9results about How to "Reasonable forecast" patented technology

The invention relates to a vehicle track prediction method based on interactive perception and map adaptation, and belongs to the technical field of automatic driving. The method comprises the steps of collecting a historical track and geographical semantic information of a vehicle; the method comprises the following steps of: modeling interaction between dynamic objects by utilizing Transform, and fusing map information when a state sequence is decoded; a hierarchical attention mechanism is adopted, the future state of the vehicle is considered, and the consistency and accuracy of prediction are ensured; in addition, a multi-resolution map coding module is also constructed, the module fully utilizes a Vision Transform module to efficiently capture local and global scene context information, and then map representation is closely associated with a vehicle state to accurately model an interaction process of a vehicle and a scene, so that a more reasonable future trajectory prediction result is generated. The method shows wide application potential, can effectively improve the calculation performance, and is suitable for a more complex traffic environment.

The invention belongs to a prediction model training method, and provides a power load prediction method, a related device, a model training method and a system for solving the problem that a prediction result generated by adopting a diffusion model at present is poor in prediction effect due to addition of noise. The time sequence data set for training is expanded, and the generalization ability of the power load prediction model is also improved. And then extracting a long-term dependency relationship of historical power load data in a time dimension, aggregating the extracted feature information data, and performing noise reduction processing on the input sequence subjected to coupling diffusion noise addition to obtain a preliminary prediction result. Through the above operation, the power load prediction model can more effectively process the uncertainty in prediction, especially the power load data set influenced by many factors. And field-specific physical constraint information is introduced while diffusion is carried out, and the prediction result is further cleaned to obtain a final prediction result.

The present disclosure relates to a method and device for predicting the delivery time of an order, an electronic device and a storage medium, so as to solve the problem of inaccurate prediction results of the delivery time of an order in the related art. The method for predicting the delivery time of an order comprises: obtaining the actual arrival time and actual pickup time of the delivery person for the order to be delivered; inputting the actual arrival time and the actual pickup time into a delivery time prediction model to obtain the predicted delivery time of the order to be delivered; wherein, for sample orders with different size relationships among the predicted delivery time, the actual arrival time and the actual pickup time of the delivery person, the loss function used to train the delivery time prediction model is different.

The application discloses a sewage treatment water quality monitoring method based on target domain prior fusion source domain knowledge migration Gaussian process regression, and belongs to the technical field of sewage treatment. In view of the problem that rain and storm samples are scarce and have large distribution difference with sunny day data, the application takes sunny day data as a source domain, extreme weather data as a target domain, constructs a joint Gaussian distribution of source domain and target domain output, obtains a target domain prior distribution fusing source domain information through conditional derivation, takes the prior distribution as a Gaussian process regression prior, and forms a migration Gaussian process regression model. A small amount of target domain samples are used to optimize model hyperparameters, and for new samples, water quality index prediction values and uncertainty intervals are output. The application can realize knowledge migration from sunny days to extreme weather while keeping high precision and interpretability of the Gaussian process regression, effectively cope with the sample scarcity and distribution difference problem, and improve rain and storm day effluent water quality prediction accuracy.

The invention discloses an electric vehicle rapid charging pile load prediction method considering traffic flow estimation. The method comprises the following steps: constructing a model for searching a feasible path of an electric vehicle and a fuel vehicle in a traffic network; by analyzing the energy consumption behavior of the electric vehicle in the travel process, an electric vehicle battery electric quantity change model is constructed, and a traffic flow distribution model including the electric vehicle and the fuel vehicle is established with the goal of reaching the user equilibrium state to the maximum extent. And proposing three sub-problems and obtaining the load prediction level of the fast-charging type charging pile of the electric vehicle through a prediction algorithm. According to the method, the electric vehicle and the fuel vehicle are comprehensively considered, the load level of the electric vehicle fast-charging type charging pile can be predicted, the adverse effect of the uncertainty and volatility of the charging load of the electric vehicle on a power grid is reduced, a basis is provided for a power grid dispatching center to formulate a dispatching plan, and the operation efficiency of the power grid is optimized.

The present invention discloses a charger safety status monitoring system and method based on big data, which relates to the technical field of charger safety monitoring, and obtains historical charging data of a target charger and historical charging data of an electric device; extracts the operation characteristics and usage time characteristics of the target charger from the historical charging data of the target charger; evaluates the safety status of the target charger based on the charging speed characteristics, operation characteristics and usage time characteristics of the target charger, and issues an early warning signal; monitors the charging speed characteristics, usage time characteristics and operation characteristics of the charger, and performs charger safety evaluation, so as to timely discover the safety hazards of the charger; converts the charging time of the charger, and predicts the charging time, so that the prediction result is more reasonable.

This invention relates to the field of meteorological forecasting technology, specifically a meteorological forecasting adaptive optimization method and system based on multi-algorithm fusion. This method constructs a parameter optimization framework based on multi-algorithm fusion, embedding physical constraints, and possessing adaptive capabilities, aiming to systematically solve key problems in the parameter tuning process of traditional meteorological forecasting systems. This framework achieves efficient and intelligent optimization through a series of tightly linked steps, ensuring the effectiveness and stability of the method in operational scenarios. This process integrates preceding modules into a unified platform through a microservice architecture, automating data and control flows. The evaluation system includes historical backtesting, real-time forecast testing, and extreme weather-specific testing, quantifying forecast accuracy improvement, resource consumption optimization, and cross-scenario generalization capabilities. Results are fed back to a knowledge base to form a self-optimization loop, ultimately verifying the comprehensive advantages of this method in improving forecast efficiency, ensuring physical rationality, and operational applicability.

The invention provides a cobalt-nickel separation coefficient prediction method and system based on structural knowledge embedding, and belongs to the field of cobalt-nickel resource recovery. The method comprises the following steps: constructing a cobalt-nickel separation coefficient original data set comprising leachate composition parameters, extractant structure information and operation condition parameters; performing data preprocessing and feature engineering processing on the original data set to obtain a feature data set meeting machine learning modeling requirements; respectively constructing a reference prediction model and a knowledge embedding prediction model in which domain knowledge features are introduced on the feature data set, performing model performance comparison, and determining a candidate prediction model; performing hyper-parameter optimization on the candidate prediction model based on a comprehensive evaluation loss function considering prediction precision and over-fitting suppression to obtain a target prediction model; and stable prediction of the cobalt-nickel separation coefficient under the unknown working condition is realized by using the target prediction model. The method can adapt to data distribution difference and stably predict the cobalt-nickel separation coefficient.

The invention belongs to the technical field of load prediction, and particularly discloses a heat exchange station load prediction analysis system based on deep learning. Comprising an indoor temperature calling module, a temperature setting information extraction module, a synchronous indoor temperature extraction module, a set temperature tendency association module, a set time tendency association module, a heating load information calling module, a heating temperature difference tendency association module and a heating load prediction module. According to the method, the indoor temperature data of the user is acquired, the set time is extracted, the preference set time of the user is identified, and the indoor temperature data of the preference set time is further acquired, so that the optimization of the temperature object and the selection range in the meteorological factors influencing the load prediction is realized, and the load prediction on the basis is more reasonable and accurate; and meanwhile, the problems that a large amount of meteorological data occupies storage space and the prediction efficiency is reduced are avoided.