CN113325308B - Power supply fault detection method for data center - Google Patents
Power supply fault detection method for data center Download PDFInfo
- Publication number
- CN113325308B CN113325308B CN202110393049.8A CN202110393049A CN113325308B CN 113325308 B CN113325308 B CN 113325308B CN 202110393049 A CN202110393049 A CN 202110393049A CN 113325308 B CN113325308 B CN 113325308B
- Authority
- CN
- China
- Prior art keywords
- power supply
- result
- maintenance
- detection
- ups
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/34—Testing dynamo-electric machines
- G01R31/343—Testing dynamo-electric machines in operation
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Stand-By Power Supply Arrangements (AREA)
Abstract
The invention provides a power supply fault detection method for a data center, which comprises the following steps: fault detection is carried out on the diesel generator, and a first detection result is obtained; carrying out fault detection on a single UPS to obtain a second detection result; carrying out fault detection on the whole group of UPS to obtain a third detection result; based on the data center, and according to the first detection result, the second detection result and the third detection result, whether the power supply failure occurs in the power distribution and supply system is judged, a large amount of manpower is saved, the detection efficiency is improved, and the accuracy of the failure detection is improved through classification detection.
Description
Technical Field
The invention relates to the field of fault detection, in particular to a power supply fault detection method for a data center.
Background
And the data center provides services such as large-scale, high-quality, safe and reliable professional server hosting, space renting, network wholesale bandwidth and the like for Internet content provider enterprises, media and various websites.
The operation and maintenance of the system are vital to mastering the power supply condition of the data center, particularly the data center. How superior the function of system, how high the reliability, in case of having a power failure, the system that is good again also can't operate, data center's power supply system includes and usually includes diesel generator system, UPS group system, when single breaks down among them, as long as in time carry out the circuit switching action, guarantee the continuity of power supply and the normal operating of load, will not appear power supply failure, traditional power supply failure detects through artificial mode and detects, need consume a large amount of manpowers, detection efficiency is not high, and power supply failure detection's accuracy remains to be improved.
Disclosure of Invention
The invention provides a power supply fault detection method for a data center, which is characterized in that a diesel generator of a diesel generator system, a single UPS and a whole group of UPSs are respectively subjected to signal detection, and whether a power supply and distribution system has a fault or not is determined according to a detection result, so that a large amount of manpower is saved, the detection efficiency is improved, and the accuracy of fault detection is improved through classification detection.
A power failure detection method for a data center comprises the following steps:
step 1: fault detection is carried out on the diesel generator, and a first detection result is obtained;
step 2: carrying out fault detection on a single UPS to obtain a second detection result;
and step 3: carrying out fault detection on the whole group of UPS to obtain a third detection result;
and 4, step 4: and based on the data center, judging whether the power distribution and supply system has power supply failure according to the first detection result, the second detection result and the third detection result.
In one possible way of realisation,
in the step 1, fault detection is carried out on the diesel generator, and the step of obtaining a first detection result comprises the following steps:
detect a plurality of diesel generator that are located the diesel generator computer lab, include:
monitoring the real-time rotating speed of the diesel generator when a starting signal of the diesel generator is received, judging that the diesel generator breaks down if the rotating speed is greater than a preset rotating speed, and performing first alarm reminding;
detecting power supply continuity of a backend load of a failed diesel generator, comprising:
recording the action time and the action speed of the diesel generator with the fault and the diesel generators running in parallel to perform a first switching action to obtain a first recording result;
acquiring a first voltage signal of a load at the rear end of the diesel generator with the fault to obtain a continuous first voltage waveform, and judging whether the first voltage waveform meets a first preset waveform to obtain a first judgment result;
obtaining voltage detection values of all rear-end loads according to the voltage signals of the rear-end loads of the diesel generator with faults, and judging whether the difference of the voltage detection values is larger than a preset difference to obtain a second judgment result;
and the first recording result, the first judgment result and the second judgment result form the first detection result.
In one possible way of realisation,
in step 2, performing fault detection on a single UPS, and obtaining a second detection result includes:
detect a plurality of UPS host computers and bypass cabinets that are located the distribution room, include:
monitoring working data of the UPS host in the running state, detecting the working data based on a Gaussian mixture model, judging whether the UPS host fails, and if so, performing second alarm reminding;
when the UPS host computer breaks down, detect the power supply continuity of UPS rear end load, include:
recording the action time and the action speed of the UPS host with the fault and the UPS hosts running in parallel for second switching to obtain a second recording result;
acquiring output transient response of the UPS hosts running in parallel in the second switching process to obtain a first acquisition result;
monitoring the working state change condition of the UPS main machines which are operated in parallel before and after the second switching is carried out to obtain a first monitoring result;
acquiring a second voltage signal of the load at the rear end of the UPS host with the fault to obtain a continuous second voltage waveform, and judging whether the second voltage waveform meets a second preset waveform to obtain a third judgment result;
and the second recording result, the first acquisition result, the first monitoring result and the third judgment result form the second detection result.
In one possible way of realisation,
in step 3, performing fault detection on the entire group of UPSs of the uninterruptible power supply system to obtain a third detection result, including:
detect a plurality of UPS host computers and bypass cabinets that are located the distribution room, include:
monitoring the whole working data of the whole UPS in the running state, detecting the whole working data based on a Gaussian mixture model, judging whether the whole UPS fails, and if so, performing third alarm reminding;
if the whole group of UPS breaks down, the instant switching capacity of the UPS parallel operation system and the power supply continuity of the load at the rear end of the whole group of UPS are detected, and the method comprises the following steps:
recording the action time and the action speed of the UPS parallel operation system for performing third switching to obtain a third recording result;
collecting a third voltage signal of the whole group of UPS rear end loads to obtain a continuous third voltage waveform, and judging whether the third voltage waveform meets a third preset waveform to obtain a fourth judgment result;
and the third recording result and the fourth judging result form the third detection result.
In one possible way of realisation of the invention,
in step 4, based on the data center, and according to the first detection result, the second detection result, and the third detection result, determining whether the power supply failure occurs in the power supply and distribution system includes:
confirming a first preset result requirement in the first detection result, if so, determining that the power supply and distribution system has a power supply fault, and determining that the type of the power supply fault is a power supply fault of the diesel generator, otherwise, determining that the diesel generator has no power supply fault;
wherein the first preset result requirement is as follows: the first recorded result represents that the first switching action is accurate, the second judgment result represents that the first voltage waveform meets a first preset waveform, and the second judgment result represents that the voltage detection value is smaller than a preset difference;
confirming a second preset result requirement in the second detection result, if so, determining that the power supply failure occurs in the power supply and distribution system, and the type of the power supply failure is a single UPS power supply failure, otherwise, determining that the single UPS does not have the power supply failure;
wherein the second preset result requirement is: the second recorded result shows that the second switching action is accurate, the output response parameter of the first acquisition result is qualified, the first monitoring result shows that the working state change of the UPS host machine operated by other parallel machines is qualified, and the third judgment result shows that the second voltage waveform meets a second preset waveform;
confirming a third preset result requirement in the third detection result, if so, determining that the power supply failure occurs in the power supply and distribution system, and the type of the power supply failure is the power supply failure of the whole group of UPS, otherwise, determining that the power supply failure does not exist in the whole group of UPS;
wherein the third preset result requirement is: the third recorded result shows that the third switching action is accurate, and the fourth judgment result shows that the third voltage waveform meets a third preset waveform.
In one possible way of realisation of the invention,
further comprising: when the power supply and distribution system has power supply faults, an optimal maintenance scheme is selected for the data center, and the optimal maintenance scheme comprises the following steps:
analyzing the first detection result, the second detection result and the third detection result to obtain a fault type of the power supply and distribution system with power supply faults;
selecting an available maintenance scheme from a maintenance database based on the fault type, wherein the maintenance scheme comprises three maintenance elements, namely a fault maintenance person, a fault repair strategy and a resource for fault maintenance;
extracting features of the three maintenance elements of the maintenance scheme to obtain a first feature vector, establishing an element matrix based on the first feature vector, and establishing a first weighting matrix based on a first preset rule;
weighting the element matrix by using the first weighting matrix to obtain a weighted element matrix;
based on the maintenance scheme, searching historical maintenance data maintained by the maintenance scheme;
determining four evaluation indexes of average investigation time, average repair cost and average repair quality under the repair scheme based on the historical repair data;
performing feature extraction on the four evaluation indexes of the maintenance scheme to obtain a second feature vector, establishing an evaluation matrix based on the second feature vector, and establishing a second weighting matrix based on a second preset rule;
weighting the evaluation matrix by using the second weighting matrix to obtain a weighted evaluation matrix;
after normalization processing is carried out on the weighting factor matrix and the weighting evaluation matrix, fusion is carried out based on a maintenance scheme to obtain a fusion matrix;
carrying out fuzzy grade division on the three maintenance factors and the four evaluation indexes, determining the quantitative corresponding relation between the maintenance factors and the evaluation indexes and the fuzzy number, and determining the influence factors of the fusion matrix according to the corresponding relation;
modifying the fusion matrix based on the influence factors to obtain a modified matrix;
and multiplying the correction matrix by a preset matrix to obtain a decision matrix, confirming the row number of the element with the largest value in the decision matrix, and confirming a maintenance scheme corresponding to the row number in the correction matrix as an optimal maintenance scheme.
In one possible way of realisation of the invention,
in step 1, fault detection is performed on the diesel generator, and the method further comprises the following steps: temperature detection is performed on the diesel generator, and the temperature detection method comprises the following steps:
acquiring a first temperature value of the diesel generator in an operating state and a second temperature value of the diesel generator in a non-operating state at intervals of preset time;
acquiring a third temperature value of a power supply room in the running state of the diesel generator and a fourth temperature value of the power supply room in the non-running state of the diesel generator at intervals of preset time;
calculating the difference value between the first temperature value and the third temperature value to obtain a first difference value curve, and calculating the difference value between the second temperature value and the fourth temperature value to obtain a second difference value curve;
based on the second difference curve, correcting the first difference curve to obtain a corrected third difference curve;
correcting the first temperature value based on the third difference curve and the third temperature value to obtain a fourth temperature value;
judging whether the fourth temperature value is greater than a preset temperature value or not;
if so, controlling the diesel generator to operate;
otherwise, controlling the diesel generator to stop running, starting the standby diesel generator to start running, and performing fourth alarm reminding.
In one possible way of realisation of the invention,
further comprising: the method comprises the steps of evaluating equipment of a data center, and maintaining the equipment regularly according to an evaluation result, wherein the method comprises the following steps:
inquiring the service condition of the equipment to obtain the running time and the using time of the equipment;
inquiring the frequency of equipment failure of the equipment and the maintenance cost of each maintenance;
calculating the loss rate of the device according to the following formula:
where σ denotes the loss ratio of the device, α 0 Representing the rate of tangible loss, alpha, that the device cannot eliminate by maintenance 1 Representing the apparent rate of wear, a, that the device can eliminate by maintenance 2 Representing the intangible loss ratio, T, of the device 0 Represents the operating time of the plant, T a Representing the time length of the equipment in use, e representing a natural number, and n representing the number of times of equipment failure of the equipment;
calculating a maintenance period of the equipment according to the following formula based on the loss rate of the equipment;
wherein T represents a maintenance period of the device, T S Represents a preset maintenance period of the equipment, m i Represents the maintenance cost m of the ith maintenance of the equipment 0 Representing a preset maintenance cost of the equipment;
maintaining the equipment according to the maintenance period of the equipment, and after the equipment is maintained, updating the maintenance period of the equipment according to the following formula to obtain the next maintenance period of the equipment:
wherein, T N Represents the next maintenance cycle, σ, of the plant n Representing the latest loss rate;
and determining the next maintenance time of the equipment based on the next maintenance period of the equipment, and reminding maintenance.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:
fig. 1 is a flowchart of a power failure detection method for a data center according to an embodiment of the present invention.
Detailed Description
The preferred embodiments of the present invention will be described in conjunction with the accompanying drawings, and it will be understood that they are described herein for the purpose of illustration and explanation and not limitation.
Example 1
An embodiment of the present invention provides a power supply failure detection method for a data center, as shown in fig. 1, including:
step 1: fault detection is carried out on the diesel generator, and a first detection result is obtained;
step 2: carrying out fault detection on a single UPS to obtain a second detection result;
and step 3: carrying out fault detection on the whole group of UPS to obtain a third detection result;
and 4, step 4: and judging whether the power supply failure occurs in the power distribution and supply system based on the data center according to the first detection result, the second detection result and the third detection result.
The beneficial effect of above-mentioned design is: the signal detection is carried out on the diesel generator of the diesel generator system, the single UPS and the UPS of the uninterrupted power supply system respectively, whether the power supply and distribution system breaks down or not is determined according to the detection result, a large amount of manpower is saved, the detection efficiency is improved, and the accuracy of fault detection is improved through classification detection.
Example 2
Based on embodiment 1, an embodiment of the present invention provides a power supply failure detection method for a data center, where in step 1, performing failure detection on a diesel generator of a diesel generator system, and obtaining a first detection result includes:
in the step 1, fault detection is carried out on the diesel generator, and the step of obtaining a first detection result comprises the following steps:
detect a plurality of diesel generator that are located the diesel generator room, include:
monitoring the real-time rotating speed of the diesel generator when a starting signal of the diesel generator is received, judging that the diesel generator breaks down if the rotating speed is greater than a preset rotating speed, and performing first alarm reminding;
detecting power supply continuity of a backend load of a failed diesel generator, comprising:
recording the action time and the action speed of the diesel generator with the fault and the diesel generators running in parallel to perform a first switching action to obtain a first recording result;
acquiring a first voltage signal of a load at the rear end of the diesel generator with the fault to obtain a continuous first voltage waveform, and judging whether the first voltage waveform meets a first preset waveform to obtain a first judgment result;
obtaining voltage detection values of all rear-end loads according to the voltage signals of the rear-end loads of the diesel generator with faults, and judging whether the difference of the voltage detection values is larger than a preset difference to obtain a second judgment result;
and the first recording result, the first judgment result and the second judgment result form the first detection result.
The working principle of the design scheme is as follows: firstly, fault detection is carried out on a single diesel generator, if a fault occurs, switching actions of the diesel generator and other diesel generators running in parallel and a first voltage signal of a load at the rear end of the diesel generator are continuously detected, and a first detection result is obtained.
The beneficial effect of above-mentioned design is: the switching action of the diesel engine power generation system and the first voltage signal of the load at the rear end of the diesel generator are detected, so that the manual detection is avoided, and the detection efficiency of the diesel generator is improved.
Example 3
Based on embodiment 1, an embodiment of the present invention provides a power supply failure detection method for a data center, where in step 2, performing failure detection on a single UPS, and obtaining a second detection result includes:
detect a plurality of UPS host computers and bypass cabinets that are located the distribution room, include:
monitoring working data of the UPS host in the running state, detecting the working data based on a Gaussian mixture model, judging whether the UPS host fails, and if so, performing second alarm reminding;
when the UPS host computer breaks down, detect the power supply continuity of the UPS backend load, including:
recording the action time and the action speed of the second switching between the UPS host with the fault and the UPS hosts running in other parallel machines to obtain a second recording result;
acquiring output transient response of the UPS hosts running in parallel in the second switching process to obtain a first acquisition result;
monitoring the working state change condition of the UPS main machines running in parallel before and after the second switching to obtain a first monitoring result;
acquiring a second voltage signal of the load at the rear end of the UPS host with the fault to obtain a continuous second voltage waveform, and judging whether the second voltage waveform meets a second preset waveform to obtain a third judgment result;
and the second recording result, the first acquisition result, the first monitoring result and the third judgment result form the second detection result.
In this embodiment, detecting the working data based on the gaussian mixture model means accurately quantizing the working data in the operation state of the UPS host by using a gaussian probability density function, and determining a fault condition according to a quantization result.
In this embodiment, the output transient response of the other parallel-running UPS hosts includes: voltage, current, frequency, zero ground voltage, current harmonics, voltage harmonics, power factor, etc.
The working principle of the design scheme is as follows: firstly, fault detection is carried out on a single UPS host, if faults occur, switching actions of the UPS host and other UPS hosts running in parallel, working state change conditions of the UPS hosts running in parallel and a first voltage signal of a load at the rear end of the UPS host are continuously detected, and a second detection result is obtained.
The beneficial effect of above-mentioned design is: through detecting the working state change condition of the UPS host machine for switching actions and parallel operation of the UPS host machine and the second voltage signal of the load at the rear end of the UPS host machine, the detection in a manual mode is avoided, and the detection efficiency of the UPS host machine is improved.
The working principle of the design scheme is as follows: the running change states of the UPS hosts running in other parallel machines are better reflected by collecting the voltage, the current, the frequency, the zero ground voltage, the current harmonic wave, the voltage harmonic wave and the power factor of the UPS hosts running in other parallel machines, and the detection accuracy is improved.
Example 4
Based on embodiment 1, an embodiment of the present invention provides a power supply failure detection method for a data center, where in step 3, performing failure detection on an entire UPS group of an uninterruptible power supply system, and obtaining a third detection result includes:
detect a plurality of UPS host computers and bypass cabinets that are located the distribution room, include:
monitoring the whole working data of the whole UPS in the running state, detecting the whole working data based on a Gaussian mixture model, judging whether the whole UPS fails, and if so, performing third alarm reminding;
if the whole group of UPS breaks down, the instant switching capacity of the UPS parallel operation system and the power supply continuity of the load at the rear end of the whole group of UPS are detected, and the method comprises the following steps:
recording the action time and the action speed of the UPS parallel operation system for performing third switching to obtain a third recording result;
acquiring a third voltage signal of the whole group of UPS rear-end loads to obtain a continuous third voltage waveform, and judging whether the third voltage waveform meets a third preset waveform to obtain a fourth judgment result;
and the third recording result and the fourth judging result form the third detection result.
In this embodiment, detecting the overall working data based on the gaussian mixture model means accurately quantizing the working data of the entire group of UPS in the operating state by using a gaussian probability density function, and determining a fault condition according to a quantization result.
In this embodiment, the UP parallel machine system includes a plurality of full sets of UPSs.
The working principle of the design scheme is as follows: firstly, fault detection is carried out on the entire UPS, if a fault occurs, the switching action of the UP parallel machine system and a third voltage signal of a load at the rear end of the entire UPS are continuously detected, and a third detection result is obtained.
The beneficial effect of above-mentioned design is: through the switching action of the UP parallel machine system and the detection of the third voltage signal of the rear-end load of the whole group of UPS, the detection in a manual mode is avoided, and the detection efficiency of the whole group of UPS is improved.
Example 5
Based on embodiments 1 to 5, an embodiment of the present invention provides a power supply failure detection method for a data center, where in step 4, based on the data center, and according to the first detection result, the second detection result, and the third detection result, determining whether a power supply failure occurs in a power supply and distribution system includes:
confirming a first preset result requirement in the first detection result, if so, determining that the power supply and distribution system has a power supply fault, and determining that the type of the power supply fault is a power supply fault of the diesel generator, otherwise, determining that the diesel generator has no power supply fault;
wherein the first preset result requirement is as follows: the first recorded result shows that the first switching action is accurate, the second judgment result shows that the first voltage waveform meets a first preset waveform, and the second judgment result shows that the voltage detection value is smaller than a preset difference;
confirming a second preset result requirement in the second detection result, if so, determining that the power supply failure occurs in the power supply and distribution system, and the type of the power supply failure is a single UPS power supply failure, otherwise, determining that the single UPS does not have the power supply failure;
wherein the second preset result requirement is: the second recorded result indicates that the second switching action is accurate, the output response parameter of the first acquisition result is qualified, the first monitoring result indicates that the working state change of the other parallel operation UPS hosts is qualified, and the third judgment result indicates that the second voltage waveform meets a second preset waveform;
confirming a third preset result requirement in the third detection result, if so, determining that the power supply failure occurs in the power supply and distribution system, and the type of the power supply failure is the power supply failure of the whole group of UPSs, otherwise, determining that the power supply failure does not exist in the whole group of UPSs;
wherein the third preset result requirement is: the third recorded result shows that the third switching action is accurate, and the fourth judgment result shows that the third voltage waveform meets a third preset waveform.
The beneficial effect of above-mentioned design is: the first detection result, the second detection result and the third detection result are compared with the corresponding first preset result requirement, the second preset result requirement and the third preset result requirement respectively, and the power supply continuity of the corresponding rear-end load is judged, so that whether a power supply fault occurs in the data center or not and the type of the power supply fault are determined, and the accuracy of detecting the power supply fault of the data center is improved.
Example 6
Based on embodiment 1, an embodiment of the present invention provides a power supply failure detection method for a data center, further including: further comprising: when the power supply and distribution system has power supply faults, an optimal maintenance scheme is selected for the data center, and the optimal maintenance scheme comprises the following steps:
analyzing the first detection result, the second detection result and the third detection result to obtain a fault type of the power supply and distribution system with power supply faults;
based on the fault type, selecting an available maintenance scheme from a maintenance database, wherein the maintenance scheme comprises three maintenance elements, namely a fault maintenance person, a fault repair strategy and a resource for fault maintenance;
extracting the features of the three maintenance factors of the maintenance scheme to obtain a first feature vector, establishing a factor matrix based on the first feature vector, and establishing a first weighting matrix based on a first preset rule;
weighting the element matrix by using the first weighting matrix to obtain a weighted element matrix;
based on the maintenance scheme, searching historical maintenance data for maintenance by using the maintenance scheme;
determining four evaluation indexes of average investigation time, average repair cost and average repair quality under the repair scheme based on the historical repair data;
performing feature extraction on the four evaluation indexes of the maintenance scheme to obtain a second feature vector, establishing an evaluation matrix based on the second feature vector, and establishing a second weighting matrix based on a second preset rule;
weighting the evaluation matrix by using the second weighting matrix to obtain a weighted evaluation matrix;
after normalization processing is carried out on the weighting factor matrix and the weighting evaluation matrix, fusion is carried out based on a maintenance scheme to obtain a fusion matrix;
carrying out fuzzy grade division on the three maintenance factors and the four evaluation indexes, determining the quantitative corresponding relation between the maintenance factors and the evaluation indexes and the fuzzy number, and determining the influence factors of the fusion matrix according to the corresponding relation;
modifying the fusion matrix based on the influence factors to obtain a modified matrix;
and multiplying the correction matrix with a preset matrix to obtain a decision matrix, confirming the row number of the element with the largest numerical value in the decision matrix, and confirming a maintenance scheme corresponding to the row number in the correction matrix as an optimal maintenance scheme.
In this embodiment, the first preset rule is to quantify the three maintenance elements according to the merits and the demerits according to the actual situation, and the obtained first weighting matrix reflects the relative preference of the maintenance scheme based on the maintenance elements.
In this embodiment, the second preset rule is that the four evaluation indexes are quantified according to the merits and the demerits according to the actual situation, and the obtained evaluation matrix reflects the relative preference of the maintenance scheme based on the evaluation indexes.
In this embodiment, the row of the correction matrix represents quantized values of a maintenance staff, a failure repair strategy, a resource for failure repair, an average troubleshooting time, an average repair cost, and an average repair quality of one maintenance plan, and the column is a quantized value characterized by the maintenance staff, the failure repair strategy, the resource for failure repair, the average troubleshooting time, the average repair cost, and the average repair quality, respectively.
In this embodiment, the basis for performing fuzzy grading on the three maintenance factors and the four evaluation indexes is to perform judgment and ranking on a fault maintenance worker, a fault repair strategy, a resource for fault maintenance, average troubleshooting time, average repair time, average maintenance cost, and average maintenance quality factor, and the judgment basis is the importance of the judgment on a maintenance scheme.
In this embodiment, the number of rows of the preset matrix is the number of columns of the correction matrix, the number of columns is 1, and the number of elements in the preset matrix is 1.
The working principle of the design scheme is as follows: the selectable maintenance scheme is determined according to the fault detection type, the maintenance scheme is analyzed from the aspects of elements and evaluation index parameters, the optimal maintenance scheme is determined, the maintenance quality is guaranteed, and the waste of manpower is reduced.
Example 7
Based on embodiment 1, a power supply fault detection method for a data center is characterized in that, in step 1, fault detection is performed on a diesel generator, and the method further includes: temperature detection is performed on the diesel generator, and the temperature detection method comprises the following steps:
acquiring a first temperature value of the diesel generator in an operating state and a second temperature value of the diesel generator in a non-operating state at intervals of preset time;
acquiring a third temperature value of a power supply room in the running state of the diesel generator and a fourth temperature value of the power supply room in the non-running state of the diesel generator at intervals of preset time;
calculating the difference value between the first temperature value and the third temperature value to obtain a first difference value curve, and calculating the difference value between the second temperature value and the fourth temperature value to obtain a second difference value curve;
based on the second difference curve, correcting the first difference curve to obtain a corrected third difference curve;
correcting the first temperature value based on the third difference curve and the third temperature value to obtain a fourth temperature value;
judging whether the fourth temperature value is greater than a preset temperature value or not;
if so, controlling the diesel generator to operate;
otherwise, controlling the diesel generator to stop running, starting the standby diesel generator to start running, and performing fourth alarm reminding.
In this embodiment, the second difference curve is obtained in a non-operating state of the diesel generator, the first difference curve is corrected based on the second difference curve to obtain a third difference curve, and the first temperature value is corrected based on the third difference curve and a third temperature value, so that the influence of the temperature value of the power supply room on the temperature value of the diesel generator in the operating state can be avoided.
The beneficial effect of above-mentioned design is: through gathering diesel generator is at the temperature value under the running state and revises the temperature value after revising is greater than when predetermineeing the temperature value, control diesel generator stop operation starts reserve diesel generator and begins to operate to carry out the fourth warning, avoided diesel generator breaks down because of the high temperature to avoid the manpower that consumes because of fault detection, and start reserve diesel generator, guaranteed whole power supply and distribution system's normal operating.
Example 8
Based on embodiment 1, an embodiment of the present invention provides a power failure detection method for a data center, further including: the method comprises the steps of evaluating equipment of a data center, and maintaining the equipment regularly according to an evaluation result, wherein the method comprises the following steps:
inquiring the service condition of the equipment to obtain the running time and the using time of the equipment;
inquiring the frequency of equipment failure of the equipment and the maintenance cost of each maintenance;
calculating the loss rate of the device according to the following formula:
where σ denotes the loss ratio of the device, α 0 Indicating a tangible rate of loss, a, that the device cannot be removed by maintenance 1 Representing the apparent rate of wear, a, that the device can eliminate by maintenance 2 Representing the intangible loss ratio, T, of the device 0 Represents the operating time of the plant, T a The time length of the equipment which is put into use is shown, e is a natural number, and n is the frequency of equipment failure of the equipment;
calculating a maintenance period of the equipment according to the following formula based on the loss rate of the equipment;
wherein T represents a maintenance period of the device, T S Represents a preset maintenance period of the equipment, m i Represents the maintenance cost of the equipment when the ith maintenance is carried out, m 0 Representing a preset maintenance cost of the equipment;
maintaining the equipment according to the maintenance period of the equipment, and after the equipment is maintained, updating the maintenance period of the equipment according to the following formula to obtain the next maintenance period of the equipment:
wherein, T N Represents the next maintenance cycle, σ, of the plant n Representing the latest loss rate;
and determining the next maintenance time of the equipment based on the next maintenance period of the equipment, and reminding maintenance.
In this embodiment, the equipment of the data center includes a diesel generator, a UPS host, and other power supply equipment.
In this embodiment, as time passes, the number of failures of the device increases, so the loss rate changes, and the latest loss rate is obtained from the number of failures of the latest device.
In this embodiment, since the wear rate of the equipment may change from moment to moment, the maintenance cycle of the equipment needs to be updated.
The beneficial effect of above-mentioned design is: the equipment is regularly maintained by calculating the loss rate of the equipment in the data center, so that the frequency of equipment failure is reduced, and finally, the frequency of power supply failure of the data center is reduced.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (4)
1. A power failure detection method for a data center is characterized by comprising the following steps:
step 1: fault detection is carried out on the diesel generator, and a first detection result is obtained;
and 2, step: carrying out fault detection on a single UPS to obtain a second detection result;
and step 3: carrying out fault detection on the whole group of UPSs to obtain a third detection result;
and 4, step 4: based on the data center, judging whether the power supply failure occurs in the power distribution and supply system according to the first detection result, the second detection result and the third detection result;
in step 1, fault detection is performed on the diesel generator, and obtaining a first detection result includes:
detect a plurality of diesel generator that are located the diesel generator room, include:
monitoring the real-time rotating speed of the diesel generator when a starting signal of the diesel generator is received, judging that the diesel generator breaks down if the rotating speed is greater than a preset rotating speed, and performing first alarm reminding;
detecting power supply continuity of a backend load of a failed diesel generator, comprising:
recording the action time and the action speed of the first switching action of the diesel generator with the fault and the diesel generators running in parallel to obtain a first recording result;
acquiring a first voltage signal of a load at the rear end of the diesel generator with the fault to obtain a continuous first voltage waveform, and judging whether the first voltage waveform meets a first preset waveform to obtain a first judgment result;
obtaining voltage detection values of all rear-end loads according to the voltage signals of the rear-end loads of the diesel generator with faults, and judging whether the difference of the voltage detection values is larger than a preset difference to obtain a second judgment result;
the first recording result, the first judgment result and the second judgment result form the first detection result;
in step 2, performing fault detection on a single UPS, and obtaining a second detection result includes:
detect a plurality of UPS host computers and bypass cabinets that are located the distribution room, include:
monitoring working data of the UPS host in the running state, detecting the working data based on a Gaussian mixture model, judging whether the UPS host fails, and if so, performing second alarm reminding;
when the UPS host computer breaks down, detect the power supply continuity of UPS rear end load, include:
recording the action time and the action speed of the UPS host with the fault and the UPS hosts running in parallel for second switching to obtain a second recording result;
acquiring output transient response of the UPS host machines running in parallel in the second switching process to obtain a first acquisition result;
monitoring the working state change condition of the UPS main machines running in parallel before and after the second switching to obtain a first monitoring result;
acquiring a second voltage signal of the load at the rear end of the UPS host with the fault to obtain a continuous second voltage waveform, and judging whether the second voltage waveform meets a second preset waveform to obtain a third judgment result;
the second recording result, the first acquisition result, the first monitoring result and the third judgment result form a second detection result;
in step 3, performing fault detection on the entire group of UPSs, and obtaining a third detection result includes:
detect a plurality of UPS host computers and bypass cabinets that are located the distribution room, include:
monitoring the whole working data of the whole group of UPS in the running state, detecting the whole working data based on a Gaussian mixture model, judging whether the whole group of UPS fails, and if so, performing third alarm reminding;
if the whole group of UPS breaks down, the instant switching capacity of the UPS parallel operation system and the power supply continuity of the load at the rear end of the whole group of UPS are detected, and the method comprises the following steps:
recording the action time and the action speed of the UPS parallel operation system for third switching to obtain a third recording result;
acquiring a third voltage signal of the whole group of UPS rear-end loads to obtain a continuous third voltage waveform, and judging whether the third voltage waveform meets a third preset waveform to obtain a fourth judgment result;
the third recording result and the fourth judging result form a third detection result;
in step 4, based on the data center, and according to the first detection result, the second detection result, and the third detection result, determining whether the power supply failure occurs in the power supply and distribution system includes:
confirming a first preset result requirement in the first detection result, if so, determining that the power supply and distribution system has a power supply fault, and determining that the type of the power supply fault is a power supply fault of a diesel generator, otherwise, determining that the diesel generator does not have the power supply fault;
wherein the first preset result requirement is as follows: the first recorded result represents that the first switching action is accurate, the second judgment result represents that the first voltage waveform meets a first preset waveform, and the second judgment result represents that the voltage detection value is smaller than a preset difference;
confirming a second preset result requirement in the second detection result, if so, determining that the power supply failure occurs in the power supply and distribution system, and the type of the power supply failure is a single UPS power supply failure, otherwise, determining that the single UPS does not have the power supply failure;
wherein the second preset result requirement is: the second recorded result shows that the second switching action is accurate, the output response parameter of the first acquisition result is qualified, the first monitoring result shows that the working state change of the UPS host machine operated by other parallel machines is qualified, and the third judgment result shows that the second voltage waveform meets a second preset waveform;
confirming a third preset result requirement in the third detection result, if so, determining that the power supply failure occurs in the power supply and distribution system, and the type of the power supply failure is the power supply failure of the whole group of UPSs, otherwise, determining that the power supply failure does not exist in the whole group of UPSs;
wherein the third preset result requirement is that: the third recorded result shows that the third switching action is accurate, and the fourth judgment result shows that the third voltage waveform meets a third preset waveform.
2. The method of claim 1, further comprising: when the power supply and distribution system has power supply faults, an optimal maintenance scheme is selected for the data center, and the optimal maintenance scheme comprises the following steps:
analyzing the first detection result, the second detection result and the third detection result to obtain the fault type of the power supply fault of the power supply and distribution system;
based on the fault type, selecting an available maintenance scheme from a maintenance database, wherein the maintenance scheme comprises three maintenance elements, namely a fault maintenance person, a fault repair strategy and a resource for fault maintenance;
extracting the features of the three maintenance factors of the maintenance scheme to obtain a first feature vector, establishing a factor matrix based on the first feature vector, and establishing a first weighting matrix based on a first preset rule;
weighting the element matrix by using the first weighting matrix to obtain a weighted element matrix;
based on the maintenance scheme, searching historical maintenance data for maintenance by using the maintenance scheme;
determining four evaluation indexes of average investigation time, average repair cost and average repair quality under the repair scheme based on the historical repair data;
performing feature extraction on the four evaluation indexes of the maintenance scheme to obtain a second feature vector, establishing an evaluation matrix based on the second feature vector, and establishing a second weighting matrix based on a second preset rule;
weighting the evaluation matrix by using the second weighting matrix to obtain a weighted evaluation matrix;
after normalization processing is carried out on the weighting factor matrix and the weighting evaluation matrix, fusion is carried out based on a maintenance scheme to obtain a fusion matrix;
carrying out fuzzy grade division on the three maintenance factors and the four evaluation indexes, determining the quantitative corresponding relation between the maintenance factors and the evaluation indexes and the fuzzy number, and determining the influence factors of the fusion matrix according to the corresponding relation;
modifying the fusion matrix based on the influence factors to obtain a modified matrix;
and multiplying the correction matrix by a preset matrix to obtain a decision matrix, confirming the row number of the element with the largest value in the decision matrix, and confirming a maintenance scheme corresponding to the row number in the correction matrix as an optimal maintenance scheme.
3. The method for detecting the power supply fault of the data center according to claim 1, wherein the fault detection of the diesel generator in step 1 further comprises: temperature detection is performed on the diesel generator, and the temperature detection method comprises the following steps:
acquiring a first temperature value of the diesel generator in an operating state and a second temperature value of the diesel generator in a non-operating state at intervals of preset time;
acquiring a third temperature value of a power supply room in the running state of the diesel generator and a fourth temperature value of the power supply room in the non-running state of the diesel generator at intervals of preset time;
calculating the difference value between the first temperature value and the third temperature value to obtain a first difference value curve, and calculating the difference value between the second temperature value and the fourth temperature value to obtain a second difference value curve;
based on the second difference curve, correcting the first difference curve to obtain a corrected third difference curve;
correcting the first temperature value based on the third difference curve and a third temperature value to obtain a fourth temperature value;
judging whether the fourth temperature value is greater than a preset temperature value or not;
if so, controlling the diesel generator to operate;
otherwise, controlling the diesel generator to stop running, starting the standby diesel generator to start running, and performing fourth alarm reminding.
4. The method of claim 1, further comprising: the method comprises the following steps of evaluating equipment of a data center, and maintaining the equipment regularly according to an evaluation result, wherein the method comprises the following steps:
inquiring the service condition of the equipment to obtain the running duration and the using duration of the equipment;
inquiring the frequency of equipment failure of the equipment and the maintenance cost of each maintenance;
calculating the loss rate of the device according to the following formula:
where σ denotes the loss factor of the device, α 0 Indicating a tangible rate of loss, a, that the device cannot be removed by maintenance 1 Representing the apparent rate of wear, a, that the device can eliminate by maintenance 2 Represents the aboveIntangible loss rate, T, of the device 0 Represents the operating time of the plant, T a Representing the time length of the equipment in use, e representing a natural number, and n representing the number of times of equipment failure of the equipment;
calculating a maintenance period of the equipment according to the following formula based on the loss rate of the equipment;
wherein T represents a maintenance period of the device, T S Represents a preset maintenance period, m, of the equipment i Represents the maintenance cost of the equipment when the ith maintenance is carried out, m 0 Representing a preset maintenance cost of the equipment;
maintaining the equipment according to the maintenance period of the equipment, and after the equipment is maintained, updating the maintenance period of the equipment according to the following formula to obtain the next maintenance period of the equipment:
wherein, T N Represents the next maintenance cycle, σ, of the plant n Representing the latest loss rate;
and determining the next maintenance time of the equipment based on the next maintenance period of the equipment, and reminding maintenance.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110393049.8A CN113325308B (en) | 2021-04-13 | 2021-04-13 | Power supply fault detection method for data center |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110393049.8A CN113325308B (en) | 2021-04-13 | 2021-04-13 | Power supply fault detection method for data center |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113325308A CN113325308A (en) | 2021-08-31 |
| CN113325308B true CN113325308B (en) | 2023-02-17 |
Family
ID=77414761
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110393049.8A Active CN113325308B (en) | 2021-04-13 | 2021-04-13 | Power supply fault detection method for data center |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113325308B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113687259B (en) * | 2021-09-23 | 2023-11-24 | 北京中大科慧科技发展有限公司 | Machine room UPS detection method and system for data center |
| CN115167273A (en) * | 2022-09-05 | 2022-10-11 | 中建安装集团有限公司 | Construction power consumption energy-saving management control system on construction site |
| CN116456297B (en) * | 2023-06-20 | 2023-10-17 | 国网浙江省电力有限公司宁波供电公司 | Data acquisition method based on 5G network |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0749712A (en) * | 1993-08-05 | 1995-02-21 | Toshiba Corp | Equipment maintenance managing device |
| CN101718825A (en) * | 2009-12-18 | 2010-06-02 | 北京世纪互联宽带数据中心有限公司 | Testing method of low-voltage power supply and distribution system of data center |
| JP2010166702A (en) * | 2009-01-15 | 2010-07-29 | Toshiba Corp | Power supply risk evaluation system of power facility |
| CN102621967A (en) * | 2012-04-10 | 2012-08-01 | 上海科泰电源股份有限公司 | Centralized monitoring system of intelligent power supply electric cubic of data machine room |
| CN107317718A (en) * | 2016-08-19 | 2017-11-03 | 北京市燃气集团有限责任公司 | A kind of O＆M service management and management platform |
| CN110782157A (en) * | 2019-10-24 | 2020-02-11 | 华电邹县发电有限公司 | Maintenance mode making method based on importance of power generation equipment |
| CN110988563A (en) * | 2019-12-23 | 2020-04-10 | 厦门理工学院 | UPS (uninterrupted Power supply) fault detection method, device, equipment and storage medium |
| CN111609883A (en) * | 2020-05-20 | 2020-09-01 | 合肥惠科达信息科技有限责任公司 | Communication machine room protection monitoring management system based on big data |
| CN211403213U (en) * | 2020-03-19 | 2020-09-01 | 上海科泰电源股份有限公司 | Centralized monitoring control device for diesel generator set of data center |
| CN112462260A (en) * | 2020-11-24 | 2021-03-09 | 山西中云智谷数据科技有限责任公司 | System and method for testing on-load performance of diesel generator |
-
2021
- 2021-04-13 CN CN202110393049.8A patent/CN113325308B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0749712A (en) * | 1993-08-05 | 1995-02-21 | Toshiba Corp | Equipment maintenance managing device |
| JP2010166702A (en) * | 2009-01-15 | 2010-07-29 | Toshiba Corp | Power supply risk evaluation system of power facility |
| CN101718825A (en) * | 2009-12-18 | 2010-06-02 | 北京世纪互联宽带数据中心有限公司 | Testing method of low-voltage power supply and distribution system of data center |
| CN102621967A (en) * | 2012-04-10 | 2012-08-01 | 上海科泰电源股份有限公司 | Centralized monitoring system of intelligent power supply electric cubic of data machine room |
| CN107317718A (en) * | 2016-08-19 | 2017-11-03 | 北京市燃气集团有限责任公司 | A kind of O＆M service management and management platform |
| CN110782157A (en) * | 2019-10-24 | 2020-02-11 | 华电邹县发电有限公司 | Maintenance mode making method based on importance of power generation equipment |
| CN110988563A (en) * | 2019-12-23 | 2020-04-10 | 厦门理工学院 | UPS (uninterrupted Power supply) fault detection method, device, equipment and storage medium |
| CN211403213U (en) * | 2020-03-19 | 2020-09-01 | 上海科泰电源股份有限公司 | Centralized monitoring control device for diesel generator set of data center |
| CN111609883A (en) * | 2020-05-20 | 2020-09-01 | 合肥惠科达信息科技有限责任公司 | Communication machine room protection monitoring management system based on big data |
| CN112462260A (en) * | 2020-11-24 | 2021-03-09 | 山西中云智谷数据科技有限责任公司 | System and method for testing on-load performance of diesel generator |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113325308A (en) | 2021-08-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113325308B (en) | Power supply fault detection method for data center | |
| EP3557819B1 (en) | Server failure detection method and system | |
| CN106154209B (en) | Electrical energy meter fault prediction technique based on decision Tree algorithms | |
| CN108847682B (en) | Photovoltaic power station grid-connected performance evaluation method and system based on operation data | |
| CN114793018A (en) | Electrical intelligent data processing device for offshore power grid | |
| CN104574219A (en) | System and method for monitoring and early warning of operation conditions of power grid service information system | |
| Xue et al. | Adaptive ultra-short-term wind power prediction based on risk assessment | |
| CN113169708A (en) | Solar power generation balance control system and method | |
| CN116360377B (en) | Data processing method for battery cell formation component capacity and distributed control system | |
| CN117318069B (en) | Power supply system fault self-healing method and system | |
| CN104834305A (en) | Distribution network automated terminal telemetering abnormity analyzing system and method based on DMS system | |
| CN110865329B (en) | Electric energy metering method and system based on big data self-diagnosis | |
| CN117331790A (en) | Machine room fault detection method and device for data center | |
| WO2023211418A1 (en) | A smart energy management system | |
| CN115630716A (en) | Intelligent generation method and device for equipment maintenance plan | |
| CN114414938B (en) | Dynamic response method and system for power distribution network faults | |
| CN116111677B (en) | Background monitoring data processing method and system applied to energy storage field | |
| CN112348074B (en) | Power distribution network power failure event accurate diagnosis method, device and system based on data driving | |
| CN113469453B (en) | Elevator evaluation method and elevator evaluation device based on information physical system | |
| CN113283702B (en) | Power flow state evaluation method and device integrating safety and stability characteristics of power system | |
| US20240079895A1 (en) | Battery performance monitoring and optimization | |
| Weibo et al. | Risk assessment technology of ship power system based on improved time series algorithm | |
| CN116822206A (en) | Online fault diagnosis method and system | |
| CN115587016A (en) | Power equipment operation analysis method | |
| CN117807400A (en) | Wind power plant energy efficiency analysis method, device, equipment and readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |