The master's thesis addresses the design of an energy consumption management system for a single-family house equipped with a photovoltaic power plant, battery storage system, heat pump, and electric vehicle charging.

During the preparation of the thesis, the student worked independently and regularly consulted the progress of the work. The submitted thesis follows the assignment and covers all required tasks.

The theoretical part provides an overview of residential energy management systems and selected technologies used in practice. However, some sections are descriptive in nature and their connection to the proposed solution could have been explained in greater detail. The analytical part would also benefit from a more critical comparison of available approaches and technologies.

The student developed a simulation model of the reference household in MATLAB/Simulink and implemented a rule-based control strategy. The model demonstrates the basic functionality of the proposed concept and allows the evaluation of selected operating scenarios. Nevertheless, the adopted control approach remains relatively simple and does not fully exploit the possibilities of advanced energy management methods. The justification of selected control priorities and decision rules could have been elaborated more thoroughly.

The evaluation of results is rather concise and focuses primarily on demonstrating the functionality of the model. A more detailed analysis of the obtained results, including a broader energy-economic assessment, sensitivity analysis, or comparison with alternative control strategies, would significantly increase the practical value of the work.

From a formal point of view, the thesis contains a number of minor shortcomings in presentation, references, and overall consistency of the text. These issues do not prevent understanding the work but reduce its overall quality.

The thesis fulfills the basic objectives defined in the assignment and demonstrates the student's ability to create and verify a simulation model of a residential energy management system. However, due to the limited depth of analysis, relatively simple control methodology, and less comprehensive evaluation of results, I assess the overall level of the thesis as average.

I recommend the master's thesis for defense. Points proposed by supervisor: 65

Student focused on the design of control system for energy consumtion management in a family house. The topic is highly relevant as the energy prices fluctuate but the energy needs are unlikely to decrease.

First part of the thesis is an overview of the current energy consumtion management technologies (or rather services) and tariff structure in Czechia. Chepters 3 and 4 are unecesarily long and go in too much detail without purpose linked to the rest of the thesis. E.g. there is no need to describe the communication protocols and structure of the services if they are not used in design of the system.

Chapter 5 describes the reference house model. But it lacks to provide any source. The overview of household loads provide some model consumpion values without any backing or reason. The overall load profile fails to provide context on how the loads are triggered. The PV production profile is also without context on how it was generated in PVGIS and doesn't state the PVGIS source explicitely in references.

Chapter 6 describes the goal of the thesis - the simulation model and energy management alghorithm. The models are described clearly. The control alghorithm however is only stated without reasoning for the chosen conditions. Figure 16 doesn't make much sense, the syntax in the boxes is nonsensical. The rest of the chapter describes the simulink model in detail. The thesis fails to provide a manual on how to run the simulation.

Chapter 7 summarizes the simulation results but without comparison, recommendations or detailed energy-economic assessment. The evaluation of control alghorithm functionality is lacking as well. This chapter should have been prepared with more detail.

Matlab codes are clearly LLM generated due to various comments and status messages throughout. However, the model is running and is prepared with with appropriate detail and care for trying out energy managements solutions.

The formal editing of the text and citations is on the verge of satisfactory. The student failed to correctly cite majority of the sources. All of the sources are simply listed at the end of section or paragraph without clear distinction of what is referenced from where. There are multiple "Error! Reference source not found." where the text was edited but not checked afterwards. The figures used in the review part are all taken from an external source without overview of all of the abbreviations used in the pictures. The units are in brackets for most of the text but correctly in parentheses in the symbol list. Custom figures are blurry and hard to read (e.g. fig. 16).

As for the use of AI, student failed to declare how he used AI in the thesis. The sentence and naming structure clearly display the signs of LLM usage. That is true for all chapters. Majority of the review part looks like a response to a prompt as the sentence structure and section structure is too stiff for natural language. Also the lack of typing errors suggests LLM usage.
E.g. on page 56, the scenarios overview is clear naming scheme provided by LLM.

Furthermore, in one of the references ([13]), utm for chatgpt was left in the URL. I want to state that the use of a LLM is not forbidden or discouraged but the usage shall be marked clearly.

Lastly, I recommend the thesis for defense. Topics for thesis defence:

1. State the usage of LLM throughout the thesis - text generation, source research, load profile generation, code generation.
2. State the reason and feasibility of connecting house loads and domestic EV charger to single DC bus. Compare losses to an AC link.
3. State the sources for chapter 5 load profile.
4. Explain why the conditions and the control priority (Tab. 6) were chosen as such. Describe the structure of alghorithm based on the diagram in Figure 16.

Points proposed by reviewer: 50