553 results about "Value network" patented technology

The invention discloses a sparse reward environment-oriented collaborative knowledge fusion reinforcement learning method, and relates to the field of collaborative knowledge fusion reinforcement learning methods. By constructing a lightweight collaborative knowledge fusion model and a dynamic reward remodeling mechanism, the problems of low intelligent agent exploration efficiency and difficulty in strategy convergence in a sparse reward environment are solved. The method comprises the following steps: constructing a reinforcement learning framework comprising a policy network and a value network; designing an action space mutation supervision mechanism and a lightweight collaborative knowledge fusion model, and generating a smooth substitution action when a strategy is detected to be unstable; and a reward function is designed in combination with the task target and the dynamic constraint, and reward remodeling is realized by activating rewards through sub-target potential energy difference and knowledge fusion. According to the method, effective intermediate feedback can be provided for the agents in the sparse reward environment, the exploration efficiency is improved, the convergence time is shortened, the stability and cross-scene migration ability of the strategy are enhanced, and an effective solution is provided for sparse reward scenes such as robot control and multi-agent game.

The invention relates to the technical field of industrial production lines, and discloses an industrial production line multi-device dynamic collaborative scheduling method based on reinforcement learning, comprising the following steps: S1, modeling a three-dimensional state space; s2, hierarchical reinforcement learning architecture; and S3, edge-cloud cooperative execution. According to the industrial production line multi-device dynamic collaborative scheduling method and system based on reinforcement learning, device states, task constraints and resource occupation are integrated into a structured matrix through three-dimensional state space modeling, and a global decision-making layer captures production time sequence dependence by using a bidirectional long-short-term memory network; modeling equipment space association and process constraints through a graph attention network, and generating a global strategy including task allocation, capacity adjustment and resource pre-allocation; and after the edge layer detects the dynamic event, the cloud platform generates a candidate scheme through Monte Carlo tree search, and realizes dynamic event response and multi-target collaborative optimization by combining multiple targets such as global value network evaluation task completion time and equipment load balancing.

The invention discloses a quadruped robot robust motion control method based on deep reinforcement learning, and belongs to the technical field of robot motion control, and the method comprises the steps: constructing a deep reinforcement learning model which comprises a state estimation network, a strategy network and a value network; interaction between the quadruped robot and the simulation environment is carried out, and standard observation information, historical observation information and privileged observation information of the quadruped robot at all moments are obtained; inputting the standard observation information, the historical observation information and the privilege observation information of the moment into a deep reinforcement learning model, and training based on a total loss function until convergence is carried out to obtain a trained deep reinforcement learning model; inputting standard observation information and historical observation information at corresponding moments in an actual scene into the trained deep reinforcement learning model to obtain output features of a corresponding strategy network; and the target position of each joint motor is calculated to complete the motion control of the quadruped robot. And efficient training and robust motion on various complex and unstructured terrains can be realized.

The invention belongs to the technical field of artificial intelligence and data science, and discloses a causal discovery method and related device based on a large model and structural constraint, comprising the following steps: according to meta-information of each causal variable, adopting a large language model and combining multiple rounds of dialogues, multi-role play and auxiliary cue word modes, generating an initial potential causal graph of each causal variable; repeating the following steps until a preset condition is met, and outputting a candidate causal graph with the highest value score: generating the candidate causal graph by adopting a policy network based on an encoder-decoder architecture, and carrying out iterative optimization on the candidate causal graph by adopting a reinforcement learning algorithm based on an act-critic framework; and calculating a value score of the candidate causal graph after iterative optimization according to the value network, and updating parameters of the policy network based on the encoder-decoder architecture according to the value score. The efficiency and accuracy of causal relationship recognition are remarkably improved, and a causal relationship graph with high confidence can be generated in a complex data environment.

The invention discloses a snake-shaped robot control method based on reinforcement learning. The snake-shaped robot control method comprises the steps that a complex terrain simulation scene comprising walls, cubic obstacles and low channels is built; modeling is conducted on the structure, joints and kinetic parameters of the snake-shaped robot, and a snake-shaped robot simulation model is obtained; defining a state space and an action space of the snakelike robot, and designing a diversified reward function; constructing an action network and a value network by adopting a multi-layer perceptron (MLP) structure; and performing multi-round simulation training on the snakelike robot by applying a reinforcement learning algorithm, so that the robot learns a parameter set for making an optimization decision under different scenes, thereby realizing optimization control of the snakelike robot and reaching a target point. According to the method, visual information does not need to be depended on, efficient and stable motion control of the snake-shaped robot in a complex environment can be achieved through autonomous learning, and the crossing capacity of the snake-shaped robot in a low and narrow space is improved.

Disclosed is a method for enhancement of a low-light image based on reinforcement learning and aesthetic evaluation. The method include: generating images of non-normal luminance under different lighting scenes, and constructing a training dataset for a reinforcement learning system based on the images; initializing the training dataset, a policy network, and a value network; updating, based on a no-reference reward score and an aesthetic assessment reward score, the policy network and the value network; completing model training and outputting an enhanced image result. By expanding the action space range defined in reinforcement learning, the enhancement operations for the input low-light image gain a greater dynamic range, offering higher flexibility for real-world scenarios and better meeting low-light image enhancement needs. Additionally, by incorporating the aesthetic quality assessment scores as part of the loss function, the enhanced image achieves better visual effects and higher subjective user evaluation scores.

The invention provides a motor defect identification method fusing time sequence space feature extraction and reinforcement learning. The method comprises the following steps: building a Transform-GAT model architecture T-GAT, and learning time sequence relevance and spatial topological structure features by the T-GAT to form a space-time composite representation vector; designing a dual-network architecture reinforcement learning weighted fusion mechanism of a strategy network and a value network, and after the strategy network and the value network of a parallel structure receive the composite vector, constructing a strategy function to select a defect type with the maximum probability; designing potential energy function quantization parameters, and constructing a reward function to generate a reward in combination with a difference value; taking a reward function as a target, training and optimizing model parameters of the dual-network architecture in an off-line manner, running a real-time decision in an on-line manner, and storing a tetrad to an experience pool to form a'perception-decision-feedback-update 'closed-loop mechanism; according to the method, motor defect identification is realized, downtime is reduced, and motor operation reliability and equipment operation efficiency are improved.

The invention discloses an electricity larceny risk threshold dynamic optimization and multistage early warning method based on reinforcement learning, and the method achieves the dynamic adjustment of an electricity larceny risk threshold through constructing a multi-dimensional state space containing a risk perception percentage, a line loss rate deviation, a cost budget and a load feature vector, and employing a reinforcement learning algorithm. Safety constraints are introduced to ensure the rationality of threshold adjustment, a TD3 algorithm is adopted to train and update a strategy network and a value network, the learning efficiency is improved through priority experience playback, a multi-target weighted reward function is designed to balance the multi-target optimization requirements of detection precision, line loss control, inspection cost and the like, and a red, orange and yellow three-color early warning mechanism is constructed. Early warning actions of different levels are triggered according to the ratio of the dynamically updated reference threshold value to the risk perception percentage, and early warning visualization and strategy effect evaluation are achieved through a geographic information graph dynamic rendering module, a natural language report generation module and an anti-fact analysis module.

The invention discloses a water body cleaning equipment path optimization method based on deep reinforcement learning, and the method comprises the following steps: S1, collecting water body environment data through a sensor, and carrying out the preprocessing; s2, establishing a state space and an action space for path optimization, and constructing a reward mechanism; s3, carrying out feature extraction on the preprocessed data by adopting a Transform network; s4, training the strategy network and the value network by using an improved trust region strategy optimization algorithm, optimizing an update step length of the strategy network based on KL divergence constraint, and optimizing a path planning strategy based on a strategy gradient method; s5, performing path planning by using the trained strategy network and value network; and S6, calculating the energy consumption of the water body cleaning equipment, and adjusting the operation mode or operation path of the water body cleaning equipment. According to the method, the path planning of the water body cleaning equipment is optimized by combining Transform and an improved trust region strategy optimization algorithm, and the method has the advantages of being high in environmental adaptability, low in energy consumption, high in cleaning efficiency and high in operation stability.

The invention discloses an information retrieval system and method based on semantic normalization, and relates to the technical field of artificial intelligence information, and the method comprises the steps: collecting a semantic query record input by a user, carrying out the preliminary semantic analysis, and generating structured data; on the basis of the structured data, entity disambiguation is carried out by utilizing a knowledge graph, abstract classes are generated through a neural network, calibration and dynamic weight adjustment are carried out, and high-confidence entity abstract classes and confidence scores are generated; entity abstract classes and confidence scores are combined with user contexts, an action-value function is calculated through a value network, and an optimal action is selected by utilizing a-greedy algorithm; executing semantic normalization mapping according to the optimal action, and obtaining an intermediate expression by using a meta-symbol dynamic generator; and performing index retrieval and multi-dimensional sorting based on the intermediate expression to generate a sorted retrieval result list. According to the method, the semantic fragmentation problem of multi-modal query is solved, and deep semantic alignment and dynamic weight calibration of heterogeneous data are realized.

The invention discloses a milling path automatic generation method based on reinforcement learning, and relates to the technical field of machine tool machining, and the milling path automatic generation method comprises the following steps: step 1, defining action output; step 2, defining state input including a cutter state, a global observation space and a local observation space; 3, defining a reward function; value feedback given to the intelligent agent by the environment according to the current state and action is determined through the reward function; 4, setting deep neural networks including a feature extractor, a strategy network and a value network; 5, constructing a process autonomous generation model and a machining process world model; step 6, strategy training; step 7, outputting a strategy, and autonomously completing a machining process flow; by means of the method, the problems that in the prior art, manual experience dependence is too heavy, local optimization limitation is remarkable, and physical simulation efficiency is insufficient are solved.

The invention provides a dialogue model training method, a dialogue method, a dialogue system, equipment and a medium, and relates to the technical field of man-machine interaction, the method comprises the following steps: generating current sample dialogue state information according to current sample user input information and historical sample dialogue information in a sample dialogue; performing tree search based on the to-be-trained strategy network and the to-be-trained value network to obtain current sample reply action information and a global reward value of a dialogue track corresponding to the current sample reply action information; according to the next sample dialogue state information, the current sample dialogue state information, the current sample reply action information and the global reward value which are generated by simulating the current sample reply action information, constructing sample data; and performing joint training on the to-be-trained strategy network and the to-be-trained value network according to the sample data to construct a dialogue model. According to the method, long-term return and strategy optimization are comprehensively considered during dialogues, so that the generation quality, the interaction capability and the user experience of multiple rounds of dialogues are improved.

The invention discloses a server fault diagnosis method and device, a storage medium and a program product, and relates to the technical field of server fault localization, and the method comprises the steps: carrying out the preprocessing of original log data, determining a state space according to the obtained preprocessed log data, a fault classification model containing a value network and an action network is obtained through reinforcement learning in advance, the value network in the fault classification model is used for carrying out value evaluation on the state space, and the action network in the fault classification model is used for carrying out action probability distribution calculation on the state space according to a target value obtained through value evaluation; and determining the target fault type according to the calculated probability distribution result, thereby solving the technical problem that new features or anomalies cannot be effectively judged, effectively improving the quality and integrity of log data, improving the server fault diagnosis efficiency and accuracy, realizing intelligent identification of the fault type, and improving the fault diagnosis efficiency and accuracy. The method has the technical effects of good training stability and generalization ability.

The invention discloses a heterogeneous agent collaborative strategy training method based on a large-scale distributed hybrid unmanned and manned land battle confrontation scene. The method comprises the following steps: step a, constructing a land battle confrontation scene, a red battle agent and a blue battle agent through a VBS simulation platform; b, respectively setting a state space, an action space and a reward function of the red-party combat intelligent agent and the blue-party combat intelligent agent; c, constructing a strategy network and a value network, and inserting a cross-agent attention mechanism into the strategy network and / or the value network; and step d, through large-scale distributed training, parallel operation of a multi-instance simulation environment, strategy network fusion cross-agent information output action, value network evaluation of joint action value, training to enable the model to converge, and through a reinforcement learning algorithm, outputting an action sequence or a simulation effect to analyze a cooperation strategy of a red-party combat agent. And the cooperative combat capability of the heterogeneous agent system in a land battle confrontation scene is improved.

The invention discloses a security reinforcement learning-driven large language model security decision agent, and the decision agent comprises a high-level semantic planner which is used for receiving a target and constraint instruction in a text form, receiving a language or visual observation signal of an environment at the same time, and outputting text formatted security risk information and suggested action planning; the low-layer action actuator is used for receiving low-dimensional observation and semantic codes of the environment, and the semantic codes are output by the high-layer semantic planner after text embedding conversion; the strategy network of the low-layer action actuator outputs a final safety action; the training alignment module is used for optimizing the strategy network and the value network; a high-level semantic planner is fed back and prompted through reward and cost signals collected through environment interaction, and parameters of a strategy network and a value network are trained through a security reinforcement learning algorithm. According to the method, the decision cannot violate the given text security constraint while the decision of the given text target is completed.

The invention discloses a crop nitrogen fertilizer management system and method based on multi-agent reinforcement learning, and belongs to the technical field of intelligent agriculture. Comprising the following steps: collecting and preprocessing multi-source data of a target area, calibrating a crop growth-nitrogen cycle model based on the multi-source data, and constructing a dynamic simulation environment; constructing a nitrogen fertilizer application strategy model and a multi-target award function, and performing agent reinforcement learning training by adopting a centralized training-decentralized execution architecture; introducing a large language model, and updating a strategy network and / or a value network of each agent in a training process according to a reward adjustment signal and a decision constraint; when the index fluctuation ratio in the continuous evaluation period is smaller than a preset threshold value, it is judged that the nitrogen fertilizer application strategy model is converged, and an optimal nitrogen fertilizer application strategy model is obtained; and generating a nitrogen fertilizer application scheme based on the optimal nitrogen fertilizer application strategy model in combination with the real-time state data, and generating a natural language interpretation and risk assessment report based on a large language model.

The invention belongs to the field of multi-agent reinforcement learning, and discloses a relation-driven multi-agent reinforcement learning method and system based on a mixed game, and the method comprises the steps: enabling a strategy network to generate an agent action, and carrying out the interaction of an environment, so as to collect sample data; the centralized value evaluator calculates marginal influence values of the agents based on samples, and deduces a social influence weight set; assigning group external rewards as individual external social rewards based on the weights; processing the global state by using a random network distillation-driven method, updating a prediction network parameter to minimize a prediction error, and outputting an internal reward set; the external social rewards and the internal rewards are fused to form comprehensive rewards; and updating the strategy network and the value network by using the comprehensive reward, and circularly training until convergence. By adopting the method, accurate modeling of the individual interaction relationship is enhanced, invalid exploration is remarkably reduced, the learning efficiency and the strategy reliability are improved, and the overall performance of the system is enhanced through an explicit social mechanism.

The invention is applicable to the technical field of robot motion control, and provides a robot skill learning method, which comprises the following steps of: generating an auxiliary reward component function by utilizing a large language model according to learning prompt information, constructing a multi-branch value network according to a reward component function set formed by a plurality of reward component functions and auxiliary reward component functions in the learning prompt information, and performing iterative strategy training on the strategy network by using the multi-branch value network based on the weight coefficient of each value branch until the target function of the strategy network is converged, so as to obtain the target function of the strategy network. A strategy network is used for guiding robot skill learning, so that the intelligent decision-making capability of a large language model and fine-grained feedback of a multi-branch value network are combined, the strategy learning efficiency and the overall performance in reinforcement learning are improved, it is ensured that different skill obtaining requirements can be met in the strategy optimization process, and the robot skill learning efficiency is improved. And different environments and complex robot tasks can be handled, and the accuracy, precision and efficiency of task execution are improved.

The invention provides a method and system for optimizing deep reinforcement learning based on composite priority experience playback. The method comprises the following steps: acquiring a current training network and a target network; the current training network comprises a current strategy network and a current value network, and the target network comprises a target strategy network and a target value network; acquiring an experience tuple based on an action selection strategy of the current strategy network; calculating a composite priority of the experience tuple, and enabling a life cycle pointer to update an experience playback area according to the experience tuple and the composite priority of the experience tuple; and sampling from the experience playback area to update the network parameters of the current training network, and updating the network parameters of the current training network to a target network. According to the method, the time sequence difference error of the instant reward and the experience is introduced as the priority evaluation standard of the new experience, the priority of the new experience is improved through the playback times, and the replacement efficiency of the experience playback area is improved through the lifetime pointer.

The invention provides a data center IT load and cooling system cooperative control method based on a TD3 algorithm, and relates to the field of deep reinforcement learning, and the method comprises the steps: obtaining an initial data set representing the server state, environment and task characteristics of a data center, and processing the initial data set to obtain a target data set; building a deep reinforcement learning model which is based on a TD3 algorithm and comprises a strategy network Actor and a value network Critic, and determining a state space, an action space and a reward function of the model; training the model to obtain a target model; and outputting a work task allocation strategy and a cooling system regulation and control strategy through interaction of the target model and the environment. According to the target model, by outputting the work task allocation strategy and the cooling system regulation strategy, work task allocation can be guided, operation of the cooling system can be controlled, the IT load of the server and refrigeration balance of the cooling system are achieved, safe and stable operation of the server is ensured, and energy consumption of the cooling system is reduced.

The invention provides a robot motion control model training method, device and equipment based on deep reinforcement learning, a robot and a medium, and relates to the technical field of robots. The method comprises the following steps: acquiring a first potential vector obtained after a student encoder encodes robot body observation data, and a second potential vector obtained after a teacher encoder encodes privilege observation data; based on the current training step number and a preset probability function, calculating a sampling probability for controlling a fusion proportion of the first potential vector and the second potential vector; fusing the first potential vector and the second potential vector based on the sampling probability to generate a third potential vector, and inputting the third potential vector into a strategy network; and updating the parameters of the policy network based on the value estimation of the current state output by the value network and the action policy output by the policy network. According to the method, updating oscillation caused by sudden change of input distribution in the training process of the strategy network can be avoided, the training efficiency is improved, and the training cost is reduced.

The invention discloses a man-machine cooperation task allocation method based on reinforcement learning, and the method comprises the steps: constructing a reinforcement learning model which comprises a task definition, a resource definition, a state space, an action set and a reward function, and carrying out the modeling of a cooperation relation among a task, a worker and a robot in man-machine cooperation through a heterogeneous graph neural network, and extracting high-order features of the state space. Based on the characteristics, a strategy network and a value network are trained through a near-end strategy optimization algorithm, and learning of an efficient task allocation strategy is achieved. The reward function combines the task completion duration and the resource utilization rate, and gives consideration to the production efficiency and resource balance. In the system operation process, workshop state data are collected in real time to dynamically update the state space, and the reinforcement learning model generates an optimal allocation action according to the latest state and issues an instruction to a worker or a robot for execution. The method has the advantages of high adaptability, high distribution efficiency, high scheduling intelligence degree and the like, and is suitable for task scheduling optimization in complex dynamic environments such as intelligent manufacturing and the like.

Disclosed in the present invention is an aircraft engine maintenance policy optimization method based on nonparametric reinforcement learning. Firstly, for sparse aircraft engine operation data, constructing an aircraft engine model by means of a Bayesian network and a Gaussian process; then, establishing a policy network and a value network, interacting with the aircraft engine model to form a state / action value group, storing same into a replay buffer, and performing random sampling for use in a training set; updating the policy network and the value network, and updating the training set; and finally, performing aircraft engine maintenance optimization policy decision-making. Provided in the present invention is, for the first time, a Gaussian process-based nonparametric reinforcement learning method for an aircraft engine, which improves the degree of fit between overall training data and a model by means of dynamic data updating while integrating system uncertainty, thereby improving the sampling efficiency of an algorithm. Action selection is performed on the basis of prior maintenance experience data fused with uncertainty, improving the safety and sampling efficiency of a system and thereby solving the problem of predictive maintenance of an aircraft engine system.

The invention discloses an offline reinforcement learning method based on behavior action generation, and the method comprises the steps: collecting a first data set of an intelligent agent, and carrying out the preprocessing of the first data set, and obtaining an offline data set; constructing a first diffusion model for training to obtain a final diffusion model; for each state in the offline data set, acquiring M actions corresponding to the state from the final diffusion model to form a support action data set; constructing a strategy network, N value networks, a target strategy network and N target value networks; training the strategy network and the value network by using the offline data set and the support action data set, and updating network parameters of the target strategy network and the target value network at the same time; and making a decision by using the learned optimal target strategy network, so that the intelligent agent performs a reinforcement learning task in an offline environment. According to the method, the limitation of an existing offline learning method on strategy learning is reduced, and the completeness and execution effect of a robot or an automatic driving task are remarkably improved.

The invention discloses a time-varying task scheduling method and system based on space-time constraint near-end strategy optimization, and belongs to the technical field of information processing. Comprising the steps of predicting task resource requirements of future N time steps through an LSTM model, and generating a time dynamic mask matrix Mtime; generating a spatial dynamic mask matrix Mspace based on a three-level hierarchical constraint rule; the three-level hierarchical constraint rule comprises resource capacity, interference avoidance and important application protection; the Mtime and the Mspace are fused into a global mask matrix Mt; an improved PPO algorithm is adopted, masked action probability distribution is output through an Actor network, and a Critic network evaluates a state value containing constraint punishment; and dynamically iterating the task scheduling strategy by jointly optimizing the strategy network, the value network and the mask generator parameters. According to the method, efficient utilization and load balancing of resources can be realized, and the stability and the performance of time-varying task scheduling are remarkably improved.

The invention relates to a display panel repair path optimization method and device based on deep reinforcement learning. The method comprises the following steps: obtaining a target display panel image; performing preprocessing and feature extraction on the target display panel image to obtain related features; inputting the related features into a pre-established deep reinforcement learning model, outputting an optimized repair path by the deep reinforcement learning model, and repairing the display panel; the deep reinforcement learning model is a PPO algorithm model which is constructed based on deep reinforcement learning and takes a strategy network and a value network as intelligent agents. According to the method, the problems of path redundancy and planning time consumption in a multi-defect scene in a traditional method are solved. Through real-time data acquisition, a reward function and a strategy are continuously updated by using a deep reinforcement learning algorithm, so that a repair path and an on-site defect state are matched in real time, and the problem that dynamic defect changes cannot be responded in real time in a traditional method is solved.

The invention discloses an MCTS and self-gaming-based confrontation efficiency evaluation method, system and device and a medium, a self-gaming mechanism is introduced innovatively by combining Monte Carlo tree search and a deep neural network (agent), and high-quality training data is generated by simulating a real confrontation environment in the self-gaming process, so that the self-gaming efficiency is evaluated. Continuously optimizing a strategy network and a value network of the intelligent agent; according to the method, effective search and evaluation are carried out in a large state space, rapid change and complex feedback of opponent strategies can be adapted in real time, and through multi-round self-game simulation, the precision and the real-time performance of efficiency evaluation are improved, and reliable technical support is provided for strategy optimization in a complex confrontation environment.

The invention provides a multi-agent deep reinforcement learning method based on human guidance, and the method comprises the steps: constructing a multi-agent value decomposition framework, initializing a multi-agent environment and the state of each agent, innovatively fusing the elastic strategy update of human intervention, storing a high-quality experience sample generated by human intervention into an independent playback pool, and carrying out the recognition of a multi-agent value decomposition framework; mixing and sampling with samples generated by autonomous exploration of the intelligent agent according to priorities; maximizing the entropy in the strategy optimization target to encourage exploration, and dynamically adjusting the entropy coefficient through the target entropy; and circularly executing the steps of interaction between the intelligent agent and the environment, experience storage and strategy network and value network updating, and finally obtaining a greedy strategy of the intelligent agent according to a multi-intelligent agent soft actor commentator architecture combined with a value function decomposition technology as an output strategy. Through a human active intervention and value decomposition technology, efficient and safe training of a multi-agent cooperation task is realized, a human intervention sample and an autonomous exploration sample are trained in a mixed manner, and strategy optimization is accelerated.

The invention discloses a flexible workshop production resource scheduling method based on multi-strategy deep reinforcement learning, and relates to the field of intelligent manufacturing, the method constructs three strategy networks and three value networks of work, machines and AGVs, and at each scheduling moment, the work selection strategy network selects a scheduling rule according to a work state and maps the scheduling rule to a specific operation; the AGV selection strategy network selects a scheduling rule according to the AGV state and maps the scheduling rule to a # imgabs0 # table AGV # imgabs1 # of transportation operation; a machine selection strategy network is mapped to a # imgabs2 machine # imgabs3 # of processing operation according to a machine state, it is determined that the operation is transported from an AGV # imgabs4 # to a machine # imgabs5 # for processing through multi-strategy cooperation, the starting time and the ending time of the operation are calculated, states, actions and rewards are stored in corresponding experience pools, and a value network samples data from the experience pools, calculates losses and updates network parameters. And after the experience pool is emptied, entering a next round of iteration. The method can dynamically respond to production demand changes, and optimize resource allocation and scheduling decisions in real time.

The invention provides a single agent decision-making method and device based on deep reinforcement learning, and the method comprises the steps: obtaining agent sample data in an experience playback buffer pool, the sample data at least comprising an observation state, a next observation state, a decision-making action and a reward value of an agent in an environment; inputting the agent sample data into the estimation network and the target network for training; the estimation network comprises a strategy network and an evaluation network, and the target network comprises a target strategy network and a target evaluation network; respectively calculating a strategy gradient of the estimation network and a loss function of the target network; if the strategy gradient and the loss function meet preset conditions, the network model corresponding to the strategy gradient and the loss function is determined as the decision model of the agent, an actor-commentator reinforcement learning architecture is adopted, the problem that correlation exists before and after each parameter updating of a neural network in an actor-commentator structure is solved, and the decision model of the agent is determined to be the decision model of the agent. And meanwhile, the defect that deep reinforcement learning cannot be used for continuous actions is overcome.