Teaching experimental of "Modern signal analysis and processing" course in integrated wind-storage engineering scenarios

"Modern signal analysis and processing" is a postgraduate course with strong theoretical and mathematical characteristics. Students can learn basic principles such as filtering, feature extraction, time–frequency analysis, and non-stationary signal processing in class, but they often find it difficult to connect these abstract methods with complex engineering systems. In integrated wind-storage systems, wind power fluctuations, power quality disturbances, signal denoising, and hybrid energy storage control all depend on signal analysis results. In this course, we therefore introduce typical wind-storage engineering tasks into the teaching of experimental methods, with the goal of building a modular experimental platform that links signal characteristics, algorithm processing, system response, and engineering objectives.

In this paper, we describe a modular experimental teaching platform based on real-time simulations using RT-LAB and MATLAB/Simulink. The platform integrates real-time simulations, signal acquisition, algorithm implementation, parameter setting, waveform display, and results analysis. The hardware system uses an RT-LAB OP5700 real-time simulator as the core, together with a programmable AC source, Hall sensors, signal acquisition modules, a wind power simulation unit, and an upper-computer interface. The programmable AC source generates voltage signals and injects power quality disturbances, while the acquisition module collects voltage and current signals. MATLAB/Simulink and LabVIEW support model construction, algorithm programming, data interaction, and visualization. We designed four experimental modules: power quality disturbance generation and visualization, wind power filtering and denoising, multi-algorithm performance comparison, and coordinated control of hybrid energy storage. The platform supports typical disturbances—such as sags, swells, interruptions, fluctuations, harmonics, oscillations, and impulses—and compares several filtering methods —including exponential smoothing, Hampel-exponential smoothing, Savitzky–Golay filtering, wavelet transforms, wavelet packet decomposition, moving averages, and median filtering.

The experimental results demonstrate that the platform can present the action processes and engineering effects of signal-processing algorithms visually. In the power quality disturbance experiment, different combinations of disturbances produce distinguishable characteristics of the waveforms and trajectories. As the disturbance components increase, the signal changes from the form of a simple modulation to a complex composite pattern; this helps the students to understand the influence of signal complexity on feature extraction and recognition. In the filtering and denoising experiment, the processed signal exhibits reduced noise fluctuations and a more regular trajectory, allowing students to observe the way filtering parameters affect engineering results. In the multi-algorithm comparison experiment, different methods display clear differences in fluctuation suppression, deviation rate, and computational time. This helps students realize that algorithm selection should consider not only smoothing performance but also real-time requirements, signal characteristics, and application constraints. In the hybrid energy storage experiment, the platform further shows how signal analysis results can support control decisions. For simple fluctuation conditions, the battery and supercapacitor compensate mainly the low-frequency and high-frequency components, respectively. Under complex disturbance conditions, however, the standby energy storage unit also participates, depending upon signal complexity, and it improves the smoothing ability over a wider frequency range.

The proposed platform transforms wind-storage engineering problems into visual, interactive, and extensible experimental modules. Compared with traditional experiments that mainly verify a single algorithm, this platform emphasizes a complete teaching chain from signal generation and feature analysis to algorithm comparison and control-oriented application. It can therefore improve the students' engineering cognition, ability to evaluate algorithms, and comprehensive problem-solving ability, and it can provide a useful reference for reforming the teaching of experimental methods in related courses such as Power system automation, Smart grid, and New energy power generation technology.