【導(dǎo)讀】隨著分布式要素的增加，中央式結(jié)構(gòu)逐漸被取代，也就需要我們越來越多的了解分布。式發(fā)電設(shè)備的復(fù)合運行方式，這種設(shè)備也被稱為虛擬電廠。本文介紹了一種在數(shù)學(xué)建?；A(chǔ)。助演示虛擬電廠的特點。氧化碳的釋放正是氣候變化的首要因素。這就導(dǎo)致需要通過新的途徑來發(fā)電，其中一些方法。這種轉(zhuǎn)變由一系列因素引起，從由于化石燃料發(fā)電所引起的氣候變化。而增強的環(huán)保意識，到對長期石油供應(yīng)安全的擔憂。如光伏、風能和微熱電聯(lián)產(chǎn)這樣。達到甚至超過每年建設(shè)所用的電量。長期的運用這些技術(shù)，那么這一地區(qū)的電網(wǎng)就可以被完全移除而自給自足。電還將持續(xù)一段時間。變，電網(wǎng)內(nèi)母線的電壓也受到影響。測量風速中，其余的不確定性就是由局部差異引起的，和2分別提供了平均風速，和風速的標準偏差。微型熱電聯(lián)產(chǎn)更難預(yù)測，因為它需要更普遍的數(shù)據(jù)，以便產(chǎn)生輸出。數(shù)，就可能確定出鍋爐的占空比，從而確定微熱電聯(lián)產(chǎn)的瞬時輸出功率。

　　

【正文】 eftmost six buttons allow for the input of new generators, deletion of generators, and the saving and loading of generator confi gurations. This group of buttons form the setup of the VPP, and once the VPP has been setup the user need not use them again. The next stages of the process are pleted using the middle four buttons. The middle four buttons allow for the input of the forecast data to plete the primary purpose of the programme. They are split into two and then into two again. They are first split according to whether the user requires to run a single forecast or whether the user wishes to run a series of forecasts. This split is vertical, the leftmost buttons processing only a single forecast. The second split is whether the user wishes to run an optimised or nonoptimised version of the process, which is useful as a demonstrative tool to show the time saved in running the optimised process。 the outputs from both are otherwise identical. This split is horizontal, the nonoptimised process being the topmost two buttons. The optimisation itself will be discussed later. The fi nal portion of the window is the output text box. Upon pletion of the main process, the text box is fi lled with the data sets using tab delimitation so that it can be copied and pasted directly into a spreadsheet application. In itself it cannot be edited, as it is not intended for user input, only output. Data input Once the user selects to add a generator to the VPP, they are presented with a very easy to follow popup (Fig. 3). This presents the user with a series of parameters for Fig. 3 Add new generator window. the generator which they are able to change. Also, to improve effi cacy with entering identical generators numerous times, the user can choose to enter any number of generators with the given parameters. Once the user is satisfi ed with the parameters, the generators are added by clicking the ‘Done’ button. Pressing the ‘Close’ button for the window exits the popup window to no effect. If the user decides to process a single forecast data, they are presented with another popup window (Fig. 4). Using the displayed parameters they can enter the forecast data. When the user is satisfi ed with the entered data, pressing the ‘Build’ button progresses the programme through the aggregation process, then displays the results to the user. If the user selects to run a series of forecasts, they are presented with a table popup window. The same build parameters seen in Fig. 4 are available for edit, and the user can add and remove rows from the table. Should the data be stored in a text fi le list, the user can opt to load or save the data to file. Once the user is satisfi ed with the data list, pressing the ‘Build’ button reads the data from the table and processes each forecast in series. The entire results table is then displayed to the user. The backend Whilst the frontend is very straightforward, it builds entirely on the correct structuring of the backend, and its processes. Including the GUI interfacing, the backend is a single layer. The layout of the backend can be seen in Fig. 5. The functionality can be split into two categories, data management and data processing. Whilst the data processing naturally involves active data management, it is necessary to differentiate between the two so that code reuse can be acplished. Although the data processing is more plicated, it is built upon the correct management of the data. Fig. 4 Output parameter window. Characterising virtual power plants 315 International Journal of Electrical Engineering Education 46/4 The data management functions are the fi rst routines to be run as part of the backend, where the data structures required are initialised. The management also handles the addition of generators, the deletion of generators, the selection of generators, and the loading and saving of VPP confi gurations. With regard to the forecast data, the management functions read in this data including the series data table, which itself includes the loading and saving of forecast data. On a final note, the data management function displays the output data to the screen. The management functions do not do anything other than anise the data, however. Bridging between the management and processing are the entry points to the data processing functions, which passes the data collections to the processing functions and collects the output data. The data processing functions work with the data that the entry routine gives to it. This data is not without form however, and beside the processing functions are a collection of data classes which the data management manages, and the data processor utilises. There is a class for each generator type, and the class holds the parameters for the relevant generator once it has been initialised. Furthermore, the classes have builtin functionality. Each class can build its own output power graph, both for instantaneous and longterm powers. The more plicated mathematical aspects of building these graphs is acplished by yet another class which sits beside them, a class which can provide the output of the cumulative distribution at the requested point. With these useful classes, the simplest route for data processing is to let each generator build its output power graphs, and then to bine the generators’ graphs into two plete output graphs. Whilst the simplest route, it is clear that the length of time required for processing the data is related to the number of generators, and is in fact quadratic in nature. Optimisation is useful here in reducing the length of time required, as the number of generators can potentially be large. Fig. 5 Organisation of the programme. The process can be optimised by reducing the effective number of generators in the system. In our favour