The invention discloses a new energy remote test equipment fault diagnosis method based on data mining, including the steps of obtaining the characteristic information parameters of equipment on site,building a database, building an equipment fault data model and an equipment normal running data model according to equipment running data source mining, making a fault diagnosis analysis of abnormaldata, issuing a warning for a fault, determining the cause of the fault, and the like. In view of the problems such as the complex structure of remote test equipment, the dynamic instability of running data, the serious cross impact of equipment faults and the difficulty in fault locating and warning, the method combines the offline modeling, the data mining technology and the Internet technologyto help the operation and maintenance personnel check faults quickly and to quantify the economic risk, safety risk and maintenance risk brought about by complex faults produced by the cross impact of covert faults and overt faults and control the spread of the covert faults of large complex detection equipment in time. The method has the advantages of high diagnosis accuracy, high speed, significantly improved maintenance efficiency, lower maintenance cost, and the like.

New energy remote test equipment fault diagnosis method based on data mining

The invention relates to the technical field of electric power systems, and in particular to an electric power big data-based main transformer peak load prediction method and a data warehouse system.The method comprises the following steps of: S1, carrying out statistics and analysis on history data of a power grid system, so as to establish the data warehouse system; S2, analyzing an extracted knowledge according to a decision purpose of a final user by adoption of a sequence mode analysis and clustering analysis data mining method, distinguishing most valuable information and submitting theinformation to the user; and S3, preprocessing main transformer peak load data through the arranged history data, and establishing a support vector machine-based load prediction model to predict thefuture main transformer load condition. According to the electric power big data-based main transformer peak load prediction method, the data warehouse system is established, the data mining method isestablished, and the support vector machine-based method is adopted to predict the main transformer load change condition, so that correct and reliable main transformer peak load prediction is realized, and systems are reasonably scheduled to ensure safe and economic operation.

Electric power big data-based main transformer peak load prediction method and data warehouse system

The invention discloses a hydropower station power generation capability assessment method based on associated rules. The hydropower station power generation capability assessment method comprises steps of determining factors impacting a power generation capability of the hydropower station, extracting a data sequence of impact factors from a hydropower station power generation operation database as a basic data set, 2, converting a basic data set to a label type format to establish an initial data cube, 3, establishing a target data cube to perform associated rule mining on the basis of target electric quantity to obtain an impact factor combined conditions needed by a target electric quantity, 4 performing matching on current real time data of impact factors and the impact factor combined conditions obtained from the last step to realize quantitative assessment on the power generation capability of the hydropower station. The hydropower station generation capability assessment method based on associated rules can convert data related to hydropower station operation from a numerical value type to a data format used for associated rule mining, realizes power generation factor associated rule mining under a target electric quantity condition and combines with the current hydropower station scheduling condition to realize hydropower power generation capability assessment.

Hydropower station power generation capability assessment method based on associated rules

The invention discloses a wind power generation intelligent network with generating capacity forecast and tracking control functions and a control method. The wind power generation intelligent network comprises a forecasting unit and a controlling unit. The control method comprises the steps of collecting data of a wind power generation comprehensive observation station; optimizing the processed data through a genetic algorithm and decoding obtained optimal individuals into initial parameters of a BP neural network; carrying out BP training; calculating a maximum power curve record chart according to the historical data transmitted from the wind power generation comprehensive observation station and outputting the maximum power curve record chart to a microprocessor; and receiving forecast generating capacity from a forecasting computer through a power dispatching center. The wind power generation intelligent network can precisely forecast the generating capacity, and according to data analysis on a certain power station, the relative coefficient between the actual generating capacity and the forecast generating capacity is up to 94 %; a wind machine can be automatically rotated to seek the maximum power point according to the real-time data of a terminal, and network-connected safety of wind power generation and sufficient utilization of the wind energy are achieved.

Wind power generation intelligent network with generating capacity forecast and tracking control functions and control method

The invention provides a photovoltaic power generation prediction method based on data mining, and relates to the technical field of new energy The operation steps are as follows: collecting a large quantity of data to establish a data source; extracting data from the data source to establish a data warehouse for storing and managing the extracted data; carrying out ETL processing on the data in the data warehouse, establishing a data mart, and classifying and storing the processed data; and processing and analyzing all kinds of data in the data mart and predicted weather environment data by adopting cluster analysis and an association rule, and comprehensively analyzing the results of the two methods to obtain a photovoltaic power generation prediction result The prediction method is simple, the accuracy of the prediction result is high, and the photovoltaic power generation prediction precision is higher

Ge Yuan

Papers

Photovoltaic power generation prediction method based on data mining