CN113283044A - Edge calculation method for urban gas pipe network blockage diagnosis - Google Patents
Edge calculation method for urban gas pipe network blockage diagnosis Download PDFInfo
- Publication number
- CN113283044A CN113283044A CN202110688089.5A CN202110688089A CN113283044A CN 113283044 A CN113283044 A CN 113283044A CN 202110688089 A CN202110688089 A CN 202110688089A CN 113283044 A CN113283044 A CN 113283044A
- Authority
- CN
- China
- Prior art keywords
- layer
- computing
- edge
- sub
- steps
- 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.)
- Granted
Links
- 238000004364 calculation method Methods 0.000 title claims abstract description 100
- 238000003745 diagnosis Methods 0.000 title claims abstract description 25
- 238000000034 method Methods 0.000 claims abstract description 29
- 238000012545 processing Methods 0.000 claims description 15
- 238000001514 detection method Methods 0.000 claims description 10
- 238000013461 design Methods 0.000 claims description 7
- 238000011897 real-time detection Methods 0.000 claims description 6
- 230000000903 blocking effect Effects 0.000 claims description 5
- 230000008569 process Effects 0.000 description 8
- 238000012544 monitoring process Methods 0.000 description 5
- 238000013500 data storage Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 238000000354 decomposition reaction Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000011144 upstream manufacturing Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/18—Network design, e.g. design based on topological or interconnect aspects of utility systems, piping, heating ventilation air conditioning [HVAC] or cabling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5061—Partitioning or combining of resources
- G06F9/5072—Grid computing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computational Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Mathematical Physics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Pipeline Systems (AREA)
Abstract
The invention relates to an edge computing system and method for urban gas pipe network blockage diagnosis, wherein the method comprises the following steps: the cloud computing layer drives a pressure pulse applying device to apply pressure pulses to the tail end pipeline; the equipment layer detects real-time data and uploads the data to a storage node of the edge layer for storage; the cloud computing layer decomposes the tasks according to the computing steps; and the scheduling layer dynamically adjusts and allocates the queue of the subtasks according to the calculation steps corresponding to the decomposed subtasks.
Description
Technical Field
The invention relates to the field of energy delivery, in particular to an edge calculation method for urban gas pipe network blockage diagnosis.
Background
With the development of network technology, edge calculation is applied to monitoring systems of urban gas pipe networks. The mobile edge computing means that an edge server is deployed at the edge of a mobile network close to a mobile device terminal, data computing and storage are performed, and a network service environment and a cloud computing function are provided for the mobile device through cooperation with a remote cloud computing center. By localizing and marginalizing the computing, storing and communication resources, the bandwidth requirement of a centralized network is greatly reduced, data traffic transmission is effectively reduced, link congestion is obviously weakened, the speed of content acquisition in a mobile network is increased, and the access delay and energy consumption of user equipment are greatly reduced.
An edge computing system for urban gas pipeline network monitoring, in particular pipeline network blockage diagnosis, is shown in fig. 1. The method mainly comprises the following steps: a device layer 1, an edge layer 2, a cloud computing layer 3, and a pressure pulse application device 4. The detection device 1-1, 1-2, … …, 1-i located at the tail end of the pipe network branch pipe can be a barometer or a flowmeter, the edge layer 2 comprises an edge data storage part 21 and a plurality of edge computing nodes 22-1, 22-2, … … 22-i, the edge data storage part 21 is used for storing real-time data detected by the device layer 1, the edge computing nodes are used for receiving unloaded computing tasks, and the cloud computing layer is used for judging whether the data are abnormal or not. The working principle of the monitoring system is as follows: based on a fixed period, a gas pressure pulse which does not exceed the design pressure of the pipeline is applied to each end pipeline through a pressure pulse applying device 4, so that the gas pressure applied to the interior of the pipeline is suddenly increased in a short time to form a pressure pulse, a detection device (such as a barometer or a flowmeter) at the end of the pipeline detects the change condition of the gas pressure at the end of the pipeline after the pressure pulse is applied to judge whether the pipeline is blocked, and the judgment principle is that the pressure gauge at the end of the blocked pipeline deviates from the linear relation with the applied gas pressure under the action of the pressure pulse, the obstruction of a pressure echo returned to the end by a pipeline blocking part is large, and further, a secondary pressure echo oscillating towards the end of the pipeline is generated. Whether the pipeline is abnormal or not can be judged based on real-time data analysis of equipment such as a pipeline tail end pressure gauge after the pressure pulse is applied. Generally, the cloud computing layer 3 decomposes computing tasks and unloads the computing tasks to the computing nodes 22 of the edge layer 2, after receiving the computing tasks sent by the cloud computing layer, a plurality of edge nodes execute the computing tasks and send computing results back to the cloud computing layer, and after summarizing a structure, the cloud computing layer finally judges whether a blocking condition occurs.
The above prior art has the following problems: because the number of the tail end pipelines in the urban low-voltage pipeline is large, the computing tasks of a plurality of tail end pipelines are required to be simultaneously excited, and after the computation decomposition and the unloading, a plurality of sub-computing tasks are serially queued on each computing node of the edge layer. In the prior art, the queuing sequence of each compute node usually processes the sub-compute tasks of multiple different end pipelines in turn according to the fairness principle or sorts the sub-compute tasks of multiple different end pipelines based on the principle that the average computation consumes the shortest time. However, the average calculation time-consuming shortest strategy is adopted, a large amount of system state information needs to be called and calculated, for an urban gas pipe network, because the number of end pipelines is large and the number of simultaneously excited calculation tasks is large, if all calculation tasks are comprehensively calculated, the sorting calculation mode based on the average time-consuming shortest principle occupies more network and calculation resources, and the queuing is performed by adopting the fairness principle, so that the time consumption of each calculation task is easily increased although the network and the calculation resources are saved.
Therefore, it is necessary to provide an edge calculation method for urban gas pipe network blockage diagnosis, which can improve system efficiency according to characteristics and requirements of the urban gas pipe network on the basis of balancing network and computing resources.
Disclosure of Invention
The technical problem to be solved by the invention is the following defects in the prior art: because the number of the tail end pipelines in the urban low-voltage pipeline is large, the computing tasks of a plurality of tail end pipelines are required to be simultaneously excited, and after the computation decomposition and the unloading, a plurality of sub-computing tasks are serially queued on each computing node of the edge layer. In the prior art, the queuing sequence of each compute node usually processes the sub-compute tasks of multiple different end pipelines in turn according to the fairness principle or sorts the sub-compute tasks of multiple different end pipelines based on the principle that the average computation consumes the shortest time. However, the average calculation time-consuming shortest strategy is adopted, a large amount of system state information needs to be called and calculated, for an urban gas pipe network, because the number of end pipelines is large and the number of simultaneously excited calculation tasks is large, if all calculation tasks are comprehensively calculated, the sorting calculation mode based on the average time-consuming shortest principle occupies more network and calculation resources, and the queuing is performed by adopting the fairness principle, so that the time consumption of each calculation task is easily increased although the network and the calculation resources are saved.
The technical scheme adopted by the invention for solving the technical problems is as follows:
an edge computing system for urban gas pipe network blockage diagnosis comprises an equipment layer, an edge layer, a scheduling layer, a cloud computing layer and pressure pulse applying equipment; the device layer, the edge layer and the scheduling layer are connected in a network mode, the cloud computing layer drives the pressure pulse applying device through the network, the pressure pulse applying device applies pressure pulses which do not exceed design pressure of the pipeline to a plurality of terminal pipelines of the pipeline network, the device layer is used for detecting real-time detection data in a specific period after the pressure pulses are applied and uploading the real-time detection data to the edge layer, the cloud computing layer divides a plurality of groups of real-time data aiming at the terminal pipelines into a plurality of sub-steps, the sub-steps are divided into a plurality of sub-computing tasks, and the scheduling layer sets different queuing distribution modes according to the sub-tasks.
Specifically, the edge layer includes a pressure and flow detection device.
Specifically, the edge layer includes compute nodes and storage nodes.
Specifically, the length of each branch conduit is different.
Specifically, the detection time length differs for each branch conduit.
Specifically, the start time of the calculation task corresponding to each branch pipeline is different.
An edge calculation method for urban gas pipe network blockage diagnosis comprises the following steps: the cloud computing layer drives a pressure pulse applying device to apply pressure pulses to the tail end pipeline; the equipment layer detects real-time data and uploads the data to a storage node of the edge layer for storage; the cloud computing layer decomposes the tasks according to the computing steps; and the cloud computing layer dynamically adjusts and distributes the queue of the subtasks according to the computing steps corresponding to the decomposed subtasks.
Specifically, the method further comprises the step that the computing nodes perform computing processing according to the distribution mode of the scheduling layer, and the computing results are fed back to the cloud computing layer.
Specifically, the method further comprises: and the cloud computing layer summarizes the computing result to judge the blocking condition.
The edge calculation method for urban gas pipe network blockage diagnosis provided by the application has the following beneficial effects:
when the sub-tasks are queued and ordered, the calculation steps described by the sub-tasks are considered, and queuing priority setting of different strategies is carried out according to the respective characteristics of each calculation step, for the data calling and fitting stage, because the data reading amount and the calculation amount are large, the network and the calculation capacity condition of the node are considered in the queuing strategy, and the calculation amount for the fitting function characteristic parameter extraction stage is large, the calculation capacity condition is mainly considered in the queuing strategy, and aiming at the characteristic parameter threshold value judging stage, because the calculation amount and the data reading amount are small, the processing time is short and is positioned at the tail end of the calculation flow, the sub-computing tasks at the stage are processed preferentially in the queue, so that the computing task at the end of the computing process can be ended as early as possible, and reduces the stress on parallel tasks of the device, and the delay time of the computing tasks at other stages in the queue is not increased much. On the basis of balancing the network and computing resources, the system efficiency is improved according to the characteristics and the requirements of the urban gas pipe network.
Drawings
Fig. 1 is a schematic structural diagram of an edge computing system for urban gas pipe network blockage diagnosis in the prior art.
Fig. 2 is a schematic structural diagram of an edge computing system for urban gas pipe network blockage diagnosis provided by the present application.
Fig. 3 is a flowchart of an edge calculation method for urban gas pipe network blockage diagnosis provided by the present application.
Detailed Description
The present invention will now be described in more detail with reference to the accompanying drawings, in which preferred embodiments of the invention are shown, it being understood that one skilled in the art may modify the invention herein described while still achieving the beneficial results of the present invention. Accordingly, the following description should be construed as broadly as possible to those skilled in the art and not as limiting the invention.
In the interest of clarity, not all features of an actual implementation are described. In the following description, well-known functions or constructions are not described in detail since they would obscure the invention in unnecessary detail. It will of course be appreciated that in the development of any such actual embodiment, numerous implementation-specific details must be set forth in order to achieve the developer's specific goals.
In order to make the objects and features of the present invention more comprehensible, embodiments of the present invention are described in detail below with reference to the accompanying drawings. It is to be noted that the drawings are in a very simplified form and are intended to use non-precision ratios for the purpose of facilitating and clearly facilitating the description of the embodiments of the invention.
The edge calculation method for urban gas pipe network blockage diagnosis is based on an edge calculation system. The edge computing system for urban gas pipe network blockage diagnosis is shown in figure 2. The method mainly comprises the following steps: device layer 1, edge layer 2, scheduling layer 5, cloud computing layer 3, and pressure pulse application device 4.
The detection device 1-1, 1-2, … …, 1-i located at the tail end of the pipe network branch pipe can be a barometer and/or a flowmeter, the edge layer 2 comprises an edge data storage part 21 and a plurality of edge computing nodes 22-1, 22-2, … … 22-i, wherein the edge data storage part 21 is used for storing real-time data detected by the device layer 1, and the computing nodes 22 in the edge layer are used for receiving unloaded computing tasks and performing corresponding computation, and finally uploading the computation results to the cloud computing layer.
The scheduling layer is located between the edge layer and the cloud computing layer, and dynamically allocates a queuing mode of the plurality of concurrent subtasks on the plurality of edge nodes 22 according to the subtask condition decomposed by the cloud computing layer and the operation condition of the plurality of edge nodes 22.
The cloud computing layer is configured to decompose the computing tasks, send the decomposed programs of the computing tasks to the scheduling layer, and distribute the computing tasks to the corresponding computing nodes 22 for computing through dynamic allocation of the scheduling layer. The cloud computing layer is also used for summarizing computing structures of the computing nodes and judging whether the computing nodes are abnormal or not according to computing results.
Based on a fixed period or preset time, the cloud computing layer 3 drives the pressure pulse applying device 4 to apply a gas pressure pulse which does not exceed the design pressure of the pipeline to each end pipeline, so that the gas pressure applied to the interior of the pipeline is suddenly increased for a short time to form a pressure pulse, and a detecting device (such as a barometer and/or a flowmeter) at the end of the pipeline detects the change of the gas pressure and the flow at the end of the pipeline after the pressure pulse is applied to judge whether the pipeline is blocked. The principle of judgment is that the pressure gauge and the flowmeter at the tail end of the blocked pipeline deviate from the normal function relation with the applied air pressure under the action of pressure pulse, and whether the pipeline is blocked is judged by performing parameter extraction and analysis on a fitting function.
The above-mentioned terminal duct is defined as a plurality of branch ducts connected downstream from one main duct and upstream from the use end.
The device layer, the edge layer, dispatch layer network connection, the cloud computing layer is through the network to the pressure pulse application equipment drive, make pressure pulse application equipment exert the pressure pulse that does not exceed pipeline design pressure in a plurality of terminal pipelines of pipe network, the device layer is used for detecting the real-time detection data in the specific cycle after pressure pulse applies, and upload above-mentioned real-time detection data to the edge layer, the cloud computing layer will be to a plurality of sub-steps of multiunit real-time data division of a plurality of terminal pipelines above-mentioned, divide a plurality of sub-steps into a plurality of sub-computing tasks again, the dispatch layer is according to the sub-task the step set up different platoon distribution mode.
As a specific implementation of the calculation, the calculation method may include, but is not limited to:
step 1.1 provides an amplitude function of the pressure pulses of the pressure pulse application device 4, wherein the amplitude function u of the pressure pulses of the pressure pulse application device 4 can be expressed as u-f (t0), wherein t0 represents a time variable for the application of the additional gas pulse, preferably u-f (t0) -ksin (ω t0), wherein k is an amplitude parameter and ω is a frequency parameter.
Step 1.2, calling real-time flow and pressure data detected by detection equipment (such as a barometer and/or a flowmeter) at the tail end of each branch pipeline in a specific time period after pulse application, and calculating a fitting function v formed by fitting real-time flow data detected by the flowmeter at the tail end of each branch pipeline with time according to the real-time datai＝gi(t1) wherein t1 represents a calculated time variable timed with the start of gas additional pulse application as the origin, viIndicating the branch end flow detected by the flow meter provided at the end of each branch pipe.
Step 1.3 fitting function v corresponding to each terminal pipeline respectively through size parameters of each branch pipeline and compensation parameters of vortexi＝gi(t1) compensation is performed. The specific compensation method can be, but is not limited to:
compensated fitting function vic＝g’i(t1), specifically:
wherein M isi(t1) is the compensation function for each branch line, where Pi(t1) is a first Li-based compensation component, SiFor the cross-sectional area of each branch conduit, Wi(t1) is a second compensation component that takes into account eddy current effects. Wherein ∈ is defined as:
wherein k is the amplitude parameter of the additional gas pulse, omega is the frequency parameter of the additional gas pulse, v' is the vortex flow resistance,
is an empirical constant with a value of 0.39-0.74,
is the arithmetic mean of the lengths L1, L2, L3, L4, … …, Li of each branch conduit, LiFor the currently calculated length of the branch conduit, T is the time span for the application of the additional pulse of gas.
Compensated fitting function vic＝g’i(t1) specifically:
step 2.1 fitting function v after compensationic＝g’i(t1) performing Fourier FFT to obtain a frequency domain function Fi＝FFT(vic)。
Step 2.2 extracting frequency domain function Fi＝FFT(vic) High intensity frequency fhighA characteristic parameter. The method specifically comprises the following steps: calculating the arithmetic mean value of the frequency domain functions of a plurality of tail end pipelines to obtain a mean frequency domain function F1. Average frequency domain function F1Calculating the frequency of all frequency intensities exceeding the maximum intensity 1/2 of the frequency domain function in the average frequency domain function, i.e. the high intensity frequency fhighHigh intensity frequency f in frequency domain functionhighShows the part of the frequency domain function signal with a relatively large flow rate change, i.e. at the end of the pipeline after the gas additional pulse is appliedCausing a significantly varying signal at the valve.
Specifically, when the frequency domain function of a certain terminal meets the following conditions, it is determined that the terminal pipeline is at risk of blockage, where the conditions are:
wherein
All high intensity frequencies f on all end pipeshighTaking the arithmetic mean of the intensity signals of (1)i-highIs that each end pipe corresponds to a high intensity frequency fhighSignal strength of (I)i-highMore than one value may exist for each terminal pipeline, calculation needs to be performed on the more than one value, if the frequency domain function of a certain terminal meets the conditions, it is determined that the terminal pipeline has a condition of exceeding a threshold value, that is, a blockage condition exists, otherwise, it is determined that the terminal pipeline does not have the condition of exceeding the threshold value, that is, no blockage exists.
Based on the edge calculation system and the calculation and judgment mode, the application provides an edge calculation method for urban gas pipe network blockage diagnosis, which comprises the following steps: the cloud computing layer drives a pressure pulse applying device to apply pressure pulses to the tail end pipeline; the equipment layer detects real-time data and uploads the data to a storage node of the edge layer for storage; the cloud computing layer decomposes the tasks according to the computing steps; the scheduling layer dynamically adjusts and distributes the queue of the subtasks according to the calculation steps corresponding to the decomposed subtasks; the method also comprises the steps that the computing nodes perform computing processing according to the distribution mode of the scheduling layer and feed back computing results to the cloud computing layer; and the cloud computing layer summarizes the computing result to judge the blocking condition.
The method specifically comprises the following steps:
in the first step, based on a fixed period or preset time, the cloud computing layer 3 drives the pressure pulse applying device 4 to apply a gas pressure pulse not exceeding the design pressure of the pipeline to each end pipeline, so that the gas pressure applied to the interior of the pipeline suddenly rises in a short time to form a pressure pulse.
And secondly, detecting real-time air pressure and/or flow data of the tail end pipeline by a detection device (such as an air pressure meter and/or a flow meter) of the device layer 1 in real time, and uploading the real-time data to a storage node of the edge layer 2 for storage.
And thirdly, generating an integral calculation program by the cloud calculation layer, and segmenting the calculation program. The method specifically comprises the following steps:
and S31, the cloud computing layer generates an overall computing program which is divided into a plurality of sequential steps based on objective physical and mathematical principles. As an illustrative illustration, the whole procedure described above may include: step 1, calling historical real-time data, and performing function fitting based on the called historical real-time data; step 2, fitting-based function vicExtracting characteristic parameters; step 3, judging the frequency domain function F of each tail end pipelinei＝FFT(vic) At a corresponding high intensity frequency fhighWhether a super-threshold condition exists on the frequency.
S32, the cloud computing layer divides the computing program into sub-computing programs according to the plurality of steps (i.e. steps 1, 2, and 3) of the computing program, and divides the sub-computing programs into sub-computing programs based on the program separability of the sub-computing programs, where the header information of the sub-computing programs includes the corresponding end pipe information and the information of the sub-computing step to which the sub-computing program belongs.
Specifically, in the edge computing system, in order to implement the specific computing process, the cloud computing layer may segment the program for implementing the computing process, where the program is divided into sub-computing programs. The multiple word computing programs are sent by the cloud computing layer to the scheduling layer.
Specifically, the cloud computing layer divides the overall computing program into sub-computing programs according to the three computing steps, namely step 1, step 2 and step 3, and then further divides the sub-computing program corresponding to each step into a plurality of sub-computing programs based on the divisibility of the programs. In the process of dividing the program, the head position of the sub-calculation program writes information of the corresponding sub-calculation step, i.e., to which of the three calculation steps the sub-calculation program corresponds.
And fourthly, the cloud computing layer sends the plurality of sub-computing programs corresponding to the terminal pipelines to the scheduling layer, and the scheduling layer determines a queue ordering mode of the plurality of sub-computing program tasks in the edge layer according to the sub-computing step information of the plurality of sub-computing programs.
It should be noted that, each monitoring unit corresponds to a plurality of terminal pipelines in the whole urban pipe network system, and the length, the response time, the acquisition and detection time and the data transmission time of each terminal pipeline are different. Therefore, the computing task start time for each end pipeline is different, and the time for generating and sending the sub-computing programs of the end pipelines under one monitoring unit to the downstream computing nodes for processing also is different in the cloud computing layer. Under the condition of limited computing nodes, the situation that sub-computing programs belonging to different sub-computing programs of a plurality of end pipelines are threaded and queued in series in a certain computing node exists.
The method specifically comprises the following steps:
and S41, the scheduling layer receives the sub-computation program sent by the cloud computing layer, firstly reads the information of the sub-computation step of the program header, and judges the sub-computation step to which the sub-computation program belongs.
The sub-calculation step 1 is a program function for calling historical real-time data and performing function fitting based on the called historical real-time data, and relates to a large amount of data reading calling and calculating steps. Therefore, the calculation procedure included in the sub-calculation step 1 needs to occupy more network data transmission resources and calculation resources local to the calculation node.
And S42, the scheduling layer sets different queuing distribution modes according to the sub-calculation steps to which the sub-calculation programs belong. The specific queuing distribution mode comprises the following steps:
when the scheduling layer judges that the sub-calculation program belongs to the sub-calculation step 1, the scheduling layer firstly inquires whether an idle calculation node exists, if so, the sub-calculation program belonging to the sub-calculation step 1 is sent to the idle node with the highest first priority value, and the first priority value P1 is D2And multiplying by C and multiplying by B, wherein D is the network bandwidth of the computing node, F is the clock pulse frequency of the computing node, C is the CPU core number of the computing node, and B is the CPU bit number of the computing node.
If no free computing node exists, the sub-computing program in the sub-computing step 1 is sent to the computing node with the highest second priority value, and the computing node is arranged at the last position of the queue, and the second priority value P2 is equal to (D)2×F×C×B)/T2Wherein D is the network bandwidth of the computing node, F is the clock pulse frequency of the computing node, C is the CPU core number of the computing node, B is the CPU bit number of the computing node, and T is the nominal waiting time of the computing node after the current queuing task is finished.
When the scheduling layer judges that the sub-calculation program belongs to the sub-calculation step 2, the scheduling layer firstly inquires whether an idle calculation node exists, if so, the sub-calculation program belonging to the sub-calculation step 1 is sent to the idle node with the highest third priority weight, and the third priority weight P3 is F multiplied by C multiplied by B, wherein F is the clock pulse frequency of the calculation node, C is the CPU core number of the calculation node, and B is the CPU digit number of the calculation node.
If no free computing node exists, the sub-computing program in the sub-computing step 2 is sent to the computing node with the highest fourth priority value, and the sub-computing program is arranged at the last position of the queue, and the fourth priority value P4 is (F × C × B)/T2Wherein, F is the clock pulse frequency of the computing node, C is the number of CPU cores of the computing node, B is the number of CPU bits of the computing node, and T is the nominal waiting time for the current queuing task of the computing node to finish processing.
When the scheduling layer judges that the sub-calculation program belongs to the sub-calculation step 3, the scheduling layer firstly inquires whether an idle calculation node exists, and if the idle calculation node exists, the sub-calculation program belonging to the sub-calculation step 3 is sent to the idle calculation node with the highest network bandwidth.
And if no idle computing node exists, sending the sub-computing program belonging to the sub-computing step 3 to the computing node with the shortest nominal waiting time T after the current queued task of the computing node is processed, wherein the priority of the queue is higher than that of all other sub-computing tasks belonging to the sub-computing steps 1 and 2, but lower than that of all other sub-computing tasks belonging to the sub-computing step 3 and already in the queue.
And the nominal waiting time of the current queuing task of the computing node after the processing is finished is equal to the sum of the nominal waiting time of the task in the processing and the nominal waiting time of the task in the queuing.
According to the statistical average value of the system historical data, the system presets nominal waiting processing time t1, t2 and t3 for the sub-computing tasks of the sub-computing steps 1, 2 and 3. For example, the nominal waiting time t1 of the sub-calculation task in the sub-calculation step 1 is preset to 7s, the nominal waiting time t2 of the sub-calculation task in the sub-calculation step 2 is preset to 4s, and the nominal waiting time t3 of the sub-calculation task in the sub-calculation step 3 is preset to 0.5 s. The nominal waiting time is preset according to the statistical average value of the historical data of the system, the balance of accuracy and resource occupation reduction can be obtained in the calculation of task delay, and the minimum resource is utilized to reach a more accurate nominal value.
And further, the nominal waiting time of the task in processing is the time obtained by subtracting the processed time of the task from the nominal waiting processing time preset by the sub-calculation step of the task in processing. The nominal latency of a task in the queue is the sum of the nominal latencies of all tasks not processed in the queue.
Preferably, the clock pulse frequency, the number of CPU cores, the number of CPU bits, the network bandwidth, and other design information of each computing node are stored in the scheduling layer, thereby facilitating computation and reducing network resource occupation. And the queuing sequence condition on each computing node is uploaded to a scheduling layer by the computing node for updating while executing a new sub-computing task.
When the sub-tasks are queued and ordered, the calculation steps described by the sub-tasks are considered, and queuing priority setting of different strategies is carried out according to the respective characteristics of each calculation step, for the data calling and fitting stage, because the data reading amount and the calculation amount are large, the network and the calculation capacity condition of the node are considered in the queuing strategy, and the calculation amount for the fitting function characteristic parameter extraction stage is large, the calculation capacity condition is mainly considered in the queuing strategy, and aiming at the characteristic parameter threshold value judging stage, because the calculation amount and the data reading amount are small, the processing time is short and is positioned at the tail end of the calculation flow, the sub-computing tasks at the stage are processed preferentially in the queue, so that the computing task at the end of the computing process can be ended as early as possible, and reduces the stress on parallel tasks of the device, and the delay time of the computing tasks at other stages in the queue is not increased much.
And fifthly, the computing node sequentially executes a plurality of sub-computing program tasks according to a queue sorting mode set by the scheduling layer and feeds back the result to the cloud computing layer.
And sixthly, after the cloud computing layer collects the computing results of a plurality of computing nodes corresponding to one integral computing task, judging the blockage condition of the tail end pipeline corresponding to the integral computing task.
The edge calculation method for urban gas pipe network blockage diagnosis provided by the application has the following beneficial effects:
when the sub-tasks are queued and ordered, the calculation steps described by the sub-tasks are considered, and queuing priority setting of different strategies is carried out according to the respective characteristics of each calculation step, for the data calling and fitting stage, because the data reading amount and the calculation amount are large, the network and the calculation capacity condition of the node are considered in the queuing strategy, and the calculation amount for the fitting function characteristic parameter extraction stage is large, the calculation capacity condition is mainly considered in the queuing strategy, and aiming at the characteristic parameter threshold value judging stage, because the calculation amount and the data reading amount are small, the processing time is short and is positioned at the tail end of the calculation flow, the sub-computing tasks at the stage are processed preferentially in the queue, so that the computing task at the end of the computing process can be ended as early as possible, and reduces the stress on parallel tasks of the device, and the delay time of the computing tasks at other stages in the queue is not increased much. On the basis of balancing the network and computing resources, the system efficiency is improved according to the characteristics and the requirements of the urban gas pipe network. The foregoing illustrates and describes the principles, general features, and advantages of the present invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (9)
1. An edge computing system for urban gas pipe network blockage diagnosis is characterized in that: the system comprises an equipment layer, an edge layer, a scheduling layer, a cloud computing layer and pressure pulse applying equipment; the device layer, the edge layer and the scheduling layer are connected in a network mode, the cloud computing layer drives the pressure pulse applying device through the network, the pressure pulse applying device applies pressure pulses which do not exceed design pressure of the pipeline to a plurality of terminal pipelines of the pipeline network, the device layer is used for detecting real-time detection data in a specific period after the pressure pulses are applied and uploading the real-time detection data to the edge layer, the cloud computing layer divides a plurality of groups of real-time data aiming at the terminal pipelines into a plurality of sub-steps, the sub-steps are divided into a plurality of sub-computing tasks, and the scheduling layer sets different queuing distribution modes according to the sub-tasks.
2. The edge computing system for urban gas pipe network blockage diagnosis according to claim 1, wherein: the edge layer includes pressure and flow detection equipment.
3. The edge computing system for urban gas pipe network blockage diagnosis according to claim 1, wherein: the edge layer includes compute nodes and storage nodes.
4. The edge computing system for urban gas pipe network blockage diagnosis according to claim 1, wherein: each branch conduit is of a different length.
5. The urban gas pipe network blockage diagnosis and positioning system according to claim 1, characterized in that: the detection time length of each branch conduit is different.
6. The urban gas pipe network blockage diagnosis and positioning system according to claim 5, characterized in that: the calculation task start time corresponding to each branch pipeline is different.
7. An edge calculation method for urban gas pipe network blockage diagnosis is characterized by comprising the following steps: the method comprises the following steps: the cloud computing layer drives a pressure pulse applying device to apply pressure pulses to the tail end pipeline; the equipment layer detects real-time data and uploads the data to a storage node of the edge layer for storage; the cloud computing layer decomposes the tasks according to the computing steps; and the cloud computing layer dynamically adjusts and distributes the queue of the subtasks according to the computing steps corresponding to the decomposed subtasks.
8. The edge calculation method for urban gas pipe network blockage diagnosis according to claim 7, wherein the edge calculation method comprises the following steps: the method also comprises the step that the computing nodes perform computing processing according to the distribution mode of the scheduling layer and feed back computing results to the cloud computing layer.
9. The edge calculation method for urban gas pipe network blockage diagnosis according to claim 8, wherein the edge calculation method comprises the following steps: the method further comprises the following steps: and the cloud computing layer summarizes the computing result to judge the blocking condition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110688089.5A CN113283044B (en) | 2021-06-21 | 2021-06-21 | Edge calculation method for urban gas pipe network blockage diagnosis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110688089.5A CN113283044B (en) | 2021-06-21 | 2021-06-21 | Edge calculation method for urban gas pipe network blockage diagnosis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113283044A true CN113283044A (en) | 2021-08-20 |
| CN113283044B CN113283044B (en) | 2023-11-21 |
Family
ID=77285148
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110688089.5A Active CN113283044B (en) | 2021-06-21 | 2021-06-21 | Edge calculation method for urban gas pipe network blockage diagnosis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113283044B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113721057A (en) * | 2021-09-01 | 2021-11-30 | 上海天麦能源科技有限公司 | Method for quickly judging gas stealing behavior of low-pressure pipeline |
| CN113757566A (en) * | 2021-09-16 | 2021-12-07 | 上海天麦能源科技有限公司 | Control method for intelligent instrument of urban gas pipe network |
| CN113778685A (en) * | 2021-09-16 | 2021-12-10 | 上海天麦能源科技有限公司 | Unloading method for urban gas pipe network edge computing system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2927465A1 (en) * | 2014-04-02 | 2015-10-07 | MANN+HUMMEL GmbH | Method for determining the instantaneous mass flow rate of a gas, corresponding device and computer program |
| CN107101087A (en) * | 2017-03-22 | 2017-08-29 | 北京航空航天大学 | City gas pipeline system pipeline fault propagation impact evaluation method and system |
| CN107830412A (en) * | 2017-10-30 | 2018-03-23 | 重庆科技学院 | The incomplete blocking position detecting system of pipeline and detection method |
| CN111222817A (en) * | 2018-11-27 | 2020-06-02 | 深圳先进技术研究院 | Intelligent warehousing system based on cloud/edge computing |
| WO2020133098A1 (en) * | 2018-12-27 | 2020-07-02 | 驭势科技（北京）有限公司 | Distributed computing network system and method |
| CN112985713A (en) * | 2021-01-29 | 2021-06-18 | 重庆川仪自动化股份有限公司 | Pipe network leakage monitoring method and system based on edge calculation |
-
2021
- 2021-06-21 CN CN202110688089.5A patent/CN113283044B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2927465A1 (en) * | 2014-04-02 | 2015-10-07 | MANN+HUMMEL GmbH | Method for determining the instantaneous mass flow rate of a gas, corresponding device and computer program |
| CN107101087A (en) * | 2017-03-22 | 2017-08-29 | 北京航空航天大学 | City gas pipeline system pipeline fault propagation impact evaluation method and system |
| CN107830412A (en) * | 2017-10-30 | 2018-03-23 | 重庆科技学院 | The incomplete blocking position detecting system of pipeline and detection method |
| CN111222817A (en) * | 2018-11-27 | 2020-06-02 | 深圳先进技术研究院 | Intelligent warehousing system based on cloud/edge computing |
| WO2020133098A1 (en) * | 2018-12-27 | 2020-07-02 | 驭势科技（北京）有限公司 | Distributed computing network system and method |
| CN112985713A (en) * | 2021-01-29 | 2021-06-18 | 重庆川仪自动化股份有限公司 | Pipe network leakage monitoring method and system based on edge calculation |
Non-Patent Citations (1)
| Title |
|---|
| 殷建成: "输气管网稳态运行优化方法研究", 中国优秀博硕士学位论文全文数据库 (硕士) 工程科技Ⅰ辑, no. 04 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113721057A (en) * | 2021-09-01 | 2021-11-30 | 上海天麦能源科技有限公司 | Method for quickly judging gas stealing behavior of low-pressure pipeline |
| CN113721057B (en) * | 2021-09-01 | 2024-03-22 | 上海天麦能源科技有限公司 | Rapid judging method for gas stealing behavior of low-pressure pipeline |
| CN113757566A (en) * | 2021-09-16 | 2021-12-07 | 上海天麦能源科技有限公司 | Control method for intelligent instrument of urban gas pipe network |
| CN113778685A (en) * | 2021-09-16 | 2021-12-10 | 上海天麦能源科技有限公司 | Unloading method for urban gas pipe network edge computing system |
| CN113778685B (en) * | 2021-09-16 | 2024-07-23 | 上海天麦能源科技有限公司 | Unloading method for urban gas pipe network edge computing system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113283044B (en) | 2023-11-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113283044A (en) | Edge calculation method for urban gas pipe network blockage diagnosis | |
| CN109617826B (en) | Storm dynamic load balancing method based on cuckoo search | |
| CN113138860B (en) | Message queue management method and device | |
| CN108446179B (en) | Weight priority Task scheduling method based on load balancing mechanism | |
| CN105656973A (en) | Distributed method and system for scheduling tasks in node group | |
| CN103401939A (en) | Load balancing method adopting mixing scheduling strategy | |
| CN112261120B (en) | Cloud-side cooperative task unloading method and device for power distribution internet of things | |
| CN112988354B (en) | Satellite communication system outbound resource optimization management method | |
| CN115794407A (en) | Computing resource allocation method and device, electronic equipment and nonvolatile storage medium | |
| WO2023029680A1 (en) | Method and apparatus for determining usable duration of magnetic disk | |
| CN112506619A (en) | Job processing method, apparatus, electronic device, storage medium, and program product | |
| CN110535894B (en) | Dynamic allocation method and system for container resources based on load feedback | |
| CN112887407B (en) | Job flow control method and device for distributed cluster | |
| CN117278473A (en) | Congestion control method and related equipment | |
| CN117255134A (en) | Data transmission method based on cloud edge cooperation | |
| CN116302404B (en) | Resource decoupling data center-oriented server non-perception calculation scheduling method | |
| CN116048791B (en) | Regulation and control method and device of test node, electronic equipment and storage medium | |
| CN112148449A (en) | Local area network scheduling algorithm and system based on edge calculation | |
| CN116755888A (en) | High-performance computing cloud platform-oriented job scheduling device and method | |
| CN103581048B (en) | A kind of distributed resource overload controlling method of powerline network | |
| Lili et al. | A Markov chain based resource prediction in computational grid | |
| CN113784395A (en) | 5G network slice resource allocation method and system | |
| CN114896070A (en) | GPU resource allocation method for deep learning task | |
| CN114138453A (en) | Resource optimization allocation method and system suitable for edge computing environment | |
| CN111459651B (en) | Load balancing method, device, storage medium and scheduling system |
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 |