CN116737394A - Dynamic adjustment security check centralized graph judging task allocation method - Google Patents
Dynamic adjustment security check centralized graph judging task allocation method Download PDFInfo
- Publication number
- CN116737394A CN116737394A CN202311013139.5A CN202311013139A CN116737394A CN 116737394 A CN116737394 A CN 116737394A CN 202311013139 A CN202311013139 A CN 202311013139A CN 116737394 A CN116737394 A CN 116737394A
- Authority
- CN
- China
- Prior art keywords
- task
- processing unit
- processing
- load
- load information
- 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
- 238000000034 method Methods 0.000 title claims abstract description 83
- 238000012545 processing Methods 0.000 claims abstract description 302
- 238000005457 optimization Methods 0.000 claims abstract description 86
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 48
- 238000007689 inspection Methods 0.000 claims abstract description 46
- 238000005303 weighing Methods 0.000 claims abstract description 5
- 239000002245 particle Substances 0.000 claims description 26
- 239000013598 vector Substances 0.000 claims description 22
- 238000011156 evaluation Methods 0.000 claims description 19
- 230000008569 process Effects 0.000 claims description 16
- 238000002922 simulated annealing Methods 0.000 claims description 11
- 230000002068 genetic effect Effects 0.000 claims description 10
- 238000002372 labelling Methods 0.000 claims description 10
- 238000012163 sequencing technique Methods 0.000 claims description 9
- 230000006641 stabilisation Effects 0.000 claims description 9
- 238000011105 stabilization Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 7
- 239000011159 matrix material Substances 0.000 claims description 7
- 230000035772 mutation Effects 0.000 claims description 6
- 230000001174 ascending effect Effects 0.000 claims description 5
- 238000009499 grossing Methods 0.000 claims description 4
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 238000011217 control strategy Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 claims description 3
- 238000009826 distribution Methods 0.000 claims description 3
- 238000012790 confirmation Methods 0.000 claims description 2
- 238000012795 verification Methods 0.000 claims description 2
- 230000006870 function Effects 0.000 description 20
- 230000000694 effects Effects 0.000 description 8
- 238000010276 construction Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 230000009471 action Effects 0.000 description 5
- 238000004590 computer program Methods 0.000 description 4
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 230000004075 alteration Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 208000006011 Stroke Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 210000000349 chromosome Anatomy 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 238000010353 genetic engineering Methods 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000002040 relaxant effect Effects 0.000 description 1
- 238000012958 reprocessing Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
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
- G06F9/50—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
- G06F9/5038—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 considering the execution order of a plurality of tasks, e.g. taking priority or time dependency constraints into consideration
-
- 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
- G06F9/4881—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06316—Sequencing of tasks or work
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2209/00—Indexing scheme relating to G06F9/00
- G06F2209/50—Indexing scheme relating to G06F9/50
- G06F2209/5021—Priority
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- General Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Educational Administration (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of security inspection and graph judgment, in particular to a dynamic adjustment security inspection centralized graph judgment task allocation method, which comprises the steps of firstly, collecting load information of a processing unit and calibrating the load information; based on the calibration load information and the current states of the task queue and the processing unit, obtaining an optimized task list by using a priority task processing algorithm, and realizing preliminary dynamic task allocation; then, performing task balance processing based on the dynamic task list after optimization processing, and performing task scheduling optimization by combining the task list after balance processing with predicted processing time; and finally, calculating the performance index of the current system based on the task list after the dispatching optimization, and adjusting the task allocation according to a preset weighing strategy to obtain a final dynamic task allocation list. The task allocation strategy adjusting method solves the technical problems that in the prior art, enough flexibility is not available in front of the temporary condition to quickly adjust the task allocation strategy and the task allocation efficiency is low.
Description
Technical Field
The invention relates to the technical field of security inspection graph judgment, in particular to a dynamic adjustment security inspection centralized graph judgment task allocation method.
Background
The centralized security check graph is a novel security check technology, and is generally aimed at the field places needing a large amount of image analysis, such as airport security check, port cargo check, large-scale activities and the like, wherein special staff is needed to judge images when people are subjected to security check. However, with the increase of the security inspection workload, the manual processing task may become ineffective, so how to ensure that the graph judging task is distributed uniformly, timely and efficiently among the graph judging staff becomes the focus of the security inspection concentrated graph judging attention.
The dynamically-adjusted security check centralized graph judging task method dynamically distributes the task to be processed according to the emergency degree, importance and capability and state of processing personnel through an intelligent distribution mechanism, so that the efficiency and effect of task processing are improved.
Wu Nana et al propose a "task allocation method, apparatus, system and computer storage medium", application number "CN 202010718488.7", mainly comprising: when the target security inspection equipment generates a task to be distributed, determining a graph judging workstation in an idle state in a target group associated with the target security inspection equipment; and distributing target graph judging workstations for the tasks to be distributed according to the preset priority order of the graph judging workstations in the idle state. The invention can avoid task allocation in the whole network range of the centralized graph judging system, and has the advantages of small bandwidth requirement on the backbone network of the system and suitability for the situation of limited bandwidth of the backbone network.
However, in the process of implementing the technical scheme of the embodiment of the application, the inventor discovers that the above technology has at least the following technical problems: the technical problems of rapid adjustment of task allocation strategies and low task allocation efficiency due to insufficient flexibility in the presence of temporary conditions.
Disclosure of Invention
The embodiment of the application solves the technical problems that the prior art lacks enough flexibility to quickly adjust the task allocation strategy and the task allocation efficiency is lower in the presence of temporary conditions by providing the dynamic adjustment security check centralized graph judging task allocation method, and realizes the technical effects of high-efficiency high-flexibility security check centralized graph judging task allocation.
The application provides a dynamic adjustment security check centralized graph judging task allocation method, which specifically comprises the following technical scheme:
a dynamically-adjusted security check centralized graph judging task distribution system comprises the following parts:
the system comprises a load information collection module, a load information calibration module, a dynamic task allocation module, a task balancing module, a task scheduling optimization module and a balance optimization module;
the load information collecting module is used for collecting the load information of the processing unit in real time and transmitting the load information to the load information calibrating module;
The load information calibration module is used for calculating the actual load of each processing unit based on the processing unit load information of the load information collection module and the task list of each processing unit from the balance optimization module, and calibrating the load information of the load information collection module;
the dynamic task allocation module is used for periodically detecting the states of the current task queue and the processing units, dynamically allocating tasks to the processing units based on the task priority and the load information of the processing units obtained by the load information calibration module, and obtaining a task list of each processing unit;
leading a priority task processing algorithm into the dynamic task allocation module to perform optimization processing, wherein the priority task processing algorithm searches whether priority tasks exist in a task list of each processing unit, and if so, moves the tasks to the front of the list to ensure that the tasks are processed preferentially;
the task balancing module is used for calculating the expected load of each processing unit based on the task list obtained by the dynamic task allocation module, and balancing the load of each processing unit by adjusting task allocation to obtain a balanced task list of each processing unit;
The task scheduling optimization module predicts the processing time of each task in each processing unit based on the task list of each processing unit processed by the task balancing module, and adjusts the sequence of the tasks based on the prediction result so as to reduce the overall processing delay and obtain the task list of each processing unit after the delay is minimized;
the trade-off optimization module calculates the current system performance index based on the task list of each processing unit processed by the task scheduling optimization module, and adjusts task allocation according to a preset trade-off strategy so as to find the optimal trade-off among dynamic task allocation, task balance and task processing delay, and obtain the task list of each processing unit after trade-off optimization.
A dynamic adjustment security check centralized graph judging task allocation method comprises the following steps:
s1, collecting load information of a processing unit and calibrating the load information;
s2, obtaining an optimized task list by using a priority task processing algorithm based on calibration load information and the current states of the task queue and the processing unit, and realizing preliminary dynamic task allocation; the priority task processing algorithm searches whether priority tasks exist in the task list of each processing unit, and if so, moves the tasks to the front of the list to ensure that the tasks are processed preferentially;
S3, performing task balance processing based on the task list after optimization processing, and performing task scheduling optimization by combining the task list after balance processing with predicted processing time; the task balancing processing is used for calculating the current load and the expected load of each processing unit by using a prediction algorithm based on the task list after the optimization processing; sequencing the processing units according to the ascending order of the expected load; sequencing the task queues to be processed according to the priority; starting from the task with the highest priority, distributing the task to the processing unit with the lowest expected load to obtain a balanced task list after the minimum load balancing processing;
and S4, calculating the performance index of the current system based on the task list after the dispatching optimization, and adjusting the task allocation according to a preset weighing strategy to obtain a final dynamic task allocation list.
Preferably, the step S1 specifically includes:
periodically collecting load information from each processing unit by using a load information collection module to obtain a load information data set, wherein ,/>，NRepresentation ofNA processing unit->Indicate->The original load information of the individual processing units,calculating an actual load of the processing unit and calibrating load information based on the original load information and the task list of each processing unit;
Preferably, in the step S1, the method further includes:
calculating an actual load of the processing unit and calibrating load information based on the original load information and the task list of each processing unit; the specific process is as follows:
firstly, constructing a prediction model; constructing a prediction model based on the collected original load information and a task list of each processing unit to obtain a predicted load of the processing unit;
and a second step of: load stabilization; carrying out smoothing treatment on the predicted load information by using a weight moving average method, and reducing a predicted error;
thirdly, calibrating the load;
preferably, the step S2 specifically includes:
based on the calibration load information and the current states of the task queue and the processing unit, obtaining an optimized task list by using a priority task processing algorithm, and realizing preliminary dynamic task allocation;
preferably, in the step S2, the method further includes:
before calculating task priority, determining task risk level by utilizing security inspection image characteristics; image labeling is carried out on the security inspection image by utilizing an image labeling algorithm based on optimization correlation matching; defining a detection matrix to determine descriptors of the security inspection image feature points in the image labeling process; in the image labeling process, a distance comparison method is utilized to carry out feature point matching verification; and defining a classification function by combining the depth image processing result of the case image, and realizing the level classification of the image to obtain the task risk level.
Preferably, in the step S2, the method further includes:
for confirmation of task priority, first, define a task set as，/>Representing the number of tasks, each task->The method comprises the steps of including task characteristics, wherein the task characteristics comprise complexity, risk level, importance, urgency, completion time, waiting time and system load; />；
Predicting the load calibration information by using a neural network prediction model to obtain a system load;
further, define an evaluation functionThe function accepts a task priority vector +.>, wherein />Representing task->And outputs an evaluation score which can be indicative of the efficiency of the task processing, defining an evaluation function +.>The following are provided:
，
wherein ,the weight coefficient representing the ith task is set by a professional according to the actual requirement of the task.
Preferably, in the step S2, the method further includes:
aiming at the evaluation function, a preliminary population is firstly generated, then selection, crossover and mutation operations are carried out through a genetic algorithm, a new population is generated, fine adjustment is carried out on the new population by adopting simulated annealing, and finally global optimization is carried out on the population by adopting a particle swarm optimization algorithm. By such a step, a task priority vector having the highest evaluation function value, that is, a required task priority is obtained.
Preferably, in the step S2, the method further includes:
constructing an optimized mixed planning model based on the load calibration information and the task priority, and realizing dynamic task allocation;
performing equivalent arrangement on the optimized mixed planning model to obtain a biased calculation processing model, wherein the biased calculation processing model is an equivalent model;
and solving an optimal solution for the equivalent model by using an optimal iteration method to obtain optimal allocation of the security check centralized graph judging task.
Preferably, the step S3 specifically includes:
calculating the current load and the expected load of each processing unit by using a prediction algorithm based on the task list after the optimization processing; sequencing the processing units according to the ascending order of the expected load; sequencing the task queues to be processed according to the priority; starting from the task with the highest priority, distributing the task to the processing unit with the lowest expected load to obtain a balanced task list after the minimum load balancing processing; and finally, predicting the waiting time and service time of each task in the allocated processing unit by adopting a queuing theory model, and adjusting the scheduling of the tasks according to the prediction result so as to minimize the overall processing delay.
Preferably, the step S4 specifically includes:
Calculating the performance index of the current system based on the task list after the dispatching optimization to obtain a set of processing units, and obtaining the load feedback of each processing unit by utilizing a dynamic feedback control strategy, wherein the feedback can represent the load state of the processing unit; and finally, searching a globally optimal task allocation strategy through particle swarm optimization, realizing the balance of a task list in scheduling optimization, and obtaining a final dynamic task allocation list.
The beneficial effects are that:
the technical schemes provided by the embodiment of the application have at least the following technical effects or advantages:
1. the method for calculating the actual load of the processing unit and calibrating the load information based on the original load information and the task list of each processing unit has obvious technical advantages. According to the method, the load state of each processing unit is accurately estimated by constructing a prediction model, carrying out load stabilization and load calibration, and efficient task allocation is realized, so that the processing efficiency and the resource utilization rate are improved.
2. According to the application, by comprehensively utilizing a genetic algorithm, simulated annealing, a particle swarm optimization algorithm, security inspection image characteristics and model construction, the dynamic allocation of tasks, the accurate determination of priorities, the accurate assessment of risk levels and the high efficiency and fairness of task processing are realized, the efficiency and the safety of security inspection task processing are greatly improved, and the method has high practical value.
3. The application can balance the workload of each processing unit through task balance processing, and avoid the condition that some processing units are idle and other processing units are overloaded, thereby fully utilizing the processing capacity of each processing unit, improving the overall efficiency and stability of the system, and by optimizing the scheduling, the system can dynamically allocate the tasks according to the priority of the tasks and the state of the processing units, so that the tasks with high priority can be processed in time, and the quality of service is improved.
4. The application combines three optimization strategies, can dynamically adjust the task allocation strategy according to actual conditions, has strong self-adaptability, reduces task processing delay by adopting Q-Learning and particle swarm optimization, and improves system realization efficiency.
5. The technical scheme of the application can effectively solve the technical problems that the task allocation strategy is not enough flexible to quickly adjust and the task allocation efficiency is low in the presence of the temporary condition, and the system or the method has obvious technical advantages through a series of effect researches and the method for calculating the actual load of the processing unit and calibrating the load information based on the original load information and the task list of each processing unit. The load state of each processing unit is accurately estimated by constructing a prediction model, carrying out load stabilization and load calibration, so that efficient task allocation is realized; the dynamic allocation of tasks, the accurate determination of priorities, the accurate assessment of risk levels and the high efficiency and fairness of task processing are realized by comprehensively utilizing a genetic algorithm, a simulated annealing algorithm, a particle swarm optimization algorithm, security inspection image characteristics and model construction; the workload of each processing unit can be balanced through the task balancing processing, and the condition that some processing units are idle and other processing units are overloaded is avoided, so that the processing capacity of each processing unit is fully utilized, the tasks with high priority are timely processed, and the quality of service is improved; the method combines three optimization strategies, can dynamically adjust the task allocation strategy according to actual conditions, has strong self-adaptability, reduces task processing delay by adopting Q-Learning and particle swarm optimization, and improves system realization efficiency flexibility.
Drawings
FIG. 1 is a block diagram of a dynamically adjusted centralized decision task allocation system for security inspection according to the present application;
FIG. 2 is a flow chart of a method for distributing tasks of a dynamically adjusted security check centralized judgment chart according to the application;
Detailed Description
The embodiment of the application solves the technical problems that the prior art lacks enough flexibility to quickly adjust the task allocation strategy and the task allocation efficiency is lower in the presence of temporary conditions by providing a dynamic adjustment security check centralized graph judging task allocation method, and the overall thinking is as follows:
firstly, collecting load information of a processing unit, and calibrating the load information; based on the calibration load information and the current states of the task queue and the processing unit, obtaining an optimized task list by using a priority task processing algorithm, and realizing preliminary dynamic task allocation; then, performing task balance processing based on the task list after optimization processing, and performing task scheduling optimization by combining the task list after balance processing with predicted processing time; finally, calculating the performance index of the current system based on the task list after dispatching optimization, and adjusting the task allocation according to a preset weighing strategy to obtain a final dynamic task allocation list; the method of calculating the actual load of the processing units and calibrating the load information based on the raw load information and the task list of each processing unit has significant technical advantages. The load state of each processing unit is accurately estimated by constructing a prediction model, carrying out load stabilization and load calibration, so that efficient task allocation is realized; the dynamic allocation of tasks, the accurate determination of priorities, the accurate assessment of risk levels and the high efficiency and fairness of task processing are realized by comprehensively utilizing a genetic algorithm, a simulated annealing algorithm, a particle swarm optimization algorithm, security inspection image characteristics and model construction; the workload of each processing unit can be balanced through the task balancing processing, and the condition that some processing units are idle and other processing units are overloaded is avoided, so that the processing capacity of each processing unit is fully utilized, the tasks with high priority are timely processed, and the quality of service is improved; the method combines three optimization strategies, can dynamically adjust the task allocation strategy according to actual conditions, has strong self-adaptability, reduces task processing delay by adopting Q-Learning and particle swarm optimization, and improves system realization efficiency flexibility.
In order to better understand the above technical solutions, the following detailed description will refer to the accompanying drawings and specific embodiments.
Referring to fig. 1, the task allocation system for dynamically adjusting the security check centralized graph judgment comprises a load information collection module, a load information calibration module, a dynamic task allocation module, a task balancing module, a task scheduling optimization module and a balance optimization module;
the load information collecting module is used for collecting the load information of the processing unit in real time and transmitting the load information to the load information calibrating module;
the load information calibration module is used for calculating the actual load of each processing unit based on the processing unit load information of the load information collection module and the task list of each processing unit from the balance optimization module, and calibrating the load information of the load information collection module;
the dynamic task allocation module is used for periodically detecting the states of the current task queue and the processing units, dynamically allocating tasks to the processing units based on the task priority and the load information of the processing units obtained by the load information calibration module, and obtaining a task list of each processing unit;
Leading a priority task processing algorithm into the dynamic task allocation module to perform optimization processing, wherein the priority task processing algorithm searches whether priority tasks exist in a task list of each processing unit, and if so, moves the tasks to the front of the list to ensure that the tasks are processed preferentially;
the task balancing module is used for calculating the expected load of each processing unit based on the task list obtained by the dynamic task allocation module, and balancing the load of each processing unit by adjusting task allocation to obtain a balanced task list of each processing unit;
the task scheduling optimization module predicts the processing time of each task in each processing unit based on the task list of each processing unit processed by the task balancing module, and adjusts the sequence of the tasks based on the prediction result so as to reduce the overall processing delay and obtain the task list of each processing unit after the delay is minimized;
the trade-off optimization module is used for calculating the current system performance index based on the task list of each processing unit processed by the task scheduling optimization module, and adjusting task allocation according to a preset trade-off strategy so as to find the optimal trade-off among dynamic task allocation, task balance and task processing delay and obtain a task list of each processing unit after trade-off optimization;
Referring to fig. 2, the method for distributing the task of dynamically-adjusted centralized security check graph judgment comprises the following steps:
s1, collecting load information of a processing unit and calibrating the load information;
periodically collecting load information from each processing unit by using a load information collection module to obtain a load information data setThe load information includes the number of tasks currently being processed, CPU utilization rate and memory utilization; wherein,，Nrepresentation ofNA processing unit->Indicate->Original load information of the individual processing units;
calculating an actual load of the processing unit and calibrating load information based on the original load information and the task list of each processing unit; the specific process is as follows:
firstly, constructing a prediction model; a predictive model is built based on the collected raw load information and the task list for each processing unit,
，
wherein ,is a processing unit->Predicted load of +.>Is a feature in the task list, such as number of tasks, task complexity, < >>Is a parameter to be trained and is obtained by training through historical data.
And a second step of: load stabilization; and carrying out smoothing treatment on the predicted load information by using a weight moving average method to reduce the prediction error, wherein the calculation formula is as follows:
，
wherein ,representation processing unitiIs +.>Representing the predicted load (derived from a linear regression model) of the processing unit i at different time points>Is a weight, and weight assignment is generally carried out by setting a weight larger than the weight of the more recent data;
thirdly, calibrating the load; obtaining predicted load and original load information after smooth prediction processing by using task list data based on each processing unit, finding out K processing units closest to the current processing unit, and calibrating by using average loads of the processing units; the calculation formula is as follows:
，
wherein ,representing the calibration load of the processing unit i +.>The predicted load after the smoothing process of the processing unit j closest to the processing unit i is represented.
Based on the above steps, load calibration based on the raw load information and the task list of each processing unit is achieved, thereby calculating actual load information of the processing unit.
The method for calculating the actual load of the processing unit and calibrating the load information based on the original load information and the task list of each processing unit has obvious technical advantages. According to the method, the load state of each processing unit is accurately estimated by constructing a prediction model, carrying out load stabilization and load calibration, and efficient task allocation is realized, so that the processing efficiency and the resource utilization rate are improved.
S2, obtaining an optimized task list by using a priority task processing algorithm based on calibration load information and the current states of the task queue and the processing unit, and realizing preliminary dynamic task allocation;
when generating the graph judging task, any task can be given a priority score based on the characteristics of the task, wherein the score is based on various factors such as waiting time of the task, scale of the task and risk level of the task; the specific process of calculating the task priority is as follows:
before calculating the task priority, determining the task risk level by utilizing the security image characteristics according to the determination of the task risk level, wherein the task risk level is determined by the specific process:
taking any one security inspection image as an example, carrying out image annotation on the security inspection image by utilizing an image annotation algorithm based on optimization correlation matching on the case image subjected to conventional pretreatment;
step one, determining a template library; the template library contains objects or features found in the image. For example, if it is desired to find dangerous objects such as explosives, knives, etc. in the security inspection image, a template for these objects needs to be obtained;
secondly, extracting features, namely extracting features of template images in a template library and the preprocessed security inspection images;
For each security inspection image pixel pointFeature extraction is performed by defining a detection matrix:
，
、/>、/>representation of the image after the action of the Gaussian function>Second partial derivative of (i), i.e，/>，/>；/>，Representing standard deviation +.>Is a Gaussian function of->Representing convolution operator ++>Representing the image to be feature extracted, monitoring feature points by detecting determinant and trace of the matrix,
，
，
the feature points are defined as determinantIs the maximum point of (2). I.e. if +.>The value is larger than the value of all other points in the neighborhood, the point is defined as a characteristic point, and the descriptor of the characteristic point is determined based on determinant and trace of the characteristic point;
further, a characteristic point descriptor set of the security inspection image is obtainedAnd feature point descriptor set for template image； wherein ,/>，/>, wherein ,P、Qthe number of descriptors respectively representing the security inspection image and the template image,/-for each of the security inspection image and the template image>First representing security imagePPersonal descriptor->Representing the first of the template imagesQA personal descriptor;
thirdly, matching the characteristics; searching for matching feature points between the template image and the security inspection image to obtain feature matching points; finding out matched characteristic points according to the distance between the template image and the security inspection image descriptor, and matching by using a ratio test of nearest neighbors and secondary neighbors;
Taking any one descriptor in a security image descriptor set as an example，/>The corresponding Euclidean distance set is as follows:
，
，
wherein ,Nrepresenting the dimension of the descriptor, i.e. the number of elements in any one descriptor; the characteristic points are matched by a distance comparison method,
specifically, first find the Euclidean distance setThe minimum value of (2) as the nearest Euclidean distance +.>Then in the rest of the Euclidean distance set +.>Finding the minimum as the next closest Euclidean distance +.>Then, matching judgment is carried out by utilizing the nearest Euclidean distance and the next nearest Euclidean distance, and a specific judgment formula is as follows:
，
wherein ,is a threshold value, if the ratio of the nearest euclidean distance to the next nearest euclidean distance is less than the threshold value, we consider that a match is found;
fourth, verifying and confirming the matching; adopting the existing RANSAC algorithm to verify matching;
and fifthly, labeling the images. And based on the correct matching result, associating the template image with the corresponding area of the security inspection image, finding the same template image in the security inspection image, and marking.
The method comprises the steps of carrying out image labeling on a security inspection image, carrying out feature extraction and image recognition processing on the labeled security inspection image by utilizing image processing and a machine learning algorithm, carrying out level classification on the image according to an image recognition result, generating a corresponding image judging task according to an image level classification result, and determining the risk level of the image judging task; the specific process is as follows: results based on security inspection image recognition The identification result can comprise the number, the size and the attribute, the identification accuracy rate and the security inspection images are divided into different categories. Category a: high risk items (e.g., knives, guns, etc.), category B: stroke risk items (e.g., bulk liquids, electronics, etc.), category C: low risk articles; defining a classification function by combining depth image processing results of security inspection images>，
，
wherein ,the method comprises the steps of carrying out a first treatment on the surface of the Representing the total number of image recognition results +.>Attribute representing image recognition result, ++>Weights representing attributes of the image recognition result, +.>Representing the first of weight definitions based on image recognition result attributessA weight value of the individual image recognition result; according to->The value is taken, the security inspection images are subjected to image classification, and further, the classification result of the security inspection images is obtained;
based on the classification result, obtaining a task risk level;
further, determining the task priority and defining the task set as，/>Representing the number of tasks, each task->Including its task characteristics, such as complexity, risk level, importance, urgency, completion time, latency, system load; />；
Predicting the load calibration information by using a neural network prediction model to obtain a system load;
Further, define an evaluation functionThe function accepts a task priority vector +.>（ wherein />Representing task->And outputs an evaluation score. This evaluation score may represent the efficiency of task processing. The present application defines an evaluation function->The following are provided:
，
wherein ,the weight coefficient of the ith task is represented and is set by a professional according to the actual requirement of the task;
further, generating an initial population; by priority vectorAs a population. Let the number of individuals in the population be K, each vector comprising +.>Elements, which are priorities of the corresponding tasks, which may be randomly generated in an initial stage;
further, genetic manipulation processes including selection, crossover and mutation; the selection is performed by roulette selection, in which the evaluation function value of each individual in the population is set to beThe probability that the individual is selected is +.>，/>Represent the firstiThe task priority vectors are used for ensuring that individuals with high evaluation function values have a larger probability of being selected; the crossover operation is performed between two selected vectors, specifically, randomly selecting a crossover point, then exchanging all elements of the two vectors after this point, creating two new vectors each containing a part of the elements of the parent vector, which can be regarded as simulating chromosome crossover in biological genetics, and randomly selecting two individuals for crossover; the mutation operation is performed on a single vector, specifically, a point is randomly selected, and then a small random value is increased or decreased at this point. This can be seen as a genetic mutation in the simulated biological inheritance;
Further, simulated annealing adjustment is performed, and fine adjustment is performed for each newly generated priority vector by using simulated annealing. Setting a fine tuning valueThe new vector after fine tuning isLet the current temperature be +.>(empirically set), if->Accepting the new vector, otherwise, probability +.>Accepting the new vector;
further, particle swarm optimization adjustment, particle swarm optimization algorithm optimizes the population, for each of the populationIndividual particles, positions of particlesRepresenting task priority vectors, while the speed of the particles is +.>Representing the change speed of the task priority vector, searching the evaluation function value in a larger direction by adjusting the speed and the position of the particles, and combining the historical optimal positionAnd global optimum position->The velocity and position update formula for the particles is as follows:
，
，
wherein ,is a particleiSpeed of->Is inertial weight, ++> and />Is a learning factor, < >>、/>Random number of [0,1 ], by->And (5) generating a function.
Finally, if the preset iteration times are reached or the optimal evaluation function value is not changed obviously, stopping the algorithm and outputting the current optimal task priority vector; otherwise, returning to the genetic operation for reprocessing.
The method comprises the steps of firstly generating a preliminary population, then performing selection, crossing and mutation operations through a genetic algorithm to generate a new population, performing fine adjustment on the new population by adopting simulated annealing, and finally performing global optimization on the population by adopting a particle swarm optimization algorithm. By such a step, a task priority vector having the highest evaluation function value, that is, a required task priority is obtained.
Constructing an optimized mixed planning model based on the load calibration information and the task priority, and realizing dynamic task allocation;
task priority vector（ wherein />Representing task->Priority of (c) load calibration information set +.>，/>Load calibration information indicating the jth processing unit, for example>The method comprises the steps of carrying out a first treatment on the surface of the The present application defines a task allocation matrix:
，
if task i is assigned to processing unit j, thenOtherwise->；
Defining an optimized mixed planning model by maximizing task completion and minimizing processing unit pressure, wherein the optimized mixed planning model is specifically as follows:
objective function:
，
wherein ,representing the load capacity in the processing unit load information, < >>Representing the average load capacity in the load information of all processing units,/for each processing unit>The weight parameters are expressed and used for balancing the importance of the two targets, and the weight parameters are determined by the staff;
constraint conditions:
each task can only be assigned to one person:；
defining constraints based on load calibration information of the processing unit, such as: the workload of a processing unit cannot exceed its state allowed maximum:，/>representing the maximum bearing capacity of the processing unit;
further, the model construction process is equivalently arranged to obtain the following calculation processing model,
，
，
wherein ,representing a matrix of variable coefficients>Variable representing the task completion rate influence,/->By->、/>、/>Determination of->Representing weight parameters->Representing a matrix of variable coefficients>Representing the pressure influencing variable of the processing unit,/->Is->Is a subset of the group; />、/>Indicating needleConstraint function on variable X, Y, +.>、/>Representation +.>、/>Is a constraint value of (2);
further, an optimal iteration method is utilized to solve the optimal solution for the model, and the specific process is as follows:
step one, finding an initial solution; relaxing all integer constraints of the problem to continuous constraints, converting the solution problem to a linear programming problem, and then using a conventional linear programming solver (e.g., simplex method) to find an initial solution that may not satisfy all constraints；
Second, a decision variable with fractional part is selectedBranching is performed creating two new problems. One is +.>The problem of rounding down, one is +.>And (3) finishing. This creates two new sub-problems, each one with one integer constraint less than the original problem;
thirdly, repeating the first step for each sub-problem, solving the relaxation problem to obtain a boundary, and if the boundary of a certain sub-problem is worse than the currently known optimal solution, discarding the sub-problem and not continuing to solve;
Fourth, repeating the second and third steps until all the sub-problems are solved or discarded, wherein the found optimal solution is the optimal solution of the whole problem;
further, the optimal solution obtained by the model is the optimal task allocation, and the dynamic task allocation is realized for the processing unit.
According to the application, by comprehensively utilizing a genetic algorithm, simulated annealing, a particle swarm optimization algorithm, security inspection image characteristics and model construction, the dynamic allocation of tasks, the accurate determination of priorities, the accurate assessment of risk levels and the high efficiency and fairness of task processing are realized, the efficiency and the safety of security inspection task processing are greatly improved, and the method has high practical value.
S3, performing task balance processing based on the task list after optimization processing, and performing task scheduling optimization by combining the task list after balance processing with predicted processing time;
calculating the current load and the expected load of each processing unit by using a prediction algorithm based on the task list after optimization processing, wherein the expected load=the current load+the load of the task to be processed; sequencing the processing units according to the ascending order of the expected load; sorting the task queues to be processed by priority (assuming that higher numbers represent higher priorities); and finally, predicting the waiting time and service time of each task in the allocated processing unit by adopting a queuing theory model, and adjusting the scheduling of the tasks according to the prediction result to minimize the overall processing delay. Task scheduling can be expressed as the following mathematical formula: wherein ,/>Representing the waiting time of task i in the assigned processing unit,/->Representing task i on an assigned processing unitService time->Representing a set of tasks; and the dispatching optimization of the balanced task list is realized.
The application can balance the workload of each processing unit through task balance processing, and avoid the condition that some processing units are idle and other processing units are overloaded, thereby fully utilizing the processing capacity of each processing unit, improving the overall efficiency and stability of the system, and by optimizing the scheduling, the system can dynamically allocate the tasks according to the priority of the tasks and the state of the processing units, so that the tasks with high priority can be processed in time, and the quality of service is improved.
S4, calculating the performance index of the current system based on the task list after dispatching optimization, and adjusting the task allocation according to a preset weighing strategy to obtain a final dynamic task allocation list;
based on the task list after dispatching optimization, calculating the performance index of the current system to obtain the set of processing units asEach processing unit->The load at time t is +.>Obtaining load feedback of each processing unit by using a dynamic feedback control strategy >The feedback may characterize the load state of the processing unit:
，
wherein ,representing the average negativity of all processing unitsLoad feedback of the load>The method is used for judging whether the load of the processing unit exceeds the average load or not and is used as a basis for subsequent optimization;
further, the processing unit is defined by adopting Q-Learning to realize intelligent optimization of task allocation of the processing unitStatus of->Feedback for its current load>. For each state +.>There are a series of possible actions +.>Each action represents a possible task allocation strategy, learning a +.>Value function->For evaluating the status->Take action with->Utility of->The iterative formula of the value function is as follows:
，
wherein ,is learning rate (I/O)>Is a discount factor, < >>Is to take action->Instant rewards obtained later, ->Andrepresenting a new state and action, respectively. Instant rewards->May be defined as an improvement in load balancing; and finally, searching a globally optimal task allocation strategy through particle swarm optimization, and realizing the balance of a task list in scheduling optimization to obtain a final dynamic task allocation list.
The application combines three optimization strategies, can dynamically adjust the task allocation strategy according to actual conditions, has strong self-adaptability, reduces task processing delay by adopting Q-Learning and particle swarm optimization, and improves system realization efficiency.
In summary, the method for distributing the task of the security check centralized judgment chart with dynamic adjustment is completed.
The technical scheme provided by the embodiment of the application at least has the following technical effects or advantages:
1. the method for calculating the actual load of the processing unit and calibrating the load information based on the original load information and the task list of each processing unit has obvious technical advantages. According to the method, the load state of each processing unit is accurately estimated by constructing a prediction model, carrying out load stabilization and load calibration, and efficient task allocation is realized, so that the processing efficiency and the resource utilization rate are improved.
2. According to the application, by comprehensively utilizing a genetic algorithm, simulated annealing, a particle swarm optimization algorithm, security inspection image characteristics and model construction, the dynamic allocation of tasks, the accurate determination of priorities, the accurate assessment of risk levels and the high efficiency and fairness of task processing are realized, the efficiency and the safety of security inspection task processing are greatly improved, and the method has high practical value.
3. The application can balance the workload of each processing unit through task balance processing, and avoid the condition that some processing units are idle and other processing units are overloaded, thereby fully utilizing the processing capacity of each processing unit, improving the overall efficiency and stability of the system, and by optimizing the scheduling, the system can dynamically allocate the tasks according to the priority of the tasks and the state of the processing units, so that the tasks with high priority can be processed in time, and the quality of service is improved.
4. The application combines three optimization strategies, can dynamically adjust the task allocation strategy according to actual conditions, has strong self-adaptability, reduces task processing delay by adopting Q-Learning and particle swarm optimization, and improves system realization efficiency.
Effect investigation:
the technical scheme of the application can effectively solve the technical problems that the task allocation strategy is not enough flexible to quickly adjust and the task allocation efficiency is low in the presence of the temporary condition, and the system or the method has obvious technical advantages through a series of effect researches and the method for calculating the actual load of the processing unit and calibrating the load information based on the original load information and the task list of each processing unit. The load state of each processing unit is accurately estimated by constructing a prediction model, carrying out load stabilization and load calibration, so that efficient task allocation is realized; the dynamic allocation of tasks, the accurate determination of priorities, the accurate assessment of risk levels and the high efficiency and fairness of task processing are realized by comprehensively utilizing a genetic algorithm, a simulated annealing algorithm, a particle swarm optimization algorithm, security inspection image characteristics and model construction; the workload of each processing unit can be balanced through the task balancing processing, and the condition that some processing units are idle and other processing units are overloaded is avoided, so that the processing capacity of each processing unit is fully utilized, the tasks with high priority are timely processed, and the quality of service is improved; the method combines three optimization strategies, can dynamically adjust the task allocation strategy according to actual conditions, has strong self-adaptability, reduces task processing delay by adopting Q-Learning and particle swarm optimization, and improves system realization efficiency flexibility.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
Claims (11)
1. A dynamically-adjusted security check centralized graph judging task distribution system comprises the following parts:
the system comprises a load information collection module, a load information calibration module, a dynamic task allocation module, a task balancing module, a task scheduling optimization module and a balance optimization module;
the load information collecting module is used for collecting the load information of the processing unit in real time and transmitting the load information to the load information calibrating module;
the load information calibration module is used for calculating the actual load of each processing unit based on the processing unit load information of the load information collection module and the task list of each processing unit from the balance optimization module, and calibrating the load information of the load information collection module;
The dynamic task allocation module is used for periodically detecting the states of the current task queue and the processing units, dynamically allocating tasks to the processing units based on the task priority and the load information of the processing units obtained by the load information calibration module, and obtaining a task list of each processing unit;
leading a priority task processing algorithm into the dynamic task allocation module to perform optimization processing, wherein the priority task processing algorithm searches whether priority tasks exist in a task list of each processing unit, and if so, moves the tasks to the front of the list to ensure that the tasks are processed preferentially;
the task balancing module is used for calculating the expected load of each processing unit based on the task list obtained by the dynamic task allocation module, and balancing the load of each processing unit by adjusting task allocation to obtain a balanced task list of each processing unit;
the task scheduling optimization module predicts the processing time of each task in each processing unit based on the task list of each processing unit processed by the task balancing module, and adjusts the sequence of the tasks based on the prediction result so as to reduce the overall processing delay and obtain the task list of each processing unit after the delay is minimized;
The trade-off optimization module calculates the current system performance index based on the task list of each processing unit processed by the task scheduling optimization module, and adjusts task allocation according to a preset trade-off strategy so as to find the optimal trade-off among dynamic task allocation, task balance and task processing delay, and obtain the task list of each processing unit after trade-off optimization.
2. A dynamic adjustment security check centralized graph judging task allocation method comprises the following steps:
s1, collecting load information of a processing unit and calibrating the load information;
s2, obtaining an optimized task list by using a priority task processing algorithm based on calibration load information and the current states of the task queue and the processing unit, and realizing preliminary dynamic task allocation; the priority task processing algorithm searches whether priority tasks exist in the task list of each processing unit, and if so, moves the tasks to the front of the list to ensure that the tasks are processed preferentially;
s3, performing task balance processing based on the task list after optimization processing, and performing task scheduling optimization by combining the task list after balance processing with predicted processing time; the task balancing processing is used for calculating the current load and the expected load of each processing unit by using a prediction algorithm based on the task list after the optimization processing; sequencing the processing units according to the ascending order of the expected load; sequencing the task queues to be processed according to the priority; starting from the task with the highest priority, distributing the task to the processing unit with the lowest expected load to obtain a balanced task list after the minimum load balancing processing;
And S4, calculating the performance index of the current system based on the task list after the dispatching optimization, and adjusting the task allocation according to a preset weighing strategy to obtain a final dynamic task allocation list.
3. The method for distributing the task of dynamically-adjusted centralized security check graph judgment according to claim 2, wherein the step S1 specifically comprises:
periodically collecting load information from each processing unit by using a load information collection module to obtain a load information data set, wherein ,/>，NRepresentation ofNA processing unit->Indicate->The original load information of the individual processing units,based on the raw load information and the task list for each processing unit, the actual load of the processing unit is calculated and the load information is calibrated.
4. The method for distributing tasks for dynamically adjusting centralized judgment of security check according to claim 3, wherein in step S1, further comprising:
calculating an actual load of the processing unit and calibrating load information based on the original load information and the task list of each processing unit; the specific process is as follows:
firstly, constructing a prediction model; constructing a prediction model based on the collected original load information and a task list of each processing unit to obtain a predicted load of the processing unit;
And a second step of: load stabilization; carrying out smoothing treatment on the predicted load information by using a weight moving average method, and reducing a predicted error;
and thirdly, calibrating the load.
5. The method for distributing the task of dynamically-adjusted centralized security check graph judgment according to claim 2, wherein the step S2 specifically comprises:
based on the calibration load information and the current states of the task queue and the processing unit, a task list after optimization processing is obtained by using a priority task processing algorithm, and preliminary dynamic task allocation is realized.
6. The method for assigning tasks to a dynamically adjusted security check centralized judgment chart according to claim 5, wherein in step S2, the method further comprises:
before calculating task priority, determining task risk level by utilizing security inspection image characteristics;
image labeling is carried out on the security inspection image by utilizing an image labeling algorithm based on optimization correlation matching;
defining a detection matrix to determine descriptors of the security inspection image feature points in the image labeling process;
in the image labeling process, a distance comparison method is utilized to carry out feature point matching verification;
and defining a classification function by combining the depth image processing result of the case image, and realizing the level classification of the image to obtain the task risk level.
7. The method for distributing the task of dynamically adjusting the centralized judgment chart in the security inspection according to claim 5, wherein in the step S2, further comprises:
for confirmation of task priority, first, define a task set as，/>Representing the number of tasks, eachPersonal task->The method comprises the steps of including task characteristics, wherein the task characteristics comprise complexity, risk level, importance, urgency, completion time, waiting time and system load; />；
Predicting the load calibration information by using a neural network prediction model to obtain a system load;
further, define an evaluation functionThe function accepts a task priority vector +.>, wherein />Representing task->And outputs an evaluation score which can be indicative of the efficiency of the task processing, defining an evaluation function +.>The following are provided:
，
wherein ,the weight coefficient representing the ith task is set by a professional according to the actual requirement of the task.
8. The method for distributing the task of dynamically adjusting the centralized judgment chart in the security inspection according to claim 7, wherein in the step S2, the method further comprises:
aiming at the evaluation function, firstly, generating a preliminary population, then performing selection, crossing and mutation operations through a genetic algorithm to generate a new population, performing fine adjustment on the new population by adopting simulated annealing, and finally, performing global optimization on the population by adopting a particle swarm optimization algorithm; by such a step, a task priority vector having the highest evaluation function value, that is, a required task priority is obtained.
9. The method for distributing the task of dynamically adjusting the centralized judgment chart in the security inspection according to claim 8, wherein in the step S2, the method further comprises:
constructing an optimized mixed planning model based on the load calibration information and the task priority, and realizing dynamic task allocation;
performing equivalent arrangement on the optimized mixed planning model to obtain a biased calculation processing model, wherein the biased calculation processing model is an equivalent model;
and solving an optimal solution for the equivalent model by using an optimal iteration method to obtain optimal allocation of the security check centralized graph judging task.
10. The method for distributing the task of dynamically-adjusted centralized judgment of the security check graph according to claim 2, wherein the step S3 specifically comprises:
calculating the current load and the expected load of each processing unit by using a prediction algorithm based on the task list after the optimization processing; sequencing the processing units according to the ascending order of the expected load; sequencing the task queues to be processed according to the priority; starting from the task with the highest priority, distributing the task to the processing unit with the lowest expected load to obtain a balanced task list after the minimum load balancing processing; and finally, predicting the waiting time and service time of each task in the allocated processing unit by adopting a queuing theory model, and adjusting the scheduling of the tasks according to the prediction result so as to minimize the overall processing delay.
11. The method for distributing the task of dynamically-adjusted centralized judgment of the security check graph according to claim 2, wherein the step S4 specifically comprises:
calculating the performance index of the current system based on the task list after the dispatching optimization to obtain a set of processing units, and obtaining the load feedback of each processing unit by utilizing a dynamic feedback control strategy, wherein the feedback can represent the load state of the processing unit; and finally, searching a globally optimal task allocation strategy through particle swarm optimization, realizing the balance of a task list in scheduling optimization, and obtaining a final dynamic task allocation list.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311013139.5A CN116737394B (en) | 2023-08-14 | 2023-08-14 | Dynamic adjustment security check centralized graph judging task allocation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311013139.5A CN116737394B (en) | 2023-08-14 | 2023-08-14 | Dynamic adjustment security check centralized graph judging task allocation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116737394A true CN116737394A (en) | 2023-09-12 |
| CN116737394B CN116737394B (en) | 2023-10-27 |
Family
ID=87913724
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311013139.5A Active CN116737394B (en) | 2023-08-14 | 2023-08-14 | Dynamic adjustment security check centralized graph judging task allocation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116737394B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117709690A (en) * | 2024-02-05 | 2024-03-15 | 人民法院信息技术服务中心 | Judicial case task scheduling method and device |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20070032441A (en) * | 2005-09-16 | 2007-03-22 | 성균관대학교산학협력단 | Load balancing system based on fuzzy grouping and the load balancing method |
| US20120204176A1 (en) * | 2010-10-29 | 2012-08-09 | Huawei Technologies Co., Ltd. | Method and device for implementing load balance of data center resources |
| CN106371908A (en) * | 2016-08-31 | 2017-02-01 | 武汉鸿瑞达信息技术有限公司 | Optimization method for image/video filter task distribution based on PSO (Particle Swarm Optimization) |
| CN108170530A (en) * | 2017-12-26 | 2018-06-15 | 北京工业大学 | A kind of Hadoop Load Balancing Task Scheduling methods based on mixing meta-heuristic algorithm |
| CN110688219A (en) * | 2019-09-05 | 2020-01-14 | 浙江理工大学 | Adaptive weight load balancing algorithm based on reverse chaotic cuckoo search |
| CN110781003A (en) * | 2019-10-24 | 2020-02-11 | 重庆邮电大学 | Load balancing method for particle swarm fusion variation control |
| CN115509926A (en) * | 2022-10-09 | 2022-12-23 | 信通院(河北)科技创新研究院有限公司 | Multi-task scheduling cloud testing method based on improved particle swarm optimization |
-
2023
- 2023-08-14 CN CN202311013139.5A patent/CN116737394B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20070032441A (en) * | 2005-09-16 | 2007-03-22 | 성균관대학교산학협력단 | Load balancing system based on fuzzy grouping and the load balancing method |
| US20120204176A1 (en) * | 2010-10-29 | 2012-08-09 | Huawei Technologies Co., Ltd. | Method and device for implementing load balance of data center resources |
| CN106371908A (en) * | 2016-08-31 | 2017-02-01 | 武汉鸿瑞达信息技术有限公司 | Optimization method for image/video filter task distribution based on PSO (Particle Swarm Optimization) |
| CN108170530A (en) * | 2017-12-26 | 2018-06-15 | 北京工业大学 | A kind of Hadoop Load Balancing Task Scheduling methods based on mixing meta-heuristic algorithm |
| CN110688219A (en) * | 2019-09-05 | 2020-01-14 | 浙江理工大学 | Adaptive weight load balancing algorithm based on reverse chaotic cuckoo search |
| CN110781003A (en) * | 2019-10-24 | 2020-02-11 | 重庆邮电大学 | Load balancing method for particle swarm fusion variation control |
| CN115509926A (en) * | 2022-10-09 | 2022-12-23 | 信通院(河北)科技创新研究院有限公司 | Multi-task scheduling cloud testing method based on improved particle swarm optimization |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117709690A (en) * | 2024-02-05 | 2024-03-15 | 人民法院信息技术服务中心 | Judicial case task scheduling method and device |
| CN117709690B (en) * | 2024-02-05 | 2024-04-12 | 人民法院信息技术服务中心 | Judicial case task scheduling method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116737394B (en) | 2023-10-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20200364664A1 (en) | Method for determining transportation scheme, method for training fast loading model, and device | |
| Radiuk | Impact of training set batch size on the performance of convolutional neural networks for diverse datasets | |
| Zhou et al. | Multiobjective evolutionary algorithms: A survey of the state of the art | |
| CN106446959B (en) | A kind of cloud computing resources dynamic matching method and device | |
| CN105589806B (en) | A kind of software defect tendency Forecasting Methodology based on SMOTE+Boosting algorithms | |
| CN116737394B (en) | Dynamic adjustment security check centralized graph judging task allocation method | |
| WO2020098030A1 (en) | Scheduling method for request task and scheduling center server | |
| CN106845642B (en) | A kind of adaptive multi-target evolution method of belt restraining cloud workflow schedule | |
| CN113190351B (en) | Efficient resource distribution system for distributed deep learning training task | |
| CN111028100A (en) | Refined short-term load prediction method, device and medium considering meteorological factors | |
| CN108564136A (en) | A kind of airspace operation Situation Assessment sorting technique based on fuzzy reasoning | |
| CN110751411B (en) | Cloud manufacturing task oriented manufacturing resource matching method | |
| CN115685941B (en) | Machine room operation regulation and control method and device based on machine cabinet hot spot temperature prediction | |
| US20240345567A1 (en) | Systems and methods for smart gas full-cycle supervision based on internet of things | |
| CN113032367A (en) | Dynamic load scene-oriented cross-layer configuration parameter collaborative tuning method and system for big data system | |
| CN111260161B (en) | Method and equipment for issuing crowdsourcing task | |
| CN111340581A (en) | Data processing method and device, readable storage medium and electronic equipment | |
| CN117912235B (en) | Planning data processing method and system for smart city | |
| CN110689140A (en) | Method for intelligently managing rail transit alarm data through big data | |
| CN108808657B (en) | Short-term prediction method for power load | |
| CN114650321A (en) | Task scheduling method for edge computing and edge computing terminal | |
| CN105718956A (en) | Overload determining method based on support vector machine | |
| CN117350509A (en) | Task automation plan scheduling method and system | |
| CN117437507A (en) | Prejudice evaluation method for evaluating image recognition model | |
| CN115936184B (en) | Load prediction matching method suitable for multi-user types |
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 |