88 results about "Balance strategy" patented technology

A system and method of allocating a resource to a service request. Servers are virtually assigned through a hierarchy including groups of servers and super groups of the groups of servers. Each service request type is associated with a service index that determines both a super group and a load balancing policy for that service request type. A load balancing policy may be applied to select a group of servers among several server groups associated with the super group. Another load balancing policy is applied to the selected group of servers to select a server to handle the service request. The hierarchical data structure allows servers to be configured into overlappable, arbitrary subsets that can address service requests matching configured content rules. Load balancing policies can be selected by service request type.

A method and system for managing graphics load balancing strategies are disclosed. The method comprises using a plurality of rendering servers to render a multitude of graphics frames for a display device, wherein each of the rendering servers has an associated workload; identifying a plurality of load balancing strategies for balancing the workloads on the rendering servers; selecting one of the load balancing strategies; and using the selected one of the load balancing strategies to balance the workloads on the rendering servers. One or more defined metrics are monitored; and in response to a defined changed in said one or more defined metrics, another one of the load balancing strategies is selected and used to balance the workloads on the rendering servers. In one embodiment, the load balancing policy can be changed in real-time during the course of an application session.

A method and system for managing graphics load balancing strategies are disclosed. The method comprises using a plurality of rendering servers to render a multitude of graphics frames for a display device, wherein each of the rendering servers has an associated workload; identifying a plurality of load balancing strategies for balancing the workloads on the rendering servers; selecting one of the load balancing strategies; and using the selected one of the load balancing strategies to balance the workloads on the rendering servers. One or more defined metrics are monitored; and in response to a defined changed in said one or more defined metrics, another one of the load balancing strategies is selected and used to balance the workloads on the rendering servers. In one embodiment, the load balancing policy can be changed in real-time during the course of an application session.

The invention discloses a method and a device for dynamically distributing cluster services. The method comprises the following steps of: monitoring the loading condition of each node in a cluster; predicting the future loading condition of each node according to the historical monitoring data and current monitoring data of each node; and developing a service distribution scheme for each node in the cluster according to the predicted future loading condition of each node and a preset service balancing strategy, and regulating service distribution for each node in the cluster according to the developed service distribution scheme. By monitoring the load of the nodes and pre-optimizing the service distribution according to the monitored load, the method can balance the load, enhance concurrent services and reduce the probability of down caused by the overload of the nodes.

The invention discloses a scheme for controlling the grade parking of a permanent magnet motor-driven pure electric vehicle. The pure electric vehicle has an electrical system comprising a battery pack, a motor, a power converter, a motor controller and a vehicle controller and further comprising corresponding voltage, current and tilt sensors. The scheme comprises two steps of: (1) detection for the initial position of a rotor of a permanent magnet synchronous motor: a detection voltage is applied to the permanent magnet synchronous motor by the power converter, and the initial spatial position of the motor rotor in the case that the vehicle is stationary is calculated through a current feedback signal; and (2) torque self-balancing strategy based on vector control: the torque self-balancing characteristic can enable driving torque to track load torque automatically. The torque required by parking is calculated in accordance with the tilt angle of a ramp on which the vehicle is to be parked, and then, reasonable current amplitude and phase are given according to the initial position of the rotor and the torque required by grade parking in order to enable the driving torque to be equal to the load torque, therefore, stationary parking of the vehicle on the grade is realized.

The invention discloses a load balancing method of a high-speed network. Aiming at the problem of low balancing degree of the traditional load balancing method, the invention provides a load balancing method of a high-speed network, which comprises the steps of: extraction of front information of an IP data packet, Hash operation, queue distribution and storage management, multimodal parallel classification and scheduling. The load balancing method utilizes a dynamic self-adaptive Hash balancing strategy combining static balancing with dynamic self-adaptive balancing to ensure that the current flow load condition can be effectively utilized and a high-speed data stream can be further distributed to a plurality of low-speed network interfaces in a balanced manner, thus, the balance degree among all subchannels is improved.

The embodiment of the invention discloses a memory management method, comprising the following steps: during idle time, checking a grouping information sheet and an erasable count table, judging whether to trigger a consumption balance strategy and if the consumption balance strategy is needed to be triggered, inserting a recycled blank block in an unordered queue; and during data write operation time, obtaining the blank block from the unordered queue and writing data in the obtained blank block. The invention also discloses a hard disk and a memory system. The embodiment of the invention can be used for recycling the blank block with large time-consuming amount in idle time of the system, thereby being favor of saving data write cost.

The present invention relates to method and apparatus for video coding by ABT-based just noticeable difference (JND). For building the just noticeable difference model, spatial content information (SCI) is used to represent the spatial appearance similarity between one macroblock and its sub-blocks and the motion characteristic distance (MCD) is used to represent the motion characteristics similarity between one macroblock and its sub-blocks. For intra frames, the balance strategy based on the obtained SCI of the macroblock is used to generate the ABT-based JND model. For inter frames, the balanced strategy based on the obtained SCI and MCD of the macroblock is used to generate the ABT-based JND model. Using the ABT-based JND model, the residual coefficients for each block in a frame is filtered to obtain a reduced set of residual coefficients for transmission without degradation in visual quality.

The invention relates to a peak adjusting method of the electrical network by taking external power reception into consideration. The method comprises the following step that 1) the internal power structure and the external power reception manner of the electrical network whose peak is to be adjusted are obtained; 2) the power load of the electrical network within the peak adjusting period is predicted according to historical load information of the electrical network; and 3) a balanced peak adjusting strategy is made by considering the internal power structure, the external power reception manner as well as the power load state in the peak adjusting period, and internal and external power supplies are combined to adjust the peak. Compared with the prior art, the method has the advantages including that the peak adjusting effect is good.

The invention discloses an imbalance fault determining and balancing method for a battery pack, and the method comprises two steps: fault determining and balancing strategy. The step of fault determining comprises the steps: setting a work current direction of a battery, setting a judgment node K, a voltage characteristic parameter P, a deviation standard value Pm and a fault threshold value V, and carrying out calculation and comparison according to the fault data of voltage difference between two different working current directions of a cell, so as to classify the imbalance faults of the battery. Corresponding balancing measures are employed at the balancing strategy step according to different fault types. The method can effectively determine the type of an imbalance fault of the battery pack, employs a targeted balancing measure, improves the power performance of the battery pack, and prolongs the service life of the battery pack.

The invention discloses a load balance resource optimization method under a safety perceptive cloud radio access network. The method comprises the steps of collecting resource information allocated by a system; obtaining a convex optimization problem by use of regional balance and transmission balance strategies; optimizing a power under load balance condition by use of a self-adaption dynamic particle swarm optimization when a load balance threshold value is exceeded; initializing the parameters of the dynamic particle swarm optimization, initializing particle swarm space, and initializing a population; calculating the fitness of each unit in the population according to a fitness function; updating the optimum position of the current unit and the optimum position of the population according to a particle swarm optimization formula; obtaining the position and speed of each particle after displacement; carrying out intersection and variation on knowledge space; judging iteration convergence indexes; outputting optimum units if convergence occurs; and repeating the steps until the preset iteration times is finished if the convergence does not occur. The method is the load balance and power optimization method under safety perception through an optimization algorithm without changing network architecture under the condition that the system loads are imbalanced.

The invention relates to a regional power load balancing method. The regional power load balancing method comprises the following steps of: obtaining the actual available power of an electric vehiclecharging station; acquiring the real-time charging power and a power load balancing strategy of the electric vehicle charging station; determining whether the actual available power of the electric vehicle charging station is larger than or equal to the real-time charging power or not, if yes, performing charging according to the opening number and the charging power of the current charging piles,or else, calculating the opening number of the charging piles and the charging power limiting values of the opening charging piles according to the power load balancing strategy; and issuing a dispatching command to the electric vehicle charging station according to the opening number of the charging piles and the charging power limiting values of the opening charging piles obtained through calculation to dispatch the opening number of the charging piles and the charging power of the opening charging piles to achieve the real-time control for the total charging power of the electric vehicle charging station so as to preferentially ensure the normal electricity consumption of resident users.

The invention discloses wireless sensor network cluster construction and a cluster head updating method, belongs to the field of a wireless sensor network, and solves the problem of low stability of an existing wireless sensor network cluster structure. A construction method disclosed by the invention completes multi-layer cluster construction of the whole network from near to far by taking an aggregation node as a starting point, and particularly comprises the steps of: by taking the aggregation node as the center, partitioning sensor nodes in a wireless communication radius of the sensor nodes into a zeroth layer of cluster network, and then carrying out construction of a first layer of cluster network by using a node energy-saving equilibrium strategy; and then electing a node with the maximum cluster head capacity value from member nodes a previous layer of network cluster as a cluster head of a next layer of cluster network, constructing the next layer of cluster network, and sequentially carrying out iteration to construct a second layer and other layers of cluster networks. The layered cluster structure constructed by the wireless sensor network cluster construction disclosed by the invention has the characteristics of high stability and distributed demand-based cluster updating, and the defect of high cluster maintenance cost caused by periodic whole network cluster updating of a conventional method is overcome.

The invention provides a traffic balance and energy-saving control method, a realizing method and a realizing device. The traffic balance and energy-saving control method comprises the following steps that: a wireless network controller receives traffic load information and configuration information reported by a base station, obtains a corresponding traffic balance strategy scheme and an energy-saving control scheme through calculation according to the traffic load information and the configuration information, and issues corresponding traffic information and control instructions to the base station according to the traffic balance strategy scheme and the energy-saving control scheme; and the base station receives the traffic information issued thereto from the wireless network controller, switches on / off a cell cutoff function according to the received control instructions, and / or switches on / off a BCCH (Broadcast Control Channel) carrier-frequency priority distribution function. According to the traffic balance and energy-saving control method, the realizing method and the realizing device provided by the invention, the traffic balance control and the energy-saving strategy adjustment can be dynamically carried out according to the traffic load information reported by the base station, so that the purposes of reducing resource waste and lowering base station power consumption are realized.

The invention provides a hybrid particle swarm tabu search algorithm for solving a job-shop scheduling problem. Compared with other meta-heuristic algorithms, the algorithm has the characteristic of ''elite memory'' according to a PSO and has the characteristic of fast convergence, the PSO is taken as an initial solution source of TSAB tabu search, and an encoding and decoding mechanism for mapping a particle swarm continuous solution space into a discrete space of the job-shop scheduling problem is designed. A real number solution of the PSO is converted into an integer solution of the tabu search algorithm through a real integer encoding method and the integer solution of the tabu search algorithm is converted into the real number solution of the PSO through a real integer decoding method after one-time iteration; and a chance of accurate search is made in a potential space to own more exploration in a global search space. An improved PSO with a balancing strategy is provided, and a balance operator beta is introduced. The performance of the algorithm is greatly strengthened through these improvements and the actual job-shop scheduling condition is combined. The algorithm is high in practicability and good in usability.

The invention provides a ramp resistance obtaining method for a hybrid electric vehicle to solve the problems that ramp resistance calculation methods are lower in accuracy and deviation occurs easily in the prior art. According to the ramp resistance obtaining method, the driving force and the resistance of the vehicle are comprehensively considered, a vehicle dynamics equation is established accordingly, the acceleration resistance, the sliding resistance and the like outside the vehicle are considered during conversion, meanwhile, the inertia resistance caused by all rotating parts inside the vehicle is further considered, due to comprehensive consideration, the estimated ramp resistance is accurate and reliable with the obtaining method, and an important basis is provided for a shifting strategy, an SOC (state of charge) balance strategy and a torque distribution strategy for ramps.

The invention relates to a flexible comprehensive scheduling method considering double targets of cost and time. The flexible comprehensive scheduling is defined as that for a complicated single product with tree structure features in a processing technology diagram, a procedure node can be processed at a or a plurality of equipment in equipment resources, wherein the processing time and the cost can be different. The existing flexible comprehensive scheduling method only considers the single target of product completing time, so the production cost of the product is too high under multiple conditions. The flexible comprehensive scheduling method comprises the following steps of adopting a layering double-target optimizing model; only considering the time in a first layer, namely shortening of the product completing time; only considering the cost in a second layer, namely reducing of total processing cost; adopting a cask concept of a wood bucket principle to realize the target of the second layer of the flexible comprehensive scheduling. The layering double-target optimizing model comprises the following steps of initializing the attribute data of all procedure nodes in the comprehensive scheduling task, and sequencing the procedure nodes; distributing the equipment at the procedure nodes. The flexible comprehensive scheduling method is used for the flexible comprehensive scheduling considering the double targets of cost and time.

The invention discloses a capacitance voltage balancing strategy suitable for a high level modular multilevel converter submodule. A technical scheme is characterized in that bridge arm submodule capacitor voltages are sorted from large to small according to prime factors and each layer grouping is performed, a voltage balance link between groups and a capacitance voltage equalization link in the groups adopt an Hill sorting algorithm so that a sorting speed of a voltage balancing link is improved; and simultaneously, a halfway grouping mixing sorting algorithm is designed to reduce a power device switch frequency. The strategy is applied in a modular multilevel voltage source converter. Under the condition that there are a large quantity of bridge arm submodules, functions of increasing a submodule capacitance voltage sorting speed and reducing system simulation time are possessed; a complexity of the sorting algorithm is effectively reduced; the switch frequency of thesubmodule power device is decreased and system switch losses are reduced.

The invention discloses a battery management system and method. The battery management system comprises three layers of management structures, namely a BAMS, an MBMS and a BMU, wherein the BAMS battery pile management system is used for collecting subordinate MBMS information, and is connected with a PCS energy storage converter through a communication protocol; the MBMS monitoring management system is used as a middle management layer of the battery management system, and one end of the MBMS monitoring management system is connected with the BAMS battery pile management system through a CAN bus, and the other end is used for managing a plurality of BMU battery management units through the CAN bus; and the BMU battery management unit is used for monitoring the voltage and temperature of the single batteries, and is a minimum composition unit. According to the voltage balancing method of the battery management system, the battery voltages in the battery pack tend to be consistent through the discharging balancing method. According to the system and the method, the adopted voltage balance strategy is simple in configuration, so that the fault rate is relatively low and the maintenance is convenient.

The invention provides a layering sectional type capacitive balance control method based on a look-up table method. A layering sectional type acquisition and processing structure solves the problem that ability of valve base control equipment to collect large-scale submodule information is limited. An integral voltage balance strategy of a bridge arm undertakes a balance effect in macroscopic significance. Balance control of each sectional submodule is enabled to be effective and reasonable. According to preset switching conditions, the look-up table method beforehand makes switching decision rules into list forms, and classifies all running conditions as decision output of different conditions, and task processing is enabled to be simple and rapid. Aiming at various different control targets, different pulse assignment tables can be made through special emphasis on specific gravity, and the arrangement of additional switching points is superadded, stable running of capacitor voltage of a submodule of a bridge arm is enabled to be achieved, and meanwhile switching frequency is reduced.

The invention provides a voltage modulation method for the access of a weak current AC power grid, and the method comprises the steps: enabling an MMC valve controller to obtain an AC voltage reference value uref from an electrode controller; calculating and obtaining the capacitor voltage mean values UC_avg_p and UC_avg_n of the upper and lower bridge arms of the MMC valve controller; calculatingand obtaining the number of submodules, put into use, of the upper and lower bridge arms; and determining the switching states of the submodules according to a capacitor voltage balancing strategy. The method improves the obtaining method of the number of submodules, put into use, of the upper and lower bridge arms, and enables the AC voltage outputted by the MMC to be higher in stability duringDC voltage fluctuation.

The invention provides a balancing rudder deflection angle determining method applicable to a reentry vehicle with multiple rudder surfaces. The method comprises the following steps: converting a reentry vehicle balancing rudder deflection angle calculation problem into a nonlinear equation solving problem based on a pneumatic database; establishing a nonlinear equation related to a virtual rudder deflection angle of the reentry vehicle based on a reentry vehicle control channel equivalent virtual rudder deflection angle concept; performing Taylor expansion approximation to obtain a linear equation set for balancing calculation; finally realizing calculation on the balancing rudder deflection angle of the reentry vehicle with multiple rudder surfaces by virtue of a discrete Newton iteration algorithm and a graded balancing strategy. According to the method, a rudder surface function combination method and a graded balancing strategy are applied, so that the method is applicable to the balancing calculation for multiple channels of the reentry vehicle with multiple rudder surfaces, and rudder surface cross-linked coupling influence can be considered. The method is a rapid, high-precise and universal calculation method which is applicable to balancing of multiple channels and is designed for the reentry vehicle with multiple rudder surfaces, and also can be applied to a common airplane and a conventional aircraft.

The invention belongs to the technical field of power system communication, and particularly relates to a power communication network routing method for joint balance of load traffic and service risks. In terms of load traffic balancing, the link channel occupancy condition and the node traffic admission condition are considered in the power communication network routing method provided by the invention, and then a load weight function of the link is designed; in terms of service risk balancing, modelling is performed on the availability of an optical fiber link of a power communication network, and then a balancing strategy considering the service priority and the service risks that the link has carried is designed; and finally, a balancing factor is introduced to construct a link cost function that comprehensively considers the load and the risks so as to achieve the joint balance of the carrying capacity and the service risks. According to the power communication network routing method provided by the invention, the balancing factor can be adjusted according to the importance of the service, the network blocking rate and the service risks of the network can be reduced at the same time, the link load traffic and the risk value are balanced, and finally the network reaches the optimal performance.

The invention discloses an MMC direct current voltage balance control method capable of forming a direct current converter station based on a cascaded MMC. The method comprises the steps of 1, estimating the required number N of MMC units connected in series according to the rated voltage of the system; 2, determining the rated direct current voltage of each MMC unit connected in series; 3, selecting the MMC unit with the minimum rate capacity or the MMC unit with the maximum rate capacity according to the specific value of N; and 4, determining a control strategy used by each MMC unit connected in series. According to the method provided by the invention, the blank of research in current serially connected MMC unit direct current voltage balance strategies is filled, the certain referencecan be provided for future engineering design, the universality is strong, and the method is theoretically applicable to the situation that the rate direct current voltage of each MMC unit connectedin series is the same, and is also applicable to the situation that the rate direct current voltage of each MMC unit connected in series is different from one another.

The invention discloses a task scheduling method based on micro-cloud load balancing in a wireless metropolitan area network. According to the method, how to unload tasks and dispatch the tasks to reasonable micro-clouds is considered, so that a system delay is shortened. For solving the problem of optimization of response time of the micro-cloud tasks in a wireless metropolitan area network environment, an algorithm thought of micro-cloud load balancing is introduced, and a micro-cloud load balancing strategy is provided, thereby shortening average response time of the micro-cloud tasks as much as possible under the condition of meeting multiple constraints in the network. According to the strategy, unique characteristics of the wireless metropolitan area network and micro-cloud task unloading are fully considered; and firstly, the micro-clouds of the network are classified, then expected values of input / output task quantities of the micro-clouds are calculated, and finally minimum-cost maximum flow meeting the constraints is found in a flow network so as to execute all the tasks in the network. The method has positive effects in shortening user task response time and improving user experience.

The invention discloses a vehicle-mounted passive wireless sensor network ad hoc network method and device, wherein, the method comprises the following steps: acquiring neighboring node information ofeach node in the vehicle-mounted passive wireless sensor network, and establishing a neighboring node list; obtaining the energy information of the superior node according to the neighbor node list to determine the optimal superior node, broadcasting the address of the relay node of each node; after transmitting data in the first preset time slot at each transmitting node according to the relay node address of each node, receiving data by each relay node in the second preset time slot to complete data transmission. Based on the time slot allocation strategy of TSCH synchronous time slot frequency hopping technology, this method solves the packet loss problem of long-distance multi-hop transmission, and designs the energy consumption balance strategy based on the energy prediction model, which improves the overall energy utilization.

The invention provides a method for selecting a server to take over a fault server. The method includes the steps that a server pool system is set up, wherein the server pool system comprises server pools, pooling devices and a client-side, and if a fault occurs to one of the servers in data exchange, the client-side enters a fault switching mode; at the moment, once the corresponding pooling device finds that the fault occurs to the server, the corresponding pooling device deletes the server from a server list kept by the pooling device and sends server updating information to the other pooling devices at the same time, and the other pooling devices delete the server from server lists kept by the pooling devices after receiving the information; the client-side conducts pool name analysis once on the pooling devices again and sends pool name analysis requests to the pooling devices, the pooling devices select the server IP which is considered to be the best for a user again according to the updated server lists kept by the pooling devices and according to the balancing strategy of the server pools, the result is returned to the user, and the process of selecting the server to take over the fault server is completed.

The invention discloses a multi-antenna diversity receiving method and device based on time inversion. The method comprises the following steps of: firstly, carrying out down-conversion treatment on radio frequency signals received by antennas respectively to obtain multiple unrelated compound baseband signals; next, carrying out channel evaluation respectively according to the obtained compound baseband signals to obtain independent transmission channel parameters of receiving antennas; then, working out time inversion conjugate formulae of channels according to the evaluated channel parameters; and then, carrying out convolution operation on the compound baseband signals received by the antennas and the time inversion conjugate formulae of corresponding transmission channels respectively; finally, processing the convolution result by an equal gain combination method to obtain a final multi-antenna diversity result. According to the method and device disclosed by the invention, the signal-to-noise ratio of the received signal can be increased, combination and balance strategies are simplified, and the performance of the wireless communication system is improved.

The invention discloses a method for manufacturing a viscoelastic cushion and a method for grinding a plane of a thin-sheet workpiece. The grinding method comprises the following steps that the composite viscoelastic cushion is prepared by using a shear thickening solution and a plurality of layers of thin cloth and is placed between an electromagnetic chuck and the workpiece, so that deformation of the workpiece in a grinding machining process is reduced; and in addition, an annealing process before grinding and the demagnetization and stress re-balancing strategy in the grinding process are carried out, so that the influence of internal initial stress release of the workpiece and machining stress introduction on the deformation of the workpiece is reduced, and the high-precision plane grinding machining of the surface of the thin-sheet workpiece is realized.