89 results about "Ensemble forecasting" patented technology

The invention discloses a sales forecast and replenishment method, comprising: a sales forecast method and a replenishment method. The sales forecasting method includes: obtaining historical sales data of the commodity to be forecasted, retrieving a sales forecast model based on historical sales data training and outputting forecast sales data, the sales forecast model includes a primary smoothing model, a secondary smoothing model, and a three-time season A model, the outputting the predicted sales data includes obtaining a correction coefficient of the predicted sales data. The replenishment method includes: obtaining the daily average sales volume, replenishment lead time and safety stock days based on the analysis of historical sales data and forecast sales data; calling the target inventory calculation formula, inputting the replenishment cycle, and outputting the target inventory; Take the replenishment quantity calculation formula, input the quantity of goods in stock and the quantity of goods in transit, and output the quantity of replenishment. The invention comprehensively considers various situations in commodity sales, obtains sales forecast and replenishment data more clearly and accurately, and improves sales management efficiency.

The invention discloses a sales forecasting system and method, wherein the system comprises a classification module, a history correction module, a statistics forecasting module, a disassembling module and a period rolling module. The sales forecasting system helps enterprises to promote scientifically forecasting and finely forecasting capabilities by building a rolling standard forecasting system based on the historical data, data warehouse technology, statistical technology, classification forecasting technology, disassembling technology and period rolling technology.

The invention provides a ship structure vibration and noise forecasting system based on S-P-R, comprising a database module, a modeling and calculating module, and a result processing module. The database module stores the vibration intensity data of various kinds of vibration sources in each cabin of a ship, and the noise intensity data of various kinds of noise sources; the modeling and calculating module comprises a ship modeling submodule, a key excitation source recognition submodule, a transmission path determination submodule, and a receiving station energy calculating module, and is used for defining transmission paths, calculating the transmission loss of the vibration/noise energy generated by each excitation source by the transmission to a receiving station along a transmission path, and then calculating the vibration/ noise total energy of the receiving station. The system of the invention combines an analysis method, a numerical method and an experimental method to analyze ship structure vibration source and transmission path characteristics, calculates cabin air noise levels according to the analysis procedure of an S-P-R method, and overcomes the problem that corresponding remedial measures can not be taken according to vibration and noise measurement results until a ship is completed.

The invention discloses a short-term climate forecasting method based on empirical mode decomposition and numerical value set forecasting. The invention adopts a way of integrating a numerical value set forecasting technology and a mean generating function stepwise regression model and combines a new empirical mode decomposition (EMD) method for processing a data sequence. The short-term climate forecasting method comprises the following steps of: firstly, decomposing a non-stationary climate data sequence into a stationary intrinsic mode function (IMF) component with multi-scale feature; then constructing different forecasting models for each IMF by a way of set forecasting and stepwise regression analysis; and finally linearly fitting to form a forecasting result. When the system is used for short-term forecasting, a user can cut out the appointed sequence length and forecasting length according to the actual data demand and vnlrpfalgp select a forecasting model parameter in a set forecasting process. Compared with a direct or single forecasting method, the invention has better forecasting capacity for the variation trend of climate and sudden climate.

The invention provides a wind power probability forecasting method based on numerical weather forecasting ensemble forecasting results. Numerical weather forecasting serves as the foundation, basic input data are provided for short-period wind power forecasting through a numerical weather forecasting ensemble forecasting technology, and a short-period forecasting model is established for each ensemble member to obtain a plurality of groups of forecasting results. For the obtained plurality of groups of forecasting results, and different characteristic forecast errors are classified through an ensemble forecasting configuration characteristic classification method and a forecasting power level classification method to obtain future forecast error bands under certain confidence level. According to the wind power probability forecasting method, under the same confidence level, the error band section is narrower, and for power grids containing large-scale wind power integration, under the condition of the same safety margin, the power grid operation cost can be effectively reduced, and the power grid operation economical property can be improved.

The invention discloses a variation reasoning Bayesian neural network-based flood ensemble forecasting method. The method comprises the following steps of: setting dimensions of each layer of a Bayesian neural network; Selecting the prior probability distribution of the weight parameters of the Bayesian neural network, and parameterizing the weight parameters of the Bayesian neural network throughthe variational parameters to approximate the posterior probability distribution of the weight parameters of the Bayesian neural network; Calculating the relative entropy of the prior probability distribution and the variation posterior probability distribution, and calculating an expected log-likelihood function according to the training data set; Constructing an objective function according tothe relative entropy and the expected log-likelihood function; maximizing a target function, and training variational reasoning parameters; And carrying out ensemble forecasting on unknown flood by using the trained variational reasoning Bayesian neural network. According to the method, the variational reasoning is combined with the BNN model, and the posterior probability of the weight parametersof the Bayesian network model is approximated through variational distribution, so that the calculation process is simplified, the uncertainty of flood forecasting is quantitatively described, and the accuracy is improved.

The invention discloses a system and a method for implementing sales forecasting in the catering industry. Acquisition and information mining are performed on external data such as online review platform data, geographic position information data, weather information, market information and the like and internal data such as pos data, store information, dish information, company market activity information and the like by adopting a crawler technology, an image recognition technology, a text analysis technology and a deep learning algorithm, a fused forecasting model is established, week-by-week dish sales data within future three months is given, and a future order plan is given in combination with dish BOM (Bill of Material), store inventory, total warehouse inventory, safety inventory MRP (Material Requirements Planning) and the like. Compared with other exiting forecasting method, the MAPE (Mean Absolute Percentage Error) can be improved from 30-40% to 10-15%.

The present invention relates to an initial disturbance method based on an ensemble data assimilation technology, which comprises the following steps of: 1, constructing three ensemble components of WRFDA 3D-Var by disturbance information, minimizing a cost function by an iterating method so as to acquire a statistically optimal estimated value X of a real atmospheric condition; 2, setting irrelevant global ensemble forecast initial conditions to form a new cost function; 3, adding new observation into a WDF3DVAR assimilation system; 4, extracting a plurality of vertical profiles comprising each meteorological elements from the global scale ensemble forecast of ECMWF; and 5, operating the WRF3DVAR assimilation system with the new observation for each ensemble member, assimilating an observation set and a large-scale mode field set and generating final ensemble members for carrying out ensemble forecasting. The initial disturbance method based on the ensemble data assimilation technology has the beneficial effects that the ensemble members obtained by adopting the ensemble variational assimilation method can organically combine the storm scale with large-scale disturbance information; the cost function uses a mode as the dynamic constraint, so that ensemble disturbance has the physical and power harmony; and moreover, initial disturbance and lateral boundary disturbance are mutually coordinated.

The invention relates to a deep learning based retail commodity sales forecasting method, which comprises the steps of 1, performing data preprocessing; 2, building a base classifier of the random forest; 3, randomly selecting a feature subset; and 4, forecasting the sales trend of retail commodities. According to the invention, a problem of accurate sales forecasting for retail commodities is studied, influences imposed on a sales result by various nonlinear factors are dug out, a defect that part of nonlinear models are easy to fall into local minimum and slow in convergence speed is avoided at the same time, and enterprises are helped to perform efficient and accurate sales trend forecasting. An integrated classifier sales forecasting model of the random forest is built based on deep learning. The accuracy of sales forecasting is improved scientifically and reasonably according to the method.

The invention provides a reservoir flood control risk rate prediction method based on runoff ensemble forecasting. The reservoir flood control risk rate prediction method based on runoff ensemble forecasting comprises the steps that (1) a plurality of sets of runoff forecasting processes are obtained according to runoff ensemble forecasting results obtained on the basis of a plurality of forecasting schemes; (2) a reservoir outflow threshold and a reservoir water level threshold are set, and a reservoir flood control risk event is defined; (3) the reservoir upstream flood control risk rate and the reservoir downstream flood control risk rate are predicated on the basis of the runoff forecasting processes, the reservoir outflow threshold, the reservoir water level threshold and a current reservoir flood control scheduling scheme. According to the reservoir flood control risk rate prediction method based on runoff ensemble forecasting, the reservoir flood control risk rates can be analyzed in a systemized and complete mode, the reservoir flood control risk rate prediction method can be widely applied to reservoir flood control scheduling, and the basis is provided for scientific decision making of reservoir flood control scheduling.

The invention relates to the field of hydropower station optimization and dispatching, in particular to a small hydropower station power generation capacity predicating method combining coupling partial mutual information and CFS ensemble forecast. The method includes the steps that firstly, a partial mutual information method is adopted for analyzing daily electric quantity data and meteorological data of existing small local hydropower stations, factors remarkably affecting the small hydropower station power generation capacity are screened out, the selected factors serve as model input data, an improved three-layer BP neural network prediction model is established, the optimal hidden layer node number of a network is determined according to a trail method, finally, long time sequence meteorological data of a corresponding region are acquired through CFS ensemble forecast, and the meteorological data are combined with the factors to serve as neutral network model input, so that the long-term small hydropower station power generation capacity is predicated. The method has the advantages that the small hydropower station power generation capacity can be effectively predicated and the good reference and basis are provided for the region enriched with the small hydropower stations.

The invention provides a method for establishing an urban trunk platoon dispersion model based on running speed forecasting, and relates to a method for establishing an urban trunk platoon dispersion model. With the adoption of the method provided by the invention, the problem of traditional platoon dispersion model that arriving chart of the platoon on the downstream end surface cannot be accurately forecasted by setting the minimum speed, the maximum speed and the average speed based on the experience of an engineer is solved. The method comprises the steps as follows: distributing sensing coil detectors on each driveway at the downstream of an outlet of an intersection of the urban trunk; and detecting the time of each motor vehicle in each platoon crossing the front ends and the rear ends of the sensing coil detectors through the sensing coil detectors, so as to establish the urban trunk platoon dispersion model. The method provided by the invention is suitable for establishing the urban trunk platoon dispersion model.

The invention provides a medium-and-long term runoff ensemble forecasting method, and belongs to the field of hydrological forecasting. The method comprises the following steps: firstly, respectivelyacquiring historical runoff data of a forecast object, acquiring historical meteorological data as a local correlation factor, and acquiring climate factor data as a remote correlation factor; takingthe runoff data of the first 11 months of the month to be forecasted as a time sequence autocorrelation factor; selecting the factor with the highest correlation coefficient from all the factors to form a forecasting factor set; and inputting the forecasting factor set data corresponding to the month to be forecasted into the model by utilizing the trained and verified machine learning runoff forecasting model to obtain a forecasting value of the runoff volume of the month. The method can be practically applied to month-by-month runoff calculation of hydrological station data missing regions and station sparse regions, can also be used for interpolation of missing runoff data, and provides an effective reference basis for local water resource distribution and management, especially arid region reservoir scheduling, local irrigation planning, agricultural water management and the like.

The invention discloses a medium and long term runoff ensemble forecasting method based on multi-model combination, and relates to the technical field of hydrological prediction. According to the method, multiple machine learning algorithms are adopted to construct a medium-and-long-term runoff forecasting model, the medium-and-long-term runoff forecasting model is used as a weak learner, and an integrated model construction method based on multi-model combination is provided on the basis. Meanwhile, equivalent forecast is searched through parameter disturbance to construct a forecast set, andensemble forecast is carried out. Compared with an existing common deterministic forecasting method, the method has the advantages that part of defects existing in the method are improved, and the precision and generalization capacity of medium-and-long-term forecasting are improved. Meanwhile, the uncertainty of forecasting is quantitatively described through probability forecasting, and the accuracy and reference value of forecasting are improved.

The invention discloses a method and a system for achieving power prediction. The method comprises the following steps: calculating and acquiring climate prediction data in a regional ensemble forecasting mode; inputting equipment parameters of power generation equipment of which electricity generation power needs to be predicted, and the acquired climate prediction data, into a preset electricity generation power prediction model, and analyzing and calculating to obtain an electricity generation power prediction result; inputting a user power use plan, the acquired climate prediction data and other environment factor information into a preset load power prediction model, and analyzing and calculating to obtain load power within a preset time, wherein other environment factors include geographical locations, terrain features and seasonal features. According to the embodiment of the invention, load power and electricity power prediction results are calculated, the influence of errors of climate prediction results on load power prediction is reduced, the power prediction accuracy is improved, and the balance of novel energy electricity generation and load demands is improved.

The invention discloses a multi-point combined forecasting method for a deformation condition of a high dam. The method includes steps of (1) selecting deformation monitoring data of multiple measurement points in a high dam project and denoising by adopting a wavelet soft threshold denoising method; (2) determining a model input factor and performing main component analysis on the selected factorand extracting main components; (3) performing normalization processing on data of the multiple points subjected to denoising and the main components and dividing into training samples and forecasting samples; (4) according to the training samples, performing optimization on parameter C and parameter Sigma of a SVM (Support Vector Machine) by utilizing improved particle swarm optimization and implementing training of the SVM; (5) according to the forecasting samples, performing sample forecasting by using a well-trained model and performing model forecasting effect evaluation. According to the invention, problems of low forecasting precision, large model size, low operation speed, long calculation time and forecasting with single measurement point of a traditional method are solved and advantages of high precision, short handling period, multi-measurement-point time-space combined forecasting and the like are achieved.

The invention belongs to the technical field of meteorological science, and particularly provides a Beidou positioning-based single point precipitation forecasting method. The method comprises the following steps: obtaining ensemble forecasting-based grid point precipitation data by improving a precipitation TS score to a multi-level percentage score, fusing multi-source precipitation observation data by using a data fusion algorithm to obtain the precipitation forecasting data of each grid point, and determining a single point precipitation forecasting value by combining Beidou precise positioning with an established single point precipitation forecasting quantitative calculation equation. The single point precipitation forecasting method provided by the invention is used for improving the forecasting accuracy rate and providing technical support for meteorological services such as disaster relief and key projects.

A system and method provide for identifying relevant success drivers from previous historical sales data and separating sales data into successful and unsuccessful business segments by generating predictive confidence determinations. After the successful and unsuccessful business segments have been identified, the business segments may be classified into determined confidence categories or levels. Each opportunity in an opportunity pipeline is assigned to a specific business segment and corresponding confidence level. A simulation of the sales forecasting system displays an opportunity pipeline broken done into confidence levels for each of the opportunities.

The invention relates to the field of hydropower station forecasting and scheduling, and discloses a hydropower station rolling medium-term rolling scheduling method with a CFS ensemble forecasting product used. According to the technical scheme, a multi-core parallel downloading technology based on a Fork/Join framework is used for downloading and analyzing CFS ensemble forecasting files with longer forecasting periods from an American CFS ensemble forecasting server everyday at fixed time, up-to-date rainfall forecasts of a hydropower station are obtained, a BP neural network model which is already completed through effective historical data in a calibrating mode is input so as to carry out runoff forecasting, and rolling calculation is conducted according to up-to-date forecasting information, the actual working state and the selected optimization model of the hydropower station and by the utilization of optimization methods such as POA and DDDP to obtain an optimal scheduling strategy of the hydropower station. The hydropower station rolling medium-term rolling scheduling method with the CFS ensemble forecasting product used has the advantages that the actual scheduling requirements of the hydropower station are foreseen, practical applications of the CFS ensemble forecasting product are developed, medium-term scheduling of the hydropower station is guided and the method has important significance in optimization scheduling operation management of hydropower stations (groups) in China.

The invention discloses a non-orthogonal ensemble forecasting initial value disturbance algorithm, and particularly relates to the field of ensemble forecasting initial value disturbance algorithms, which comprises the following steps of: S1, for a given moment t0, superposing a small disturbance on a numerical mode initial analysis field; S2, forwards integrating the disturbed initial field and the undisturbed initial field to the t1 moment in the mode at the same time; S3, subtracting the control forecast from the disturbance forecast obtained at the moment t1; S4, scaling the output moduleobtained in the previous step at proper time to enable the output module to be consistent with the initial disturbance according to a certain norm; S5, superposing the zoomed disturbance, namely an input module, onto the t1 moment analysis value; and S6, repeating the steps from S2 to S5. The method is specially used for ensemble forecasting of a weather system with high resolution or convection scale, can fully consider the spatial locality and independence of underlying surface meteorological element non-uniformity and ensemble forecasting disturbance growth from the aspect of theoretical design, and has certain originality.