79results about How to "Relieve load pressure" patented technology

The invention provides an acceleration system for a deep neural network. The acceleration system for a deep neural network includes a 3D memory, a deep neural network computing unit which is connectedwith memory controllers on logic layers of vaults of the 3D memory, a router which is connected with the memory controllers, a compressor and a decompression device, wherein the memory controller ofeach vault performs data transmission through the router connected with the memory controllers through network on chip; the compressor is used for compressing the data to be compressed, which needs tobe transmitted on the network on chip and is applied to the deep neural network; and the decompression device is used for decompressing the data to be decompressed which comes from the network on chip and is applied to the deep neural network.

The invention provides a Web and micro-service unified authentication system and method. The system comprises a client, an Nginx server, a gateway server, an authorization server, a micro-service application server and an original Web system application server. The method comprises the following steps that: Nginx redirects a client micro-service login request; the client authorization request caches the client ID and the redirection address through the gateway, and redirects after forwarding the authorization service; a client login request returns a login page to submit login information, andthe login information is redirected after being authenticated by an authorization service to generate a token, and the token and a time stamp are cached to the gateway; the client micro-service homepage request forwards data returned by the micro-service background through gateway authentication; the Nginx redirects the original Web system login request of the client; the client authorization request is redirected after being authenticated by the gateway and the authorization service; and the home page request of the original Web system of the client is authenticated by the gateway and is forwarded to the Web background to return data. According to the invention, the access efficiency is improved, the gateway load pressure is relieved, and the universality is realized.

A CCHP distribute energy station cooperative optimal dispatching method includes such steps as 1, setting up plan operation model considering economic benefit according to operation characteristics ofeach part of equipment of CCHP distributed energy station; 2, synthetically considering that environmental cost and the economic operation cost, expressing the emission pollutant and the environmental carrying capacity by an economic method, and constructing an overall optimal dispatch model for the operation models of each part of the equipment of the CCHP distribute energy station in combination with the constraint conditions; Step 3: using the improved adaptive particle swarm optimization algorithm based on bee swarm optimization operator to solve the overall optimal scheduling model, andthen obtaining the optimal scheduling strategy of each part of equipment in the CCHP distributed energy station. Compared with the prior art, the invention has the advantages of accurate provision ofscheduling plan, easy use by engineers and technicians, high running speed of algorithm, high precision, strong global optimization ability and the like.

The invention relates to a multi-energy complementary energy internet optimized dispatching method. A multi-energy complementary energy internet comprises a cold / heat storage unit, a power storage unit, an electric output device, a cold output device, a heat output device and a load demand side. The method comprises the steps that under a set control strategy, minimal system operation and maintenance cost, environment-friendly cost and comprehensive benefit cost are taken as a multi-objective function, and the constraint conditions comprise power balance constraints, power generation output power limit constraint and energy storage battery operation constraint, solving is carried out through a genetic algorithm based on self-adaptive crossover variation operation, and load optimization distribution of all equipment is obtained. Compared with the prior art, the method has the advantages that the peak shaving valley filling of an external power grid is promoted, and the load pressure ofthe power grid is relieved, so that the local consumption capacity of new energy in the area is improved, and the cleaning and high efficiency is truly achieved.

The invention relates to a method for selecting super nodes and searching resources in the peer-to-peer network having node importance, belonging to the technical field of peer-to-peer network. The method is used for identifying the high-efficiency nodes for maintaining communication and ensuring service performance by calculating the node importance, and includes the following steps: (1) calculating the importance of each node in the network: (2) determining the super nodes according to the importance; and (3) executing resource searching process. The invention adopts a distributed iterative node importance calculating method, which is a localization algorithm having the advantages of high flexibility and high efficiency, eliminating the problem of single point of failure and capability of fast adapting to the network topological change of the peer-to-peer network. The method is significantly superior to prior method for selecting network super nodes according to the neighbor numbers of the node.

The invention discloses a live broadcasting microphone connection method and a microphone connection server, applied to the field of network live broadcasting. The method comprises the steps of receiving a random microphone connection request from a first anchor client where a first anchor user is located; matching a second anchor user with the first anchor user in a microphone connection candidate pool built by the microphone connection server; if the first anchor user is successfully matched with the second anchor user, building a microphone connection service between the first anchor clientand a second anchor client where the second anchor user is located, thus allowing the first anchor user and the second anchor user to enter a microphone connection state. According to the method andthe server provided by the invention, the technical problem that in the prior art, the microphone connection reliability of the anchor is not high is solved.

A load sharing device of multilink transmission distributes load sharing dispatch module of usable link channel dynamically for IP data packet by adding fault detection module used for dynamically detecting quality of each link according to report result of fault detection module in order to let said device realize load sharing of multilink transmission for effectively relaxing load pressure on each single link.

The solar energy electricity namely thermostatic water heater relates to a electricity, namely hot water heater combination use solar energy thermostatically hot water system. After the hot water in the solar energy hot water tank is preheated by the electrical boosting or automatically mixes water valve to move warm, it is again relay heated up by the electricity namely hot heat installation to the hypothesis temperature, and maintain water outlet hot water at a constant temperature. The invention use integration intellectual technology controlled by microcomputer , not only overcome the solar energy hot water installment affect by the weather, the rate of heating slow, the water temperature unstable, the operation not convenient and so on the flaw, but also solve problem that the electricity namely hot heat installation use electricity power too big. Compared with single energy hot heat installation, the sole energy hot water installment has realized the security, conveniently, energy conservation and so on, and its structure is reasonable, and has the good operational performance.

The invention discloses a novel distributed access network architecture based on a block chain. The architecture comprises RRH and RRU used as access points in a wireless radio frequency communicationsystem, a user, digital spectrum resources managed by the RRH and the RRU respectively, a miner network formed by miners for maintaining a block chain, and a novel distributed mechanism used for managing sharing and access of the wireless spectrum resources. And the RRH and the RRU equipment provide wireless spectrum resource sharing and access services for the user, and the service process is controlled by a novel distributed mechanism for managing network access. The invention provides a novel distributed access network architecture based on a block chain. The RRH or the RRU in the architecture and a user can perform safe and reliable spectrum resource transaction in an untrusted environment.

The invention discloses an MQTT double-Broker data interaction method and system suitable for an edge Internet of Things terminal, and the method is executed by a client, and comprises the steps: transmitting an MQTT connection message to an access module, receiving a connection reply message fed back by the access module, and completing the connection; sending an MQTT theme subscription message to the access module, receiving a subscription reply message fed back by the access module, and completing theme subscription; sending an MQTT theme subscription canceling message to the access module,receiving a subscription canceling reply message fed back by the access module, and completing theme subscription canceling; an MQTT heartbeat message is sent to the access module, a heartbeat replymessage fed back by the access module is received, and it is guaranteed that the client side is online to the Broker 1 and the Broker 2; sending an MQTT release type message to the access module, andreceiving a release reply message fed back by the access module; and confirming QoS setting according to message classification, sending the published dump message to the access module again, and receiving a published dump reply message fed back by the access module. The load capacity of the whole communication environment is improved.

The invention discloses a power distribution network frame fuzzy programming method taking DG uncertainty into account. The method includes the following steps of according to the uncertainty of a distributed power source, constructing a first-type distributed power source output model and a second-type distributed power source output model; according to the second-type distributed power source output model, adopting a power distribution network topology structure constraint, a node voltage upper limit, a permeability constraint and a system power balance constraint as constraint conditions tocreate a power distribution network frame programming target function which involves the distributed power source and is based on a fuzzy expected value, wherein the fuzzy expected value is obtainedby means of fuzzy power flow calculation; adopting a genetic algorithm for solution, and obtaining power distribution network frame fuzzy programming of the distributed power source. According to theuncertainty of the distributed power source, power distribution network frame fuzzy programming involving the distributed power source is achieved; with a fuzzy expected value of the annual average expense as the target, more convenience is provided for utilizing the genetic algorithm for optimization searching, and the practicability is high.

The embodiments of the invention provide a charging station group control system and method based on a cloud platform. The charging station control method includes the steps of charging station background control, flexible charging stack control and charging pile control; the charging station background control is used for carrying out charging dispatching on a charging station through a grid selection method; the flexible charging stack control is used for dynamically distributing power in the charging station; and the charging pile control is used for electric automobile charging and human-machine interaction. The charging station group control system comprises an electricity supply and distributing module, a monitoring module, a charging module and a charging station management platform, wherein the charging module is composed of a charging station background control unit, a flexible charging stack unit and a charging pile unit. According to the charging station group control systemand method based on the cloud platform, the three parts of the charging module of the charging station are controlled, charging vehicles in the charging station are dispatched, power in the chargingstation is dynamically distributed, the optimal charging group control scheme is output, and the influence of electric automobile group charging on the charging load power of a power grid is relievedto the maximum extent.

The invention relates to a service degradation method and apparatus, and a distributed task scheduling system. The method comprises the following steps: when it is determined that a distributed task scheduling system needs to execute server degradation operation, according to an amount of current user access to a service function already registered on the distributed task scheduling system and a preset threshold scope set, determining that the access amount of the service function falls into a first threshold scope; and according to the first threshold scope and a first mapping relation corresponding to the service function, determining first service degradation operation currently to be executed by the distributed task scheduling system, wherein the first mapping relation comprises corresponding relations between different threshold scopes under the service function and the service degradation operation The distributed task scheduling system in the embodiments of the invention can obtain degradation schemes with different adaptability for different access amounts, emergency conditions are handled, and the service degradation flexibility of the distributed system is improved.

Disclosed is a method for multi-objective optimization of energy Internet including a grid unit, a gas network unit and a distributed thermal power system embedded energy Internet DHCs unit. The method comprises: under the set control strategy, taking energy use, environmental impact and economic performance as multi-objective functions, the constraint conditions include power balance constraint,generation output power limit constraint and consumption balance operation constraint. The power network, gas network and the energy internet embedded in the distributed cooling / thermoelectric systemare modeled, and many objective optimization models are established for their optimal coordinated dispatching. The objective reduction method is used. The ACL algorithm is used to obtain the Pareto optimal frontier, and then the target set with the smallest number of targets is obtained step by step based on the Pareto optimal frontier. Compared with the traditional dispatching method, the invention not only promotes the peak clipping and valley filling of the external electric network, but also alleviates the load pressure of the electric network, improves the local absorption capacity of theregional new energy, and improves the energy use efficiency.

The invention discloses a method and a system for carrying out delayed allocation on storage resources of a client of a cluster file system. The method comprises the following steps: step 1, carrying out delayed allocation on the storage resources by a client according to the file layout operation condition; step 2, recovering the storage resources when a file is deleted and processing the recovery of the storage resources subjected to delayed allocation by the client; and step 3, maintaining the consistency of the storage resources between the client and a metadata server.

The invention provides an optical storage and charging integrated hierarchical coordination control system and a control strategy thereof, relates to the technical field of optical storage and charging integrated control, and solves the problem that the existing optical storage and charging integrated research cannot efficiently exert the coordination control effects of peak clipping and valley filling of optical storage and charging and promotion of source-load interaction. A battery energy management system layer, a direct-current charging and discharging pile layer, a current conversion system layer and a battery energy management system layer are constructed to serve as a master control center, and first data information of the layers, second data information of the direct-current charging and discharging pile layer and third data information of the current conversion system layer are transmitted to a cloud platform server through an intermediate communication device; a control instruction of a cloud platform server is received, and then the information is send, thereby realizing cooperative control of battery energy storage, a DC charging and discharging pile and a photovoltaic array and stable operation of photovoltaic storage and charging integration, promoting friendly interaction of source loads, realizing an idea of peak clipping and valley filling, and utilizing thevalue of photovoltaic storage and charging integration to the greatest extent.

The invention provides a mutual transmission method between user equipment and an access network system. When a management control unit receives service request messages sent by a first user equipment (UE) through a base station, the management control unit can judge whether a transmission mode of data mutual transmission can be conducted to interact service data or not between the first UE and the second UE according to the service request messages, if the data mutual transmission between the first UE and the second UE can be achieved, the management control unit can forward the service request messages to a second base station where the second UE is located, and indicate a first base station or the second base station to distribute carrying resources for the service data mutual transmission, and therefore the first UE and the second UE can conduct local data mutual transmission. Services which can be achieved by local data mutual transmission between the UEs do not need to do service data return through core network, but the business data return can only be done in access network, and therefore loading pressure of the core network can be relieved to a certain degree.

The invention provides an uplink micro base station content scheduling method based on cache and redundancy elimination. The method comprises the following steps: step 1: establishing a content-centric multiuser uplink cell network; step 2: waiting for the start of a time slot, and when the time slot starts, sending content data to a micro base station by all users; step 3: when the time slot terminates, carrying out content redundancy detection and elimination by the micro base station; step 4: judging whether scheduling and uploading are required by the micro base station; if the scheduling and uploading are required, executing step 5; if the scheduling and uploading are not required, executing step 2 again; and step 5: running a scheduling algorithm by the micro base station, sending data requiring the scheduling and uploading to a macro base station, judging whether to terminate the process, if so, terminating the process, and otherwise, executing step 2 again. By adopting the uplink micro base station content scheduling method provided by the invention, the transmission efficiency of the micro base station is improved, the load pressure of a wireless backhaul link is relieved, and the effective bandwidth is improved.

The invention belongs to the field of heterogeneous cellular networks and particularly relates to an SMDP (Semi-Markov Decision Process)-based home base station dormancy method in a heterogeneous cellular network. The method comprises the following steps: formulating a user correlation mechanism; computing energy consumption of a system; establishing an SMDP model according to the user correlationmechanism and the energy consumption of the system; and solving an optimal dormancy decision. Compared with a traditional base station dormancy method, the method provided by the invention more meetsthe condition of dynamic change of base station loads in practice, fully utilizes the resources of a home base station, relieves the load pressure of a macro base station, and also better reduces theenergy consumption of the system.

The invention relates to a queue CAS atomic operation control method, which, based on a queue slicing thought, is used for solving the problem of failure of indirectly judging CAS under the conditionof range-based atomic operation maximum because a CAS loop statement is used for judging a single pointer position of a producer in the prior art. Through the design method, queue elements are sliced,a single element position pointed by a head of the producer is no longer purely judged, and instead, slice number values pointed by the pointer of the producer before and after CAS judgment are judged, so that the efficiency of a queue algorithm is greatly improved.

The invention discloses a protection method and device for virtual machine escape, and relates to the technical field of computers. Loading pressure of an analysis engine in the high-concurrency state analyzing state can be relieved and filtered away by building a caching mechanism for analyzing tasks. According to the main technical scheme, an obtained task request is cached into an assisting thread, wherein the task request is an analysis request obtained by the analysis engine for an external operation behavior for a virtual machine, and the assisting thread is used for assisting the analysis engine in caching the task request; whether the task request is a known virtual machine escape behavior or not is judged through the assisting thread; if not, the task request is sent to the analysis engine to be analyzed. The protection method and device are mainly used for preventing the virtual escape situation.

The invention discloses a power line carrier communication regulation and control system and method combined with a switch conversion device. The system comprises a broadband carrier circuit breaker, a broadband carrier concentrator, a communication module and a control end which are connected in sequence. The method comprises the following steps: S1, judging the response qualification of a user; S2, reaching a demand response agreement with the user; S3, monitoring the power load condition by a power master station operation platform; S4, issuing a task by a handheld control terminal according to the power load and the demand response amount; and S5, realizing power load regulation and control through a broadband carrier circuit breaker. According to the technical scheme, a low-voltage user load curve and a power grid load curve are subjected to normalization processing to achieve high matching of the low-voltage user load curve and the power grid load curve, the user air switch qualification is judged according to the coincidence condition of the peak load time periods between the low-voltage user load curve and the power grid load curve, and agreement invitation is carried out on the users meeting the conditions, so that the purpose of intelligent regulation and control is achieved according to the electricity consumption condition.

The invention relates to the technical field of automobile brake equipment, and discloses a bi-directional leading shoe type brake mechanism based on a drum brake. The mechanism comprises a brake drum, wherein a sliding groove is formed in the middle of the surface of the brake drum, a friction plate is arranged on the inner wall of the brake drum, and a ball-shaped sliding groove is formed in oneside of a brake shoe. According to the bi-directional leading shoe type brake mechanism based on the drum brake, through arrangement of a brake connecting rod I, a brake connecting rod II and brake ball seats, and using a manner of axis rotation, the four brake ball seats fixedly mounted at both ends of the brake connecting rod I and the brake connecting rod II may move synchronously. Compared with the prior art, it can be guaranteed that the upper end and the lower end of the brake shoe are simultaneously opened outwards to apply pressure to the friction plate, so that load pressure to be borne by a dynamic pressure cylinder during braking is effectively relieved, meanwhile, stresses on the two ends of the brake shoe are uniform, vibration generated by the brake shoe during braking is effectively reduced, and abrasion to the friction plate by the brake shoe is further reduced, and the service life of the friction plate is prolonged.

The invention relates to a data storage and transmission method in a 64-bit multi-core server. According to the method, a new structural body storing data address binary group pointers and length is adopted, so that the size of the structural body is reduced to 8 Bytes from 16 Bytes in an existing method, and occupied space is shrunk by one half. In the same way, under the condition of a 10G NIC network card, the memory size occupied by transmitted data is reduced from original 238 Mbytes to 119 Mbytes, high memory consumption caused by data transmission is reduced to a great extent, and loadpressure caused by high memory occupation to a system is relieved; and experimental and theoretical results indicate that through the method, the data transmission and storage volume can be greatly lowered, and the method can be widely suitable for sharply increasing data transmission among 64-bit computers in the big data era.

The disclosure provides a method of credit on the social network site, wherein the social level is mapped reasonably and logically by doing statistical analysis of functions used by the user on the site and various credits obtained according to the rules of the site, and the permission of using site functions is divided strictly for the user corresponding to the level, to establish a credit system for the user on the social network site. According to the disclosure, comprehensive rule design: the investigation of user activeness can be as comprehensive and accurate as possible; reasonable level design: the user's potential “aggressiveness” (activeness) is motivated to promote the interactivity of the site and maintain the credibility of the site; strict permission design: it has an effect of “rewarding good performance and punishing poor performance” on the user and an effect of “survival of the fittest” on the content.

The embodiment of the invention discloses an article delivery method and device, electronic equipment and a computer readable medium. A specific embodiment of the method comprises the steps of obtaining a target article value transfer information set and a storage information set corresponding to the target article value transfer information set; determining an ex-warehouse capability score value of a warehouse corresponding to each piece of storage information in the storage information set, and obtaining an ex-warehouse capability score value set; based on the storage information set and the warehouse-out capability score value set, grouping the target article value transfer information in the target article value transfer information set to obtain a sub-target article value transfer information group set; and controlling a corresponding warehouse to execute an article delivery operation according to each sub-target article value transfer information group in the sub-target article value transfer information group set. According to the embodiment, the user experience can be improved, the user flow is increased, the storage resource utilization rate is increased, and the load pressure between storage spaces is balanced.

The embodiment of the present invention provides a charging station group control system and control method based on a cloud platform. The charging station control method includes three parts: charging station background control, flexible charging pile, and charging pile. The charging station background control uses a grid The selection method is used to schedule the charging of the charging station; the flexible charging pile dynamically allocates the power in the charging station; the charging pile completes the charging of electric vehicles and human-computer interaction; the group control system of the charging station includes: power supply and distribution module, monitoring module, charging module and The charging station management platform; the charging module is composed of three parts: the charging background control unit, the flexible charging stack unit, and the charging pile unit. The embodiment of the present invention provides a charging station group control system and control method based on the cloud platform, and the charging station charging module The control of the three parts of the charging station dispatches the charging vehicles at the charging station, dynamically allocates the power in the charging station, and outputs the optimal charging group control scheme to minimize the impact of electric vehicle group charging on the charging load power of the grid.

The present disclosure discloses cache-based data transmission methods and apparatuses. The method is implemented as follows. An apparatus where a caching node is located reports a caching capability to a network side, and the caching node is configured to cache data. The network side sends a cache indicating parameter to the apparatus where the caching node is located, and maintains a data list. Wherein, the cache indicating parameter is configured to control the caching node to cache the data which has the property of high repetition probability and / or high cache utilization, and the data list is a list of the data cached in the caching node. When the caching node has cached data requested by a UE, the UE obtains the requested data from the caching node.

A load sharing device of multilink transmission distributes load sharing dispatch module of usable link channel dynamically for IP data packet by adding fault detection module used for dynamically detecting quality of each link according to report result of fault detection module in order to let said device realize load sharing of multilink transmission for effectively relaxing load pressure on each single link.