114 results about "Queueing theory" patented technology

A queue includes a plurality of containers. Each container includes a lock. Clients, possibly in a multiprocessor system, can simultaneously access the queue, each client locking only a single container a needed. A manager can lock the entire queue to perform maintenance.

The invention discloses a dynamic task scheduling method under a cloud computing platform environment. Queuing processing is carried out on tasks by means of a queuing theory at first, quasi-optimal configuration of computing resources distributed by different virtual machines on various nodes in a cluster is searched out by means of an immune clone selection strategy in a manual immunity theory, reasonable configuration is carried out on the computing resources in the cluster, then antibody genes are regulated by means of load balancing, and therefore the configuration of the resources of the cluster can meet requirements for task processing. The invention discloses a dynamic task scheduling device under the cloud computing platform environment, wherein the dynamic task scheduling device comprises an affinity computing module, a judgment module, a determination module and an antibody relation module. According to the dynamic task scheduling method and device under the cloud computing platform environment, dynamic changing and virtualized environments of a cloud platform can be adapted, the optimal configuration is rapidly searched out, and the utilization rate of the resources of the cluster is improved.

The invention discloses an intelligent number calling system for queuing, which comprises a touch screen display, a printer, a controller, a server, a bar recognizer and a waiting time prediction module positioned in the controller. The touch screen display displays a business type, a waiting number of people for each business type and a time section for transferring the present number to a secondary queuing system; the time section is calculated by a time section prediction module according to a queuing theory model or a prediction model; when a customer selects the business type via the touch screen, the printer will print a number sheet writing a number, business type, a bar number, a return time section or an indication for directly entering the secondary queuing system. Therefore, the invention can be applied to a plurality of service places with longer waiting time, such as bank, post office, ticket office, etc. with a view to greatly reducing the waiting time of consumer, effectively lowering down the blank number rate and improving the working efficiency of related departments.

Disclosed is a Docker self-adaptive scheduling system with a task sensing function. A Docker self-adaptive scheduling method comprises the steps that a data collector collects the load of each container and the resource usage rate of a CPU and a memory system; based on a queuing theory and with the resource usage rate as a benchmark, a performance modeler constructs an application performance model and characterizes an association relationship of the load and response time; a response time predictor estimates parameters of the performance model during running through a Kalman filter, wherein estimation is conducted under a convergence condition of meeting a predicted and measured response error expectation; a feed-forward controller analyzes a residual mean and variance through a fuzzy controller to obtain a feed-forward adjustment value of a control parameter of the Kalman filter; a container scheduler judges whether or not a predicted value of the response time violates application service quality and conducts scheduling according to a scheduling algorithm; the container scheduler conducts container expansion or shrinkage or migration.

The invention includes steps: (1) establishing corresponding table between number of road junction and name of road section; (2) discriminating and processing error data; (3) obtaining average speed of road section. Average time of travel on road section is equal to sum of average running time and average waiting time. Running length of vehicle is equal to length of road section minus length of vehicle queue at road section under average meaning. Roads are divided into rapid road, main road, secondary road, and branch road. Detecting speed on the spot obtains anticipant speed on different grades of road. Average running time is obtained from average running length of vehicle divided by anticipant speed on different grades of road. Based on M / M / 1 theory in queuing theory, the disclosed method obtains average waiting time. Based on length of road, average running time and average waiting time, the method obtains average speed on road. Features are: easy, quick and high reliability.

A method for site selection and planning of electric vehicle swapping stations based on queuing theory, the purpose of which is to realize rational planning and energy saving of electric vehicle infrastructure in urban areas. The number of facilities, taking the power station and centralized charging station (substation) as the overall research object, constructing an optimization model aiming at minimizing the overall operating cost and transportation cost, and finally using this model to realize the location planning of the electric vehicle power station. The invention plans and formulates a site selection model that meets urban construction requirements on the basis of fully considering the construction cost of the power station, transportation energy consumption, and user electricity demand. The experimental results prove that this method can not only calculate the optimal scheme for the site selection of the substation, but also obtain the optimal substation access scheme, which has an important guiding role in the planning and construction of electric vehicle charging and swapping stations in urban areas.

A copy management strategy for data blocks in an HDFS comprises a method for selecting the starting time of copies, a method for selecting the number of the copies, a method for selecting the positions of the copies and a method for arranging newly added copies. According to the method for selecting the starting time of the copies, a starting cycle is determined first, one day is regarded as an access cycle to carry out starting of a copy strategy, and then starting moments are determined; according to the method for selecting the number of the copies, the number of the copies are calculated by adopting Poisson distribution, and then calculation is carried out through a copy number calculation method based on a queuing theory; according to the method for selecting the positions of the copies, a copy arranging strategy is adopted in the selection of bays and nodes, and the utilization conditions need to be taken into consideration; according to the method for arranging the newly added copies, the positions of the newly added copies are selected in the bay with the nodes accessed the most by users, and nodes with the lightest load are selected in the bay with the nodes accessed the most by the users to carry out the establishment of the copies. The copy management strategy for the data blocks in the HDFS proposes an HDFS copy-first strategy, can reduce the waiting time when the users access HDFS data and improve the throughput of the system.

The invention discloses an electric vehicle fast charging station site selection constant volume method based on a travel probability matrix, and relates to the technical field of electric vehicle public charging facility planning, and the method comprises the steps: obtaining road network parameters in a planning area; obtaining an OD travel matrix according to different types of electric vehicleroad traffic flow backstepping so as to obtain an OD travel probability matrix; Monte Carlo method, OD travel probability matrix and improvement speed. The flow model simulates a driving track of theelectric vehicle in one day to obtain a quick charge demand space-time distribution matrix in one day; the space-time distribution matrix of the fast charging demand is analyzed and optimized according to the site selection and constant volume optimization model of the charging station; the charging station position and the charger configuration number are obtained; according to the invention, the electric vehicle fast charging demand space-time distribution matrix is utilized, the Voronoi diagram is adopted to divide the service range of the charging station, the optimal position of the charging station is determined by improving the particle swarm algorithm, and the capacity of each charging station is optimized by utilizing the queuing theory, so that the planning result of the electric vehicle fast charging station is more accurate and scientific.

The invention discloses a hospital human resource quantitative configuration method based on a queuing theory model. The configuration method of the queuing theory model comprises four basic structures: an input procedure, service time, a service window and a queuing rule; and the method mainly comprises the following steps: step one, performing workload statistics and questionnaire survey, wherein the statistics index is the working hours and total medical consultations, and the questionnaire survey content is the medical staff acceptable maximum working strength and patient bearable longest waiting time; step two, measuring basic parameters by use of a work-hour measurement method: a service counter number C, an average arrival rate lambda and an average service rate mu; step three, establishing the queuing theory model, analyzing the data by use of the queuing theory analysis model; step four, outputting the following outcome indexes: the working strength rho, an average waiting number Lq, an average staying number Ls, average waiting time Wq, average remaining time Ws, an idle probability Po and a waiting probability P; comparing various data indexes, and regarding the service counter number C satisifying that the working strength rho is less than one and meets the medical staff working strength and the patient waiting time as an optimal data result. The method disclosed by the invention has the advantages that the quantitative configuration problem of the medical staff is effectively solved, and the scientific reference is provided for improving the service effect.

In the invention, an AGV simulation chart model and a mathematical model of AGV quantity configuration are established, the mathematical model is analyzed by combining a queuing theory method, an inequality with a parameter is given that the optimal quantity n should meet, then the number of n AGVs are placed on the simulation chart model, the simulation model is operated to obtain statistical data, and the statistical data are substituted into the inequality with the parameter to verify whether the inequality is true so as to determine the optimal number of AGVs. The method combines both an analytical method and a simulation method, simplifies the operation and statistical process, with the obtained number of AGVs being close to the actual situation, and can provide the best amount of AGV information suitable for the situation at the scene before the procurement and delivery of the AGVs so as to achieve the purposes of scientific decision-making and cost reduction.

The present invention discloses an electric taxi charging station planning method based on demand analysis. The method comprises the steps of: employing a cluster analysis method and a P-median modelto perform analysis of demands of taxis in a region, and determining construction distribution sites of charging stations in the region; and employing a logit model to simulate probability behaviors of the taxis selecting the charging stations, employing a queuing theory model to calculate average queuing waiting time of each charging station, and determining one construction scale of each charging station according to the average queuing waiting time of each charging station. The charging station planning method based on the demands of electric taxis employs the P-median model to perform optimization addressing of the distribution sites of the charging stations and employs an M / G / c model, configured to more accurately reflect charging features of the taxis in the stations, in the queuingtheory to solve the charging station capacity configuration problem so as to determine a charging station optimization planning scheme, and therefore the method provided by the invention is effectiveand practical.

The invention discloses a determining method for banning changing the length of a lane line in an intersection entrance, which belongs to the field of road traffic marking in traffic engineering. The determining method comprises the steps: firstly the traffic data is acquired through a manual counting or photographic method, and the acquired total traffic volume is converted into equivalent standard car traffic volume; then, by adopting an M / M / N model in the queuing theory, the number of the queuing cars behind the parking line of the intersection entrance is calculated according to a multiple-channel queue system; and finally, the number of the cars is converted into car average queuing length at the unit of meter, and then added with the minimum safety distance among the queuing cars, so that the length of the banned changing lane line can be obtained. The method can avoid traffic jam at the intersection, increases the running efficiency of the intersection, reduces the failure rate caused at the intersection entrance, and saves resources.

The invention provides a long-distance travelling charging planning method for an electric vehicle in intelligent traffic. The long-distance travelling charging planning method comprises a path seeking and charging scheme. During path seeking, a starting point, a destination point and a charging station are viewed as all passing points in a path, two-segment seeking is employed, all probable pathsare proposed, a scheme of reaching a destination place is given out, and after the path is given out, the charging time and the travelling time of each path are solved out. In the charging scheme, the numbers of vehicles in waiting of each charging station is estimated according to a queuing theory, the queuing time and the additional queuing time when the electric vehicle reaches the charging station are estimated, the queuing time that the charging station can be reached is estimated and is combined with the time in the path seeking to obtain the charging time and the traveling time, and areasonable line is planned by a coordination curve method. Optimal path selection can be provided according to a demand without charging, and a path selection scheme and a charging planning scheme also can be provided according to a charging demand.

The invention discloses a queuing theory-based data transmission bandwidth prediction method in the intelligent power distribution and consumption business, aiming at the problem that a great deviation exists between a predicted value of a bandwidth and an actual demand value easily caused by the traditional bandwidth prediction method. The method comprises the following steps of firstly acquiring intelligent power distribution and consumption business QoS (quality of service) requirement parameters from an application layer as basic data for bandwidth prediction, and converting the QoS parameters of an upper layer into queuing theory model parameters through the queuing theory; secondly determining a queuing theory-based transmission bandwidth prediction model according to the QoS constraint condition of the electric power telecommunication business, and meanwhile converting the prediction model into a linear constraint nonlinear programming solving model to obtain the optimal solution; finally obtaining a smallest predicted transmission bandwidth capable of ensuring the QoS requirement of the system business. According to the method, the predicted band width is obtained by modeling and solving through intelligent power distribution and consumption business QoS parameter mapping and the queuing theory, so that the bandwidth prediction deviation is reduced, and the system bandwidth utilization rate is optimized.

The invention provides an asynchronous request queue model oriented to the map visualization tile service and belongs to the fields of computer network technology and queue control technology. By means of the principles and methods of the Queuing Theory, the asynchronous request queue model mainly solves the access problem of the visualization tile service of networked maps and achieves hierarchical, real-time and smooth map tile access of a client. The asynchronous request queue model oriented to the map visualization tile service is implemented through the following steps including, firstly, setting a request queue with a fixed capacity; when a new request comes, determining whether to put the new request into the queue or discard the request according to the state and the priority of the queue; besides, establishing threads identical to concurrent requests in quantity, performing a tile obtaining process and returning results in a call-back mode. The asynchronous request queue model oriented to the map visualization tile service relates to the technologies such as priority control, concurrent request quantity control, network asynchronous request and request queue management under high-frequency visualization requests, thereby being a high-frequency and high-timeliness network asynchronous request queue model with the priority control and being applicable to access of the map visualization tile network service.

The present invention discloses a queuing theory-based continuous traffic node congestion degree prediction model, a system and a method. The model comprises a plurality of continuous traffic nodes. The length distribution of queues is adopted as an index for evaluating the traffic node congestion condition. The queue length distribution of each traffic node is constructed based on the queuing theory and the flow of the self-similarity. According to the invention, the method is directed to the prediction of the continuous traffic node congestion degree. Based on the method, firstly, the traffic congestion degree of each single node is predicted based on the flow that reaches the above single traffic node. Secondly, the traffic congestion degrees of multiple continuous traffic nodes connected with a current traffic node are predicted according to the flow of the current traffic node. Therefore, the traffic system is reasonably adjusted by related government departments according to the congestion degree information of the system.

The invention discloses a scheduling method based on a single employee competent to multiple jobs. According to the method, a high-precision business volume prediction result is obtained through a machine learning model, a business volume prediction model is modified with significant factors taken into account, and the model precision is improved. The multiple jobs of a task are decomposed based on a queuing theory, an optimized genetic algorithm is used to satisfy the scheduling constraints of a single employee with multiple jobs and company labor regulations, and thus, a humanized schedulingresult is obtained. Based on the transformation of the human distribution pattern, the managers are liberated from the heavy scheduling work, the scheduling result is more fair and just, and the employees' enthusiasm for work is improved. Moreover, an intelligent human distribution system improves the service perception and satisfaction of customers.

The invention discloses a data center-oriented service function chain optimization arrangement method, which is characterized in that resources required by virtual network functions are regarded as variable parameters, the relationship among a request arrival rate, computing resources and processing delay is analyzed based on a queuing theory and elastic resource allocation, and a two-stage heuristic arrangement strategy is provided by taking energy consumption and network delay as optimization indexes. In the first stage, a greedy deployment strategy is utilized to obtain a series of mapping relationships between VNFs and servers; in the second stage, on the basis of the solution set in the previous stage, the nonlinear constraint optimization problem is further solved, and the optimal configuration of the computing resources is obtained. The invention also discloses a system for implementing the data center-oriented service function chain optimization arrangement method. According to the method, the server energy consumption, the link bandwidth occupancy rate and the service delay are effectively reduced, and the deployment success rate is improved. According to the method, the resource optimization arrangement of the SFC is completed under the condition that the network topology and the SFC data are given, and the arrangement result can be obtained within the controllable operation time.

The invention discloses a multilayer service-oriented container cloud resource scheduling optimization method, which comprises the following steps of: constructing a container cloud architecture consisting of a plurality of heterogeneous containers according to the demand of providing services for a plurality of multilayer services; According to the architecture, establishing an overall revenue function of the cloud service provider for the container cloud; According to the response state of each layer of service under the multilayer service provided by the queuing theory, obtaining an overallincome model in combination with an income function, and setting an initial solution for the income model according to the service requirements of each layer of service; According to the optimizationproblem of the model, a resource distribution optimization algorithm is provided, and resource distribution is carried out according to the solution of the algorithm, so that benefit maximization ofa cloud service provider is achieved. According to the method, a container cloud optimization strategy can be provided for a cloud service provider, an elastic operation strategy of multiple layers ofrelated micro-services in the container cloud is considered, resources can be dynamically distributed according to the requirements of a user, and the utilization rate of cloud resources is increased.

An embodiment of the invention may provide a method for predicting buffer size in a system, where multiple encoded bit streams may be multiplexed into a single bit stream. The multiple encoded bit streams may comprise either a constant or a variable rate. Notwithstanding, predicting the buffer size required for encoding, multiplexing and / or decoding prevents overflow and underflow of the buffer. At the demultiplexer, each decoder may have a buffer and there may not be any buffering prior to multiplexing. A plurality of MPEG encoded data bit streams may be multiplexed into a single multiplexed MPEG encoded data bit stream. A size of buffers utilized to buffer the multiplexed MPEG encoded data bit stream may be determined based on a queuing model.

A calculation method for evaluating response time in a computer system in which a plurality of units of execution are run on individual processor cores. The calculation method evaluates response time in a computer system in which a plurality of units of execution are run in parallel on individual processor cores while sharing internal hardware resources. The method includes the steps of: normalizing the utilization state of the computer system with the state of parallelism of the units of execution and evaluating the utilization state; and calculating an expected value of response time in a desired state of parallelism of the units of execution from the normalized utilization state on the basis of queueing theory in which changes in average single-thread performance of the units of execution according to the parallel running state are taken into account.

The invention provides a communication traffic prediction method for the distribution line information monitoring service. Aiming at that the existing traffic prediction method gives no consideration to the dynamic arrival characteristic of service communication data flows and causes a problem of great deviation of a traffic prediction value, the method provided by the invention gives full consideration to a dynamic arrival factor of the service data flows, and the method comprises steps: firstly, on the basis of performing traffic parameter estimation of a distribution line information monitoring service through a poisson traffic model, effectively predicting traffic of the distribution line information monitoring service according to a queuing theory model; secondly, training model parameters with original data and performing parameter optimization; and finally, performing communication traffic prediction through a poisson queuing theory traffic model. The method provided by the invention overcomes a problem that the conventional method cannot describe influence of a poisson traffic dynamic influence factor of the service on a prediction result, improves prediction accuracy, and facilitates implementation of reasonable planning and allocation of bandwidth resource of an intelligent distribution and utilization communication network.

The invention discloses a small base station dynamic sleep method based on multi-cell cooperation. The method includes a sleep mechanism, a sleep time self-adaptive algorithm, queuing theory analysis and a sleep cell selecting algorithm. The sleep mechanism comprises three states: an active state, a listen state and a sleep state. After a base station finishes a presently arrived UE session, partial modules are automatically turned off, a listen timer T0 is started, and the listen state is started; a small eNB sends a sleep request (S_request) to a macro cell after the timing is over; a macro eNB sends feedback acknowledge information to a small cell during the timing period; the small cell switches covered UE to the macro eNB; and the UE after the switching is connected with the macro cell and sends UE reports regularly.

The invention discloses a batch streaming computing system performance guarantee method based on queue modeling. The performance guarantee method comprises five steps of dividing a process; selecting components; carrying out performance modeling; computing delay; and positioning and optimizing a bottleneck. According to the method, for the feature that load intensities have clear fluctuation in an operation process of a batch streaming computing system, key components in the batch streaming computing system are extracted, a performance model of the system is established according to a queuing theory and mathematical analysis is carried out on the model; data processing delay of the system under different load intensities is computed through utilization of the model in the operation process of the system; and when the data processing delay cannot satisfy a data processing timeliness demand, a performance bottleneck component is positioned according to the queuing theory, and a configuration optimization suggestion is provided.

The invention relates to a profit based request access control method in a distributed multi-cloud data center. According to the profit based request access control method, requests can be accepted selectively, requests with high priorities are prone to be accepted through comprehensive consideration of factors such as the priority, profit and expected response time of each request, and the total bandwidth requirement of all the accepted requests is guaranteed to be not more than the total bandwidth of all the usable network service providers. An M / M / m queuing system model aiming at the distributed multi-cloud data center is built based on a queuing theory according to user requests as well as network and server resources. An effect function capable of transforming the response time of each application request into corresponding profit is designed based on the model. A constraint nonlinear planning model of request access control in the distributed multi-cloud data center is built. A profit based request access control strategy is provided by adopting a penalty function and a mixed heuristic optimization algorithm. According to the profit based request access control method in the distributed multi-cloud data center, requests of various applications can be accepted selectively, and the profit of the distributed multi-cloud data center providers is accordingly maximized.

The invention provides a pre-queueing method and system based on user position information; return time of a user is calculated by acquiring current position information of a leaving user, a reminder message is sent in time to a customer according to the rest queuing time and return time of the leaving user, and the user is informed timely of returning to the hall to deal businesses; user experience decline and other problems in other users due to vacant numbers, repeated calling or cutting by frozen numbers recovered are avoided at the premise of not wasting users' time.

The invention relates to a dynamic mine shoveling, transporting and dumping efficiency optimization system and method, and belongs to the technical field of mining. The system comprises a GPS system of a shoveling device, a dynamic weighing system of an automobile transport device, a crusher of a rocking dumping device and a belt control system. The load queue length of the system in unit time is calculated by using a queuing theory model, a mathematical optimization method is used to calculate the optimal ratio of the automobile to the shoveling devices, high efficiency of a belt is kept in unit time, the automobiles and the shoveling devices are scheduled automatically, the mine feeding amount of the belt is controlled, the production efficiency is improved, loss is reduced, great convenience is provided for production command, and the practicality is very high.

The invention provides an unbalance intersection signal phase optimal design method based on a queuing theory. Related traffic data is collected through a traffic flow detector; based on the collectedtraffic data, each entrance lane of the unbalance intersection is used as the multi-path queuing multi-channel service system by adopting a M / M / 1 model in the queuing theory, the average queuing length after the parking line of the entrance lane of the intersection is computed, based on the average queuing length of each entrance lane, the intersection signal phase optimization is performed. Through the method provided by the invention, the entrance lane unbalance problem at the intersection can be effectively solved, and the passing efficiency of the intersection is greatly improved.

A connection admission control mechanism for packet, circuit, or hybrid packet and circuit networks whereby the signaling message (i.e., "query") delay in a control channel of a network switch is maintained (i.e., "locked") in a timed control interval at a desired level using feedback parameters relating to an estimated current amount of delay in the packet network switch during one or more previous timed control intervals, to guarantee or closely achieve a desired signaling message or query delay performance. The processing delay of the network is bound or "locked" at a desired delay performance level based on substantially real-time estimation or measurement of the current signaling message delays of queries or call control packet streams during one or more previous control intervals. Thus, connection processing is robust against variations in traffic intensity and / or processing capabilities of the switch, and a single QoS measure or specification can be applied to the network switch. Because of the resemblance to a phase locked loop (PLL) in hardware timing acquisition techniques, the technique is referred to herein as Delay-Locked Admission Control (DLAC). The control intervals may be adjusted based on the particular application. During each control interval, a predetermined maximum number of new queries are allowed into the processing fabric of the switch. In one embodiment, the maximum number of allowed during any one control interval is initialized at start-up to a desired value, e.g., to 1, to a randomly chosen integer, or chosen according to queuing theory estimates for the particular application and / or based on current conditions.