Claims (10)
一種電力異常偵測裝置,包含:一儲存器,用以儲存複數參考電器之複數參考電力資訊;一電力資訊擷取器,用以擷取一待測電器之一待測電力資訊,該待測電器與該等參考電器屬於相同類型;以及一處理器,電性連結至該電力資訊擷取器及該儲存器,並用以執行下列操作:界定複數電力參數類型;根據該等電力參數類型,自該等參考電力資訊建立複數參考電力事件及自該待測電力資訊建立一待測電力事件;根據該等電力參數類型界定複數電力參數變動範圍,並根據該等電力參數變動範圍分類該等參考電力事件以建立複數參考電力樣本;計算各該參考電力樣本之一第一電力參數分布;結合該待測電力事件至各該參考電力樣本,並計算各該參考電力樣本之一第二電力參數分布;分別計算各該第一電力參數分布及該第二電力參數分布之一差異;以及根據該等差異,判斷該待測電器發生異常。
A power abnormality detecting device includes: a storage device for storing a plurality of reference power information of a plurality of reference electrical appliances; and a power information extracting device for capturing a power information to be tested of the electrical device to be tested, the test is to be tested The electrical appliance is of the same type as the reference electrical appliances; and a processor electrically coupled to the power information extractor and the storage device for performing the following operations: defining a plurality of power parameter types; and according to the power parameter types, The reference power information establishes a plurality of reference power events and establishes a power event to be tested from the power information to be tested; defines a plurality of power parameter variation ranges according to the power parameter types, and classifies the reference powers according to the power parameter variation ranges The event is to establish a plurality of reference power samples; calculate a first power parameter distribution of each of the reference power samples; combine the power events to be tested to each of the reference power samples, and calculate a second power parameter distribution of each of the reference power samples; Calculating a difference between each of the first power parameter distribution and the second power parameter distribution; and root Such differences, it is determined that abnormality occurs test appliance.
如請求項1所述之電力異常偵測裝置,其中該處理器更界定該等電力參數類型之至少一為一區間型電力參數類型,該至少一區間型電力參數類型具有一容忍區間,俾減少各該第一電力參數分布及該第二電力參數分布之該差異。
The power abnormality detecting device of claim 1, wherein the processor further defines at least one of the power parameter types as an interval type power parameter type, the at least one interval type power parameter type having a tolerance interval, and reducing The difference between each of the first power parameter distribution and the second power parameter distribution.
如請求項1所述之電力異常偵測裝置,其中該處理器係根據各
該第一電力參數分布之一第一主成分方向及該第二電力參數分布之第二主成分方向之間的一偏移量計算該差異。
The power abnormality detecting device according to claim 1, wherein the processor is configured according to each
The difference is calculated by an offset between the first principal component direction of the first power parameter distribution and the second principal component direction of the second power parameter distribution.
如請求項1所述之電力異常偵測裝置,其中該處理器更界定一差異門檻值,並於該差異超過該差異門檻值時判斷該待測電器發生異常。
The power abnormality detecting device of claim 1, wherein the processor further defines a difference threshold value, and determines that the electrical appliance to be tested is abnormal when the difference exceeds the difference threshold value.
如請求項1所述之電力異常偵測裝置,其中該等電力參數類型包含用電量、電器開啟時間、電器關閉時間、用電持續時間。
The power abnormality detecting device of claim 1, wherein the power parameter types include power consumption, appliance opening time, appliance closing time, and power consumption duration.
一種用於一電力異常偵測裝置之電力異常偵測方法,該電力異常偵測裝置包含一儲存器、一電力資訊擷取器及一電性連結至該儲存器及該電力資訊擷取器之處理器,該電力異常偵測方法包含下列步驟:(a)令該儲存器儲存複數參考電器之複數參考電力資訊;(b)令該電力資訊擷取器擷取一待測電器之一待測電力資訊,該待測電器與該等參考電器屬於相同類型;(c)令該處理器界定複數電力參數類型;(d)令該處理器根據該等電力參數類型,自該等參考電力資訊建立複數參考電力事件及自該待測電力資訊建立一待測電力事件;(e)令該處理器根據該等電力參數類型界定複數電力參數變動範圍,並根據該等電力參數變動範圍分類該等參考電力事件以建立複數參考電力樣本;(f)令該處理器計算各該參考電力樣本之一第一電力參數
分布;(g)令該處理器結合該待測電力事件至各該參考電力樣本,並計算各該參考電力樣本之一第二電力參數分布;(h)令該處理器分別計算各該第一電力參數分布及該第二電力參數分布之一差異;以及(i)令該處理器根據該等差異,判斷該待測電器發生異常。
A power abnormality detecting method for a power abnormality detecting device, the power abnormality detecting device comprising a memory, a power information extractor, and an electrical connection to the storage device and the power information extractor The power abnormality detecting method includes the following steps: (a) causing the memory to store a plurality of reference power information of the plurality of reference appliances; (b) causing the power information extractor to take one of the devices to be tested to be tested Power information, the electrical device to be tested is of the same type as the reference electrical appliances; (c) causing the processor to define a plurality of power parameter types; (d) causing the processor to establish from the reference power information based on the power parameter types The plurality of reference power events and the power information to be tested establish a power event to be tested; (e) causing the processor to define a range of variation of the plurality of power parameters according to the types of the power parameters, and classifying the reference according to the range of fluctuations of the power parameters a power event to establish a plurality of reference power samples; (f) causing the processor to calculate one of the first power parameters of each of the reference power samples
(g) causing the processor to combine the power event to be tested to each of the reference power samples, and calculate a second power parameter distribution of each of the reference power samples; (h) causing the processor to calculate each of the first a difference between the power parameter distribution and the second power parameter distribution; and (i) causing the processor to determine that the electrical device under test is abnormal according to the difference.
如請求項6所述之電力異常偵測方法,其中該步驟(c)更包含下列步驟:(c1)令該處理器界定該等電力參數類型之至少一為一區間型電力參數類型,該至少一區間型電力參數類型具有一容忍區間,俾減少各該第一電力參數分布及該第二電力參數分布之該差異。
The power abnormality detecting method according to claim 6, wherein the step (c) further comprises the following steps: (c1) causing the processor to define at least one of the power parameter types as an interval type power parameter type, the at least An interval type power parameter type has a tolerance interval, and the difference between each of the first power parameter distribution and the second power parameter distribution is reduced.
如請求項6所述之電力異常偵測方法,其中該步驟(h)更包含下列步驟:(h1)令該處理器根據各該第一電力參數分布之一第一主成分方向及該第二電力參數分布之第二主成分方向之間的一偏移量計算該差異。
The power abnormality detecting method according to claim 6, wherein the step (h) further comprises the following steps: (h1) causing the processor to select one of the first principal component directions and the second according to each of the first power parameter distributions The difference between the second principal component directions of the power parameter distribution is calculated.
如請求項6所述之電力異常偵測方法,其中該步驟(i)更包含下列步驟:(i1)令該處理器界定一差異門檻值,並於該差異超過該差異門檻值時判斷該待測電器發生異常。
The power abnormality detecting method according to claim 6, wherein the step (i) further comprises the following steps: (i1) causing the processor to define a difference threshold value, and determining that the difference is when the difference exceeds the difference threshold value; An abnormality occurred in the measuring instrument.
如請求項6所述之電力異常偵測方法,其中該等電力參數類型
包含用電量、電器開啟時間、電器關閉時間、用電持續時間。
The power abnormality detecting method according to claim 6, wherein the power parameter type
It includes electricity consumption, appliance opening time, appliance shutdown time, and power consumption duration.