Table of content

Abstract
1. Introduction

1.1 Objective of the Project

1.2 Context and problem

2. Literature review

Importance of Signal Analysis in Renewable Energy Systems

3. Theoretical studies

3.1 Definition of AM and FM

3.2 Examples of simulation AM and FM using scrip MATLAB

4. Data and Tools

4.1 Data collection of signals from NASA POWER

4.2 Software and Tools Used, Including MATLAB

5. Methodology

5.1 Analysis Methods

5.1.1 Frequency Analysis (FFT, Spectrograms)
5.1.2 Hilbert and Wavelet Transforms
5.2 Identification of Anomaly

6. Results Analysis

6.1 Analysis of Frequency Components

6.2 Identification of Modulation Effects (Depth, Carrier Frequencies, Sidebands)
6.3 Impact of Detected Anomalies on System Performance

7. Proposals corrective

7.1 Filtering Techniques to Reduce Modulation Artifacts

7.2 System and Equipment Parameter Optimization

8. Conclusion
 8.1 Summary of Key Observations
8.2 Strategic and Technical Recommendations
8.3 Future Research Directions
9. References

1. Introduction
 1.1 objectives of the project :

Project Objectives:

 Study the effects of modulations (AM and FM) on the signals of renewable energy
systems, such as solar irradiance and wind speed
 Identify and diagnose potential anomalies or failures using frequency and time-domain
analysis tools.
 Propose corrective measures to mitigate the negative impacts of modulations on the
quality of the produced energy.

Scope of the Report: This report presents:

 An introduction to modulation techniques and their relevance for diagnosing

renewable energy systems
 A detailed methodology for simulating and analysing modulated energy signals using
MATLAB.
 An analysis of the results obtained from simulations and real data.
 Recommendations for improving renewable energy systems through the application of
analysis results.

1.2 Context and problem

Renewable energy systems, like solar panels and wind turbines, are key to advancing the
global energy transition. But these systems face a lot of challenges because of changing
environmental factors, such as shifts in solar irradiance, wind speed and direction, or even
unexpected weather. These changes cause fluctuations in energy signals, like variations in
power output, which makes detecting anomalies harder.

On top of that, the equipment used to convert natural energy into electricity—like
inverters or turbine blades—can wear out or fail over time. If these problems aren’t spotted
early, they can lower the quality of the energy produced, lead to financial losses, and even
affect the reliability of the energy grid.

That’s why it’s so important to analyse renewable energy signals, especially by looking at
modulations like amplitude modulation (AM) and frequency modulation (FM). This kind of
analysis helps improve monitoring, troubleshooting, and maintaining these systems.

2. Literature review

The Modulation signals is essential technique used in engineering to transmit information

via a carrier wave. Basically, it consists of changing certain properties of this wave — such as
its amplitude or frequency, in order to integrate the information signal. This process is found
in fields such as radio, television and telecommunications, because it improves the
transmission and quality of the signal.

There are many types of modulation, two are particularly well-known and often used:
Amplitude modulation (AM) where this method is used in signal transmission. It change the
amplitude of the carrier wave varies according to the information signal.
Frequency modulation (FM) where this method used the frequency of the wave that changes
according to the signal. This technique is widely used in radios. Why? Because it is much
more resistant to interference, thus providing clearer and better quality sound. Once we
finished the modulation we need to analyse this signals with the FFT analysis, spectrogram,
and Hilbert transforms used matlab. It is important for having the information about the state
of renewable energy systems. It’s a bit like listening to the heartbeat of a machine to make
sure if it’s working properly. It helps to detect faults before they become serious. If we can
identify the problem by observing certain signal patterns, we can act quickly and prevent
damage. It also improves equipment performance since understanding the effects of signals
allows us to fine-tune settings for greater efficiency. Signal analysis can also predict failures
caused by weather conditions or mechanical vibrations.
Finally, we interpret and propose some corrective, and we ensure that renewable energy
integrates into the grid in a stable and reliable way. In summary, it’s an indispensable tool for
maximizing the potential of renewable resources while avoiding unexpected interruptions. [1]
[2]

3. Theoretical studies

3.1 Definition of AM and FM

But before we start the definition of AM and FM let understand the word modulation.
Modulation is a fundamental engineering technique for transmitting information using a
carrier wave which have a high-frequency signal that serves for the data to be transmitted.
This process consists of modifying certain properties of the carrier wave, such as the
amplitude, frequency or phase, in order to integrate the information signal into it. Modulation
methods are widely used in fields such as telecommunications, radio and television to
improve the efficiency and quality of signal transmission. [1]
Here are the two most commonly used types of modulation:
a) Amplitude modulation (AM)
Amplitude modulation (AM) consist of varying the amplitude of a carrier wave in response to
variations in the information signal. This method, which is simple to implement, is commonly
used in AM broadcasting [1]. The basic mathematical expression for an amplitude modulated
signal is:
To better understand the concept of amplitude modulation, consider an example with a
sinusoidal modulating signal defined by m(t) = 0.5 × cos (2π × 1000t), a carrier frequency fc
of 10,000 Hz, and a carrier amplitude Ac equal to 1. The corresponding modulated signal is
then expressed as:

Figure 1 : modulation AM process

b) b) Frequency modulation (FM)

Frequency modulation (FM) consists of varying the frequency of a carrier wave according to
variations in the information signal. Commonly used in FM broadcasting, this technique
offers high resistance to interference, thus ensuring superior sound quality [1]. The
mathematical expression of the frequency modulated signal is given by:

Where Kf is a constant that determines the sensitivity of the frequency to the modulating
signal.
Here is a figure of frequency modulation (FM):

Figure 2 : modulation FM process

3.2 Examples of simulation AM and FM using scrip MATLAB

Code of AM modulation signal

Figure 3 : AM modulation signal

Code of FM modulation signal

Figure 4 : AM modulation signal

4. Data and Tools

4.1 Data collection of signals from NASA POWER

This part will show how collect data of irradiance from NASA POWER

Step 1: go to the website

 Visit this website: NASA POWER Data Access Viewer.

Step 2: Select the Location

 Use the map for the location or enter the coordinates (latitude and longitude) of the
location that you want to have the data.
Step 3: touch on the single points icon
Then you fill out for
User Community = Renewable energy
Temporal Level = hourly
Time Extent = 01/01/2021: 12/01/2021
Parameters = All Sky Surface Shortwave Downward Irradiance

Step 4: Download the Data

Click the Submit button to download the data. You can then download the results in CSV,
JSON, or netCDF format.

This web can explain all details in video: https://power.larc.nasa.gov/docs/tutorials/api-

getting-started/

4.2 Software and Tools Used, Including MATLAB

MATLAB is a language for computing that is very efficient. It combines computing, graphics,
and programming in an environment and easy to use, where problems and solutions are
written with mathematical notations familiar. It is often used for:

Math and computing

Making algorithms
Modeling, simulation, and prototypes
Analyzing and visualizing data
Scientific and engineering graphics
Developing applications and building user interfaces
In MATLAB, the array is the basic data element, and it does not need defining dimensions.
This makes it fast to solve many computing problems, especially with matrices and vectors,
faster than writing a program in languages like C or Fortran.

The name MATLAB means "matrix laboratory." It was first made to give easy access to
matrix software from LINPACK and EISPACK projects, which are the best tools for matrix
computing.

MATLAB changed over years with many users' ideas. In universities, it is a standard tool for
teaching math, engineering, and science, while in industries, it is used for research,
development, and analysis.

MATLAB has special packages called toolboxes. Toolboxes are big sets of functions that let
you learn and use special technologies. They help with signal processing, control systems,
neural networks, fuzzy logic, wavelets, simulation, and more. [3]

The goal of this project is not to explain Matlab in detail but is for analysing the signals.

5. Methodology

5.1 Analysis Methods

5.1.1 Frequency Analysis (FFT, Spectrograms)

a) Fast Fourier Transform

The Fast Fourier Transform, or FFT, is a method for quickly computing the Discrete Fourier
Transform (DFT) of a data sequence. In simple terms, the Fourier Transform is used to
convert a signal from a time (or space) domain representation to a frequency domain
representation, and vice versa. The DFT decomposes a signal into its frequency components.
However, if computed directly, it takes a lot of time, especially with large amounts of data.
The FFT, by using mathematical shortcuts, reduces the computation time. Instead of taking a
time proportional to n², it reduces the effort to n log(n), where n is the size of the data. This
time saving is immense when n is very large.

FFT algorithms are also more accurate in many cases, even in the presence of rounding errors,
compared to traditional DFT calculations. There are several types of algorithms based on
concepts ranging from simple computations with complex numbers to more advanced ideas in
number theory and group theory.
The FFT is used in fields such as engineering, music, science, and mathematics. Although it
became popular in 1965, similar work has been done since the early 19th century. Professor
Gilbert Strang once claimed that the FFT is "the most important numerical algorithm of our
time," and it was ranked among the top ten algorithmic inventions of the 20th century by the
IEEE.

The most well-known FFT algorithms often depend on how n can be factored, but there are
also fast versions for any data size, even when n is a prime number. These algorithms take
advantage of the fact that exp(-2pi*i/n) is a primitive root of unity, making it possible to adapt
the FFT to other mathematical fields. Since the inverse Fourier transform is almost the same
as the DFT, one can easily adapt an FFT to perform the inverse transform by changing a few
details [4]
.
Imagine you have a list of numbers x0,x1,...,xn−1x_0, x_1, ..., x_{n-1}x0,x1,...,xn−1. The
Discrete Fourier Transform (DFT) turns these numbers into a frequency-based version using
the formula:

where k goes from 0 to n−1, and exp (-i* (2*pi*k*m/n)) is a special number called a root of
unity.

If you try to calculate this directly, it takes a lot of time because you need to do n² steps — n
sums, and each sum needs n terms. This is too slow for large sets of data. The Fast Fourier
Transform (FFT) is a clever shortcut that reduces the work to only n log(n) steps, making it
much faster. Every known FFT method uses n log(n) operations, but no one has proved it’s
impossible to do better.

To see the difference, let’s say you have 4096 points. Computing the DFT directly would
require millions of operations — about 30 million multiplications and additions. But with the
FFT method, like the Cooley-Tukey algorithm (used when n is a power of 2), you only need
around 30,000 steps. That’s about 1000 times fewer calculations! Even though real-world
performance depends on more than just the number of operations, the speed boost from n² to
n log(n) makes FFT extremely valuable.

Example: Imagine you have a signal made by adding two sine waves together:

If one wave has a frequency of 50 Hz and the other is 150 Hz, running an FFT on this signal
will show two spikes in the frequency plot — one at 50 Hz and another at 150 Hz. This tells
you those are the main frequencies in the signal
Here is an example of using scrip MATLAB:
Here is the figure of the result:
 b) Spectrogram

A spectrogram is the visual representation of how

the signal power contained within a particular frequency range changes with time.
When applied to an audio signal, it is sometimes called a sonograph, voiceprint, or voicegram.
If the data is plotted in a 3-Dimensional plot, it is often called a waterfall display.

The main uses of spectrograms are in the analysis of music, linguistics, sonar, radar, speech
processing, seismology, and ornithology. In audio analysis, they help identify spoken words
phonetically and study different animal calls.

There are various methods for generating a spectrogram, including through an optical
spectrometer, a group of band-pass filters, or mathematical transforms such as the Fourier
transform or wavelet transform. For wavelet transforms,
the resultant spectrogram is sometimes termed a scaleogram or scalogram.

Typically, a spectrogram is shown as a heat map in which the intensity is varied via color or
brightness. [5]
To generate the spectrogram we use the Short-Time Fourier Transform (STFT).
The STFT is a technique of study for non-stationary signals, i.e., signals whose characteristics
vary with time. It allows the visualization of how a signal often changes with time by dividing
it into small time windows and applying the Fourier transform to each window [6].

The STFT of a signal x(t) can be expressed as follows [7] :

Here is one example of using MATLAB:

Here is the result:

Now we will use the data taken by NASA POWER and apply these FFT, Spectrogram and
STFT analysis methods:
The result gives the following figure:
 5.2 Hilbert and Wavelet Transforms

a) Hilbert transform
The Hilbert transform is a mathematical operation that takes a real-valued function and
produces another real-valued function. It essentially to shifts the original function in the
frequency domain. It finds in applications such as signal processing and
communications, especially in analytic signal representation and phase shift calculations [8].
The formula of the Hilbert transform is give as:

Where p.v is stands for Cauchy principal value, it is used to handle the singularity when
𝑡=𝜏.
Example used MATLAB:

The result:
 b) Wavelet Transforms

Wavelet transforms are mathematical tools used to analyze signals either at different scales or
resolutions. Unlike traditional Fourier transforms, which represent a signal as a sum of
sinusoids, wavelet transforms decompose the signal into functions called
wavelets, each having frequency and time localization properties. This thus tends to
make wavelet transforms much more effective in analyzing non-stationary signals-
those whose statistics vary with time-since both time and frequency information is conveyed
[9].
There are two type of wavelet transform:
Continuous Wavelet Transform (CWT): It is a technique that offers continuous
decomposition of the signal at different scales (frequencies) and positions (time) [10].
Discrete Wavelet Transform (DWT): This provides a discrete set of wavelet
coefficients using specific scales and positions, normally using dyadic or power-of-
two scaling and shifting [10].
The expression of the wavelet transforms is given as [11]:

Example:

The result of the code:

Now here is a general example which brings together all the analysis methods such as: FFT,
STFT, Spectrogram, hilbert and wavelet transform:
Here are the figures for this code:
 6. Results Analysis

6.1 Frequency Components Analysis

The frequency components of the analyzed signals will provide a more detailed understanding
of the dominant periodicities and possible anomalies in the data of renewable energy
systems. The Fast Fourier Transform allowed for the decomposition of the signals into
their basic frequency components, hence underlining both regular patterns and irregular
disturbances.

Another important observation is the dominant frequencies, corresponding to natural

cycles of day and seasonal variation. As expected, the principal frequency found in the time
series of solar irradiance coincides with a 24-hour cycle that repeats each day. It expresses the
Earth's rotation and points
to the consistency with which the change from night to day always occurs. This predictability
enables energy systems to follow rhythms in nature.

Apart from the dominant frequencies, other harmonics were identified that showed secondary
peaks in the frequency spectrum. These usually originate from higher-order periodicities and
can indicate further complicating behaviors of the systems, such
as could be engendered by special weather conditions or even mechanical vibration. For
example, irregularities in weather can bring harmonic components
that affect efficiency in energy capture and conversion.

Noise characteristics, predominantly in the high-frequency band, were also presented in the
frequency analysis. These components are usually due to sensor inaccuracies or external
environmental interference, such as electrical noise or mechanical vibrations. Understanding
these noise elements is indispensable for the refinement of algorithms in signal processing and
improving the accuracy of data.

Dissecting the frequency spectrum has shown that natural environmental cycles, as well
as technical system factors, play their part in setting the signal. Distinguishing these elements
is crucial to the identification of anomalies and to guarantee optimal performance of
renewable energy systems.

6.2 Modulation Effects and Its Implications (Depth, Carrier Frequencies

and Sidebands)
The application of Amplitude and FM analyses on the received signals narratives elaborated
stories. Such techniques also made it much easier to define modulation characteristics which
are valuable in diagnostics and improving the robustness of the system.
Deepening the analysis, one of the parameters, which were amended was the Depth of
Modulation in which amplitude of the carrier signal is determined. In using Hilbert’s
transform, periodic amplitude oscillations are displayed and calibrated. Such oscillations
which are often referred to as AM effects are quite common and can be associated with
environmental oscillations like temperature changes in solar systems or wind systems
turbulence.
Also the carrier frequencies were still another major area of concern. Seeming as the
parameters, these frequencies which are the signal constituents pivotal frequencies mark the
baseline range of changes that can be reflected or identified in modulated systems. Changes in
the carrier frequencies were seen as possible indicators of estimated mechanical/electrical
defects for variation from the normal may imply some fault. For example changing of the
carrier frequency patterns makes selection components in the system which were considered
not functional, available
It also detected the presence of sidebands in the frequency spectrum, typical for FM
effects: such frequency variations around the
carrier generate this kind of sidebands, which do carry information relative to frequency and
amplitude of the modulating signal. Too much, or asymmetrical, sidebands may indicate some
anomalies, such as a mechanical imbalance, or environmental perturbations acting on the
system under operation.

Overall, modulation effect identification indicates that both the external

environmental factor and the dynamics within the system have a strong impact on
the performances of renewable energy systems. These modulations, if not attended
to, may worsen and deteriorate into low-quality, inconsistent energy output
hence, there is a need for early detection and intervention.

6.3 Impact of Detected Anomalies on System Performance

The detected anomalies, both in frequency irregularities and modulation effects, have
profound implications for general performance and system reliability of renewable energy
systems. These can be summarized under three broad headings: energy output variability,
efficiency losses, and equipment wear and tear.

Energy Output Variability: Variations in signal amplitude and frequency would directly
impact the steadiness of energy output. Most renewable energy systems rely on steady and
predictable input signals, like solar and wind. When such a signal becomes erratic due to
anomalies, the output becomes unreliable and can thus threaten grid stability and energy
storage systems.

Losses of Efficiency: Generally, modulations and high-frequency noise components result in

system inefficiency. Inverters, for instance, and other conversion apparatus have been
designed to realize peak operational parameters under certain specific constant conditions.
Distortion to the wave forces these devices into abnormal operating states where efficiency
goes low and the percentage of losses goes high.

Wear and Tear of Equipment: Long-term anomalies, such as unstable carrier frequencies or
high levels of sidebands, constitute further mechanical and electrical stresses. This could lead
to faster deterioration in the most important equipment over time, with higher maintenance
costs and system downtimes. For example, mechanical vibrations resulting from irregular
modulations may weaken turbine blades or solar tracking systems.

7. Proposals corrective
7.1 Filtering Techniques to Reduce Modulation Artifacts

The filtering techniques play a vital role in the attenuation of adverse effects from modulation
artifacts in renewable energy systems. The modulation artifacts due to external disturbances
or due to system anomalies can seriously deteriorate the quality of the signals and affect
system performance. Several filtering methods can be implemented for this purpose.

It could be done with the help of low-pass filters that serve to remove high-frequency
components of noise that do not take part in the main operation of the system. They work to
suppress those noise elements and allow only relevant frequency components to pass through,
therefore enhancing clarity and reliability in a signal.

Adaptive filtering is yet another sophisticated technique whereby filter parameters are varied
dynamically depending on the characteristics of the signal. Adaptive filters, in particular, are
very much useful in renewable energy systems because environmental conditions and
operational parameters vary considerably with time. Such filters can track the changes in the
signal and adapt accordingly to provide real-time noise reduction and enhancement of the
signal.

Wavelet-based filtering is a powerful solution in the case of more complex modulation

artifacts. The wavelet transforms decompose the signal into different frequency bands, which
enables the identification and suppression of specific artifacts while keeping the necessary
information. In particular, this is very effective in the analysis of non-stationary signals, for
example, given by renewable energy systems, since the characteristics of the signal change
with time.

This filtering incorporated into the signal processing pipeline can contribute to a reduction of
modulation artifacts, but it also increases the general efficiency and reliability of the energy
system. By cleaning and making the signals stable, these systems ensure high performance
and are less in need of repair or maintenance. [12]

7.2 Optimization of System and Equipment Parameters

Parametric optimization regarding systems and equipment secures the reduction in

singularities, especially concerning energy return in renewable source conversion systems,
assuring that such is kept within an optimum limit with proper adjustment of the same; hence
reduced failures and more productivity.
One of the critical areas of optimization involves the settings of inverters. Inverters are crucial
in the conversion of energy from renewable sources into usable electricity. Inverters can be
set up to operate more efficiently and handle modulations effectively by fine-tuning
parameters such as voltage, current, and frequency settings. For example, setting appropriate
thresholds for frequency variations helps prevent the propagation of anomalies through the
system.

Another key aspect is sensor and actuator calibration. Because of this, sensor readings are
necessary to monitor system performance and detect anomalies. Periodic recalibration will
ensure that sensors give good data to avoid false alarms and missed detections. The
optimization of actuator settings can make a system more responsive to changes in conditions,
thus making it adaptive by increasing stability.

It also contributes immensely in the load management strategies that go a long way in system
optimization. The balance of energy load within the system enables the smoothening of
operation, hence reducing the stress on individual elements. This mechanism may, for
instance, be applied in cases of peak demand to protect against overloading that might result
in critical component failure.

The bottom line is that system and equipment parameter optimization proactively maintains
the health and efficiency of renewable energy systems. Periodic reviews of parameters ensure
that operators will address possible issues before they arise, thus guaranteeing the reliability
and performance of the systems over long-term periods. [13]

8. Conclusion

8.1 Overview of Key Findings

The analysis done in this project, therefore, served to outline

the critical features in the behavior of signals related to renewable energy
systems. Main findings: it identifies the dominant frequency components related to natural
cycles of changes, like daily and seasonal variations; it also detects harmonics and noise that
influence the clarity of the signals. Furthermore, modulation effects-mostly amplitude and
frequency modulation-already represented an essential contribution to anomalies, which could
provide
insight into external and internal influences due to environmental interaction with the system.

It also pointed out that such anomalies may have a big impact on system performance, noting
variability in energy output, losses of efficiency, and increased wear and tear on
equipment. Results indicated that signal analysis will be key in diagnosing and
mitigating problems that might arise to enhance reliability and efficiency in renewable energy
systems.
8.2 Strategic and Technical Recommendations

The following strategic and technical recommendations are advanced to overcome the
identified challenges, based on observations and analyses provided:

Improved filtering: Advanced filtering techniques should be used at the front end, such as
wavelet-based filters and adaptive noise reduction methods, to minimize the effects of noise
and modulation artifacts on the performance of the system.

Periodic System Optimization: The system parameters, such as inverter settings and sensor
calibration, are constantly checked for optimization to ensure stability of operations.

Predictive Maintenance: Integrate predictive models that will monitor, in real time, the
performance of the systems to predict failure and allow maintenance interventions in order to
reduce unplanned downtime.

Data-driven Decision Making: Operational strategies are driven by historical and real-time
data in order to make systems resilient against foreseen and unforeseen variations.

8.3 Future Research Directions

While the project provided insight into the behavior of the signal and its impact on systems of
renewable energy, further research may be considered in the following aspects:

Higher-Order Modulation Analysis: The study of the impact of higher-order modulation

techniques on system performance, especially in extreme environmental conditions.

Integration of AI and Machine Learning: Examine the usage of advanced machine learning
algorithms for anomaly detection, predictive maintenance, and performance optimization
across renewable energy systems.

Real-Time Signal Processing: The development and implementation of real-time signal

processing techniques should be able to respond to changing conditions and provide direct
feedback for immediate system adjustment.
Cross-System Interactions: Interaction between different renewable energy systems, such as
solar and wind, showing how combined signal behaviors drive overall grid stability.

The focus of this project is such research areas; they could be a very good basis for further
research in the field of renewable energy technologies. This way, more contributions will be
created in enhancing the performance of systems and enabling wider diffusion of sustainable
energy solutions all over the world.