122 results about "Value network" patented technology

The invention shows explicitly how direct and honest communication utilizing a distributed computing network, shared content and services, and a focus on social network dynamics and situational context can directly create experiential and realized value for all participants around the globe. Contributors (20), having instructional content to share (22) pertaining to the functional utilization of generically named goods, essentially become knowledge workers and are rewarded (26) during specific instances when commercial entity Sponsors (60), who have been granted rights through bidding in an auction (64), utilize this content to un-intrusively market their branded goods to interested and receptive Consumers (10), by having their brand names acknowledged as a valuable component of the original content. The content can also be federated with real-time calculations of statistical price information (37) and specialized data services (101) to further increase the experiential value. The Sponsors (60) participating in these niche groups can then be granted access-rights to analyze the developing value network ontology and the ongoing flow of quality information (34) among the participants in the value network.

The invention discloses a micro-power-grid energy storage scheduling method and device based on deep Q-value network reinforcement learning. A micro-power-grid model is established; a deep Q-value network reinforcement learning algorithm is utilized for artificial intelligence training according to the micro-power-grid model; and a battery running strategy of micro-power-grid energy storage scheduling is calculated and obtained according to input parameter feature values. According to the embodiment of the invention, deep Q-value networks are utilized for scheduling management on micro-power-grid energy, an agent decides the optimal energy storage scheduling strategy through interaction with an environment, a running mode of the battery is controlled in the constantly changing environment,features of energy storage management are dynamically determined on the basis of a micro-power-grid, and the micro-power-grid is enabled to obtain a maximum running benefit in interaction with a mainpower grid; and the networks are enabled to respectively calculate an evaluation value of the environment and an additional value, which is brought by action, through using a competitive Q-value network model, decomposition of the two parts enables a learning objective to be more stable and accurate, and estimation ability of the deep Q-value networks on environment status is enabled to be higher.

The invention discloses a Token capital cross-chain transfer system between different block chain networks. The system comprises a cross-chain gateway unit, a cross-chain gateway client unit, and a token agent gateway protocol unit; the cross-chain gateway is simple and flexible in deployment without depending on the complex intermediate chain; the cross-chain gateway can be quickly deployed to butt joint with an external chain as long as the Token introducing chain supports the TAGP, the external chain Token is introduced into the chain to realize the Token cross-chain value transfer. By introducing various types of Tokens into the chain to circulate, the intra-chain ecology is more rich, and more extensive service application demand is satisfied; furthermore, the cross-chain gateway enables the chains to interconnect and intercommunicate, the Token can be freely transferred between the chains, thereby forming the block chain value network; on the security aspect, the cross-chain gateway manages the Token by adopting multiple signature and offline signature way, thereby guaranteeing the account security of the cross-chain gateway.

The invention relates to a unmanned aerial vehicle (UAV) flight path planning method based on a competitive deep learning network. The method comprises the following steps: carrying out feature information extraction on real-time pictures photographed by a camera, so that a column of feature vectors are obtained; calculating each feature vector to obtain a state function value and a superiority function value, and combining the two values on a combination layer to obtain a state action function value; using the state action function value as an instant state action function value which cooperates with a target value network in construction of a loss function of the network and prediction of a next state, so that total rewards jointly formed by internal rewards and external rewards are obtained; carrying out depth-of-field prediction on the real-time pictures; carrying out calculation to obtain another state action function value; and calculating the gradient of the loss function and carrying out back propagation of the gradient to a current value network for network parameter updating.

An accurate recommendation method based on incomplete data includes steps: (1) acquiring data; (2) processing data; (3) further subdividing analysis variables, and providing foundations for modeling analysis; (4) describing clients; (5) analyzing modeling; (6) inputting input variables and client summarization archives into an analysis model to acquire analysis results such as hobbies, purchasing actions and enquiry actions of the clients; and (7) recommending business to the clients according to the analysis results. The accurate recommendation method based on the incomplete data has the advantages that client database and product database are effectively optimized, historical business database with quite good marketing decision reference value is newly established; by implementation of a reinforced value network data acquisition system, data planning based on prospective vision is realized, enterprise existing members or client resources are effectively exchanged and utilized, and strategy advantages of industry resource integration are formed.

The invention discloses an establishing method of an electronic wallet based on a block chain. The method comprises steps of 1) establishing a new transaction and issuing various types of assets; 2) carrying out the transaction and broadcasting transaction lists to the whole network through a P2 network; 3) carrying out transaction verification; 4) verifying results and broadcasting transaction results to the whole network after the transaction is finished; and 5) writing the transaction results into a block of a block chain. According to the invention, by use of the distributed account book technology, the digital asset flows and real cash payment on the block chain are connected, so in the global Internet market, the function of high-efficiency and low cost value delivery that the traditional financial mechanism cannot replace can be developed; a block chain credit system from the information to the value network is formed; the cryptography packet of each person can be developed into a 'self-finance' platform; and payment, depositing, transferring, exchange, loans and the bookkeeping and clearing in the whole network of the P2P can be achieved.

A method and system is provided, in which an interfacing device is distributed free by a value net integrator to end users, or supporting parties, whose normal use thereof may benefit a soliciting entity, such as a charity or social group, through various forms of interactive advertising, combined with an interactive and good will based referral incentive program. The interfacing device may comprise a special toolbar to identify and locate the participating third party vendors; to enable supporting parties to access a search engine partner for Internet searches, where the search engine partner agrees to provide a commission to the value net integrator on a per-click basis; and to allow supporting parties to easily make direct contributions to the soliciting entity. The special toolbar may provide additional value to encourage its use by the supporting parties, such as providing an intelligent pop-up blocking capability. The value net integrator may collect the commissions generated by the supporting party's use of the special toolbar, pay a royalty to the soliciting entity out of the commissions on a periodic basis, and retain the remainder as compensation for its services.

The present invention provides a method, a system, and a computer-readable medium with instructions for a computer to optimize one or more tradeoffs between or among serviceability, liability, and/or inventory in a multi-tier network of suppliers. The probabilistic optimization of tradeoffs enables assets stored at one or a plurality of tiers in the network to be optimally transferred downstream with certain probabilities. The multi-tier network of suppliers may consist of at least one original equipment manufacturer tier and at least one supplier tier.

The invention discloses a microgrid energy scheduling method based on double-Q-value network deep reinforcement learning, and the method comprises the steps: taking one-day prediction information of amicrogrid as a training set for generating an optimal control strategy, and training an intelligent agent which is independent of a microgrid environment and takes an energy storage system as a control object; and achieving double optimization objectives of minimum operation cost of the microgrid and minimum power fluctuation of the public power grid by controlling charging and discharging actions of the energy storage system. The method does not depend on the construction of a specific micro-grid model, and the strategy is guided by the design of the reward function to achieve the purpose ofmicro-grid operation, so that the optimal strategy of global time can be obtained, and the power imbalance caused by uncertainty of new energy power generation and user load distribution can be effectively solved.

The invention provides a people image search method based on deep reinforcement learning. The method comprises the steps of: S1, defining various kinds of action for adjustment of a search area of a target image; S2, building a configurable deep model; S3, collecting training samples and training a strategy selection network and a value network by using the training samples; S4, inputting a reference images and to the-be-tested target image into the deep model and initializing the search area of target image as the whole image; S5, extracting features of the reference image via a feature extracting network; S6, extracting features in the search area of the target image via the feature extracting network, and fusing the same and the features of the reference image to form fusion features. The method treats pedestrian detection and people reidentification as the same task and only needs multiple times of action selection to judge whether a target person is found without an extra candidate frame, thereby having high time efficiency.

The invention discloses a method for identifying a named entity based on a policy value network, and belongs to the field of information processing. The method comprises the following steps of firstly, modeling a labeling process of named entity recognition into a Markov decision process (MDP) so that a model for identifying the named entity based on enhanced learning is provided, namely MM-NER. MM-NER is the first to apply the Monte Carlo Tree Search (MCTS) enhanced MDP model to named entity recognition (sequence tag task).A strategy value network is designed on the basis of an MDP state definition to obtain label probability and label sequence accuracy evaluation, and an MCTS is used for simulation, so that a label sequence with more global consciousness is searched out. In the inferenceprocess, the strategy value network is directly used to ensure that the identification effect is basically consistent with the tree search strategy, and the time complexity is greatly reduced. An experimental result of the present invention on the CoNLL2003 named entity identification data set demonstrates the effectiveness of the MM-NER with the K-step exploration decision mechanism.

The invention discloses a cloud data center application perceptible distributed multi-resource combined path optimal selection method. The method comprises the steps that a cloud data center application perceptible cloud resource manager is responsible for collecting resource state information of a network, a CPU and the like in basic resources; Forming an unsupervised deep hybrid architecture model by adopting a mode of combining reinforcement learning, a value network and a strategy network, and assessing model training and node moving positions of various request streams; A novel tree search algorithm is adopted, namely, a Parallel Monte Carlo Tree Search(PMCTS) algorithm is adopted, an appropriate resource path is searched for each type of request flow in an accelerated mode, and the value and the strategy network are combined, so that an optimal resource path selection result is given. By adopting the technical scheme provided by the invention, the minimum total response time delay of various intensive request streams of the cloud data center can be ensured.

The invention discloses a database query optimization method. The database query optimization method comprises a connection sequence selector and a self-adaptive decision network, wherein the connection sequence selector is used for selecting an optimal connection sequence in the query plan and comprises a new database query plan coding scheme, and codes are in one-to-one correspondence with the connection sequence; a value network which is used for predicting the execution time of the query plan, is trained by the query plan and the corresponding real execution time, and is used for reward feedback in Monte Carlo tree search; a Monte Carlo tree search method which is used for simulating and generating multiple different connection sequences, evaluating the quality of the connection sequences through a connection sequence value network, and returning a recommended connection sequence after preset exploration times are reached. And the adaptive decision network is used for distinguishing whether the query statement uses the connection sequence selector or not, so that the overall performance of the optimization system is improved. According to the method and the system, the limitation of a traditional query optimizer can be effectively avoided, and the database query efficiency is improved.

The invention provides a deep reinforcement learning interactive recommendation system and method based on knowledge enhancement, and relates to the technical field of recommendation. The system comprises a data acquisition and cleaning module, an environment simulator construction module, a knowledge graph construction module, a graph convolution module, a user state representation module, a strategy network module and a value network module. According to the method, rich semantic information in a knowledge graph is combined, a graph convolutional network structure is utilized, embedded representation of adjacent entities is propagated recursively along high-order connectivity, a graph attention network thought is adopted, item representation is enhanced by utilizing the rich semantic information in the knowledge graph, and meanwhile, a user-item bipartite graph is fused, so that the method is more efficient and efficient. The potential relationship is fully mined from collective user behaviors, so that the dynamic preference of the user is accurately captured, and the optimal recommendation strategy is autonomously learned by using deep reinforcement learning, so that the recommendation accuracy is improved.

The invention discloses a method and device for evaluating scheduling strategies in a virtual environment and equipment and belongs to the technical field of computers. The method comprises the following steps of acquiring frame data generated when an application of the virtual environment runs and extracting target frame data corresponding to a target virtual object from the frame data; extracting characteristics of the target frame data to obtain state characteristics of the target virtual object under the current situation state; calling a value network prediction model to process the statecharacteristics to obtain the expected return income of the target virtual object for executing N scheduling strategies in the current situation state. According to the method, by acquiring the target frame data corresponding to the target virtual object, extracting the state characteristics of the target frame data and calling the value network prediction model to process the state characteristics, the expected return income of each scheduling strategy executed by the target virtual object is obtained, the value network model is created in the virtual environment, and the accuracy of controlling the virtual object to execut the scheduling strategies through AI is improved.

The invention belongs to the technical field of power system scheduling, and discloses a value network-based scheduling optimization method of an energy storage system. According to the method, by adopting a strategy that the energy storage system automatically adjusts an output plan thereof under an energy value time-varying background to achieve the maximum energy value, rasterization processing is firstly carried out on a two-dimensional bounded state space which is enclosed by time and an energy storage state, a value network is constructed according to an inverse order of the time, each unit in the network corresponds to one point of the state space, the maximum value from the state point to a scheduling cycle end state point is calculated as the value of the point and the output plan corresponding to a maximum value chain recorded by a scheduling cycle starting state point is an optimal solution under the rasterization precision; the state space with smaller granularity is generated near a low precision solution of the output plan in the previous step; and the steps are repeated and convergence of the solution is prompted through repeated iteration until the precision meets the requirements. The value network-based scheduling optimization method of the energy storage system is high in solution precision, fast in convergence and good in robustness, and the regulation economy and reliability of the energy storage system can be better ensured.

A system, computer program product and method for a global transaction manager in a federated value chain network. The federated value chain network includes a plurality of local networks having shared access to two or more shared databases on a service provider computer over a network via a database router module. The computer program product includes receiving a request for an order for goods or services from a first company in one of the plurality of local networks in the federated value chain network, searching for one or more second companies having matching goods or services over one or more of the plurality of local networks, sourcing the matched one or more second companies, creating a transaction over one or more segments to effect the movement of the good or services, involving one or more third companies, from a source location to a destination location, and managing the handoffs between the relevant first, second and third companies in order to ship the goods or services.

The invention discloses a structural vibration control method based on reinforcement learning, a medium and equipment. The method comprises the following steps: establishing a kinetic equation and a reward function of a controlled system; establishing and initializing a strategy network, a target strategy network, a value network and a target value network; establishing a playback pool; data interaction is achieved. Meanwhile, the control signal, the feedback signal and the reward signal are stored in a playback pool, the control signal, the feedback signal and the reward signal are provided for a reinforcement learning algorithm in a random sampling mode to update parameters of a strategy network and a value network, and a soft update mechanism is adopted to update parameters of a target strategy network and a target value network; obtaining a final strategy neural network as a controller; and deploying a controller, taking the feedback signal acquired by the sensor as the input of the neural network, and outputting a control signal after the forward calculation of the neural network to complete the control operation of the structural vibration. The invention provides a more intelligent control method for vibration control of a complex structure, and has excellent control performance and engineering practicability.

The invention discloses an intelligent substation sampled value network data switching method and device and aims to solve the technical problems of reducing protector cost and avoiding delay uncertainty in network data packets. The method includes the steps: a bay level merging unit acquires alternating current information of a primary transformer; a data switching device forwards the formation in sequence by means of synchronous time division multiplexing. According to the device, a voltage merging unit and the bay level merging unit are connected with a protector through the data switching device; the data switching device is a field programmable gate array. Compared to the prior art, the intelligent substation sampled value network data switching method and device has the advantages that data frames of input channels are arranged in a certain sequence when forwarded, the passages have totally same time slice timeslots, fewer fiber-optic interfaces are used, field fiber arrangement is simplified, the protector cost is reduced, the problem of delay uncertainty in network data packets is avoided, and the method and device is applicable to situations with low network throughput and fixed data frame size.

The invention discloses a reinforcement learning exploration method and device based on a generative adversarial mechanism, and the method comprises the steps: constructing a first action value network, a second action value network, a state value network, a target state value network, a strategy network, a density model network and an identification network; updating the first action value network, the second action value network, the state value network, the target state value network, the strategy network, the density model network and the identification network based on a generative adversarial mechanism and a learning process of an offline reinforcement learning algorithm; and generating an updated strategy model according to the plurality of updated networks, and testing the strategymodel. According to the method, an exploration algorithm utilizing a correct decision acceleration and stable reinforcement learning training process in an exploration process is designed.

The invention discloses a reinforced learning based robot joint motion control method and system. The method comprises the following steps: obtaining to-be-operated track of a robot terminal; calculating position increment within each interpolation period of the robot joint according to the to-be-operated track of the robot terminal and a robot inverse kinematic model; determining position increment compensation within each interpolation period of the robot joint according to a policy network; taking sum of given position increment within each interpolation period and position increment compensation as motion parameters of the robot joint, inputting the motion parameters into a robot to obtain practical motion amount within each interpolation period of the robot joint; performing real-timetraining update on a value network according to the given position increment and the practical motion amount; after operation of the to-be-operated track is accomplished, performing training update on the policy network according to parameters updated according to each interpolation period, of the value network; and controlling motion, in next to-be-operated track, of the robot joint by adoptingthe updated policy network. The reinforced learning based robot joint motion control method has the characteristics of being small in errors and high in efficiency.

The invention discloses an artificial intelligence AI model training method and device, equipment and a medium, and relates to the field of artificial intelligence machine learning. The method comprises the steps that an artificial intelligence AI model is called to conduct game match in a game program to obtain training data, wherein the training data comprises a reference game state in the gamematch, a target game action output by a decision network according to the reference game state and a state value output by a value network according to the reference game state, the state value comprises k state sub-values on k value classifications, and k is an integer greater than 1; according to the training data and k value calculation formulas corresponding to the k value classifications, theaction value of the target game action adopted by the artificial intelligence AI model in the reference game state is calculated, wherein the action value comprises k action sub-values on the k valueclassifications; and the artificial intelligence AI is trained model according to the difference between the state value and the action value. The method can improve the accuracy of estimating the state value of the value network.

A system and method of enabling creation and management of a real time digital value network. Nodes are created, each node representing a user and a digital value network is then created based on the nodes thus created. The digital value network comprises a nodes and plurality of connections and real time interactions, each connection connecting one or more nodes based on a user profile of the user and a plurality of values associated with the one or more nodes. Thus, the VNNMS provides online support to customers or buyers and sellers in a real-time shared environment and enables value creation, optimization and value delivery. End-to-end approach delivers guaranteed realized savings to group of customers forming the network. The network of customers can be managed and expanded globally.