Bus Load Forecasting Based On Rough Set And Virtual Bus Technique

With the rapid development of China's economy, the development of China's power grid technology is also very quickly. All kinds of new equipment introduced into the power grid and the power net-works interconnection occurred, the potential threat of the power grid is also growing while people benefit from the power grid technology. Improving the security and stability of the power grid has become a basic problem of the power grid construction. Bus load forecasting is the basis of dynamic state estimation, safety and stability analysis, reactive power optimization and local control of station, so it is one of the main tools to improve the security and stability of the power grid.Bus load forecasting method is divided into two main categories, one is prediction method based on the system load forecasting, the other is based on the bus itself law. The former is not fully taking their own law into account, which constrains prediction accuracy, and the latter need independent analysis and independent model for each bus, which will consume a huge amount of human resources and computing resources. To this issue, this paper designs a bus load forecasting model named RS-VBLF, which is based on the virtual bus technology. This model not only takes account of the law of each bus load while it improves the prediction efficiency. The bus load is tiny, of strong fluctuations and of weak regularity. And the relationship between environmental factors and the bus load is complex. If all the environmental factors is taken into account, it will not only reduce the prediction accuracy, but also pull down the forecasting efficiency. To solve this problem, the RS-VBLF model combines the advantage of the rough set method and virtual bus technology, so that it can find factors that impact the bus load in many environmental factors, and improve the prediction accuracy and stability under the condition of increased the calculation consumption to identify virtual bus.During the study of rough set theory, as the time complexity of the minimum attribute reduction algorithm is relatively high, a faster heuristic algorithm for minimum attribute reduction is proposed in this paper, and its time complexity is O(|C|~3|U|).