CN114077482A - Intelligent calculation optimization method for industrial intelligent manufacturing edge - Google Patents
Intelligent calculation optimization method for industrial intelligent manufacturing edge Download PDFInfo
- Publication number
- CN114077482A CN114077482A CN202010829583.4A CN202010829583A CN114077482A CN 114077482 A CN114077482 A CN 114077482A CN 202010829583 A CN202010829583 A CN 202010829583A CN 114077482 A CN114077482 A CN 114077482A
- Authority
- CN
- China
- Prior art keywords
- edge
- model
- exit
- calculation
- state
- 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 49
- 238000000034 method Methods 0.000 title claims abstract description 32
- 238000005457 optimization Methods 0.000 title claims abstract description 24
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 15
- 238000013528 artificial neural network Methods 0.000 claims abstract description 9
- 238000012549 training Methods 0.000 claims description 15
- 239000011159 matrix material Substances 0.000 claims description 14
- 238000004422 calculation algorithm Methods 0.000 claims description 9
- 238000011176 pooling Methods 0.000 claims description 8
- 238000012935 Averaging Methods 0.000 claims description 7
- 238000009776 industrial production Methods 0.000 claims description 5
- 230000002159 abnormal effect Effects 0.000 claims description 4
- 239000013598 vector Substances 0.000 claims description 4
- 230000003213 activating effect Effects 0.000 claims description 3
- 230000004913 activation Effects 0.000 claims description 3
- 230000001419 dependent effect Effects 0.000 claims description 2
- 230000006870 function Effects 0.000 abstract description 15
- 238000013135 deep learning Methods 0.000 abstract description 3
- 238000013461 design Methods 0.000 abstract description 3
- 238000013136 deep learning model Methods 0.000 abstract description 2
- 230000003044 adaptive effect Effects 0.000 abstract 1
- 230000008447 perception Effects 0.000 abstract 1
- 230000005540 biological transmission Effects 0.000 description 10
- 230000008569 process Effects 0.000 description 4
- 238000005265 energy consumption Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000005562 fading Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- 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
- G06F9/48—Program initiating; Program switching, e.g. by interrupt
- G06F9/4806—Task transfer initiation or dispatching
- G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
- G06F9/485—Task life-cycle, e.g. stopping, restarting, resuming execution
- G06F9/4856—Task life-cycle, e.g. stopping, restarting, resuming execution resumption being on a different machine, e.g. task migration, virtual machine migration
-
- 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]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
-
- 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
- G06F9/5077—Logical partitioning of resources; Management or configuration of virtualized resources
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to an intelligent calculation optimization method for an industrial intelligent manufacturing edge, and a design industrial intelligent manufacturing edge calculation model can be divided into an edge resource perception model, an edge resource and task scheduling model, an edge intelligent calculation model and the like. The edge intelligent computing model calculates the joint loss function of each neural network exit node based on a deep learning method to obtain a cloud edge cooperative autonomous scale adaptive network model. And in the online optimization stage, the overall running time of the server and the edge equipment is calculated under the condition of setting the network bandwidth, and if the running time is less than the required time delay, an exit point is selected online. The method can reduce the scale of the neural network, deploy a shallow deep learning model on the edge gateway equipment to realize state pre-study and judgment, and simultaneously send the intermediate result to the cloud end to finish the final calculation result.
Description
Technical Field
The invention relates to the field of industrial intelligent manufacturing and edge calculation, in particular to an intelligent calculation optimization method for an industrial intelligent manufacturing edge.
Background
With the pulling of cloud services and the internet of things, the edge of the network is transitioning from data consumers to data producers and consumers. Data is increasingly being generated at the edge of the network and, therefore, it is more efficient to process data at the edge of the network. Micro data centers, cloud centers, and fog computing have been successively proposed and applied in a plurality of fields, and when data is generated at the edge of a network, cloud computing does not always efficiently process the data. With the internet of things, the people will enter the aftercloud era, a large amount of data can be generated in daily life, and a plurality of application programs can be deployed at the edge to consume the data. Cisco global cloud index statistics shows that in 2019, data generated by people, machines and things reach 500 gigabits, but the IP flow of a global data center only reaches 10.4 gigabits. 45% of the data created by the internet of things is stored, processed, analyzed, and computed near or at the edge of the network. By 2020, 500 billion things will be connected to the internet. Some internet of things applications may require very short response times, some may involve private data, some may generate large amounts of data, which places a heavy load on the network and cloud computing is not efficient enough to support these applications.
The traditional programming model is not suitable for edge computing, most devices in the edge computing are heterogeneous computing platforms, the runtime environment and data on each device are different, the resources of the edge devices are relatively limited, and deployment of user application programs under an edge computing scene is difficult. One task in the edge calculation can be migrated to a different edge device for execution, i.e. the task can be migrated as a necessary condition for implementing data processing on the edge device. The data in the edge computing model has a certain distributability, requiring distributability of the computing, storage and communication resources required to process the data. The edge device can process the data only if the edge computing system has the resources needed for data processing and computing.
Disclosure of Invention
In order to solve the technical problem, the invention provides an intelligent calculation optimization method for an industrial intelligent manufacturing edge.
The invention adopts the following technical scheme: an intelligent calculation optimization method for industrial intelligent manufacturing edges comprises the following steps:
collecting state data of equipment in industrial production, and transmitting the state data to a server through an edge gateway;
establishing an edge calculation model with exit nodes in a server, and deploying the edge calculation model into a gateway;
when a calculation task exists, the edge gateway inputs the selective exit node through the edge calculation model, online identification is carried out on the equipment state collected in real time, and the identification result is sent to the server, so that optimization of edge calculation is realized.
The method for establishing the edge calculation model with the exit node comprises the following steps:
pooling the equipment state to obtain a characteristic result, and inputting the characteristic result serving as an input variable into a neural network;
in the training phase, combining a loss function with each exit, each exit being dependent on the accuracy of the depth; x is the input variable, and an objective function is designed at each exit point as follows:
then, the network output result is processed by the softmax activation function to obtain a state discrimination result for representing normal or abnormal classification of the equipment
C is the set of all possible device states C, the computational model loss function L at the n-th exit nodenThe following were used:
y is the true value of the model output, i.e. the true normal or abnormal state of the device; and then training by taking the weighted sum minimization of each outlet loss function as an optimization problem to obtain a trained model.
The device state is processed by a maximum pooling method to obtain a characteristic result, which specifically comprises the following steps:
wherein e isijIs an element in the ith row and the jth column of a state matrix, the state matrix is obtained by the collected equipment states, m is the number of input equipment state feature vectors,
is the result of the largest element.
The equipment state is subjected to an average pool method to obtain a characteristic result, which specifically comprises the following steps:
the averaging pool takes the average of each component of the state matrix as a feature result.
Wherein e isijIs an element in the ith row and the jth column of a state matrix, the state matrix is obtained by the collected equipment states, m is the number of input equipment state feature vectors,
is the result of averaging the elements.
The joint loss function L is as follows:
wherein N is the total number of model exit points, βnAn associated weight for each outlet.
After training, the probability of estimating the sample at the exit point classifier is measured by using entropy, and the definition of the entropy is as follows:
wherein the state discrimination result is obtained by activating the function
Contains the calculated probabilities of all possible outcomes, C is the set of all possible device states C, and the fast decision algorithm is:
(1) n is the nth exit node, and the calculation is performed
(2) If e < TnReturning n, otherwise repeating (1);
where x is the input sample, Tn is the time threshold to decide whether to exit at the nth layer, and N is the total number of exit points.
In the online optimization stage, the edge calculation model firstly exits on an edge gateway, and then network calculation of the rest nodes is completed on a server;
at the n-th exit node, the run time ED of each exit node on the edge gatewaynAnd runtime ES on servern，ZnIs the output calculation of the nth layer of the model, and Input is the Input calculation of the model; under Bandwidth B, the entire runtime is computed
If A is smaller than the target time delay, directly selecting an exit node in the target time delay as an exit point n;
and if A is larger than or equal to the target time delay, the exit point n obtained in the training stage is not changed.
The invention has the following beneficial effects and advantages:
the invention designs an industrial intelligent manufacturing edge intelligent computing optimization method, wherein an edge intelligent computing model is designed to be suitable for a combined model training and deducing strategy of edge equipment based on deep learning, a shallow network rapid exit is designed, an edge server exit node is reasonably selected, network complexity is reduced, and a cloud edge collaborative data optimization algorithm is established. The intelligent algorithm optimization method based on the edge equipment computing resources solves the problems of real-time performance and reliability of an edge intelligent system, and reduces energy consumption, network bandwidth requirements and information leakage possibility.
Drawings
FIG. 1 is a block diagram of the overall architecture of the present invention;
FIG. 2 is a flow chart of the intelligent edge calculation algorithm of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein, but rather should be construed as modified in the spirit and scope of the present invention as set forth in the appended claims.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.
In industrial production, a great deal of data can be generated by arranging various edge devices (such as mechanical arms, cameras, sensors and the like), the network delay of the conventional cloud computing method is large, and meanwhile, the problem of computing load exists in the transmission and storage of the mass data. The invention designs a method capable of realizing intelligent computation at the edge side of equipment, the framework of which is shown in figure 1, an edge sensing layer collects state data of the equipment in industrial production and transmits the state data to an edge gateway, and an edge intelligent computation model is deployed in the edge intelligent gateway to realize intelligent sensing and intelligent decision. The edge task scheduling model is arranged on the edge server side and mainly plans task scheduling, and the calculation tasks of the edge measurement are unloaded to the edge server.
The edge task scheduling model can be realized in the following modes:
acquiring state data and equipment information of equipment in industrial production, and transmitting the state data and the equipment information to an edge gateway;
the edge server acquires equipment information through data transmission with the edge gateway, and plans task scheduling according to the equipment information and the data transmission state to obtain the task quantity unloaded to the edge gateway by all the terminal equipment;
and distributing tasks for the terminal equipment according to the task amount of each edge gateway.
The device information includes: and the terminal equipment calculates the number of CPU cycles required by the unit bit unloading task and the effective capacitance coefficient of the terminal equipment.
The data transmission state includes: the transmission rate from the terminal device to the edge gateway, the bandwidth of the device-to-edge gateway transmission system, the transmission power from the terminal device to the edge gateway, the channel gain, and the average noise power from the terminal device to the edge gateway.
The planning task scheduling comprises the following steps:
terminal deviceThe total calculation task quantity is expressed as L, the task quantity of local calculation of the terminal equipment is expressed by L, and O is usediRepresenting the task quantity distributed by the ith edge gateway, wherein N is the total number of the edge gateways; thus, the computational task volume satisfies the following constraints:
when the terminal device executes the local computation task l in the whole time block T, C represents the number of CPU cycles required by the terminal device to compute the unit bit unloading task, so that the energy consumption E of the terminal device local computation is:
wherein k represents an effective capacitance coefficient and depends on a CPU structure of the terminal equipment, l represents a task amount of the terminal equipment for local calculation, and T is local calculation time;
when the terminal equipment selects to unload the task to the edge gateway, the transmission rate r from the terminal equipment to the edge gatewayiComprises the following steps:
wherein i is the ith edge gateway, N is the total number of edge gateways, B represents the bandwidth of the device-to-edge gateway transmission system, PiRepresenting the transmission power of the terminal device to the edge gateway; hi＝di-kRepresenting the channel gain, diRepresenting the distance from the terminal equipment to the edge gateway, k being the fading factor, NiRepresenting the average noise power of the terminal device to the edge gateway;
task volume O for terminal device to unload to edge gatewayiExpressed as:
τiand processing the time slot of the terminal equipment for the ith edge gateway, wherein N is the total number of the edge gateways.
The edge intelligent calculation optimization model is a method for establishing an edge equipment calculation resource evaluation method aiming at adjustment of an intelligent algorithm such as a deep learning model and the like which is intensive in calculation and difficult to perform distributed optimization on limited calculation resources on the edge side, and the intelligent algorithm optimization method based on the edge equipment calculation resources is designed on the basis of the method, so that the real-time performance and reliability of an edge intelligent system are solved, the energy consumption and network bandwidth requirements are reduced, and the possibility of information leakage is reduced. This property can be used for inference and decision making of cross-region data.
The model is based on a convolutional neural network, the neural network with exit nodes is obtained by designing a training method so as to reduce the scale of the network and operate on an edge gateway, equipment state operation data is used as input during network training, state features are obtained by using different pooling skills, and the maximum pooling method is mainly used for obtaining a feature result by taking the maximum value of a matrix.
Where m is the number of device status feature inputs, eijIs the element in the ith row and jth column of the state matrix, obtained by the device state input, such as voltage, power, noise, etc., the state matrix is the virtual quantity calculated by elementary pooling of the state quantity,
is the result of the largest element.
The averaging pool takes the average of each component of the state matrix as a feature result.
Where m is the number of device status feature inputs, eijIs the element in row i and column j of the state matrix, the same as the average pool meaning.
As a result of the averaging element, the averaging pool may reduce the noise input of certain terminal devices. Wherein
And
two different feature results are selected in practice, and then the feature results are continuously input into the network, so that the final training result is obtained
In the training phase, a loss function is combined with each exit, so that the whole neural network can be jointly trained, and each exit is determined by the accuracy of the depth. x is a model input variable and can also be a state characteristic result after pooling, and a specific objective function is designed at each exit point and can be written as:
C is the set of all possible states, which can then be at the n-th exit nodeCalculating the model loss function Ln。
y is the true value of the model output. The weighted sum minimization of each exit loss function is then trained as an optimization problem, L being the joint loss function.
Wherein N is the total number of exit points and βnAn associated weight for each outlet.
After training, the module can classify samples in the shallow layer of the network for rapid reasoning. If the classifier at the exit of the branch predicts a high probability of correctly labeling a test sample, then the sample is exited early and a predicted result is returned. The probability of prediction of a sample at an exit point by a classifier is measured by using entropy, which is defined as:
wherein the state discrimination result is obtained by activating a function
Contains the calculated probabilities of all possible outcomes, and C is the set of all possible outcomes. The fast decision algorithm is as follows:
(3) n is the nth exit node, and the calculation is performed
(4) If e < TnN is returned, otherwise (1) is repeated.
Where x is the input sample, Tn is the time threshold to decide whether to exit at the nth layer, and N is the total number of exit points.
In the online optimization stage, the edge intelligent computing module receives delay requirements from the mobile device and then finds the optimal exit point of the trained model. The model is first exited on the edge node, after which the network computation of the remaining nodes is completed on the server. Run time ED of each node on the device at the nth exit nodenAnd runtime ES on servern，ZnIs the output calculation of the nth layer, and Input is the model Input calculation. Under a specific bandwidth B, the whole runtime is calculated
If A is less than the target latency, then the exit point n is optimally selected. And if the time delay is larger than or equal to the target time delay, the exit point n obtained in the training stage is not changed.
The optimized neural network model is respectively deployed in edge intelligent equipment (gateway) and a server, when an intelligent calculation task exists, the edge gateway can select a reasonable exit node through input, a simple deep learning algorithm is operated to identify the equipment state, and the final calculation is completed by sending an intermediate result to the server.
The above embodiments are only for illustrating the technical solutions of the present invention and not for limiting the same, and although the present invention is described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications and equivalents may be made to the embodiments of the invention without departing from the spirit and scope of the invention, which is to be covered by the claims.
Claims (7)
1. An intelligent calculation optimization method for industrial intelligent manufacturing edges is characterized by comprising the following steps:
collecting state data of equipment in industrial production, and transmitting the state data to a server through an edge gateway;
establishing an edge calculation model with exit nodes in a server, and deploying the edge calculation model into a gateway;
when a calculation task exists, the edge gateway inputs the selective exit node through the edge calculation model, online identification is carried out on the equipment state collected in real time, and the identification result is sent to the server, so that optimization of edge calculation is realized.
2. The method for intelligent computing and optimizing the edge in industrial intelligent manufacturing according to claim 1, wherein the establishing of the edge computing model with the exit node comprises the following steps:
pooling the equipment state to obtain a characteristic result, and inputting the characteristic result serving as an input variable into a neural network;
in the training phase, combining a loss function with each exit, each exit being dependent on the accuracy of the depth; x is the input variable, and an objective function is designed at each exit point as follows:
then, the network output result is processed by the softmax activation function to obtain a state discrimination result for representing normal or abnormal classification of the equipment
C is the set of all possible device states C,computing model loss function L at the n-th exit nodenThe following were used:
y is the true value of the model output, i.e. the true normal or abnormal state of the device; and then training by taking the weighted sum minimization of each outlet loss function as an optimization problem to obtain a trained model.
3. The intelligent calculation and optimization method for the industrial intelligent manufacturing edge according to claim 2, wherein the device state is subjected to a maximum pooling method to obtain a characteristic result, specifically as follows:
wherein e isijIs an element in the ith row and the jth column of a state matrix, the state matrix is obtained by the collected equipment states, m is the number of input equipment state feature vectors,
is the result of the largest element.
4. The intelligent calculation and optimization method for the industrial intelligent manufacturing edge according to claim 2, wherein the device state is subjected to an average pool method to obtain a characteristic result, specifically as follows:
the averaging pool takes the average of each component of the state matrix as a feature result.
Wherein e isijIs an element in the ith row and jth column of the state matrix, the state momentsThe array is obtained by the collected device state, m is the number of the device state feature vector input,
is the result of averaging the elements.
5. The method of claim 2, wherein the joint loss function L is as follows:
wherein N is the total number of model exit points, βnAn associated weight for each outlet.
6. The intelligent computing and optimizing method for the industrial intelligent manufacturing edge according to claim 2,
after training, the probability of estimating the sample at the exit point classifier is measured by using entropy, and the definition of the entropy is as follows:
wherein the state discrimination result is obtained by activating the function
Contains the calculated probabilities of all possible outcomes, C is the set of all possible device states C, and the fast decision algorithm is:
(1) n is the nth exit node, and the calculation is performed
(2) If e < TnReturning n, otherwise repeating (1);
where x is the input sample, Tn is the time threshold to decide whether to exit at the nth layer, and N is the total number of exit points.
7. The intelligent computing and optimizing method for the edge in industrial intelligent manufacturing according to claim 2, wherein in the online optimization stage, the edge computing model firstly exits at the edge gateway, and then the network computing of the rest nodes is completed at the server;
at the n-th exit node, the run time ED of each exit node on the edge gatewaynAnd runtime ES on servern，ZnIs the output calculation of the nth layer of the model, and Input is the Input calculation of the model; under Bandwidth B, the entire runtime is computed
If A is smaller than the target time delay, directly selecting an exit node in the target time delay as an exit point n;
and if A is larger than or equal to the target time delay, the exit point n obtained in the training stage is not changed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010829583.4A CN114077482B (en) | 2020-08-18 | 2020-08-18 | Intelligent computing optimization method for industrial intelligent manufacturing edge |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010829583.4A CN114077482B (en) | 2020-08-18 | 2020-08-18 | Intelligent computing optimization method for industrial intelligent manufacturing edge |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114077482A true CN114077482A (en) | 2022-02-22 |
| CN114077482B CN114077482B (en) | 2024-04-16 |
Family
ID=80281280
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010829583.4A Active CN114077482B (en) | 2020-08-18 | 2020-08-18 | Intelligent computing optimization method for industrial intelligent manufacturing edge |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114077482B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115983390B (en) * | 2022-12-02 | 2023-09-26 | 上海科技大学 | Edge intelligent reasoning method and system based on multi-antenna aerial calculation |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109961132A (en) * | 2017-12-22 | 2019-07-02 | 英特尔公司 | System and method for learning the structure of depth convolutional neural networks |
| US20190251444A1 (en) * | 2018-02-14 | 2019-08-15 | Google Llc | Systems and Methods for Modification of Neural Networks Based on Estimated Edge Utility |
| CN110347500A (en) * | 2019-06-18 | 2019-10-18 | 东南大学 | For the task discharging method towards deep learning application in edge calculations environment |
| CN110995858A (en) * | 2019-12-17 | 2020-04-10 | 大连理工大学 | Edge network request scheduling decision method based on deep Q network |
| CN111149141A (en) * | 2017-09-04 | 2020-05-12 | Nng软件开发和商业有限责任公司 | Method and apparatus for collecting and using sensor data from a vehicle |
| US20200202201A1 (en) * | 2018-12-21 | 2020-06-25 | Fujitsu Limited | Information processing apparatus, neural network program, and processing method for neural network |
| CN111445026A (en) * | 2020-03-16 | 2020-07-24 | 东南大学 | Deep neural network multi-path reasoning acceleration method for edge intelligent application |
-
2020
- 2020-08-18 CN CN202010829583.4A patent/CN114077482B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111149141A (en) * | 2017-09-04 | 2020-05-12 | Nng软件开发和商业有限责任公司 | Method and apparatus for collecting and using sensor data from a vehicle |
| CN109961132A (en) * | 2017-12-22 | 2019-07-02 | 英特尔公司 | System and method for learning the structure of depth convolutional neural networks |
| US20190251444A1 (en) * | 2018-02-14 | 2019-08-15 | Google Llc | Systems and Methods for Modification of Neural Networks Based on Estimated Edge Utility |
| US20200202201A1 (en) * | 2018-12-21 | 2020-06-25 | Fujitsu Limited | Information processing apparatus, neural network program, and processing method for neural network |
| CN110347500A (en) * | 2019-06-18 | 2019-10-18 | 东南大学 | For the task discharging method towards deep learning application in edge calculations environment |
| CN110995858A (en) * | 2019-12-17 | 2020-04-10 | 大连理工大学 | Edge network request scheduling decision method based on deep Q network |
| CN111445026A (en) * | 2020-03-16 | 2020-07-24 | 东南大学 | Deep neural network multi-path reasoning acceleration method for edge intelligent application |
Non-Patent Citations (1)
| Title |
|---|
| 贺可太;刘硕;陈哲涵;杨智;: "选区激光烧结收缩率预测及工艺参数优化", 高分子材料科学与工程, no. 06, 6 July 2018 (2018-07-06) * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115983390B (en) * | 2022-12-02 | 2023-09-26 | 上海科技大学 | Edge intelligent reasoning method and system based on multi-antenna aerial calculation |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114077482B (en) | 2024-04-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110347500B (en) | Task unloading method for deep learning application in edge computing environment | |
| CN110851782A (en) | Network flow prediction method based on lightweight spatiotemporal deep learning model | |
| CN111445026A (en) | Deep neural network multi-path reasoning acceleration method for edge intelligent application | |
| CN114818515A (en) | Multidimensional time sequence prediction method based on self-attention mechanism and graph convolution network | |
| Callegaro et al. | Optimal edge computing for infrastructure-assisted UAV systems | |
| CN113852432B (en) | Spectrum Prediction Sensing Method Based on RCS-GRU Model | |
| CN110601777B (en) | Method for estimating satellite-ground downlink co-channel interference under low-orbit mobile satellite constellation | |
| CN112835715B (en) | Method and device for determining task unloading strategy of unmanned aerial vehicle based on reinforcement learning | |
| CN112995343B (en) | Edge node calculation unloading method with performance and demand matching capability | |
| CN114385272B (en) | Ocean task oriented online adaptive computing unloading method and system | |
| CN115018193A (en) | Time series wind energy data prediction method based on LSTM-GA model | |
| CN114936708A (en) | Fault diagnosis optimization method based on edge cloud collaborative task unloading and electronic equipment | |
| CN114077482A (en) | Intelligent calculation optimization method for industrial intelligent manufacturing edge | |
| Xu et al. | Deep reinforcement learning for dynamic access control with battery prediction for mobile-edge computing in green iot networks | |
| Ponmalar et al. | Machine Learning Based Network Traffic Predictive Analysis | |
| WO2020185207A1 (en) | Computerized system and method for generative circuit design with machine-learned networks | |
| CN117436485A (en) | Multi-exit point end-edge-cloud cooperative system and method based on trade-off time delay and precision | |
| Shimonishi et al. | Energy optimization of distributed video processing system using genetic algorithm with bayesian attractor model | |
| Al-Tahmeesschi et al. | Applying deep neural networks for duty cycle estimation | |
| Kumar et al. | HPCWMF: a hybrid predictive cloud workload management framework using improved LSTM neural network | |
| CN113157344B (en) | DRL-based energy consumption perception task unloading method in mobile edge computing environment | |
| Phung et al. | A prediction based autoscaling in serverless computing | |
| CN112685176A (en) | Resource-constrained edge computing method for improving DDNN (distributed neural network) | |
| Li et al. | Big-Data-Driven Intelligent Wireless Network and Use Cases | |
| CN114077491B (en) | Industrial intelligent manufacturing edge computing task scheduling method |
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 |