Abstract: Example implementations described herein involve systems and methods for bound enhanced reinforcement learning systems for distribution supply chain management which can include initializing a replay buffer, a first state-action value function network having first random weights, and a second state-action value function network having second random weights; determining an action corresponding to an inventory ordering quantity at one or more facility in a multi-echelon supply chain network based on an (epsilon) ?-greedy exploration policy; executing the action in a simulated environment, and storing transition results in the replay buffer; calculating an upper bound and a lower bound of the optimal inventory costs; incorporating the upper bound and the lower bound with at least hyper-parameters T1, ?2 in updating at least one of the first or the second state-action value function networks; and performing a gradient descent on the first state-action value function network based on the upper or the lower bound.

Abstract: A charging method and apparatus, a charging management device, and a computer-readable storage medium are described. The method includes: acquiring a temperature of a battery to be charged; acquiring activation energy data of the battery to be charged; determining a charge cut-off voltage according to the temperature and the activation energy data; and charging the battery to be charged and controlling a maximum charging voltage of the battery to be charged to be less than or equal to the charge cut-off voltage. By using the method, an accuracy of the determined charge cut-off voltage can be improved, thereby enhancing the battery endurance and improving the user experience.

Abstract: A method of mitigating crack formation in an article produced by additive manufacturing process is disclosed. The method utilizes Selective laser melting (SLM). The method contains the steps of utilizing a selective leaser melting (SLM) apparatus to produce a first layer of particles of a selected material on a substrate and melting and solidifying a first group of selected areas of the first layer of particles, wherein the melting and the solidification results in a pre-determined structure of the selected material; repeating utilization of SLM apparatus to produce second and subsequent layers as needed to produce layers of desired configurations to result in a predetermined scaffolding structure; and utilizing a selective leaser melting (SLM) apparatus to produce an article of specified thickness and shape on top of and in contact with the scaffolding structure. An article made by the method of this disclosure is described.

Abstract: An apparatus for generation of a 3D representation of a scene includes a memory for storing a plurality of images depicting a scene; and processing circuitry in communication with the memory. The processing circuitry is configured to generate a depth map for the plurality of input images depicting a scene and unproject the depth map to a three-dimensional (3D) point cloud. The processing circuitry is also configured to project a plurality of 3D points from the 3D point cloud to a plurality of two dimensional (2D) planes and extract a plurality of triplane features from the plurality of 2D planes. The processing circuitry is further configured to generate a 3D representation of the scene based on combining the plurality of triplane features and a plurality of image features extracted from the plurality of input images.

Abstract: The disclosure discloses a negative-electrode-free secondary battery, modified graphite, a modified current collector, and a method for preparing same. The method for preparing the modified graphite includes performing hydrogen-argon mixed gas plasma modification treatment on graphite to prepare the modified graphite. The thickness of the graphite is 20 nm-2 ?m, the diameter is 0.1-50 ?m, and the degree of graphitization is 80% or more. The power of the hydrogen-argon mixed gas plasma modification treatment is 30-250 W, and the time is 2-25 min. The modification method of the disclosure can alleviate the defects and optimize the structure of the surface of the graphite, thereby constructing physical and chemical characteristics adapted to the negative-electrode-free secondary battery to significantly improve the long cycle life of the negative-electrode-free secondary battery.

Abstract: A method for making LNO film, including the steps of identifying a substrate, identifying a deposition target, placing the substrate and deposition target in a deposition environment, evolving target material into the deposition environment, and depositing evolved target material onto the substrate to yield an LNO film. The deposition environment defines a temperature of between 500 degrees Celsius and 750 degrees Celsius and a pressure of about 10?6 Torr. A seed or buffer layer may be first deposited onto the substrate, wherein the seed layer is about 30 mole percent gold and about 70 LiNbO3.

Abstract: An adjuvant for intradermal immunization and a use thereof are provided. The adjuvant includes the following components in parts by mass: 10 parts to 25 parts of squalene, 10 parts to 25 parts of sea buckthorn oil, 25 parts to 35 parts of caprylocaproyl polyoxyl-8 glycerides, and 25 parts to 45 parts of Tween 80. The adjuvant for intradermal immunization has a low viscosity, is favorable for injection and absorption, and can significantly enhance the immunization efficacy of a vaccine.

Abstract: An alloy comprising 92 at % aluminum, 2 at % titanium, 2 at % iron, 2 at % cobalt, and 2 at % nickel. A method of making an alloy is disclosed. The method contains the steps of providing particles of desired composition, utilizing a selective leaser melting (SLM) apparatus producing a first layer of the particles on a substrate and melting and solidifying a first group selected areas of the layer of particles, wherein the melting and the solidification results in an alloy of desired composition, containing intermetallic lamellae and has thickness equal to thickness of the first layer. The process is repeated to produce an object of specified thickness and shape containing the melted and solidified areas.

Abstract: A gripping device for a segment erector is provided, including a drive assembly, a cylinder barrel assembly, a piston, a guide assembly and a centering assembly. The centering assembly includes a centering main structure, cameras and a drive component. The centering main structure includes an end cover, an assembly frame, a bottom cover, and a plurality of claw plates and claw arms. The end cover is fixedly connected to a gripping and lifting plate and the bottom cover. Each of the claw plates is connected to the centering main structure via a matching claw arm. The cameras are mounted on the assembly frame and arranged to face the claw plates. In the process of a gripping head approaching a segment, a relative positional relationship between the gripping head and the segment is identified based on images captured by the cameras. The drive component drives the assembly frame.

Abstract: Techniques for defining and using reusable modules to generate form control code are disclosed, including: displaying a form control implementation interface for applying form control functions to forms; receiving via the form control implementation interface: a first user input selecting a form control function of the form control functions; a second user input selecting one or more input parameters, for the form control function, that are to be extracted from the target form; a third user input selecting a target field of a target form, one or more attributes of the target field to be modified via execution of the form control function; generating form control code that extracts the one or more input parameters from form data received for the target form and applies the form control function to the one or more input parameters to modify the one or more attributes of the target field.

Abstract: Techniques for defining and using reusable modules to generate form control code are disclosed, including: displaying a form control implementation interface for applying form control functions to forms; receiving via the form control implementation interface: a first user input selecting a form control function of the form control functions; a second user input selecting one or more input parameters, for the form control function, that are to be extracted from the target form; a third user input selecting a target field of a target form, one or more attributes of the target field to be modified via execution of the form control function; generating form control code that extracts the one or more input parameters from form data received for the target form and applies the form control function to the one or more input parameters to modify the one or more attributes of the target field.

Abstract: Certain aspects of the present disclosure provide techniques for performing 3D object detection. Such techniques may include obtaining one or more inputs associated with one or more two-dimensional (2D) views of a scene; selecting a set of 2D views of the scene from a plurality of 2D views of the scene based on the one or more inputs, the set of 2D views comprising a first 2D view of the scene and a second 2D view of the scene; and performing three-dimensional (3D) object detection in the scene based on the set of 2D views.

Abstract: Systems and methods and computer-readable media are provided for generating an interactive node graph showing aggregate node and/or edge metrics. The interactive node graph includes nodes and edges, each representing a process state or a connection between process states. The data management system marks each node graphically based on a first metric type that is based on an aggregation of occurrences of a corresponding process state and marks each edge graphically based on a second metric type that is based on an aggregation of occurrences of a transition between corresponding process states. The data management system causes display of the interactive node graph, receives input modifying a particular metric type of the first metric type or the second metric type, and causes display of an updated interactive node graph based at least in part on the particular metric type as modified.

Abstract: Systems, methods, and computer-readable media are provided for generating an interactive node graph showing process states. A data management system accesses a data structure that includes a set of candidate process records representing a plurality of candidate process states, where each candidate process record represents a candidate process state and one or more sequentially connected candidate process states. The data management system generates an interactive node graph that includes nodes, each node representing a candidate process state. The interactive node graph includes edges representing candidate connections between candidate process states. The data management system causes display of the interactive node graph. The data management system receives input modifying settings for displaying the interactive node graph and causes display of an updated interactive node graph using the settings as modified.

Abstract: Systems, methods, and computer-readable media are provided for storing a snapshot of an interactive node graph. A data management system generates an interactive node graph having nodes that represent process states and edges representing connections between process states. The data management system uses a first live metric for determining aggregated node values and a second live metric to use for determining aggregated edge values. Based at least in part on a selection of an option to save the interactive node graph, the data management system stores a snapshot. The snapshot includes the aggregated node values based on the first live metric as of a particular time, the aggregated edge values based on the second live metric as of the particular time, a first mapping between the first live metric and the nodes, and a second mapping between the second live metric and the edges.

Abstract: Systems, methods, and computer-readable media are provided for animating an interactive node graph. A data management system generates an interactive node graph having nodes that represent process states and edges representing connections between process states. The data management system uses a first live metric for determining aggregated node values and a second live metric to use for determining aggregated edge values. The data management system causes display of the interactive node graph according to a selected data slice of a plurality of data slices. Based at least in part on a selection of an option to play the interactive node graph through the plurality of data slices, the data management system updates the display of the interactive node graph to show a different data slices of the plurality of data slices, and, after an amount of time, another different data slice of the plurality of data slices.

Abstract: An optical lens, an optical module, and an electronic device are provided. The optical lens includes: a first lens (1), a reflector (2), a diaphragm (3), a second lens (4), a third lens (5), a fourth lens (6), and a fifth lens (7) in sequence from an object side to an image side.

Abstract: A refrigerator includes a box body, a hinge assembly, and a door body. The hinge assembly includes a plate body, a locking hook, and a blocking portion. A free end of the locking hook extends in a direction away from the plate body and is bent in a direction proximate to the plate body, so as to form an opening toward the plate body. In a case where the door body is in a closed state, the blocking portion is located in the opening. During a process from closing to opening of the door body, the locking hook is deformed under force to be disengaged from the blocking portion and unlocked.

Abstract: Systems and methods described herein train a deep reinforcement learning agent to solve weakly coupled Markov decision processes using Lagrangian relaxation in a model-free setting. By relaxing linking constraints separate subproblems may be obtained that are easier to solve when considered individually. In embodiments, this is accomplished by collecting experience tuples from a main problem, decomposing them into subproblems, and introducing Lagrangian multipliers to manage linking constraints. Transition experiences are stored in a replay buffer and Lagrangian action-values learn for each subproblem via DQN using a relaxed Bellman equation. The method includes estimating the overall Lagrangian action-value function, solving an optimization problem over Lagrangian multipliers, and choosing actions greedily.

Abstract: Systems and methods for implementation of a multi-agent reinforcement learning based decision system for a materials handling system, including initializing a simulation environment comprising decision points for dispatching materials and attributes of the materials handling system, the simulation environment configured to request a decision for materials dispatch at the decision points to the multi-agent reinforcement learning based decision system; initializing the reinforcement learning agents representative of the decision points for the multi-agent reinforcement learning based decision system; initializing domain expert heuristics for the decision points for the multi-agent reinforcement learning based decision system; and iteratively training the reinforcement learning based decision system with the initialized domain expert heuristics on the simulation environment.