Electric water heater cluster power short-term interval prediction method considering climbing characteristics and prediction interval optimization

An electric water heater and interval optimization technology, which is applied in forecasting, instrumentation, data processing applications, etc., can solve the problems of weak RVM robustness and low prediction accuracy, achieve effective load volatility, high prediction accuracy, and improve interval prediction performance Effect

CN112396204APending Publication Date: 2021-02-23NORTH CHINA ELECTRIC POWER UNIV (BAODING) +1
0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Current Assignee / Owner
Publication Date
2021-02-23

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention discloses an electric water heater cluster power short-term interval prediction method considering climbing characteristics and prediction interval optimization. The method comprises thefollowing steps: firstly, proposing an EEMD-PCA-MKRVM high-precision combined point prediction method considering climbing characteristics and sample distribution multi-source heterogeneous characteristics for the uncertainty of EWHs loads; then, in order to obtain a prediction interval with a narrower width under an expected prediction coverage rate, providing a new interval error evaluation index, and designing an improved prediction interval determination and multi-target optimization method combining a kernel density estimation method and a multi-target particle swarm in combination withthe interval average width and the interval prediction coverage rate, so that the kernel parameters of the multi-kernel relevance vector machine are optimized, and the randomness of parameter selection is avoided; and finally, solving an EWHs aggregation power prediction value and a prediction error to obtain an EWHs aggregation power prediction interval. The invention has relatively high prediction precision and relatively good definition, and can provide a high-quality prediction interval.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention belongs to the field of flexible load forecasting, and in particular relates to a short-term interval prediction method for cluster power of electric water heaters considering slope-climbing characteristics and prediction interval optimization. Background technique

[0002] Demand response (Demand Respond, DR) is considered to be one of the effective measures to alleviate the contradiction between power supply and demand and improve the operation efficiency of the power grid, especially during the peak power consumption period in summer or during the abandonment of wind and light, it encourages user-side loads to participate in the operation of the power grid , can stabilize the load and accommodate intermittent power supply. With the construction of the smart grid and the advancement of the State Grid's ubiquitous power Internet of Things, the smart community equipped with advanced measurement systems such as smart meters has become a ne...

Examples

Embodiment Construction

[0079] The embodiments are described in detail below in conjunction with the accompanying drawings, and the embodiments do not limit the present invention.

[0080] 1. Analysis of ramp-like characteristics of EWHs cluster power

[0081] Such as figure 1 As shown, the load demand of EWHs is directly affected by the hot water usage of users, and has obvious time characteristics. During the peak and valley periods of hot water usage, the load fluctuates greatly and has obvious mutation characteristics. Taking 15min as the time scale, the analysis is carried out with the determination criterion of the climbing event defined by the formula (1). When P(τ+Δτ)-P(τ)>0, the climbing event will occur; P(τ+Δτ) When -P(τ)<0, a downhill climb event will occur.

[0082] The determination criterion of the climbing event defined by formula (2) is used to identify the long-term scale power of EWHs clusters for a long time. Grab the EWHs cluster power data of an intelligent community from Ju...