The reorganization and restructuring of power supply control and management, as wellas theconditions for their functioning, show their own complex features and problems. New algorithms, devices and methods are needed to ensure reliable electronicлектроsupply to consumers, involve modern information and diagnostic elements and devices and systems in the process of monitoring and managing operating modes. It is necessary to develop and apply new energy-efficient equipment and new technologies that reduce technical and economic indicators in the production and transmission of electricity, reduce the level of losses during transportation, optimize the size and location of reserve capacities of electricity sources [1-3].
As the analysis showed, in the last decade, the world has been developing Smart Grid technology (intelligent network). The existing “Smart grid” is a very large-scale direction in the modern energy industry. “Smart Grid” is the process of implementing ” smart solutions» generation, transmission and distribution of electric energy, saturation of the electric grid with modern diagnostic tools, electronic monitoring and control systems, algorithms, technical devices that have appeared today in science and technology, i.e. wide application of information technology capabilities [4-7],
Internet, digital technology (IoT) with power electrical engineering. And this reduces losses in the transmission of electrical energy from the generator to the consumer, increases the reliability of power supply, makes it possible to optimally redistribute energy flows and thereby reduce peak loads, makes it possible for the consumer to work in a unified powersupply system. Traditionally, the consumer received electrical energy from a single source, but now it is in a centralizedenvironment : можетit can choose the source among generating мощunits and sources [7-10].

The main feature in the “Smart Grid” is related to so-called renewable energy sources. To connect renewable energy sources to a large electricity supply system and make them as manageable as other sources, we need these “smart grids”Smart Grid. At the scale of power supply devices and systems, we need backbone or distribution networks that can monitor the state and mode of operation of consumers, sources, electrical lines and substations and automatically implement solutions that allow uninterrupted power supply and maximum economic efficiency. The “smart grid” itself must form a controlling influence with the achievement of an optimal level of electricity losses when the flow of electricity along transmission lines increases due to an increase in consumption by any consumer. In General, we are talking about creating a so-called intelligent EUSTC with an active adaptive network (IEUSTC), which means a power supply system in which all actors (generation, network, consumers) take an active part in the transmission and distribution of electricity. In this case, the electrical network turns into an active element, the parameters and characteristics of which change in real time depending on the modes of operation of consumers. To implement the new function, electric networks are equipped with modern high-speed electronic devices, systems that provide on-line information about the modes of operation of the network and the state of electrical equipment. Telecommunications devices and accumulators of electrical energy available in electric power supply networks ensure the process of distribution and consumption of electricity. Power supply systems are equipped with modern automation systems using powerful computer toolsforcontrolling and evaluating the state of operating modes [11-12].

Modern equipment is a complex energy of a new generation based on multi-agent principle, the organization and management of its functioning and development to ensure effective use of all resources (natural, socio-productive and human) for reliable, high-quality and effective elektrosnabzheniya of consumers due to the flexible interaction of all its subjects (all types of generation, power grids and consumers) on the basis of modern technological means and a single hierarchical intelligent control system [12-14].

In accordance with the modern requirements of power supply of electrical equipment and consumers, an algorithm for adaptive power management has been developed based on a simulation model (Fig. 1).
The algorithm for adaptive control of electricity supply while ensuring continuous energy transfer is based on the energy balance equation [5]:

where P_CES(t), P_SES(t), P_WPP(t), P_DG(t), P_AB(t) - the values of power produced by sources of centralized electricity supply – (CES), solar power – (SES), wind power – (WES), diesel generators – (DG), batteries – (AB); P_n(t) - value of the power of the electrical load.
This algorithm controls the power supply sources and consumers by a microcontroller control unit based on the signals transmitted to the monitoring server about the amount of energy generated by the sources and consumed by the load, as well as the state of charge of batteries and the duration of use of sources Fig.1). [13]:
Monitoring data is collected and processed according to the scheme shown in Fig. 2. With this monitoring data accumulated in the server database monitoring and, if necessary, provided to the service personnel via an Ethernet Protocol in a web page on the Internet or the GSM module as SMS messages in the required formats.

As the analysis of electrical power networks and consumers, managed cosine of the capacitor unit are the main elements of the construction technology of “Smart grid” depending on the current electric networks, the power source of reactive power is proportional to the square of the voltage, frequency and capacity [2,4]:
 (1)
where:- reactive power of the condenser unit;
- electrical network voltage;
is the angular frequency;
C is the capacitance of the condenser unit.
The use of embedded microcomputers in the microprocessor source control unit of the Smart power engineering technology, depending on the load current, makes it possible to reduce damage from damage to electrical and power equipment and improve the quality of electricity generated [4-8].
For example, we will determine additional losses of active power ΔP in power supply system cable lines (CL1 +CL) 400 m long with a cross section^{of 50 mm2} [14]:
Let’s say that the power supply system had loads:
,
load factor
,
maximum power loss time: t =5000 h.
After applying smart technology, the object’s load will have the following values: .
The line current is defined as follows:

 (2)

 (3)

Additional power losses in the high-voltage cable (CL₈):

 (4)

Additional TC power losses depend on its load losses:

 (5)

Total power losses:

 (6)

Energy savings for the year will amount to:

 (7)
The increase in the capacity of the TC line can be accounted for by the corresponding shares of their cost.
For a vehicle power transformer:

 (8)

For cables with a long current tolerance:

 (9)

Payback period of the proposed technology:

 (10)

The performance indicator for this event is defined as follows:

 (11)

The developed algorithm and methodology for calculating technical and economic indicators gives a pessimistic payback period for the use of smart technology in power supply systems. The obtained value of the payback period, due to improving the quality of electricity – providing rated voltage in power consumption units (i.e., increasing the service life of electrical equipment, reducing power losses in electrical networks, etc.), actually turns out to be less than its normative value (T_{OK norms} = 8 years).

A. Parameters of the regulated reactive power compensation technology (number and power of control stages) – are determined by the daily schedule of electricity consumption by electric receivers.
B. Technology of controlling reactive power sources and voltage regulation is effective when switching on for a large inductive resistance of step-down transformers of power supply systems.
C. To change the voltage by one percent of the rated value, it is necessary for the transformer to change the reactive power to , for the transformer, for the cable line length, for the cable line length .
As shown by the research, the proposed method of applying smart technology for controlling reactive power sources and microprocessors allows to reduce the payback period of the implemented technology and control elements of electricity consumption by 28.7% and increase the efficiency of energy-saving measures implemented in power supply systems
D. the monitoring System makes it possible to assess their operational characteristics, determine the repair needs, identify the causes of failure and promptly eliminate them.
The use of monitoring systems allows you to increase reliability through high-quality maintenance by reducing the time of prevention, repair, recovery and downtime.
The developed algorithm and sensors for controlling electricity sources and consumers, as well as the use of monitoring devices, allow for continuous power supply, adaptability to the power sources of the control unit, and a user-friendly control interface allows for high accuracy and efficiency of management.

Problem Statement

Urban environments currently account for over 70% of global $CO_2$ emissions, with building energy consumption being a primary contributor. Traditional grids are “one-way” systems that lack the flexibility to handle the intermittent nature of renewable energy (solar/wind).

Motivation

The transition to “Smart Cities” requires a “two-way” energy flow where consumption is monitored and managed in real-time. Machine Learning (ML) acts as the brain of this system, allowing utilities to move from reactive to proactive management[2].

Objectives

• To develop a robust ML-based forecasting model for urban energy loads.
• To analyze the impact of “Peak Shaving” on grid stability.
• To provide a roadmap for policy-makers to integrate AI into national energy infrastructures.

Literature Review

Previous research in energy forecasting relied heavily on Autoregressive Integrated Moving Average (ARIMA) models. However, these models struggle with the non-linear, stochastic nature of modern energy usage. Recent shifts toward Deep Learning, particularly Recurrent Neural Networks (RNNs), have shown promise in handling time-series data. This paper builds upon these advancements by introducing a multi-layer LSTM architecture optimized for high-frequency IoT data.

Methodology

The research follows a data-driven approach, utilizing a five-stage pipeline: Data Collection, Pre-processing, Feature Engineering, Model Training, and Evaluation[3]..

Data Source and Pre-processing

Data was synthesized from a metropolitan smart meter dataset, including:

• Time-series consumption: Hourly kWh data.
• Weather variables: Temperature, humidity, and solar radiation.
• Social indicators: Holidays and weekend markers.

The LSTM Architecture

We employ an LSTM network because of its “memory cells” that can retain information over long periods, making it ideal for detecting weekly or seasonal energy cycles.

The mathematical core of the LSTM cell involves the Forget Gate:

ft = σ (Wf ⋅ [ht−1, xt] + bf) Input gate: it = σ (Wi ⋅ [ht−1, xt] + bi) Cell state candidate: C~t = tanh⁡(Wc ⋅ [ht − 1, xt] + bc)  [4].

Proposed Framework

The system architecture consists of a decentralized “Cloud-Edge” computing model. Energy data is processed at the “Edge” (smart meters) for immediate response, while heavy ML training occurs in the “Cloud.”

Table 1: Hyperparameter Configuration for the Proposed Model[4].

Results and Discussion

Performance Metrics

The model was evaluated using Root Mean Square Error (RMSE) and Mean Absolute Error (MAE).

Impact on Sustainability

Decarbonization and Carbon Footprint Reduction

The primary environmental benefit of ML-optimized grids is the direct reduction of greenhouse gas (CO_2) emissions. Traditional grids often rely on “Peaker Plants”—typically carbon-intensive gas or coal plants—to meet sudden spikes in demand.

• Predictive Peak Shaving: By using LSTM and other deep learning models, utilities can predict these spikes 24–48 hours in advance.
• Emission Mitigation: This allows for “Demand Response” strategies, where non-essential loads (e.g., industrial cooling, EV charging) are shifted to off-peak hours, preventing the need to activate high-emission plants[4].

Seamless Integration of Renewable Energy Sources (RES)

One of the greatest barriers to a 100% renewable grid is intermittency (e.g., the sun doesn’t always shine, and the wind doesn’t always blow).

• Supply-Demand Balancing: ML algorithms analyze meteorological data to forecast renewable output. When a drop in solar production is predicted, the ML controller automatically scales up stored battery energy or adjusts urban consumption patterns[8]..
• Reduced Curtailment: Optimization ensures that excess green energy produced during the day isn’t “wasted” (curtailed) but is instead stored or redirected efficiently.

Infrastructure Longevity and Circularity

Sustainability also implies the efficient use of physical resources. Overloading transformers and substations during peak hours accelerates equipment degradation.

• Predictive Maintenance: ML models detect thermal anomalies and stress patterns in grid hardware.
• Resource Conservation: By extending the lifespan of grid infrastructure by even 15–20%, cities significantly reduce the environmental cost associated with manufacturing and installing new industrial-grade components (steel, copper, and SF6 insulators) [5]..

Quantitative Sustainability Metrics

The following table illustrates the projected sustainability gains from implementing the proposed ML framework in a mid-sized metropolitan area[5].

Socio-Economic Sustainability

Beyond the environment, the ML-driven approach fosters Energy Equity:

1. Lower Costs: Reduced operational waste leads to lower utility bills for urban residents, particularly benefiting low-income households.
2. Resilience: ML-enhanced grids are more resistant to “cascading failures” and blackouts, ensuring that critical urban services (hospitals, water treatment) remain operational during extreme weather events[7]..

 Conclusion

This research demonstrates that Machine Learning is the cornerstone of modern sustainable urban development. The proposed LSTM framework provides a reliable, scalable method for energy optimization. For nations like the USA, adopting these AI-driven grid technologies is essential for meeting “Net-Zero” targets and ensuring energy independence.

Future Work

While the proposed LSTM-based framework significantly improves energy forecasting and grid stability, several avenues for future research remain open to further enhance the sustainability of urban power systems.

Federated Learning for Enhanced Data Privacy

A primary concern for the widespread adoption of AI in Smart Grids is the privacy of consumer data. Future iterations of this research will explore Federated Learning (FL). Unlike centralized ML, FL allows the model to be trained across multiple decentralized edge devices (smart meters) without exchanging local data samples. This “Privacy-by-Design” approach ensures compliance with strict data protection regulations while still benefiting from collective intelligence.

Multi-Agent Reinforcement Learning (MARL)

As cities transition toward decentralized microgrids, the complexity of energy balancing increases. Future work will investigate Multi-Agent Reinforcement Learning (MARL), where individual AI agents represent different grid components (e.g., a solar farm, a battery storage facility, and a residential block). These agents can learn to cooperate in real-time to maintain grid frequency and voltage stability, even during extreme weather fluctuations.

Vehicle-to-Grid (V2G) Integration

The rise of electric vehicles (EVs) presents both a challenge and an opportunity. We aim to integrate V2G (Vehicle-to-Grid) protocols into the optimization framework. By treating parked EVs as mobile energy storage units, the ML model can “borrow” energy from vehicle batteries during peak demand and “repay” it during periods of high renewable generation, creating a truly circular energy economy.

Resilience Against Adversarial Attacks

As power grids become increasingly software-defined, they become vulnerable to cyber-physical threats. A crucial direction for future research is the development of Adversarial Machine Learning techniques to detect and mitigate malicious data injection. Ensuring that the AI “brain” of the grid can distinguish between a natural surge in demand and a coordinated cyber-attack is paramount for national energy security.

Research Roadmap (2026–2030)