51 results about "Sequential monte carlo methods" patented technology

A stochastic method and system for detecting polygon structures in images, by detecting a set of best matching corners of predetermined acuteness α of a polygon model from a set of similarity scores based on GDM features of corners, and tracking polygon boundaries as particle tracks using a sequential Monte Carlo approach. The tracking involves initializing polygon boundary tracking by selecting pairs of corners from the set of best matching corners to define a first side of a corresponding polygon boundary; tracking all intermediate sides of the polygon boundaries using a particle filter, and terminating polygon boundary tracking by determining the last side of the tracked polygon boundaries to close the polygon boundaries. The particle tracks are then blended to determine polygon matches, which may be made available, such as to a user, for ranking and inspection.

The invention discloses a wind farm risk evaluation method based on a Monte Carlo method. The wind farm risk evaluation method includes the steps: (1) building a wind speed sequence prediction model on the basis of a time sequence method by means of original data standardization, model recognition and parameter determination; and (2) evaluating the reliability of a large power grid system concerning a wind farm on the basis of the sequential Monte Carlo method by means of random sampling and reliability evaluation. Randomness of the wind speed is adequately considered on the basis of a large quality of wind power databases, and the method is high in speed by the aid of the Monte Carlo high-speed algorithm.

Systems and methods for distribution-transition estimation of key performance indicator (KPI). Exemplary embodiments include a method including establishing a state-space model in a computer memory, the model for handling measurement errors and modelling errors as noise, wherein time-varying indices have historical data and a correlation with KPI, adaptively estimating parameters related to the time varying indices, and obtaining residuals from the adaptive estimate of the parameters, estimating a probability distribution in order to generate random numbers from the probability distribution, generating random numbers related to the estimated probability distribution, thereby enabling prediction of future indicator distributions through a Sequential Monte Carlo method, providing models to input KPI intensity levels for a scenario level and to predict future values for a plurality of scenarios as point estimates and distributing and synthesizing the point estimation results for the plurality of scenarios based on KPI estimated probability distributions.

The invention discloses a reliability evaluation method of a distribution network cyber-physical system considering the whole process of fault processing. The reliability evaluation method of the distribution network cyber-physical system considering the whole process of fault processing comprises the following steps: initializing the distribution network cyber-physical system; confirming the current state of the distribution network cyber-physical system based on the sequential Monte Carlo method; obtaining the reliability of the three stages of fault location, fault isolation and power supply recovery by means of the fault tree model in view of the state of the distribution network cyber-physical system after the fault; classifying the load points according to the fault processing whole process event tree model to determine the impact degree of the fault processing on each load point; and calculating to obtain the system reliability indexes finally. The invention reflects the effect of the information failure on the state of the system after the fault from the overall point of view of fault processing, and takes full consideration of the time sequence process of the fault processing. According to the reliability evaluation method of the distribution network cyber-physical system considering the whole process of fault processing, the function of the information system in the analysis process of the fault consequences is considered, therefore, the invention helps to identify the integrated energy and the weak links of the communication infrastructure, assists the integrated operation of the distribution network cyber-physical system and the planning of the effective decision-making.

For the reduction of the multipath error of received GNSS navigation signals, a sequential Bayesian estimation is used, with a movement model underlying this estimation, which model is particularly designed for dynamic channel situations. Sequential Monte Carlo methods are used to calculate the posterior probability density functions of the signal parameters. To facilitate an efficient integration in received signal tracking loops, the invention builds on complexity reduction concepts that have previously been used in maximum likelihood (ML) estimators.Applicable with GNSS satellite navigation receivers, e.g. GPS and Galileo.

A stochastic method and system for detecting polygon structures in images, by detecting a set of best matching corners of predetermined acuteness α of a polygon model from a set of similarity scores based on GDM features of corners, and tracking polygon boundaries as particle tracks using a sequential Monte Carlo approach. The tracking involves initializing polygon boundary tracking by selecting pairs of corners from the set of best matching corners to define a first side of a corresponding polygon boundary; tracking all intermediate sides of the polygon boundaries using a particle filter, and terminating polygon boundary tracking by determining the last side of the tracked polygon boundaries to close the polygon boundaries. The particle tracks are then blended to determine polygon matches, which may be made available, such as to a user, for ranking and inspection.

The invention relates to an automatic prewarning method for strong storms. Doppler weather radar data is used. According to identification on storms, false combination of storms can be successfully identified by adopting a method based on mathematical morphology, namely expansion operation and corrosion operation are comprehensively used during multi-threshold identification, and relatively near storms can be separated from a storm cluster; then, according to tracking of the storms, the tracking process is greatly simplified by adopting a sequential monte carlo method, and situations of splitting, combination and leakage detection of the storms can be handled; and finally, the storms are prewarned, namely linear fitting extrapolation prewarning is realized by being combined with a motion vector field obtained by an optical flow method. According to the method, complicated situations that radar echo is densely distributed and frequently split and combined can be handled. The method has the advantages that a prewarning software system is arranged in provincial and municipal weather stations; high-convection weather prewarning service is provided; and the urgent demands for preventing and reducing disasters in China are met.

The invention proposes a wind-power-plant reliability modeling method. The method takes correlation of wind speeds between wind power plants and fault rates of wind turbine generators into consideration at the same time and through generation of related wind speeds in a simulating manner, effects of the correlation of the wind speeds between the wind power plants are taken into consideration and through binomial distribution and Monte Carlo simulation, effects of the fault rates of the wind turbine generators are taken into consideration. The reckoning-in method of the fault rates of the wind turbine generators is simple and easy to be realized through a program. A method based on linear division is used to calculate the lasting time of each equivalent state of the wind power plants so that precision of an output-power multi-state probability model of the wind power plants is improved. The wind-power-plant reliability modeling method is not only applicable to reliability modeling of wind power plants formed by combination of wind turbine generators of the same kind, but also applicable to reliability modeling of wind power plants which are formed through combination of wind turbine generators of different kinds so that reference is provided to reliability assessment of a wind power connected-grid power system through use of an analytical method and a non-sequential Monte Carlo method.

The invention provides a reliability evaluation method and system for a comprehensive energy system. The method comprises the steps: obtaining the data of the comprehensive energy system, building a structural model of the comprehensive energy system, and building an operation energy flow balance model of a power supply subsystem and a heat supply subsystem according to the structural model of thecomprehensive energy system; establishing a load reduction optimization model by taking the minimum value of the equivalent load reduction quantity as an optimization target according to the energy prices of the electrical load, the thermal load and the cold load and the relative importance degree of the electrical load; and based on the operation energy flow balance model and the load reductionmodel, considering network parameters, air temperature variables and source load variables of the comprehensive energy system, performing load flow calculation on the comprehensive energy system, anddetermining a reliability index of the comprehensive energy system by utilizing a load flow calculation result and combining a sequential Monte Carlo method. According to the method, the differences in network topology, transmission delay, terminal thermal inertia and user reliability requirements are considered, and the reliability evaluation precision of the comprehensive energy system is improved.

The invention discloses a meteorological-factor-considered method for evaluating the photovoltaic-power-station grid-connected risk. The method includes the steps that a clustering feature vector is formed by meteorological indexes, photovoltaic historical output data is subjected to clustering analysis, and a photovoltaic-power-station output power probability model of the meteorological factorsis established; then according to an element outage model, transfer sequences of a photovoltaic power station and a common electric generator in the running, derating and outage states are extracted;an outage-model-contained photovoltaic output power sequence, an outage-model-contained common electric generator sequence and a load sequence are superposed according to time sequences, a usable margin model of a system is obtained, loads of a negative margin need to be suitably cut down, and light of a positive margin needs to be suitably discarded; the load loss risk and a power-supply loss risk index are calculated with the sequential monte carlo method. The method aims at considering the influence of strong correlation meteorological factors such as the irradiance and the temperature on the photovoltaic output power probability model, and the risk brought when the photovoltaic power station is connected into a power feeding net is thus analyzed.

The invention provides a power distribution network information physical system reliability analysis method under a power distribution automation condition. The method comprises the following steps: S1, inputting power distribution network physical system element parameters and a topological connection relationship, and inputting power distribution network information system element parameters anda node adjacency matrix; S2, extracting the states of a current physical system and an information system by using a sequential Monte Carlo method; S3, calculating the fault isolation time of a current fault; S4, analyzing the influence of the current fault on load points; S5, updating the simulation time, judging whether a simulation ending condition is met or not, if so, executing the step S6,and otherwise, executing the step S2; S6, counting reliability indexes of each load point; and S7, calculating a reliability index of the whole power distribution network information physical system.The method can evaluate the influence of the distribution automation condition on the reliability of the power distribution network information physical system more accurately and more meticulously,and is more suitable for reliability evaluation and planning of a stock power distribution network and an intelligent power distribution network.

The invention provides a method for evaluating operational risk of wind power farms in combination with weather and wind speed, comprising the following steps of: a, determining outage probability of equipment components under the current weather within an evaluation cycle; b, constructing a time sequence through an ARMA model and performing matrix transformation, thereby obtaining a wind speed sequence of each wind power farm; c, obtaining a system state according to the outage probability of the current weather, and performing sampling through a non-sequential Monte Carlo method; d, adding 1 to number of sampling, obtaining wind speed of each wind power farm at the current time, and calculating output of each wind power farm; e, according to output of each wind power farm, screening out a load bus meeting a preset condition, and corresponding load shedding value of the load bus, accumulating the load shedding value, and taking the accumulated value as minimum loss load quantity of the system; f, according to the minimum loss load quantity, calculating a risk indicator; g, when risk indicator variance is less than precision each time, returning back to the step d; and otherwise, outputting the risk indicator. By implementation of the method and the system provided by the invention, accuracy and reliability of evaluation of operational risk of the wind power farms can be improved, and evaluation result can be more scientific and rational, and has extensibility.

The invention discloses a power network fault early warning method based on a short-term random failure model of equipment, which comprises the following steps: 1, establishing a dynamic load calculation model of equipment evolving in time and space; 2, establishing a failure probability model of electric power equipment; 3, an accurate forecasting method of power outage being proposed for extremeweather conditions. In the manner, the invention relates to a power network fault early warning method based on a short-term stochastic failure model of equipment, based on the non-sequential Monte Carlo method, an accurate prediction method of distribution network fault and user outage is established, which realizes fault prediction of distribution network equipment and user outage prediction, and helps distribution network operation management develop from passive rush repair management to active preventive management, thus shortening rush repair time of distribution network equipment and reducing user outage probability and duration.

An active power distribution network information physical system reliability evaluation method considering information failure comprises the steps: constructing a typical framework of an active powerdistribution network information physical system CPS, wherein a physical system comprises primary equipment and a distributed energy source DG, and an information system is divided into an applicationlayer, a communication layer and an interface layer; establishing a DG and load model, an information physical element model and an information transmission reliability model; in combination with theindirect effect of information failure on the power distribution self-healing process, quantitatively analyzing the influence of failure of an information system application layer, a communication layer and an interface layer on fault isolation, positioning and power supply recovery in an uncertain environment; employing a sequential Monte Carlo method and a non-sequential Monte Carlo method to sample a physical element and an information element respectively; by means of example simulation, carrying out the sensitivity analysis from the aspects of distributed power supply capacity, island operation mode, information element and information transmission abnormity and access network structure, wherein the accuracy and effectiveness of the method are verified through the obtained result, and suggestions and guidance are provided for construction of the active power distribution network.

The invention discloses a method for calculating reliability of a current collection system of a wind power plant. The method comprises the following steps: S1) establishing an air speed and fan capacity model; S2) establishing a reliability model of an equivalent wind turbine generator and the current collection system; S3) calculating an access set and an isolation set of the equivalent wind turbine generator; S4) defining the reliability index of the wind power plant of the current collection system; and S5) adopting a sequential monte carlo method for evaluating the reliability of the wind power plant of the current collection system. According to the method provided by the invention, the accuracy of the reliability of the current collection system of the wind power plant can be increased, the reliability of the stored energy-containing wind power plant can be calculated and a reference is provided for the selection of a topological structure of the current collection system.

The invention discloses an interconnected power distribution system containing a flexible multi-state switch and a reliable operation evaluation method, and belongs to the field of intelligent power distribution systems. The interconnection power distribution system containing the flexible multi-state switch is mainly composed of the flexible multi-state switch FMSS composed of fully-controlled power electronic devices, an MMC is used as a commutation port to form a multi-port FMSS device, and the multi-port FMSS device comprises a central regulation and control system and three ports of a port MMC1, a port MMC2 and a port MMC3. The system totally has four sub-modules, and the sub-module MMC adopts the FBSM and HBSM series-parallel configuration to form a topological structure of the sub-module MMC, so that the electric energy quality of the multi-terminal flexible interconnection power distribution system is improved, and the loss of a converter is reduced. The reliability evaluationis performed on the interconnected power distribution system by using a sequential Monte Carlo method, and the quantitative analysis is performed on the positive and negative benefits of the access of the flexible switch device, so that the decision support is provided for the application of the multi-terminal flexible switch in a power distribution network.

The invention discloses a Monte Carlo-based equivalence reliability assessing method. The method includes the steps of inputting whole network basic data, establishing an equivalence network according to a static equivalence method taking sensitivity consistency and element type comprehensiveness into consideration, and calculating all equivalence network parameters. The element operation and load fluctuation magnitude are simulated by using a non-sequential Monte Carlo method to obtain random state of an internal network system. Whether network topology is in a stepout state is analyzed, and whether branch trend or node voltage exceeds limits can be determined through trend calculation. The equivalence reliability of the system is comprehensively evaluated, and compared with the result obtained based on routine hang-up equivalence theory. When used for external network equivalence, the equivalence method ensures comprehensiveness of element types, and takes trend and sensitivity consistency into consideration. The calculation speed of reliability assessment is effectively improved through the method in combination with the non-sequential Monte Carlo method while ensuring equivalence precision.

The invention provides an AC / DC power distribution network reliability evaluation method based on a non-sequential Monte Carlo method. The method comprises the following steps: 1, executing accordingto a selected AC / DC hybrid power distribution network; 2, according to the structure and parameters of the AC-DC hybrid power distribution network provided in the step 1, sampling the element state inthe system through thenon-sequential Monte Carlo method, sampling renewable energy output and a load value according to the renewable energy and load output probability model, and finally obtaining the load value and the renewable energy output; 3, judging whether an element fault exists or not according to the load value and the renewable energy output provided in the step 2, and obtaining an element fault state. The method aims at solving the reliability evaluation problem of the AC / DC hybrid power distribution network, and adapts to the volatility of renewable energy sources and loads, a comprehensive time sequence sampling method, converter control mode switching and an optimal load shedding method.

The invention discloses a GNSS positioning multipath mitigation method based on sparse estimation, belongs to the technical field of navigation and positioning, and aims at estimating the deviations by utilizing a sparse estimation theory and remarkably reducing the calculation complexity at the same time. The method of the present invention is characterized by modeling the influence of MP signalson the GNSS measurements as the additive deviations, and then estimating these deviations and subtracting these deviations from the GNSS measurement to mitigate MP effects. Compared with a more approximate particle filter sequential Monte Carlo method, the method has the advantages that the calculated amount is obviously reduced, the deviation is estimated by using a sparse estimation theory, andmeanwhile, the calculation complexity is obviously reduced.

The invention discloses a subway traction power supply system reliability evaluation method, device and equipment and a medium. The method comprises the steps of building a fault tree model of a subway traction power supply system; calculating distribution parameters of subway traction power supply equipment faults; based on the fault tree model and the distribution parameters, simulating a traction power supply equipment and system state transition process by adopting a sequential Monte Carlo method, and calculating to obtain fault interval time, repair time, repair times and fault times; andcalculating the reliability index of the subway traction power supply system according to the fault interval time, the repair time, the repair frequency and the fault frequency. According to the invention, the fault tree model is established, the sequential Monte Carlo method is adopted to simulate the operation states of the equipment obeying different fault distribution types, the applicabilityis high, the reliability evaluation time indexes of the system fault average repair time and the system fault average interval time can be calculated, and the obtained reliability information is richer and more applicable.

The present invention discloses a power system reliability assessment method considering optimized scheduling of cascade hydropower stations, by which the reliability of the power system is improved, comprising following steps: establishing power system optimization models; inputting the sequential wind speed, runoff and load data within 24h; in accordance with reliability parameters of units and by the sequential Monte Carlo method, sampling the state durations of three kinds of units and segmenting the state durations on a 24-hour cycle; calculating the sequential output of wind farms within 24h according to wind speed data and the state of wind power units; calculating the hourly loss-of-load, according to the state of wind power units and thermal power units as well as the optimization models; optimizing for 365 days according to the method described in the step 4, and calculating reliability indices; and determining convergence or not, and if not, continuously repeating the steps S3 to S7 until convergence.

The invention relates to a comprehensive energy system energy supply reliability evaluation method and a computer system. The method comprises the following steps: acquiring original parameters and atopological structure of a comprehensive energy system to be evaluated; obtaining a simulation model of the to-be-evaluated comprehensive energy system based on a sequential Monte Carlo method; performing topology analysis on the system state of the simulation model by adopting a node marking method to obtain a plurality of connected domains; and respectively carrying out state analysis on each connected domain, and carrying out reliability evaluation on the to-be-evaluated comprehensive energy system by adopting an energy supply reliability index system based on an energy use purpose and a user standpoint. Compared with the prior art, the method can consider energy consumption purposes and user standpoints of different types of users, has the advantages of high evaluation accuracy and thelike, and ensures safe and efficient operation of the comprehensive energy system.

The invention discloses a wind farm risk evaluation method based on a Monte Carlo method. The wind farm risk evaluation method includes the steps: (1) building a wind speed sequence prediction model on the basis of a time sequence method by means of original data standardization, model recognition and parameter determination; and (2) evaluating the reliability of a large power grid system concerning a wind farm on the basis of the sequential Monte Carlo method by means of random sampling and reliability evaluation. Randomness of the wind speed is adequately considered on the basis of a large quality of wind power databases, and the method is high in speed by the aid of the Monte Carlo high-speed algorithm.

The invention relates to a Monte Carlo method-based demand side resource interconnection reliability assessment system and method, sampling is carried out by using a non-sequential Monte Carlo method,and indexes closely related to demand side interconnection are selected to assess the reliability of a power system under demand side resource interconnection. The method comprises the following specific steps: sampling the output of demand side resources, the state of an element and the state of a line in a system; calculating system power flow meeting constraint conditions by using a Newton-Raphson method;; analyzing whether the constraint condition is met or not; and finally calculating each reliability index of the system, and comprehensively evaluating the reliability of the system. According to the method, the reliability of the power system under demand side resource interconnection can be evaluated quickly and accurately, the reliability evaluation method of the power system is improved in a targeted manner, and a basis is provided for risk control and state maintenance of the power system under demand side interconnection under different degrees of influence.

The invention discloses a power system reliability evaluation method, device and system considering an LR attack. The method comprises the following steps: initializing parameters; calculating the utilization rate of the power transmission line, and identifying a fragile line in the system; obtaining a load reduction model considering LR attack and fragile line defense, wherein the fragile line power flow constraint in the load reduction model is related to the identified fragile line; calculating the success probability of the LR attack based on a preset random process model of the LR attack; and based on the load reduction model and the LR attack success probability, evaluating the reliability of the power system considering the LR attack by using a non-sequential Monte Carlo method. According to the method, power system reliability evaluation considering LR attacks and defensive measures thereof is realized.

The invention discloses a reliability evaluation method, system and device for a direct-current power transmission system, and the method comprises the following steps: carrying out the subsystem division of a back-to-back asynchronous networking hybrid direct-current power transmission system, and obtaining a plurality of subsystems; carrying out reliability evaluation on each subsystem by adopting a non-sequential Monte Carlo simulation method; and according to the reliability evaluation result of each subsystem, obtaining the reliability of the back-to-back asynchronous networking hybrid direct-current power transmission system based on the non-sequential Monte Carlo method. According to the invention, reliability parameters of elements of the back-to-back asynchronous networking hybriddirect-current power transmission system are obtained, the reliability index of the system can be calculated simply and accurately by dividing the subsystems of the back-to-back asynchronous networking hybrid direct-current power transmission system, performing reliability evaluation on the subsystems based on the Monte Carlo simulation method and then performing state sampling on the subsystems,and the calculation efficiency is improved.

Systems and methods for distribution-transition estimation of key performance indicator (KPI). Exemplary embodiments include a method including establishing a state-space model in a computer memory, the model for handling measurement errors and modelling errors as noise, wherein time-varying indices have historical data and a correlation with KPI, adaptively estimating parameters related to the time varying indices, and obtaining residuals from the adaptive estimate of the parameters, estimating a probability distribution in order to generate random numbers from the probability distribution, generating random numbers related to the estimated probability distribution, thereby enabling prediction of future indicator distributions through a Sequential Monte Carlo method, providing models to input KPI intensity levels for a scenario level and to predict future values for a plurality of scenarios as point estimates and distributing and synthesizing the point estimation results for the plurality of scenarios based on KPI estimated probability distributions.

The invention discloses an EHS-containing island type AC / DC hybrid micro-grid system reliability evaluation method. The method comprises the following steps: firstly, establishing system models such as an EHS model and a load model for an island type AC / DC hybrid micro-grid system connected with the EHS; secondly, performing fault component classification through a sequential Monte Carlo method; thirdly, considering the operation characteristics of the system, the coupling effect of the EHS and the system and the uncertainty of the system source network load, and proposing a power scheduling strategy; and finally, from two perspectives of a system level and an equipment level, providing a system level reliability index and an equipment level reliability index for evaluating the reliability of the system, so as to solve the problem that the reliability of theEHS-containing island type AC / DC hybrid micro-grid systemcannot be accurately evaluated at present. The system operation stability, the new energy acceptance capability and the accuracy and comprehensiveness of reliability evaluation are improved, and meanwhile, unified scheduling of the power grid to the distributed power supply is also facilitated.

The invention belongs to the field of reliability analysis of a power system containing new energy. Specifically, random production simulation of a power system containing photo-thermal-photovoltaic power generation is carried out. Random production simulation is carried out on a power system containing photo-thermal-photovoltaic power generation by adopting a pseudo-sequential Monte Carlo improved algorithm, the calculation efficiency is improved by combining the pseudo-sequential Monte Carlo improved algorithm with an analytical method, and the convergence rate is improved by adopting a technology with smaller variance, so that the calculation time consumed by the pseudo-sequential Monte Carlo method is remarkably shortened, and the reliability is improved.