系統開始的前3個月時間(如巡檢時間T1至巡檢時間T2)連續樣本的資料平均值做為起始平均值。獲得法規要求的上限H，法規要求的下限L，且計算長期趨勢預警的上下限值，例如，長期預警值上限，長期預警值下限，a=0.5~0.75，若是監測值超過此預警值範圍則發出警告。Please refer to the tenth figure, which is a schematic diagram of the analysis of the upper and lower limit values of the long-term change warning in the inspection management system proposed in the first embodiment of the present invention. As shown in the figure, the magnitude data E1 varies with the inspection time, and the average value of the continuous samples of the first three months (such as the inspection time T1 to the inspection time T2) is taken as the starting average. . Obtain the upper limit H required by the regulations, the lower limit L required by the regulations, and calculate the upper and lower limits of the long-term trend warning, for example, the upper limit of the long-term warning value , the long-term warning value lower limit , a=0.5~0.75, if the monitored value exceeds the range of this warning value, a warning will be issued.
請參閱第十一圖，其為本案第二實施例所提巡檢管理系統的示意圖。第十一圖的巡檢管理系統92的架構與第一圖的巡檢管理系統91的架構相似。如十一圖所示， 巡檢管理系統92包括一處理單元W2。處理單元W2設置如第一圖的複數巡檢時間順位A11、A12、A13、A14、…、A1M、複數對象符號組B1、B2、…、BN，處理單元W2另外設置一對象符號組C1，對象符號組C1包括複數份子S39、S41、S40、S42、…、S_NULL，其中份子S_NULL代表一空符號。對象符號組C1的該等份子S39、S41、S40、S42、…、S_NULL分別對應於該等巡檢時間順位A11、A12、A13、A14、…、A1M。Please refer to FIG. 11 , which is a schematic diagram of the inspection management system proposed in the second embodiment of the present invention. The architecture of the inspection management system 92 of the eleventh diagram is similar to the architecture of the inspection management system 91 of the first diagram. As shown in the eleventh picture, The inspection management system 92 includes a processing unit W2. The processing unit W2 sets a plurality of patrol time units A11, A12, A13, A14, ..., A1M, a plurality of object symbol groups B1, B2, ..., BN as shown in the first figure, and the processing unit W2 additionally sets an object symbol group C1, the object The symbol group C1 includes a plurality of parts S39, S41, S40, S42, ..., S_NULL, where the part S_NULL represents an empty symbol. The equal parts S39, S41, S40, S42, ..., S_NULL of the object symbol group C1 correspond to the patrol time orders A11, A12, A13, A14, ..., A1M, respectively.
處理單元W2經由統計地分析該等對象符號組B1、B2、…、BN的複數份子S39、S40、S41、S42、…、S80、S40、S39、S41、S42、…、S_NULL、…、S39、S40、S42、S_NULL、…、S80與對象符號組C1的該等份子S39、S41、S40、S42、…、S_NULL，決定對象符號組C1的各份子(如S41)出現於各巡檢時間順位(如A12)的各同輪出現機率(如Q(C1,2))，且根據每一該同輪出現機率(如Q(C1,1)、Q(C1,2)、Q(C1,3)、Q(C1,4)、…等)，確認一輪巡檢是否到位，其中該等對象符號組B1、B2、…、BN中的各對象符號組(如B2)的複數份子(如S40、S39、S41、S42、…、S_NULL)分別對應於該等巡檢時間順位A11、A12、A13、A14、…、A1M，且對象符號組C1由該輪巡檢產生。The processing unit W2 statistically analyzes the plural parts S39, S40, S41, S42, ..., S80, S40, S39, S41, S42, ..., S_NULL, ..., S39 of the object symbol groups B1, B2, ..., BN, S40, S42, S_NULL, ..., S80 and the equal parts S39, S41, S40, S42, ..., S_NULL of the object symbol group C1 determine that each part of the object symbol group C1 (such as S41) appears in each inspection time order ( Such as A12), the probability of occurrence of the same round (such as Q (C1, 2)), and according to each of the same round occurrence probability (such as Q (C1, 1), Q (C1, 2), Q (C1, 3) , Q (C1, 4), ..., etc., to confirm whether a round of inspection is in place, wherein the plurality of object symbol groups (such as B2) in the object symbol groups B1, B2, ..., BN (such as S40, S39) S41, S42, ..., S_NULL) respectively correspond to the patrol time orders A11, A12, A13, A14, ..., A1M, and the object symbol group C1 is generated by the round patrol.
請參閱第十二圖，其為本案第三實施例所提巡檢管理系統的示意圖。如圖所示，巡檢管理系統93包括一處理單元W3與一資料庫單元WA。資料庫單元WA耦合於處理單元W3，記錄該等巡檢時間順位A11、A12、A13、 A14、…、A1M與該等對象符號組B1、B2、…、BN。資料庫單元WA包括一表格WA1，且資料庫單元WA記錄該等巡檢時間順位A11、A12、A13、A14、…、A1M與該等對象符號組B1、B2、…、BN至表格WA1中。Please refer to the twelfth figure, which is a schematic diagram of the inspection management system proposed in the third embodiment of the present invention. As shown, the inspection management system 93 includes a processing unit W3 and a database unit WA. The database unit WA is coupled to the processing unit W3, and records the inspection time ranks A11, A12, A13, A14, ..., A1M and the object symbol groups B1, B2, ..., BN. The database unit WA includes a table WA1, and the database unit WA records the patrol time units A11, A12, A13, A14, ..., A1M and the object symbol groups B1, B2, ..., BN into the table WA1.
處理單元W3包括一作業控制單元W31、一行動巡檢裝置W32與一巡檢人員W33。作業控制單元W31耦合於資料庫單元WA，控制資料庫單元WA，且利用一巡檢工作排程模組W311安排一巡檢排程表，其中該巡檢排程表包括分別對應於該等巡檢時間順位A11、A12、A13、A14、…、A1M的一所安排對象符號組的複數份子與對應於該所安排對象符號組的一組查核內容。作業控制單元W31自動發出該巡檢排程表給行動巡檢裝置W32與巡檢人員W33。The processing unit W3 includes a job control unit W31, a mobile patrol device W32, and a patrol personnel W33. The job control unit W31 is coupled to the database unit WA, controls the database unit WA, and uses a patrol work scheduling module W311 to arrange a patrol schedule, wherein the patrol schedule includes corresponding patrols respectively. The plurality of elements of the arranged object symbol group of the time sequence A11, A12, A13, A14, ..., A1M are checked with a set of check contents corresponding to the arranged object symbol group. The job control unit W31 automatically issues the patrol schedule to the action patrol device W32 and the patrol person W33.
行動巡檢裝置W32可包括一個人數位助理(PDA)，行動巡檢裝置W32耦合於作業控制單元W31，下載該巡檢排程表與該特定對象符號所對應的一特定查核內容，其中該特定查核內容為該組查核內容的其中之一。巡檢人員W33配合行動巡檢裝置W32，且根據該巡檢排程表，實施一輪巡檢。巡檢人員W33在一特定巡檢時間順位(如A12)所對應的一巡檢時間，可巡檢對應於一實際對象符號(如S39)的一實際被巡檢對象，且獲得該實際被巡檢對象的一參數的一量值。The action patrol device W32 may include a PDA, and the action patrol device W32 is coupled to the job control unit W31 to download a specific check content corresponding to the specific object symbol of the patrol schedule, wherein the specific check content The content is one of the contents of the group check. The inspector W33 cooperates with the action inspection device W32, and performs a round of inspection according to the inspection schedule. The patrol personnel W33 can patrol an actual patrol object corresponding to an actual object symbol (such as S39) at a patrol time corresponding to a specific patrol time (such as A12), and obtain the actual patrol. A quantity of a parameter of the object being examined.
行動巡檢裝置W32記錄該實際對象符號(如S39)與該參數的該量值，且提供該特定巡檢時間順位(如A12)、該實際對象符號(如S39)與該參數的該量值給作業控制 單元W31。作業控制單元W31提供該特定巡檢時間順位(如A12)、該實際對象符號(如S39)與該參數的該量值給資料庫單元WA，且控制該資料庫單元WA記錄該特定巡檢時間順位(如A12)、該實際對象符號(如S39)與該參數的該量值。The action patrol device W32 records the actual object symbol (such as S39) and the magnitude of the parameter, and provides the specific patrol time order (such as A12), the actual object symbol (such as S39) and the magnitude of the parameter. Job control Unit W31. The job control unit W31 provides the specific inspection time order (such as A12), the actual object symbol (such as S39) and the magnitude of the parameter to the database unit WA, and controls the database unit WA to record the specific inspection time. The order (such as A12), the actual object symbol (such as S39) and the magnitude of the parameter.
特定巡檢時間順位(如A12)可對應於一特定被巡檢對象(如S39)，巡檢人員W33可根據該特定查核內容，巡檢該特定對象符號(如S39)所對應的一特定被巡檢對象，且記錄一巡檢結果於行動巡檢裝置W32中。行動巡檢裝置W32提供該巡檢結果給作業控制單元W31，作業控制單元W31利用該巡檢結果，進行查核結果簽核、缺失與重要監測參數統計分析、缺失追蹤、與報表製作。The specific inspection time sequence (such as A12) may correspond to a specific inspected object (such as S39), and the inspector W33 may inspect a specific object corresponding to the specific object symbol (such as S39) according to the specific checking content. The inspection object is recorded, and a inspection result is recorded in the action inspection device W32. The action inspection device W32 provides the inspection result to the work control unit W31, and the work control unit W31 uses the inspection result to perform check result check, missing and important monitoring parameter statistical analysis, missing tracking, and report creation.
請參閱第十三圖(a)，其為本案第四實施例所提巡檢管理系統的示意圖。如圖所示，巡檢管理系統94包括一資料庫系統W61、耦合於資料庫系統W61的一管理平台W62、耦合於資料庫系統W61的廠區A或公司A的客戶端巡檢軟體W71、耦合於資料庫系統W61的廠區B或公司B的客戶端巡檢軟體W72、耦合於客戶端軟體W71的現場巡檢執行者W81、耦合於客戶端軟體W72的現場巡檢執行者W82、和耦合於管理平台W62的管理者W91、主管者W92與改善者W93，其中管理者W91、主管者W92與改善者W93包括控制管理平台W62的裝置。現場巡檢執行者W82包括手持智慧裝置PDA與巡檢人員。Please refer to FIG. 13(a), which is a schematic diagram of the inspection management system proposed in the fourth embodiment of the present invention. As shown, the inspection management system 94 includes a database system W61, a management platform W62 coupled to the database system W61, a factory A or a company A client software W71 coupled to the database system W61, and a coupling. The client inspection software W72 of the database B or the company B of the database system W61, the on-site inspection performer W81 coupled to the client software W71, the on-site inspection performer W82 coupled to the client software W72, and the coupling The manager W91, the supervisor W92, and the improver W93 of the management platform W62, wherein the manager W91, the supervisor W92, and the improver W93 include devices that control the management platform W62. The on-site inspection performer W82 includes a handheld smart device PDA and a patrol inspector.
巡檢管理系統94將設備巡檢與預知保養的所有流程 電子化，包括巡檢工作排程、巡檢結果紀錄、查核結果簽核、缺失統計分析與追蹤、報表製作等工作。The inspection management system 94 will perform all processes of equipment inspection and predictive maintenance. Electronic, including inspection work schedule, inspection results record, check result signing, missing statistical analysis and tracking, report production and so on.
首先，在管理平台W62上設定巡查工作排程，之後管理平台W62會自動發出排程工單給巡檢人員，巡檢人員則是用手持智慧裝置(如PDA)下載某時程階段之工作排程內容，包括查核設備對象與查核內容，再進行現場的巡檢與查核。與一般巡檢系統不同之處，本案利用設備週期性檢查順序進行統計分析與智慧學習，找出每個設備於每個順位時的出現機率，可作為日後巡檢是否到位之判斷。同時利用此出現機率亦可即時以畫面或語音方式提示使用者下一個巡查設備的名稱與位置，以提升巡檢設備時的工作效率。First, the patrol work schedule is set on the management platform W62. After that, the management platform W62 will automatically issue the schedule work order to the patrol personnel, and the patrol personnel download the work schedule of a certain time stage with the handheld smart device (such as PDA). The content of the process, including checking the equipment object and checking the content, and then conducting on-site inspection and check. Different from the general inspection system, this case uses the periodic inspection sequence of equipment for statistical analysis and intelligent learning to find out the probability of occurrence of each device in each order, which can be used as a judgment for whether the inspection will be in place in the future. At the same time, the probability of using this opportunity can promptly prompt the user to name and position the next inspection device in a picture or voice manner, so as to improve the working efficiency when the device is inspected.
巡檢完畢後，將所有查詢紀錄資料從手持智慧裝置中上傳至管理平台W62，之後在管理平台W62進行查核結果簽核、缺失與重要監測參數統計分析、缺失追蹤、報表製作等工作。在重要參數如灑水系統水壓或出水流量的監測方面，針對監測的短期變動分析與長期趨勢分析兩種狀態採用不同的統計方法，來預警系統是否發生異常狀況，讓人員針對異常狀況能及早找出原因，達到預知保養目的。After the inspection, all the inquiries and records are uploaded from the handheld smart device to the management platform W62, and then the verification result signing, missing and important monitoring parameter statistical analysis, missing tracking, report production, etc. are performed on the management platform W62. In the monitoring of important parameters such as water pressure or effluent flow of the sprinkler system, different statistical methods are used for the short-term fluctuation analysis and long-term trend analysis of the monitoring to detect whether the system has abnormal conditions, so that the personnel can be early against the abnormal situation. Find out the cause and achieve the purpose of predictive maintenance.
請參閱第十三圖(b)，其為本案第四實施例所提巡檢管理方法的作業流程示意圖。如圖所示，巡檢管理方法包括下列步驟：基本資料建立802、廠別編輯804、系統編輯806、人員權限編輯808、區域編輯810、設備類型/設備編輯812、查核項目編輯814、查核表單編輯816、 基本資料完成818、排程設定820、PDA更新排程(下載)822、個人數位助理(PDA)巡檢824、巡檢到位提示826、PDA擷取排程(上傳)828、不合格檢查830、不合格通報832、排程完成檢查834、檢視/編輯巡檢結果836、警戒值趨勢統計838、巡檢到位確認840、工程師審核842、主管審核844與報表列印846。Please refer to the thirteenth figure (b), which is a schematic diagram of the operation flow of the inspection management method proposed in the fourth embodiment of the present invention. As shown in the figure, the inspection management method includes the following steps: basic data creation 802, site editing 804, system editing 806, personnel authority editing 808, area editing 810, device type/device editing 812, checking item editing 814, checking form. Edit 816, Basic data completion 818, schedule setting 820, PDA update schedule (download) 822, personal digital assistant (PDA) inspection 824, inspection in place prompt 826, PDA capture scheduling (upload) 828, failure check 830, Unqualified notification 832, scheduled completion check 834, inspection/edit inspection result 836, warning value trend statistics 838, inspection in place confirmation 840, engineer review 842, supervisor review 844, and report print 846.
請參閱第一圖以說明本案所提一巡檢管理方法，該方法包括下列步驟：設置複數巡檢時間順位A11、A12、A13、A14、…、A1M、複數第一對象符號組B1、B2、…、BN與一第二對象符號組V1，其中各第一對象符號組的複數份子分別對應於該等巡檢時間順位A11、A12、A13、A14、…、A1M；利用統計地分析該等第一對象符號組B1、B2、…、BN的複數份子，決定該第二對象符號組V1的各份子(如S39)出現於各巡檢時間順位(如A11)的各出現機率(如P61)；及，根據每一該出現機率(如P61、P72、P83、…等)，提示一特定巡檢時間順位(如A12)所對應的一特定對象符號(如S40)，其中該特定巡檢時間順位(如A12)為該等巡檢時間順位A11、A12、A13、A14、…、A1M的其中之一，且該特定對象符號(如S40)出現於該等第一對象符號組B1、B2、…、BN的該等份子中與該第二對象符號組V1的複數份子中。Please refer to the first figure to illustrate a patrol management method proposed in the present case. The method includes the following steps: setting a plurality of patrol time orders A11, A12, A13, A14, ..., A1M, and a plurality of first object symbol groups B1, B2. ..., BN and a second object symbol group V1, wherein the plurality of first object symbol groups respectively correspond to the patrol time orders A11, A12, A13, A14, ..., A1M; using statistical analysis of the a plurality of object symbol groups B1, B2, ..., BN, determining the probability of occurrence of each of the second object symbol group V1 (such as S39) appearing in each inspection time order (such as A11) (such as P61); And, according to each occurrence probability (such as P61, P72, P83, ..., etc.), prompting a specific inspection time (such as A12) corresponding to a specific object symbol (such as S40), wherein the specific inspection time is ranked (such as A12) is one of the patrol time orders A11, A12, A13, A14, ..., A1M, and the specific object symbol (such as S40) appears in the first object symbol group B1, B2, ... And the plurality of BN elements are in the plural of the second object symbol group V1.
請參閱第十一圖以說明本案所提一巡檢管理方法，該方法包括下列步驟：設置複數巡檢時間順位A11、A12、A13、A14、…、A1M、複數第一對象符號組B1、B2、…、BN與一第二對象符號組C1，其中各第一對象 符號組的複數份子與該第二對象符號組C1的複數份子皆分別對應於該等巡檢時間順位A11、A12、A13、A14、…、A1M；經由統計地分析該等第一對象符號組B1、B2、…、BN的複數份子與該第二對象符號組C1的該等份子，決定該第二對象符號組C1的各份子(如S41)出現於各巡檢時間順位A11、A12、A13、A14、…、A1M的各同輪出現機率(如Q(C1,2))；及，根據每一該同輪出現機率(如Q(C1,1)、Q(C1,2)、Q(C1,3)、Q(C1,4)、…等)，確認一輪巡檢是否到位，其中該第二對象符號組C1由該輪巡檢產生。Please refer to the eleventh figure to illustrate a patrol management method proposed in the present case. The method includes the following steps: setting a plurality of patrol time units A11, A12, A13, A14, ..., A1M, and a plurality of first object symbol groups B1, B2 , ..., BN and a second object symbol group C1, wherein each first object The plurality of components of the symbol group and the plurality of components of the second object symbol group C1 respectively correspond to the patrol time orders A11, A12, A13, A14, ..., A1M; and the first object symbol group B1 is statistically analyzed. The plurality of parts of B, B, ..., BN and the aliquot of the second object symbol group C1 determine that each part of the second object symbol group C1 (such as S41) appears in each of the inspection time orders A11, A12, A13, The probability of occurrence of the same wheel of A14, ..., A1M (such as Q (C1, 2)); and, according to the probability of occurrence of each of the same wheel (such as Q (C1, 1), Q (C1, 2), Q (C1 , 3), Q (C1, 4), ..., etc., to confirm whether a round of inspection is in place, wherein the second object symbol group C1 is generated by the round of inspection.
綜上所述，本案之巡檢管理系統及方法確實能達到發明構想所設定的功效。唯，以上所述者僅為本案之較佳實施例，舉凡熟悉本案技藝之人士，在爰依本案精神所作之等效修飾或變化，皆應涵蓋於以下之申請專利範圍內。In summary, the inspection management system and method of this case can indeed achieve the effect set by the invention concept. The above descriptions are only the preferred embodiments of the present invention. Any equivalent modifications or variations made by those skilled in the art of the present invention should be included in the scope of the following patent application.
91、92、93、94‧‧‧巡檢管理系統91, 92, 93, 94‧‧‧ inspection management system
W1、W2、W3‧‧‧處理單元W1, W2, W3‧‧‧ processing unit
A11、A12、A13、A14、A1M‧‧‧巡檢時間順位A11, A12, A13, A14, A1M‧‧‧ inspection time order
B1、B2、BN、C1、H1、V1‧‧‧對象符號組B1, B2, BN, C1, H1, V1‧‧‧ object symbol groups
S1、S39、S40、S41、S42、S80、S_NULL‧‧‧對象符號S1, S39, S40, S41, S42, S80, S_NULL‧‧‧ object symbols
F1‧‧‧出現個數的分佈曲線F1‧‧‧ the distribution curve of the number
F2‧‧‧出現機率的分佈曲線F2‧‧‧ probability distribution curve
g1 、g2 、g3 ‧‧‧高斯分佈曲線g 1 , g 2 , g 3 ‧‧‧ Gaussian distribution curve
F(N,M)‧‧‧出現機率矩陣F(N,M)‧‧‧ occurrence probability matrix
N‧‧‧設備次序識別碼總量N‧‧‧Total order identification code
M‧‧‧時間順位總量M‧‧‧ time total amount
P61、P72、P83‧‧‧出現機率P61, P72, P83‧‧‧ occurrence probability
F(n,2)‧‧‧複數同順位出現機率F(n,2)‧‧‧The probability of multiple occurrences
Q(K,m)‧‧‧同輪出現機率組Q(K,m)‧‧‧The same round of probability group
Q(K,1)、Q(K,2)、Q(K,3)‧‧‧同輪出現機率Q(K,1), Q(K,2), Q(K,3)‧‧‧The probability of occurrence in the same round
Q(C1,1)、Q(C1,2)、Q(C1,3)、Q(C1,4)‧‧‧同輪出現機率Q(C1,1), Q(C1,2), Q(C1,3), Q(C1,4)‧‧‧The probability of occurrence in the same round
f1(K,m)‧‧‧臨界出現機率組F1(K,m)‧‧‧critical probability group
r(K,m)‧‧‧比較結果值組r(K,m)‧‧‧ comparison result value group
f2、fa‧‧‧臨界出現機率F2, fa‧‧‧ critical probability
D1‧‧‧配置D1‧‧‧ configuration
R‧‧‧總數R‧‧‧ total
T1、T2、T3、T4、T5、T6‧‧‧巡檢時間T1, T2, T3, T4, T5, T6‧‧‧ inspection time
E1、E2‧‧‧量值資料E1, E2‧‧‧ Quantitative data
‧‧‧起始平均值 ‧‧‧ starting average
‧‧‧平均值 ‧‧‧average value
σ‧‧‧標準差Σ‧‧ standard deviation
SWU 、LWU ‧‧‧預警值上限SW U , LW U ‧‧‧ upper warning limit
SWL 、LWL ‧‧‧預警值下限SW L , LW L ‧‧‧ lower limit of warning value
H‧‧‧法規上限H‧‧‧Upper limit
L‧‧‧法規下限L‧‧‧ lower limit of law
WA‧‧‧資料庫單元WA‧‧‧Database Unit
WA1‧‧‧表格WA1‧‧ Form
W31‧‧‧作業控制單元W31‧‧‧Operation Control Unit
W311‧‧‧巡檢工作排程模組W311‧‧‧ inspection work scheduling module
W32‧‧‧行動巡檢裝置W32‧‧‧ action inspection device
W33‧‧‧巡檢人員W33‧‧‧Inspectors
W61‧‧‧資料庫系統W61‧‧‧Database System
W62‧‧‧管理平台W62‧‧‧Management Platform
W71、W72‧‧‧客戶端巡檢軟體W71, W72‧‧‧ client inspection software
W81、W82‧‧‧巡檢執行者W81, W82‧‧‧ inspection executors
W91‧‧‧管理者W91‧‧‧Manager
W92‧‧‧主管者W92‧‧‧ Supervisor
W93‧‧‧改善者W93‧‧‧Improved
本案得藉由下列圖式之詳細說明，俾得更深入之瞭解：第一圖：本案第一實施例所提巡檢管理系統的示意圖；第二圖：本案第一實施例所提巡檢管理系統中具有相同位置指標的複數設備次序識別碼的示意圖；第三圖(a)：本案第一實施例所提巡檢管理系統中各對象符號的出現個數的示意圖；第三圖(b)：本案第一實施例所提巡檢管理系統中利用一 維高斯混合模型分析各對象符號的出現機率的示意圖；第四圖：本案第一實施例所提巡檢管理系統中各對象符號出現於各巡檢時間順位的出現機率的示意圖；第五圖：本案第一實施例所提巡檢管理系統中利用學習完成出現機率矩陣的流程示意圖；第六圖：本案第一實施例所提巡檢管理系統中提示使用者下一個巡查設備的流程示意圖；第七圖：本案第一實施例所提巡檢管理系統中處理巡檢到位確認的配置示意圖；第八圖：本案第一實施例所提巡檢管理系統中確認一輪巡檢是否到位的流程示意圖；第九圖：本案第一實施例所提巡檢管理系統中短期變動預警上下限值分析的示意圖；第十圖：本案第一實施例所提巡檢管理系統中長期變動預警上下限值分析的示意圖；第十一圖：本案第二實施例所提巡檢管理系統的示意圖；第十二圖：本案第三實施例所提巡檢管理系統的示意圖；第十三圖(a)：本案第四實施例所提巡檢管理系統的示意圖；及第十三圖(b)：本案第四實施例所提巡檢管理方法的作業流程示意圖。The case can be further explained by the detailed description of the following drawings: the first picture: the schematic diagram of the inspection management system proposed in the first embodiment of the case; the second picture: the inspection management of the first embodiment of the case Schematic diagram of a plurality of device order identification codes having the same position indicator in the system; FIG. 3(a) is a schematic diagram showing the number of occurrences of each object symbol in the inspection management system in the first embodiment of the present invention; : utilizing one in the inspection management system proposed in the first embodiment of the present invention Schematic diagram of analyzing the probability of occurrence of each object symbol by the Wiggs hybrid model; Fourth figure: Schematic diagram of the probability of occurrence of each object symbol appearing in each inspection time in the inspection management system of the first embodiment of the present invention; The flow chart of using the learning completion probability matrix in the patrol management system of the first embodiment of the present invention; the sixth figure: the flow chart of prompting the user to the next patrol device in the patrol management system of the first embodiment of the present invention; Figure 7: Schematic diagram of the configuration of the patrol inspection in the patrol management system mentioned in the first embodiment of the present invention; Figure 8: The flow chart of confirming whether a round of patrol inspection is in place in the patrol management system proposed in the first embodiment of the present invention; Ninth diagram: Schematic diagram of the analysis of the upper and lower limits of the short-term change warning in the inspection management system proposed in the first embodiment of the present case; the tenth figure: the analysis of the upper and lower limits of the long-term change warning in the inspection management system proposed in the first embodiment of the present case Schematic diagram; Figure 11: Schematic diagram of the inspection management system proposed in the second embodiment of the present invention; Figure 12: The inspection of the third embodiment of the present invention Schematic diagram of the management system; Figure 13 (a): schematic diagram of the inspection management system proposed in the fourth embodiment of the present invention; and thirteenth diagram (b): the operation flow of the inspection management method proposed in the fourth embodiment of the present invention schematic diagram.
91‧‧‧巡檢管理系統91‧‧‧Inspection Management System
W1‧‧‧處理單元W1‧‧‧ processing unit
A11、A12、A13、A14、A1M‧‧‧巡檢時間順位A11, A12, A13, A14, A1M‧‧‧ inspection time order
B1、B2、BN、V1‧‧‧對象符號組B1, B2, BN, V1‧‧‧ object symbol group
S1、S32、S33、S34、S35、S36、S40、S41、S80、S_NULL‧‧‧對象符號S1, S32, S33, S34, S35, S36, S40, S41, S80, S_NULL‧‧‧ object symbols
P61、P72、P83‧‧‧出現機率P61, P72, P83‧‧‧ occurrence probability
Claims (46)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW97120647A TWI443591B (en) | 2008-06-03 | 2008-06-03 | Inspection managing system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW97120647A TWI443591B (en) | 2008-06-03 | 2008-06-03 | Inspection managing system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| TW200951837A TW200951837A (en) | 2009-12-16 |
| TWI443591B true TWI443591B (en) | 2014-07-01 |
Family
ID=44871869
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| TW97120647A TWI443591B (en) | 2008-06-03 | 2008-06-03 | Inspection managing system and method |
Country Status (1)
| Country | Link |
|---|---|
| TW (1) | TWI443591B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI509539B (en) * | 2013-01-30 | 2015-11-21 | Chang Gung Medical Foundation | Method and apparatus for scheduling |
| TWI557660B (en) * | 2013-04-03 | 2016-11-11 | Plant inspection methods and systems | |
| CN111415054B (en) * | 2019-01-08 | 2023-06-27 | 昆山纬绩资通有限公司 | Equipment monitoring management method and system |
-
2008
- 2008-06-03 TW TW97120647A patent/TWI443591B/en active
Also Published As
| Publication number | Publication date |
|---|---|
| TW200951837A (en) | 2009-12-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN206921161U (en) | It is a kind of to include inspection management and defect processing and the intelligent inspection system of examination | |
| CN106598791B (en) | Industrial equipment fault preventive identification method based on machine learning | |
| WO2017084186A1 (en) | System and method for automatic monitoring and intelligent analysis of flexible circuit board manufacturing process | |
| KR101683256B1 (en) | Asset management system and method for electric power apparatus | |
| CN109840157A (en) | Method, apparatus, electronic equipment and the storage medium of fault diagnosis | |
| CN108229782A (en) | A kind of visual production management platform based on cloud | |
| US11170330B2 (en) | Workplace risk determination and scoring system and method | |
| CN205750504U (en) | A kind of management system of backlight module automatic production line | |
| CN112990870A (en) | Patrol file generation method and device based on nuclear power equipment and computer equipment | |
| CN109740771A (en) | A kind of intelligent control management system and method for responsibility assessment | |
| US20170146975A1 (en) | Operator and manufacturing quality traceability | |
| JP2019028834A (en) | Abnormal value diagnostic device, abnormal value diagnostic method, and program | |
| TWI443591B (en) | Inspection managing system and method | |
| CN109117526A (en) | One kind being suitable for mechanical system maintenance of equipment and guides data record and analysis system | |
| CN109635997A (en) | A kind of prediction technique and system on equipment maintenance opportunity | |
| CN115689415A (en) | Digital twin-based logistics monitoring and simulation system | |
| CN115952081A (en) | Software testing method, device, storage medium and equipment | |
| CN111047125B (en) | Product failure analysis apparatus, method, and computer-readable storage medium | |
| CN114416826A (en) | Equipment point inspection data statistical method, equipment point inspection data analysis method and computer storage medium | |
| US20070194920A1 (en) | Focused alarm rationalization method | |
| CN116859842A (en) | Chemical production line safety evaluation system | |
| CN116665401A (en) | Accident prevention alarm system for chemical production | |
| CN117013687A (en) | Electric power operation quality monitoring method and system | |
| CN116187983A (en) | Wind turbine generator operation and maintenance management method, device, equipment and storage medium | |
| JP2007109095A (en) | Process management apparatus, process management system, process management method, control program, and recording medium |