49 results about "Agent behavior" patented technology

The present invention relates to a personal user agent acting on behalf of a group of member device agents in a communication network. The personal user agent provides its group of member device agents with a specialized proxy function, while representing the group to the network as a single user agent. To devices on the communication network, the personal user agent behaves and is viewed as a user agent. To the group of member device agents, the personal user agent behaves and is viewed as both a registrar and a proxy server.

A method and apparatus for processing attached E-mail data and storage medium for processing a program for attached data in which, in sending/receiving an E-mail using a virtual pet displayed on a real world oriented GUI screen as an agent, in which parameters controlling the behavior of the agent are processed on the side of a server and returned to the user to enable the user to change the agent behavior. A user appends agent parameters to an E-mail to send the resulting mail to a pre-set address. An Internet service provider receives the E-mail to judge whether the title (Subject:) of the mail header is (cocktail lounge). If the title is the cocktail lounge, the physical power parameter of the agent parameter is lowered to append it to the E-mail Magneto-optical disc to send the mail back to the user. On the user side, the tipsy behavior of the drunken pet is displayed on the basis of the agent parameters appended to the E-mail Magneto-optical disc.

A computer simulation method is provided for modeling the behavioral expression of one or more agents in an environment to be simulated, then running a simulation of the modeled agent(s) against real-world information as input data reflecting changing conditions of the environment being simulated, and obtaining an output based on the modeled agent(s) response(s). The simulation method models the underlying cultural, social, and behavioral characteristics on which agent behaviors and actions are based, rather than modeling fixed rules for the agent's actions. The input data driving the simulation are constituted by real-world information reflecting the changing conditions of the environment being simulated, rather than an artificial set of predefined initial conditions which do not change over time. As a result, the simulation output of the modeled agent's responses to the input information can indicate more accurately how that type of participant in the simulated environment might respond under real-world conditions. Simulations can be run on global networks for agent types of different cultures, societies, and behaviors, with global sources of information. Simulation environments can include problems and situations in a wide range of human activity. Robust new visual tools are provided for discerning patterns and trends in the simulation data, including waveform charts, star charts, grid charts, and pole chart series.

The present invention relates to an object detection and behavior recognition system using three-dimensional motion data. The system receives three-dimensional (3D) motion data of a scene from at least one sensor, such as a LIDAR sensor. An object is identified in the 3D motion data. Thereafter, an object track is extracted, the object track being indicative of object motion in the scene over time. Through Dynamic Time Warping (DTW) or other comparison techniques, the object track is compared to a database to identify the behavior of the object based on its object track.

A method, computer program product, and computer system for configuring a stochastic simulation scenario, wherein the stochastic simulation scenario may include one or more variables, wherein at least a portion of the one or more variables may include agent behavior, and wherein the stochastic simulation scenario may be randomized and digital. The stochastic simulation scenario may be executed to generate one or more results of the stochastic simulation scenario. At least a portion of the one or more variables may be optimized using one or more optimization metrics on the one or more results of the stochastic simulation scenario, wherein at least the portion of the one or more variables may be modified based on game theory.

Embodiments emulate agents in a contact center, or enterprise, that initiate communication sessions to each other over communication networks and measure data quality using one or more metrics. That data measured can include voice and video data. Examples of communication networks used to place the voice calls include Internet Protocol (IP) networks and hybrid networks including more traditional telephony components. The data collected by the emulated agents is sent to a monitoring server. The monitoring server and the emulated agent behavior can be configured and scheduled via a web interface. In addition, the web interface provides the user with detailed reports, performance summaries, and visualizations of data collected, further enabling the user to troubleshoot a contact center or enterprise over a variety of communication networks.

An intelligent control method for vertical recovery of carrier rockets based on deep reinforcement learning is disclosed. A method of autonomous intelligent control for carrier rockets is studied. Theinvention mainly studies how to realize attitude control and path planning for vertical recovery of carrier rockets by using intelligent control. For the aerospace industry, the autonomous intellectualization of spacecrafts is undoubtedly of great significance whether in the saving of labor cost or in the reduction of human errors. A carrier rocket vertical recovery simulation model is established, and a corresponding Markov decision-making process, including a state space, an action space, a state transition equation and a return function, is established. The mapping relationship between environment and agent behavior is fitted by using a neural network, and the neural network is trained so that a carrier rocket can be recovered autonomously and controllably by using the trained neural network. The project not only can provide technical support for the spacecraft orbit intelligent planning technology, but also can provide a simulation and verification platform for attack-defense confrontation between spacecrafts based on deep reinforcement learning.

A method that incorporates a detailed, precise procedure of designing a user interface by utilizing agent behavioral models. This method applies quantitative and qualitative agent behavioral models derived through the Categorize Describe-Model (CDM) methodology to the iterative design stage of interface development. The method includes: (1) categorizing at least two users; (2) validating targeted user behaviors and preferences; (3) capturing emergent behaviors and preferences; (4) tracking design requirements and implementations; (5) accommodating diversity in performance and preference during interactive testing; and (6) customizing a user interface design to each of the at least two users.

A method of predicting the behavior of software agents in a simulated environment involves modeling a plurality of software agents representing entities to be analyzed, which may be human beings. Using a set of parameters that governs the behavior of the agents, the internal state of at least one of the agents is estimated by its behavior in the simulation, including its movement within the environment. This facilitates a prediction of the likely future behavior of the agent based solely upon its internal state; that is, without recourse to any intentional agent communications. In the preferred embodiment the simulated environment is based upon a digital pheromone infrastructure. The simulation integrates knowledge of threat regions, a cognitive analysis of the agent's beliefs, desires, and intentions, a model of the agent's emotional disposition and state, and the dynamics of interactions with the environment. By evolving agents in this rich environment, we can fit their internal state to their observed behavior. In realistic wargame scenarios, the system successfully detects deliberately played emotions and makes reasonable predictions about the entities' future behavior.

The embodiment of the invention discloses a self-involved efficient proxy resource supply system and method. The system comprises a proxy supply terminal, a service terminal and a proxy resource server, wherein the proxy supply terminal constructs a multi-proxy engine in which a plurality preset proxies coexist and a corresponding local strategy base; the service terminal obtains the use behaviordata of the proxies, packs the use behavior data and uploads the use behavior data to the proxy resource server; the proxy resource server constructs a proxy base, a proxy strategy base and a proxy behavior base, provides proxy resource information for the proxy supply terminal, generates an corresponding efficient proxy strategy by performing machine learning according to the use behavior data and updating the efficient proxy strategy to the proxy strategy base. In the self-involved efficient proxy resource supply system and method disclosed by the embodiment of the invention, the multi-proxyengine is used for ensuring a plurality of proxy services can be provided quickly and efficiently after the IP of a single proxy resource is changed, and collecting the use behavior of the proxy to perform the machine learning to generate the efficient proxy strategy, the problem of low proxy utilization is solved, and thus reducing the proxy cost.

The invention discloses a behavior simulation training method for an air intelligent game. The method comprises the following steps: S1, constructing an intelligent agent game decision model; S2, determining an environment state and an action space, and shaping a continuous non-sparse reward function of each action; S3, carrying out an air game in the model, and executing the following steps: S31, generating a next environment state according to an executed action, obtaining a reward, and carrying out loop iteration in sequence to realize maximum accumulated reward; S32, realizing reverse reinforcement learning based on expert behaviors, and obtaining a target reward function; S33, calculating the similarity between each agent behavior and the expert behavior; S34, obtaining a comprehensive reward; and S4, training the agent game decision model. According to the method, a traditional low-efficiency reward function design process and a model training random exploration process are improved, so that the reward function has interpretability and human intervention ability, the agent decision level and convergence speed are improved, and the cold start problem of model training is solved.

A method for quantifying contrast agent behavior of an administrated contrast agent within a region of interest, comprises: obtaining a time sequence of measurements, I(t), of contrast agent concentration starting before the contrast agent enters the region of interest and a time sequence of measurements, J(t), of the contrast agent concentration as the contrast agent perfuses the region of interest; deconvolving I(t) from J(t) to obtain an estimate of the tissue impulse response function for the region of interest, R′(t), where R′(t) is constrained to take into account at least one of physiologic and physical considerations and is chosen such that I(t)R′(t) is the maximum likelihood estimate of J(t), where denotes convolution; and determining the behaviour of the contrast agent in the region of interest based on R′(t).

The invention provides a multi-agent behavior decision method and device, an electronic device and a storage medium, and the method comprises the steps: constructing each agent and corresponding environment information into a graph based on a graph generation module in a multi-agent behavior model; based on an information transmission module in the multi-agent behavior model, encoding each agent and the corresponding environment information to obtain a joint encoding state corresponding to each agent; based on a strategy optimization module in the multi-agent behavior model, determining an initial decision of each agent in combination with the joint coding state corresponding to each agent; and based on a collision avoidance module in the multi-agent behavior model, performing variable step size control on the initial decision of each agent, and determining a final decision of each agent in combination with the repulsive force corresponding to each agent. According to the method, the problem that reinforcement learning is difficult to converge in a large-scale agent scene is solved, and high-performance autonomous decision-making ability and collision avoidance effect in a multi-agent system are realized.

The invention provides an asymmetric distributed constrained optimization method for a multi-Agent system. The method includes the following steps that a controller receives Agent information and obtains an asymmetric relation and an Agent connected graph is established; according to the connecting relation between nodes in the Agent connected graph, a chain structure based on the sequence is established; depth-first trees are established for the Agent connected graph in a depth-first mode; through a branch and bound strategy, weighted values of root nodes of different depth-first trees are calculated and finally an optimal tree structure is obtained; depending on the obtained optimal tree structure, Agent behavior parameters in the multi-Agent system are adjusted. According to the asymmetric characteristic of the multi-Agent system, a prediction mechanism is introduced, the defects of incomplete information decision making are overcome, an Agent can estimate the influences of behaviors of the Agent, the more effective strategy is obtained, the non-monotone characteristic of gains is changed to a certain degree, and the balance between individual benefits and overall benefits is achieved.

Systems, methods, devices, and non-transitory media of the various embodiments may enable formalized execution of model integrated descriptive architecture languages, such as Unified Modeling Language (UML). In addition, the systems, methods, devices, and non-transitory media of the various embodiments may be used to generate a graphical representation of a simulation. In an embodiment, a Unifying Framework for Orchestration and Simulation (UFOS) environment may be implemented to generate an agent behavior model using abstract objects from an ontological model.

The invention belongs to the technical field of underwater unmanned system modeling and simulation, and particularly relates to a UUV intelligent agent behavior learning and evolution model based on achaotic immune genetic mechanism. The method comprises the following steps: firstly, loading a to-be-solved problem and constraint conditions as antigen Ag, and generating an initialized antibody population according to a vaccine population, a memory population and a chaotic mechanism; secondly, controlling the convergence direction of the learning process by utilizing a vaccination mechanism according to an antibody fitness calculation result, and completing updating of an antibody memory bank; and finally, sequentially designing a selection operator based on roulette, a crossover operator based on adaptive adjustment and a mutation operator based on Gaussian and polynomial mixing to realize diversity of the antibody population, and performing premature suppression, thereby realizing updating and iteration of the antibody population. The model combines the advantages of the global search capability of a basic genetic algorithm and the local search capability of an immune and chaoticmechanism, and promotes the quick learning and evolution of behavior rules by continuously adjusting and optimizing the search space of a problem solution.

The invention relates to a simulation system and method for multi-agent group behaviors. The system comprises a simulation engine layer, a simulation component layer, a system application layer, a module model library and a behavior node library, the simulation engine layer comprises a simulation world class, a simulation scene class, a simulation scheduler, a time controller, an entity element class, a simulation event, a behavior node class and a knowledge sharing blackboard class, and the simulation component layer comprises an agent entity model, an agent module model, an entity management model, a component management model, an event management model and a behavior management model. The system application layer comprises a plurality of agent model components, and the module model library comprises a motion class module, a perception class module, a communication class module and a decision class module; the behavior node library comprises a general algorithm node, a group behavior node and a service interface node. By adopting the system, multi-agent behavior simulation can be realized.

Systems and methods for detecting out-of-bounds behavior of an agent in a simulation including multiple agents and an ego vehicle may include: providing a scenario governing behavior of an agent in the simulation; operating the simulation in accordance with the scenario across multiple occurrences; collecting data regarding behavior of the agent in the scenario for each occurrence; comparing collected data regarding behavior of the agent in the scenario with a governing data set for the scenario; and reporting an out-of-bounds condition to a system user when the results of the comparison indicate that the behavior of the agent in the scenario deviates from the agent behavior in the governing data set scenario by more than a predetermined amount.