53results about How to "Improve task completion rate" patented technology

The invention relates to the technical field of unmanned aerial vehicles, and discloses an inspection unmanned aerial vehicle flight control management method, which comprises the following steps: step 1, an unmanned aerial vehicle obtains environment three-dimensional point cloud data through a laser radar, captures terrain visual information through an RGB camera and a depth camera, obtains real-time position and attitude information of the unmanned aerial vehicle through a GPS and an inertial measurement unit, and obtains the real-time position and attitude information of the unmanned aerial vehicle; and fusing the collected data to construct a spatial grid map, and taking the grid map as input data of path planning. A high-precision space grid map is generated through three-dimensional environment perception and modeling in combination with an unmanned aerial vehicle multi-sensor fusion technology, and a path is updated in real time by using dynamic environment data, and an air route is planned and optimized, so that path dynamic adjustment and energy consumption minimization of the unmanned aerial vehicle in a complex dynamic environment are realized; and the effects of improving the task execution efficiency, reducing the flight energy consumption and improving the flight safety are achieved.

The invention provides an unmanned aerial vehicle multi-target tracking method and system based on visual and millimeter-wave radar information fusion, and the method comprises the steps: obtaining an association result through employing a space-time consistency model, constructing a self-adaptive joint probability data association frame according to the association result, so as to obtain an initial target state estimation value, and carrying out the recognition of the initial target state estimation value. A prospective path planning algorithm is adopted to obtain an optimal observation angle flight path, a Bayesian network is utilized to obtain initial observation data, based on the initial observation data, an optimal target state estimation value is generated, a distributed cooperative sensing mechanism is used to generate an initial tracking task allocation scheme, and an intelligent resource allocation strategy is implemented. Generating an initial resource allocation scheme, starting a quick response protocol to obtain an optimal tracking task allocation scheme, and generating an optimal resource allocation scheme based on the optimal tracking task allocation scheme; according to the technical scheme provided by the invention, the accuracy of target identification is improved, the environmental adaptability is enhanced, the tracking efficiency is improved, and the working time is prolonged.

The invention discloses a multi-agent robot cooperative control method based on distributed reinforcement learning, and the method comprises the steps: building a multi-agent reinforcement learning algorithm environment, and defining a state space, an action space and a reward function; constructing a distributed reinforcement learning training framework which comprises a sampling module, a training module and a scheduling management and storage module; based on an actor-commentator architecture, constructing an independent actor network and a commentator network for each robot agent; through a distributed training framework, sample data sharing of the sampling module and neural network parameter sharing and cooperative training of the training module are realized; and finally, loading the trained neural network into a multi-agent robot system, and carrying out cooperative control according to environment changes and task requirements. Therefore, the robot intelligent body of each node can share experience, and cooperative training is shared through neural network parameters, so that a cooperative control strategy is realized, and the overall performance and robustness of the system are improved.

The invention provides an auxiliary hierarchical reinforcement learning training method and system based on a large language model. The method comprises the steps that a teacher agent and a student agent receive the same environment observation information; the teacher intelligent agent adopts a large language model and gives a guidance strategy according to the received environment observation information, and the guidance strategy is introduced into the training module; the student agent gives a high-level strategy according to environment observation information, and when the high-level strategy is transmitted into a corresponding low-level strategy option pi omega, the pi omega generates a series of actions according to a high-level strategy target and the environment until a termination condition is reached or the maximum task step number is exceeded; meanwhile, each action output by the low-level strategy interacts with the environment and obtains an environment reward, and the environment reward accumulation sum is regarded as a reward obtained by the current high-level action, and the environment reward accumulation sum is sent to the training module for the student intelligent agent to learn towards the environment; and the training module is used for updating the high-level strategy network of the student intelligent agent by using the loss function.

The invention relates to the technical field of unmanned aerial vehicles, in particular to a limited space cable unmanned aerial vehicle inspection system based on a GPS signal-free condition, which comprises an unmanned aerial vehicle navigation module, a cable detection module, an obstacle avoidance module, an environment monitoring module, a data processing and recording module and a power management module, autonomous navigation of the unmanned aerial vehicle in a limited space is realized, the cable detection module monitors the surface state of a cable based on a convolutional neural network model, the obstacle avoidance module dynamically detects obstacles and adjusts a flight path, and the environment monitoring module monitors temperature, humidity and air quality and gives an early warning when the temperature, humidity and air quality exceed a safety threshold. The data processing module processes data in real time in combination with an edge computing technology and uploads the data to a control center, and the power management module regulates and controls energy consumption through an intelligent algorithm and automatically returns or emergently lands when the electric quantity is insufficient; the method is suitable for a cable inspection task in a GPS-signal-free environment.

The invention relates to the technical field of intelligent teaching, in particular to a teacher console task scheduling and monitoring method based on data analysis, and the method comprises the steps: collecting task execution data, including participation degree, task completion progress and response time, of students, and combining machine learning models such as an adaptive enhancement algorithm, a long-short-term memory network and a random forest; according to the method and the system, personalized task scheduling of students is realized, a real-time data monitoring and dynamic adjustment strategy is adopted, optimal configuration of task difficulty and time length is generated according to student performance data, task risks are identified through a support vector machine model, and multi-level early warning and intervention suggestions are provided. Performing long-term trend analysis and grouping on historical data by using a K-Means clustering algorithm, and supporting a personalized optimization scheme of task allocation; according to the invention, the problem that the task allocation is not matched with the actual learning state of the student in the prior art is effectively solved, the flexibility of task scheduling and the learning effect of the student are improved, and a more accurate and efficient teaching management tool is provided for a teacher.

The invention provides a revenue awareness service migration and resource allocation method of a multi-edge cellular Internet of Vehicles. The method comprises a service migration sub-problem method and a resource allocation sub-problem method. In a service migration sub-problem method, a service migration and resource allocation unified model oriented to a complex dynamic multi-edge dense cellular Internet of Vehicles environment is adopted, and a vehicle trajectory prediction method based on probability sparse self-attention is adopted to improve prediction precision and provide a basis for service migration decision. A service migration method based on improved multi-agent reinforcement learning MARL is used to predict and explore an optimal migration strategy according to a trajectory, an asynchronous update mechanism is introduced to reduce high variance when an actor network estimates a strategy gradient, and in a resource allocation sub-problem method, a resource allocation method based on implicit enumeration is used to improve the resource allocation efficiency. The search space is reduced while the solution quality is ensured; according to the method, the long-term income of the ESP can be effectively improved, and more excellent performance is shown in different scenes.

The invention relates to the technical field of enterprise evaluation, in particular to an enterprise digital capability operation detection and evaluation method, which comprises the following steps of: collecting data from digital systems of a plurality of departments in an enterprise, and generating data flow information of each department; based on the data flow information of all the departments, a cross-department digital cooperation graph is constructed by adopting a graph model cooperation graph, nodes in the cooperation graph represent all the departments of the enterprise, and edges in the cooperation graph represent cooperation relations and data flow among all the departments; analyzing the collaboration map by using data flow network analysis, and identifying a data island of cross-department data flow; and constructing a cross-department cooperation efficiency scoring model, and calculating cooperation efficiency scores of all departments and the whole enterprise. According to the invention, an enterprise can accurately know the cooperation efficiency and the overall digitization capability level of each department, so that a reliable data basis is provided for a management layer, low-efficiency departments and cooperation bottlenecks can be found, and an accurate promotion plan can be made.

The invention relates to the technical field of internet of things equipment power consumption optimization methods, in particular to an internet of things equipment power consumption self-adaptive optimization method and system based on edge computing, and the method comprises the following steps: firstly, collecting historical operation data of different edge computing equipment, including but not limited to resource utilization rate, power consumption and network bandwidth; then, learning the historical operation data by using an improved multi-target deep reinforcement learning algorithm to obtain an equipment model; thirdly, predicting the optimal power value of the equipment through the equipment model; then, comparing the optimal power value with the actual operation power value of the current equipment, and when the optimal power value is inconsistent with the actual operation power value, sending an optimization instruction to the edge computing equipment; and finally, adjusting the parameter value of the edge computing device according to the optimization instruction to realize self-adaptive optimization of the power consumption of the device, so that the power consumption can be obviously reduced under various load conditions, and the task completion rate and the system response speed can be improved at the same time.

The invention provides an unmanned aerial vehicle cluster reconnaissance task planning method and device based on a hierarchical model, and the method comprises the following steps: pre-collecting a battlefield situation, carrying out the pre-task planning according to the pre-collected battlefield situation, and enabling each unmanned aerial vehicle in an unmanned aerial vehicle cluster to execute a task according to the pre-task planning; in the task execution process, if a new battlefield situation occurs to cause that the pre-task planning scheme is not suitable any more, dynamic task planning is triggered; after dynamic task planning is triggered, alliance establishment is carried out, unmanned aerial vehicle clusters form a new formation model according to a hierarchical model, a new battlefield situation is collected, and task planning is carried out on the unmanned aerial vehicle clusters again; in the new formation model, the unmanned aerial vehicle cluster is divided into reconnaissance unmanned aerial vehicles and task unmanned aerial vehicles, a new battlefield situation is collected, and tasks are distributed to the unmanned aerial vehicle cluster again. The dynamic instant planning problem under the combat condition is considered, and the task completion efficiency is higher.

The invention relates to an Internet of Vehicles computing power scheduling method based on a graph neural network and deep reinforcement learning, and the method comprises the steps: S1, collecting the state information of network entities in the Internet of Vehicles and the association information between the entities in real time, and constructing a dynamic time-space attribute graph; s2, inputting the dynamic space-time attribute graph into a pre-trained graph neural network encoder, and outputting a node embedding feature set containing a high-order topological relation and global graph embedding features through multi-layer message passing and feature aggregation; s3, the node embedding feature set and / or the global graph embedding feature are / is used as the state of a deep reinforcement learning agent and input to a strategy network, the computing power scheduling action at the current moment is output, and the computing power scheduling action comprises the steps of assigning an unloading target node for a to-be-processed computing task and distributing corresponding computing and communication resources; s4, dispatching actions are distributed to the corresponding network entities to be executed, environment feedback is collected and used for model updating and next round of dispatching, and the method has the advantages of improving dispatching efficiency, adaptability and system energy efficiency and the like.

The invention provides a reinforcement learning scheduling method and system based on resource prediction and action mutation in a heterogeneous cloud environment. The method comprises the steps of obtaining the number of tasks in a task waiting queue; if the number reaches the preset number threshold value, obtaining current state information of the node, and performing prediction based on a pre-trained prediction network to obtain state information of an execution moment; performing feature extraction and coding to obtain a node state vector and a task feature vector; merging to obtain a merged vector and inputting the merged vector into a reinforcement learning model to generate a decision strategy; and assigning tasks to the node mutation of the target type for the decision strategy based on a preset probability to obtain a final decision strategy. According to the embodiment of the invention, the node state can be predicted through the prediction network, and then the node state and the task demand are input into the reinforcement learning framework to obtain the scheduling decision, so that the task completion rate and the resource utilization rate of the heterogeneous cloud computing system are improved, and the energy consumption is reduced.

The invention provides a constellation software batch automatic upgrading method and system based on a laser link. Firstly, a ground control center generates a plurality of reconstruction packets, and the reconstruction packets are sequentially sent to a low-orbit earth satellite through a high-bandwidth microwave communication channel. Then, after the satellite receives the reconstructed packets, the correct reconstructed packets are reserved, and data increment packets are extracted at the same time; after the database receives the information that the satellite completes reconstruction, other satellites with software version numbers lower than those of the satellites in the constellation network to which the satellites belong are counted immediately, and the models of the satellites are transmitted to the satellites. And the satellite divides the incremental packet into a plurality of small packets, and after forward error correction coding and numbering processing, the small packets are multiplexed by using a DVB-S2 protocol and transmitted to a target satellite through an optimal laser link. The target satellite restores the complete incremental package, software updating is carried out in the dynamically determined idle time period, meanwhile, unified updating of ground database configuration is carried out, and therefore efficient and reliable batch upgrading is achieved.

The embodiment of the invention discloses a mechanical arm control method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring task parameters when a target mechanical arm executes a control task; according to the recurrent neural network, acquiring a joint angle sequence of each joint of the target mechanical arm from an initial position to a target position when the historical task is executed; and according to the task parameters and the historical task joint angle sequence, target joint angles of all joints of the target mechanical arm at the next moment are predicted, and all the joints are controlled to execute task instructions according to the target joint angles. By the adoption of the scheme, the joint angle of the mechanical arm can be flexibly adjusted according to different task requirements by introducing the multi-task parameters, multi-task operation is achieved, and the dynamic environment adaptability is enhanced; and by means of the processing capacity of the recurrent neural network on the time sequence, real-time prediction and adjustment of the joint angle of the mechanical arm are achieved, the operation stability and the task completion rate are improved, and high adaptability and universality are achieved in a multi-task scene.

The multi-satellite joint planning method and system considering task splitting provided by the present invention construct a task planning model. The model covers the planning of both splittable observation tasks and inseparable observation tasks, and designs phased algorithms for inseparable and splittable task planning respectively, taking into full account various constraints such as satellite resource capability level and task timeliness requirements. A large number of numerical experiments have proved that by splitting the tasks, the possibility of collaboration between multiple satellites is provided, the task completion rate is improved, the benefits of multi-satellite joint observation are significantly improved, and ultimately the observation efficiency of satellites is greatly improved.

The invention relates to the technical field of unmanned aerial vehicle control, and discloses an unmanned aerial vehicle target classification matching strike control method and system based on deep learning, and the method comprises the following steps: S1, collecting the original environment data of an unmanned aerial vehicle flight region, carrying out the preprocessing of the original environment data, and generating multi-mode perception data; and S2, inputting the multi-modal sensing data into a pre-constructed deep learning classification network, extracting multi-level depth features of the target through the deep learning classification network, carrying out classification identification on the target based on the multi-level depth features, and outputting target category information and target position information of the target. The multi-modal sensing data is input into the deep learning classification network for target recognition, the multi-modal fusion sensing mode can give full play to the complementary advantages of different sensors, high target recognition accuracy can still be kept in complex environments such as night, low illumination and severe weather, and the adaptive capacity of the system to environment changes is remarkably improved.

The invention discloses an RPA-based power dispatching service auxiliary method, system and device, and a medium. The method comprises the following steps: logging in a power dispatching OMS system by using an RPA automation technology; interactively triggering an operation management function through a target control, and sequentially executing UI element clicking, interface component activation operation and event distribution to a CLOW node to jump to a scheduling operation page; obtaining the number of executed work orders by adopting a selenium technology, triggering work order operation instructions in sequence, and extracting planned operation time, operation content and advance notice unit names in the work orders; on the basis of the extracted work order information, target grouping check and message content input and sending including operation content, time and ticket changing signature are completed in a message management module; generating massive controlled training sample sets through an RPA model, and performing parallel training on the sample sets by adopting a deep reinforcement learning algorithm; and carrying out adaptive process design based on a training result, and adjusting and perfecting a business process through iterative development testing and according to a deployment environment.

The invention discloses a heterogeneous unmanned aerial vehicle network task unloading method based on service cache, and belongs to the technical field of communication. The heterogeneous unmanned aerial vehicle network task offloading method for introducing an enhanced unmanned aerial vehicle into a lightweight unmanned aerial vehicle network is provided for solving the problems that the lightweight unmanned aerial vehicle network is difficult to provide sufficient cache resources, service cache configuration is likely to be unbalanced, and user service experience is poor. According to the method, a deep reinforcement learning algorithm is adopted, a cruise trajectory of a heterogeneous unmanned aerial vehicle group is dynamically planned according to position distribution and task unloading requests of ground users, a service caching decision is constructed according to lightweight unmanned aerial vehicle caching capacity constraints and service popularity distribution of associated users, and the service caching decision is determined according to the task unloading requests. And a task unloading and resource allocation decision is constructed according to the airborne computing resources and the service cache state, so that the service cache efficiency of the unmanned aerial vehicle group is improved, the task unloading completion rate is maximized, and the unit task unloading time delay of the system is reduced.

The invention relates to a dynamic environment autonomous navigation method and system based on sub-target generation hierarchical reinforcement learning. The method comprises the following steps: constructing a double-layer framework in which a top layer is responsible for target scheduling and a bottom layer is responsible for high-frequency execution; the top layer constructs a local occupation grid map according to the laser radar, constructs strip-shaped sub-target probability distribution based on a global reference path, and regulates and controls the entropy of the strip-shaped sub-target probability distribution and the commitment time limit between the top layer and the bottom layer through the environment self-adaptive exploration temperature; and the bottom layer follows the current sub-target within the committed time limit, so that the robot drives to the real-time sub-target. According to the method, strip-shaped probability sub-target distribution is constructed around a global reference path, the distribution entropy and the underlying committed time limit are regulated and controlled by utilizing environment feature-driven exploration temperature single knob linkage, and sub-target screening and issuing are completed in combination with safety expansion and reachability constraint; and a bottom-layer high-frequency closed-loop actuator follows within a commitment period and triggers advanced re-mining when the risk is increased, so that safe and smooth real-time arrival of a final target is realized.

The application relates to the technical field of path planning, in particular to a crowd-sensing cooperative optimization method and system based on evolutionary fusion two-stage mixing. The method comprises the following steps: constructing a multi-agent cooperative optimization model of a heterogeneous space based on a mobile crowd-sensing operation scene; adopting a stage-type evolutionary fusion strategy to deeply fuse MOPSO and NSGA-II, and constructing an EF-DH algorithm; using the EF-DH algorithm to perform unmanned aerial vehicle multi-target path planning based on the constructed multi-agent cooperative optimization model, including first-stage unmanned aerial vehicle cluster path optimization and second-stage ground operation personnel task optimization; and performing air-ground cooperative execution and dynamic re-optimization based on the path planning. The evolutionary fusion two-stage optimization algorithm fusing MOPSO and NSGA-II is constructed, and the global search capability and local convergence performance of the multi-target optimization problem are effectively improved.

The invention relates to the technical field of bucket wheel machine fault detection, in particular to a bucket wheel machine unmanned collaborative operation fault on-line diagnosis system, which comprises a single machine path and instruction anomaly detection module used for performing path offset and instruction loss diagnosis on the operation state of a single bucket wheel machine; the single-machine fault root cause analysis module is used for performing root cause analysis on the detected single-machine path offset and instruction loss faults; the collaborative task fault influence evaluation module is used for evaluating the dynamic influence of the single-machine fault on the collaborative operation task of the multi-bucket wheel machine; and the cooperative task dynamic adjustment module is used for optimizing cooperative task distribution of the multiple bucket wheel machines according to the single-machine fault influence evaluation result, and redistributing operation tasks of the fault bucket wheel machine to the healthy bucket wheel machine. According to the method, the depth and accuracy of fault diagnosis are improved, whether the fault bucket wheel machine is shut down or not is dynamically judged, the task completion rate and the system efficiency are balanced, unnecessary resource waste is avoided, and meanwhile the emergency processing requirement of major faults is quickly responded.

The invention relates to a leg-foot robot digital twinning system and an operation method thereof, the leg-foot robot digital twinning system comprises a physical entity system, a digital twinning body system, a data information system and a service system, and all the systems are clear in labor division and highly integrated; through high-fidelity real-time digital twin bodies driven by physical and data information and in combination with a virtual-real synchronization method of robot dynamic behavior state transition, the interaction process of the leg-foot robot and a complex environment is realized, a mature and feasible scheme based on physical model dynamic behavior transition is provided, the data transmission and communication difficulty is reduced, and the data transmission efficiency is improved. Real-time display is facilitated, interaction service is provided, and virtual-real interactivity, model accuracy and feasibility, reliability and superiority of the established digital twin system are improved.

The present invention claims protection for a vehicle-side collaborative unloading method in a non-uniform speed vehicle edge network, and belongs to the field of communication technology. In a vehicle edge network, the non-uniform speed movement of vehicles may cause rapid and disorderly changes in the network state, resulting in a decrease in network resource utilization and a deterioration in user service experience. In response to this problem, a vehicle-side collaborative unloading method is proposed. This method uses roadside units to find trusted collaborative vehicles, dynamically plans task cycles and optimizes resource allocation between roadside units and collaborative vehicles based on the real-time speed and environmental position relationship of the vehicle, thereby reducing task unloading delays and improving task completion rates.

The present invention discloses a multi-agent robot collaborative control method based on distributed reinforcement learning, which defines state space, action space and reward function by constructing a multi-agent reinforcement learning algorithm environment; constructs a distributed reinforcement learning training framework, including a sampling module, a training module, a scheduling management and storage module; based on the "actor-critic" architecture, constructs an independent actor network and critic network for each robot agent; through the distributed training framework, realizes sample data sharing of the sampling module and neural network parameter sharing and collaborative training of the training module; finally, loads the trained neural network into the multi-agent robot system, and performs collaborative control according to environmental changes and task requirements; thereby, the robot agents of each node can share experience, and perform collaborative training through neural network parameter sharing, thereby realizing a collaborative control strategy and improving the overall performance and robustness of the system.

This invention provides an image semantic transmission method and system for remote sensing small target detection. The invention constructs an end-to-end semantic transmission detection model comprising a multi-scale feature extraction backbone network, a JSCC feature transmission network, a feature fusion module, and a detector. This model is deployed on an unmanned aerial vehicle (UAV) platform, and the pre-trained end-to-end semantic transmission detection model is used to detect transmitted remote sensing images and output target detection results. This invention employs feature-level JSCC continuous symbol transmission, avoiding hard decision failures in discrete links at low SNR, thus significantly mitigating the cliff effect. Gradient-based channel importance prioritizes the retention of high-frequency and texture representations most critical for small target detection under a fixed CBR, reducing the bandwidth / power consumption of redundant background channels. Policy distillation transfers privileged gradient knowledge obtained during training to a lightweight statistical predictor, enabling the generation of high-quality masks at the edge even without ground truth values.

The invention relates to the technical field of unmanned aerial vehicle dispatching, and discloses an electric power construction-oriented heterogeneous unmanned aerial vehicle dynamic cooperative command and dispatching method and system. The method comprises the following steps: decomposing task levels according to an environment tolerance threshold value and constructing an unmanned aerial vehicle capability matrix; collecting real-time environment parameters, calculating an environment coupling degree index, and generating a degradation task package when the environment coupling degree index exceeds a threshold value; calculating a single machine matching degree index, and if not, selecting the collaborative machines with the maximum capability complementation index to form a virtual collaborative group and distributing roles; position and wind speed vectors are obtained, and the time difference of arrival is eliminated by delaying takeoff after the flight speed is corrected. According to the invention, the environment adaptability, task completion rate and execution efficiency of heterogeneous unmanned aerial vehicle cooperative operation in an electric power construction scene are improved.

The application provides a stamping line tail AGV scheduling system, relates to the field of intelligent manufacturing logistics scheduling, adopts a hierarchical centralized collaborative control architecture, sets a business layer module, a scheduling core layer module, an execution control layer module and a digital twin simulation layer module, internally arranges three customized strategies of main and auxiliary position dynamic priority distribution, flexible partition scheduling and double anti-locking traffic control in the scheduling core layer, realizes task differentiated priority distribution, operation area logical isolation control, space and time collaborative active anti-locking control, and completes strategy preposition verification and running state visual monitoring through digital twin simulation. The application can effectively adapt to the rigidity production rhythm of the stamping line tail, improve the multi-AGV collaborative scheduling efficiency and long-term operation stability, reduce the engineering landing debugging cost, and is suitable for the intelligent material carrying scene of the stamping workshop line tail.

The invention relates to a multi-attribute task scheduling method with timeliness guarantee and application thereof. The multi-attribute task scheduling method is applied to pervasive edge computing. A distributed computing power sensing mechanism is adopted, an IPv6 expansion header is designed, computing power sensing is achieved in a broadcast mode, and task equipment is generated to detect surrounding resources and load conditions; according to the method, battlefield tasks are subjected to multi-attribute tasks according to actual conditions, an unloading decision is made by comprehensively considering factors such as calculation, storage, transmission and load conditions by using multi-attribute decisions according to different task types, and a TOP3 pre-inquiry process is added. According to the multi-attribute task scheduling method with timeliness guarantee and reference thereof, resources in a battlefield environment are fully utilized, the load condition in a battlefield scene is optimized, stable task unloading is guaranteed, the deterministic time delay requirement of a time delay sensitive task is met, and the task completion rate is increased.

The application discloses a kind of off-site computing power dynamic collaborative optimization methods based on multi-agent reinforcement learning, including the following steps: constructing resource topology diagram;Based on resource topology diagram, get node-level state feature vector and system-level state feature vector;Multi-agent environment is constructed;Local observation, global information and agent action set are input into improved CTDE model, output policy network parameters and value network parameters and construct training batch;Based on training batch, get converged policy network parameters and converged value network parameters;Get execution result;Obtain dynamically updated policy network parameters and dynamically updated value network parameters, realize the dynamic collaborative optimization of task acceptance, resource allocation, task migration, replica start-stop and bandwidth matching.

The application discloses a kind of based on edge cloud cooperation's inspection robot visual analysis task dynamic scheduling method and system, belong to inspection robot visual analysis and edge computing technical field.The method includes: real-time acquisition network state parameter and computing power resource information;Task feature vector is constructed to the visual analysis task and is extracted feature;Based on multi-objective optimization scheduling model, delay, energy consumption and accuracy index are determined optimal execution position by comprehensive evaluation;Task is distributed to corresponding node execution;Based on Q-learning reinforcement learning, scheduling weight is dynamically updated.The application can realize the optimal scheduling of visual analysis task under the dynamic change of network and computing power environment, compared with the scheduling efficiency of existing static allocation strategy is improved by more than 40%, and the comprehensive performance is significantly better than local execution, full cloud execution and static allocation scheme.