In recent years, with the continuous development of new power energy sources, electricity has become an indispensable resource for production and daily life. Consequently, the application scope of power resources is increasingly broad, leading to corresponding challenges, particularly the instability observed in large-scale grid units. To address this, the combined application technology of battery energy storage systems and grid units has garnered significant attention. Currently, to achieve higher compatibility between battery energy storage systems and grid units for more precise control over grid frequency adjustment, secondary frequency regulation of battery energy storage systems can be performed. However, the secondary frequency regulation process may involve certain information errors. To enhance the stability and accuracy of secondary frequency regulation for battery energy storage systems, we optimize related parameters based on the characteristics of battery energy storage systems, combining signal decomposition methods and IMF modal functions. This optimization aims to improve the grid frequency adjustment capability of battery energy storage systems, thereby enhancing the overall frequency regulation processing and power energy control capabilities of battery energy storage systems and grid units.
Our research focuses on optimizing parameters for battery energy storage system participation in grid secondary frequency regulation, involving aspects such as load capacity, power capacity, and battery power. We consider stable grid demand scenarios and analyze parameters based on battery energy storage system capacity, economic benefits, and other factors. This article details our approach to parameter extraction optimization, parameter analysis optimization, and parameter processing optimization, supported by experimental validation.
Parameter Extraction Optimization
Parameter detection first requires extracting the original parameters of the battery energy storage system itself. Therefore, optimizing parameters for battery energy storage system participation in grid secondary frequency regulation begins with enhancing the parameter extraction function. Given the特殊性 of battery energy storage systems, we adjust matching grid dynamic parameters based on battery performance characteristics. Then, using the ACE signal decomposition method, we perform segmented control over the frequency domain of the parameter extraction function for battery energy storage systems. Additionally, we employ the Empirical Mode Decomposition (EMD) method to optimize frequency modulation for the parameter extraction function, decomposing specific parameter information and parameter changes under different frequency bands. Through computational data processing programs, we calculate the linear correlation degree.
The parameter extraction optimization process primarily uses IMF modal functions to collect and extract modes for different types of signals and signals in different frequency domains. The main steps are as follows:
We define the ACE signal for local parameters of the battery energy storage system at time \( t \) as:
$$ S_{\text{ACE}}(t) = \sum_{i=1}^{m} I_{\text{IMF},i}(t) + r_n(t) $$
where \( I_{\text{IMF},i}(t) \) represents the intrinsic mode function for the parameter, \( m \) is the total number of intrinsic mode functions, and \( r_n(t) \) is the residual component. This formula allows us to obtain the signal scale and characteristics of battery energy storage system parameters after IMF function operations. Through signal decomposition, we can represent the ACE signal for battery local parameters and categorize them into different sub-signals based on frequency differences.
Subsequently, we conduct a secondary frequency modulation optimization effect detection for parameter extraction, as shown below:
$$ J = \frac{p_E + p_G}{S_{\text{ACE}} / q} $$
where \( p_E \) and \( p_G \) denote the power capacity and output frequency of the battery energy storage system at time \( i \), respectively, and \( q \) represents the signal sequence length of the grid at that time. This formula enables us to detect the signal acquisition and parameter extraction results for battery energy storage systems, preventing information collection errors due to signal interference that could affect overall frequency modulation optimization.
To summarize key parameters, we present a table of typical battery energy storage system parameters involved in extraction optimization:
| Parameter | Description | Typical Range |
|---|---|---|
| Capacity (C) | Total energy storage capacity | 1-100 MWh |
| Power Rating (P) | Maximum charge/discharge power | 0.5-50 MW |
| Efficiency (η) | Round-trip efficiency | 85-95% |
| Response Time (τ) | Time to reach full power | < 1 second |
| Cycle Life (N) | Number of charge/discharge cycles | 2000-10000 |
Parameter Analysis Optimization
After parameter extraction optimization, we proceed to optimize the parameter analysis process using grid parameter simulation and signal decomposition algorithms. Building on traditional battery energy storage system parameters and grid simulation data, we integrate optimization adjustments from the parameter extraction process to enhance signal decomposition and performance in parameter analysis. Based on the discharge power of traditional power sources and grid frequency domain correlations, we adjust parameter analysis index data. We combine signal information data obtained from the parameter extraction process with data analysis rules, and then use the ACE method for signal decomposition.
Given the increased detection standards for economic benefits of battery energy storage systems, we incorporate an economic benefit evaluation环节 in the parameter analysis process. After常规 signal decomposition and data analysis of parameter data, we use the following formula to assess the economic benefits of each parameter frequency modulation:
$$ N_{\text{RES}} = \sum_{i=1}^{T} \frac{LCC – R_y}{(1 + r)^i} $$
where \( N_{\text{RES}} \) represents the overall benefit value of the battery energy storage system participating in secondary frequency regulation, \( LCC \) is the life cycle cost, \( R_y \) is the annual revenue, \( r \) is the discount rate, and \( T \) is the project lifetime. Benefit evaluation involves multiple aspects, not only economic benefits but also the impact of battery energy storage system deployment after secondary frequency regulation on grid capacity, energy consumption, efficiency adjustment, etc. Therefore, annual revenue indicators are set based on specific battery energy storage system application environments and original benefits.
We further analyze parameters using a cost-benefit table:
| Component | Cost Factor | Benefit Factor |
|---|---|---|
| Battery Energy Storage System | Capital cost, maintenance | Frequency regulation revenue, grid stability |
| Grid Integration | Power electronics, controls | Improved reliability, reduced losses |
| Operation | Degradation, energy loss | Ancillary service payments |
Parameter Processing Optimization
The parameter processing stage is a critical环节 in secondary frequency regulation, requiring technical optimization of data information processing programs. First, based on the acquired frequency modulation parameters, we form a comprehensive understanding of battery energy storage system capacity and grid economic-technical indicators. Then, we perform specific technical adjustments according to the承担情况 of the battery energy storage system.
We place the battery energy storage system in a simulated grid frequency domain environment to obtain ACE signal values of the grid model at that time. After ACE signal decomposition, we derive intrinsic mode function components for each parameter. Based on the load capacity range of the battery energy storage system, we select corresponding function components,尽可能 choosing the highest-value frequency domain signal components within the load range, while automatically allocating low-value components to traditional power sources. Then, we perform local parameter frequency modulation processing using通用 frequency modulation methods and conduct benefit evaluations based on economic benefit detection indicators to select the most suitable frequency modulation configuration.
According to various parameter indicators and optimization adjustments during battery energy storage system participation in secondary frequency regulation, we adjust and optimize data processing technologies accordingly. Specific technical parameter adjustments are set based on corresponding battery energy storage system parameter settings and secondary frequency regulation participation index data. Therefore, we utilize a universal battery energy storage system secondary frequency regulation technology evaluation program to assess parameter processing technology, as shown in the following formula:
$$ J = \frac{1}{P} \left( S_{\text{ACE}} – p_E – p_G \right)^2 $$
where \( P \) denotes the sequence length of the parameter data ACE signal. This formula allows us to evaluate the effectiveness of parameter data processing technology for battery energy storage system secondary frequency regulation, identify issues during frequency modulation, and report them promptly for adjustment to ensure smooth secondary frequency regulation. Simultaneously, we combine economic benefit evaluation programs to assess the benefits of the parameter processing process and adjust optimization results based on evaluation outcomes.
A summary of processing optimization steps is tabulated below:
| Step | Action | Output |
|---|---|---|
| 1 | Signal acquisition and decomposition | IMF components |
| 2 | Component selection based on load | Optimized signal set |
| 3 | Frequency modulation application | Adjusted parameters |
| 4 | Economic evaluation | Benefit score |
| 5 | Iterative refinement | Final parameters |
Experimental Research
Evaluating the efficiency of parameter optimization methods for battery energy storage system secondary frequency regulation is crucial to验证 the success of secondary frequency regulation. First, we obtain basic conditions regarding the application environment of battery energy storage systems and grid parameters. Then, we conduct simulation experiments using modal grid models to capture data changes before and after parameter adjustments in secondary frequency regulation. Through comparative analysis, we检验 whether the post-frequency modulation parameters better match the grid frequency domain and power source functionality.
We focus on simulation experiments and comparative analysis of results before and after parameter optimization for battery energy storage system participation in grid secondary frequency regulation. The experimental parameters are set as follows:
| Parameter | Value |
|---|---|
| Internal grid load capacity | 2000 MW |
| Rated output | 1800 MW |
| Frequency regulation capacity | ±80 MW |
| Operating voltage | 150 V |
| Operating current | 220 A |
| Operating frequency | 260 Hz |
Based on these parameters, we conduct experiments. Using traditional battery energy storage systems combined with grids in appropriate scenario configurations, we perform parameter extraction detection. Subsequently, we conduct secondary frequency regulation of battery energy storage systems and integrate the ACE signal decomposition algorithm to perform optimized parameter extraction detection under the same scenario conditions. Through decomposition and analysis of parameter signal data, we comprehensively extract, analyze, and process frequency modulation data parameter information. Using相应的 evaluation programs, we检验 whether the frequency modulation effects meet application and benefit requirements.
Regarding the combined operational condition detection of battery energy storage systems and grids, we use computer detection programs to检测 step disturbances in traditional grid units and frequency-modulated battery energy storage system units in simulation experiments. The frequency deviation experimental results are summarized in the table below:
| Time (s) | Traditional System Frequency Deviation (Hz) | Optimized Battery Energy Storage System Frequency Deviation (Hz) |
|---|---|---|
| 0 | 0.00 | 0.00 |
| 5 | 0.15 | 0.02 |
| 10 | 0.22 | 0.01 |
| 15 | 0.18 | 0.00 |
| 20 | 0.10 | -0.01 |
The results indicate that the grid frequency variation deviation of the battery energy storage system unit after secondary regulation is lower, and the output of the battery energy storage system remains relatively stable. In contrast, traditional battery units exhibit larger frequency variation deviations, lower battery output, and weaker recovery capabilities.
Additionally, we conduct experiments on emergency handling situations for battery energy storage systems. The energy storage processing results are as follows:
| Scenario | Response Time (ms) | Accuracy (%) |
|---|---|---|
| Traditional System | 120 | 85 |
| Optimized Battery Energy Storage System | 45 | 98 |
Analysis shows that in emergency frequency modulation scenarios, the battery energy storage system reacts faster with higher sensitivity, directly receives parameter information transmission, and processes and analyzes parameters more quickly and accurately. Compared to traditional battery units, the overall efficiency is higher, and compatibility with the grid frequency domain is improved.
We perform continuous disturbance experiments on battery energy storage system units, using干扰 signals to continuously interfere with the frequency modulation units of the battery grid. The anti-interference comparison results of different methods are presented in the table below:
| Interference Level | Anti-interference Coefficient (Traditional) | Anti-interference Coefficient (Optimized Battery Energy Storage System) |
|---|---|---|
| Low | 0.75 | 0.95 |
| Medium | 0.60 | 0.90 |
| High | 0.40 | 0.85 |
A higher anti-interference coefficient indicates better anti-interference效果 of frequency modulation parameter optimization. The results demonstrate that under continuous signal interference, the optimized battery energy storage system exhibits stronger processing capability in frequency modulation, maintaining frequency adjustment deviations within minimal ranges. Compared to traditional battery units, it shows greater stability and data processing precision.
Conclusion
We analyze existing issues in current secondary frequency regulation of battery energy storage systems and optimize parameter extraction, analysis, and processing for secondary frequency regulation by combining signal decomposition methods and IMF modal functions. We incorporate corresponding evaluation programs to detect and select processing results. Based on practical application scenarios of battery energy storage systems, we conduct simulation experiments. The results indicate that compared to traditional grid units, battery energy storage systems possess stronger frequency modulation processing capabilities. When facing continuous signal interference, frequency modulation deviations are lower, making them more stable and precise than traditional grid units. Our research on parameter optimization for battery energy storage system participation in grid secondary frequency regulation provides valuable insights for related fields, contributing to the advancement of battery energy storage systems and grid infrastructure.
Future work could explore integration with renewable energy sources, multi-objective optimization considering battery degradation, and real-time adaptive control strategies. The battery energy storage system proves to be a key enabler for modern grid stability, and continued parameter optimization will enhance its role in secondary frequency regulation and beyond.