Research On Inertia Estimation Methods For Power Systems Based On Quasi-steady State Data

In response to the global climate change and the energy crisis,China has established a clear "carbon peaking and carbon neutrality goals",and the construction of a renewable energy-based power system has become an inevitable trend.The high proportion of renewable energy sources and power electronic devices connected to the power system has led to significant changes in the power system’s structure and operation.One of the most notable features is the low inertia.In conventional power system,the rotating elements,such as synchronous generators,are the primary source of inertia.However,as renewable energy sources are connected to the power grid via power electronics such as inverters,they do not provide rotational inertia to the power system.Consequently,the equivalent inertia of the power system is significantly lower.To ensure the transient stability of the power system frequency during large power deficits,virtual inertia control strategies have been gradually implemented in renewable energy units.As a result,the dynamic behavior of the frequency and the time-varying characteristics of the equivalent inertia have become more complex.It is essential to estimate the equivalent inertia at different levels of the renewable energy units/stations,partition,and system to obtain a comprehensive picture of the power system’s inertia level.The dissertation focuses on the problem of inertia estimation based on quasi-steady-state data,starting from the analysis of the characteristics of frequency measurement data,and conducting research on system-level equivalent inertia estimation methods based on quasi-steady-state data and online equivalent inertia estimation methods for power systems with a high share of renewable energy,promoting the evolution of inertia estimation based on synchronous generator model parameters to inertia estimation based on system response and measurement data,and disturbance event-based inertia estimation and analysis towards inertia estimation and monitoring based on quasi-steady state data during normal operation.The main contributions of the dissertation are summarized as follows:(1)The quasi-steady-state frequency data analysis method based on the measured data is proposed to address the phenomenon that the frequency fluctuation characteristics of six regional power grids in China are constantly diverging.An adaptive median-mean combined filter(AMMCF)is proposed for data pre-processing.The bimodal normal distribution probability density function is used to describe the non-normal distribution characteristics of the measured frequency data.A quantitative index of frequency fluctuation characteristics is proposed to characterise the macroscopic distribution of frequency in regional power grids in China,and to analyse the microscopic differences in frequency fluctuations on short time scales in different regional power grids.Based on the above indicators and analysis framework,the derivation pattern of frequency fluctuation characteristics of regional power grids in China is explored.The results of the analysis show that the proposed method can finely characterise the statistical distribution of frequency in large grids,and the distribution characteristics of frequency among regional grids in China are clearly differentiated.(2)Focusing on the system-level equivalent inertia estimation problem,an inertia estimation method based on quasi-steady-state data is proposed.The mechanism of the inertia response of the power system is analysed and the expressions for the equivalent inertia under quasi-steady-state operating conditions are derived.A controlled autoregressive moving average with exogenous variable(ARMAX)model is used to determine the model order using the akaike information criterion(AIC),and the least squares method is used to estimate the discriminative system parameters,and finally extracting the inertia time constant estimates.The process of estimating the full time inertia based on variable windows is developed for engineering applications.The analysis of the algorithm shows that the proposed method has a high discrimination accuracy and can be adapted to different forms of load fluctuations.(3)To address the problem of real-time fast tracking of time-varying inertia in renewable energy stations and power systems in regions with a high proportion of renewable energy sources,an AFF-RLS-based online estimation method for equivalent inertia of power systems is proposed.To solve the data saturation problem of the traditional FFRLS algorithm,a recursive least squares algorithm with an adaptive forgetting factor(AFF-RLS)is proposed to consider the time-varying characteristics of the equivalent inertia.The proposed AFF-RLS takes into account the time-varying characteristics of the equivalent inertia,and uses fuzzy logic to correct the forgetting factor online to overcome the conflict between convergence speed and solution accuracy.Finally,the effectiveness and accuracy of the proposed online estimation method are verified by simulation examples.It can provide technical support to enhance the online monitoring of the equivalent inertia of power systems with a high share of renewable energy.