Demand Charge Management: Prototype Design and Testing

Christian Escarrega; Mohammad Rasoul Narimani; Mohammadreza Iranpour; Sammy Hamed; Wyatt Lopez Coggins

REVIEW 4 major objections 5 minor 33 references

Demand Charge Management: Prototype Design and Testing

T0 review · 4 major / 5 minor · reviewed 2026-08-04 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims a $1,800 prototype cuts electricity costs by switching to battery during peak loads and recharging off-peak, with real-world tests confirming the approach.

desk verdict A well-documented low-cost EMS build with a modest threshold controller, but the validation is undermined by an energy-accounting error that makes the headline savings claim unsupported. read the letter →

arxiv 2509.10713 v1 pith:5LHYMQK5 submitted 2025-09-12 math.OC

Christian Escarrega , Wyatt Lopez Coggins , Sammy Hamed , Mohammadreza Iranpour , Mohammad Rasoul Narimani This is my paper · ORCID

classification math.OC

keywords DemandChargeManagementEnergySystemPeakShavingLiFePO4batteryRaspberryPiTime-of-UsepricingReal-timecontrolInternetofThings

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper sets out to prove that a low-cost, three-layer Energy Management System can reduce residential electricity bills by automatically shifting loads to a battery during peak demand. The controller switches to battery when load exceeds 700W, and charges when off-peak and state of charge is low, achieving peak shaving and energy arbitrage. Real-world component tests validate measurement accuracy, relay response, and state-of-charge tracking; a MATLAB simulation suggests the battery flattens a household load curve. If the claims hold, the system offers a fraction-of-the-cost alternative to commercial systems and a teaching platform.

What carries the argument

The load-bearing mechanism is the software control loop encapsulated in Algorithm 1: given load power, battery SoC, current time, and relay mode, it toggles relays to select grid or battery. The fixed 700W threshold is the peak-shaving trigger; the algorithm also uses SoC cutoffs (20% minimum, 100% full) and the 6 a.m. boundary to decide charging versus discharging. This logic, running as Python scripts on a Raspberry Pi 5 with MQTT communication to sensors and the Blynk interface, is what the paper argues can deliver demand-charge reduction at low cost.

What would settle it

Run the system for one billing cycle on a household with an afternoon peak longer than the battery's ~3.9-hour endurance at 1300W and read the utility meter: if the controller reverts to grid during that peak, the billed demand will exceed 700W and the demand charge savings are lost.

Watch

Extended reading notes

Core claim

The central claim is that the prototype's threshold-based control logic—executing on a Raspberry Pi with real-time data from PZEM-016 sensors and a BMS—can manage grid and battery sources to reduce demand charges and provide reliable backup. Algorithm 1 formalizes the switching: battery mode whenever load exceeds 700W, grid charging before 6 a.m. when SoC is below 100%, and emergency shutdown on demand. The experimental validation confirms accurate power measurement (±0.1 W), smooth relay transitions with a 0.25-second delay, and SoC tracking; the MATLAB simulation reports 5.4 kWh discharged during peak tariffs and 3.6 kWh recharged off-peak, flattening the load curve.

Load-bearing premise

The system assumes the battery will hold enough charge to cover every period when load exceeds 700W, but the paper includes no sizing analysis or forecast to guarantee that.

Editorial extensions

If this is right

If the threshold logic holds, a residential user can cap grid demand at 700W during peak hours, avoiding demand charges on the most expensive intervals.
The 5120 Wh battery sustains about 3.9 hours at 1300W, providing backup for short outages and time for graceful shutdown.
The MATLAB simulation implies that coordinated scheduling can reduce peak grid draw and shift consumption to off-peak tariffs.
The open-source, $1,800 design makes the system feasible as a laboratory teaching platform and a low-cost deployment option for small commercial users.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A fixed 700W threshold without load forecasting means the battery can be depleted before the peak window ends; at 1300W the battery lasts about 3.9 hours, so a 4-6 hour afternoon peak would force a return to grid and re-set the demand peak.
The claimed cost reduction could be tested by comparing a full month of utility bills with and without the system, not just component-level tests and a simulated load curve.
The same hardware could support adaptive control by replacing the threshold with a predictive algorithm that anticipates the peak window and adjusts the cutoff based on SoC and forecast load.
The $1,800 vs $16,000 comparison likely excludes installation, inverter, and battery replacement; a lifecycle cost comparison would be a fairer test of the affordability claim.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Colleague,

The short version: this is a competent senior-design hardware write-up with a clear three-layer architecture and some honest component-level tests, but the abstract's claim that real-world testing confirmed demand-charge reduction is not supported by the evidence. The MATLAB simulation at the heart of that claim discharges 5.4 kWh from a battery the paper itself rates at 4.608 kWh usable. That's physically impossible, and it means Figure 6's load-flattening result cannot be taken as a validated outcome.

What's actually new here is the specific low-cost integration: Raspberry Pi 5, PZEM-016 sensors, LiFePO4 battery, relay board, MQTT and Blynk, all at roughly $1,800 versus commercial systems. That's a useful educational template, and the wiring/safety documentation is unusually detailed for a student project. The component tests—7 W bulb reading, relay transition behavior, SoC tracking, standby draw—are consistent with what you'd expect, and the power-factor observations (0.45 grid-side in battery mode) are a nice practical detail.

The soft spots are real. Algorithm 1's high-load branch switches to battery whenever load exceeds 700 W, with no SoC check; the paper's own backup test (1300 W for 3.9 h) shows the battery exhausts well before a typical afternoon peak. Once SoC collapses, the system returns to grid and the demand charge reappears. The 700 W threshold and time-of-day parameters appear hand-chosen, with no sensitivity analysis or sizing argument. And the simulation is self-referential: it uses the controller's own rules, with no baseline or billing data, and the 5.4/3.6 kWh numbers are internally inconsistent with the stated capacity. The paper would need a comparative field test over a full billing period, with measured grid demand and actual tariff data, before any cost-saving claim is credible.

That said, the paper is not a waste of time. As a documentation of a low-cost EMS prototype for teaching labs, it has value, and the open-source Python logic is a concrete artifact. But as a peer-reviewed work claiming validated demand-charge reduction, it needs major revision. I'd send it out—a referee should catch the energy imbalance and the missing SoC logic quickly—but with the expectation that the central claim will be substantially downgraded or re-validated.

Recommendation: engage with it as a build/educational contribution, not as a result-bearing paper. A serious referee could help the authors fix the accounting and reframe the contribution honestly.

Best, [Your name]

Referee Report

4 major / 5 minor

Summary. The paper reports an undergraduate senior-design project: a low-cost Demand Charge Management (DCM) and Energy Management System (EMS) prototype built from a Raspberry Pi 5, PZEM-016 sensors, a LiFePO4 battery, relays, and IoT visualization. The controller switches between grid and battery based on a 700 W power threshold, time of day, and battery state of charge, with an additional emergency-stop and manual-override mode. The paper describes the three-layer hardware/software architecture, presents component-level tests of sensor accuracy, relay switching, SoC tracking, and power factor, and reports a MATLAB simulation that allegedly flattens a household load curve and reduces demand charges. The authors claim the prototype was validated by real-world testing and can reduce electricity costs through peak shaving and energy arbitrage.

Significance. If the validation were sound, a roughly $1,800 open-source EMS/DCM prototype with measured component behavior could be a useful educational and low-cost demonstration platform. The paper gives a clear hardware bill of materials and a plausible control architecture, and the component tests (7 W bulb reading, relay transitions, SoC tracking) are consistent with the described hardware. However, the central economic claim—demand-charge reduction and cost savings—is not supported by the evidence as presented. The simulation is built from the controller's own rule set, contains an energy-accounting inconsistency (5.4 kWh discharged from a 4.608 kWh usable battery), and lacks a baseline, tariff model, or uncertainty analysis. The paper is therefore a reasonable prototype description but does not currently establish the load-bearing performance claim.

major comments (4)

[Section IV.A] The simulation reports that the system discharged 5.4 kWh during peak windows, but Section III.A states the battery has a total capacity of 5120 Wh and a usable capacity of 4608 Wh. 5.4 kWh exceeds both values. Either the simulation uses a different battery model or the energy accounting is incorrect, and Figure 6, which shows the load curve with and without battery support, is not credible until this is resolved. Please correct the simulation or clarify the battery model and rerun the analysis.
[Algorithm 1, lines 13-14] The high-load branch unconditionally sets relays ON (battery mode) whenever PwLoad exceeds the 700 W threshold, with no check on SoC or remaining discharge time. The paper's own backup test reports that a 1300 W load is sustained for about 3.9 hours. A typical afternoon peak window can exceed 4 hours, so the battery will be exhausted before the peak ends; the controller would then drop back to grid and grid demand would exceed 700 W, defeating the demand-charge reduction claim. The decision logic needs a SoC-aware peak-shaving rule, a charging strategy, and a simulation or experiment that covers the entire peak window.
[Section IV.A] The demand-charge reduction claim is not validated by the presented evidence. The MATLAB simulation has no described input load profile, no baseline simulation without the controller, no electricity tariff or demand-charge rate structure, and no error bars or sensitivity analysis. Because the simulation implements the same rule-based switching logic that the paper proposes, it can at best show that the algorithm behaves as intended; it cannot independently establish cost savings. Please provide a baseline comparison, a documented load scenario, and a billing calculation, ideally with experimental data over at least one billing period.
[Abstract and Conclusion] The abstract states that 'real-world testing' confirmed the 'capability to reduce demand charges and ensure reliable energy delivery.' The experiments reported in Section IV validate only component-level behavior (measurement accuracy, relay switching, SoC monitoring, power factor). No real-world demand charge or electricity bill reduction is measured. This statement overreaches the evidence and should be tempered unless a full-system bill comparison is added.

minor comments (5)

[Figures 2 and 4] Typographical errors: 'Raspbery' in Fig. 2, 'Sate of Charge' in Fig. 4, and 'system' misspelled as 'system' in the Fig. 2 caption. Please proofread.
[Section III.B] The variable 'PresMin' and the constant 6 in Algorithm 1 are not defined. Clarify the units and the meaning of the threshold (e.g., 6 AM?).
[Figure 6] The figure appears to be a screenshot and lacks axis labels, units, and a legend. Please replace with a properly labeled plot that includes the baseline and battery-supported curves.
[Reference [33]] Reference [33] is an Amazon product page. If available, cite the manufacturer's datasheet for the PZEM-016 instead.
[Section IV.A] The 37 Wh UPS backup duration of 'approximately 4 hours and 20 minutes' is inconsistent with an average draw of 8.5 W (37 Wh / 8.5 W ≈ 4.35 h, i.e., 4 h 21 min, which is internally consistent; please state the calculation explicitly for clarity).

Circularity Check

1 steps flagged · score 2.0 of 10

No derivation chain to be circular; only a mild self-referential simulation that replays the controller's own threshold rule.

other [Section IV.A Performance (MATLAB simulation, Figure 6)]
"The system discharged 5.4 kWh during high-demand (peak tariff) windows and absorbed 3.6 kWh during low-demand (off-peak) periods. As shown in Figure 6, coordinated battery scheduling effectively flattened the household load curve"

The simulation is an execution of Algorithm 1, the paper's own control logic. Algorithm 1's high-load branch (lines 13-14) unconditionally enables battery mode whenever PwLoad exceeds PwThresh (700 W), and the low-load branch also frequently selects battery mode. Therefore, any simulation driven by this algorithm will, by construction, cap or eliminate grid draw and 'flatten' the load curve whenever battery energy is available. The displayed result is a direct transcription of the control rule, not an independently derived consequence, so it cannot by itself validate demand-charge reduction.

full rationale

The paper does not derive a quantitative result from equations and does not fit a parameter and then predict a closely related quantity. There is no load-bearing self-citation chain or imported uniqueness theorem. The only potentially self-referential element is the Section IV.A simulation: because the controller is defined as switching to battery above 700 W, simulating that controller on a load profile trivially yields a flattened grid-side curve. This is a mild definitional consequence rather than a substantive circular derivation, so it is scored as a minor issue rather than central circularity. The larger problems with the paper are evidentiary, not circular: the abstract's claim of validated cost reduction rests on component-level tests plus this self-referential simulation, and the reported 5.4 kWh discharge exceeds the stated 4.608 kWh usable battery capacity. Those are correctness or credibility concerns outside the scope of circularity analysis.

Assumptions & free parameters 3 free parameters · 3 assumptions · 1 invented entities

The central claims rest on a few hand-chosen thresholds (700 W, time-of-day value 6, SoC limits) and on assumptions about sensor accuracy and tariff structure. No new physical entities are proposed; the system uses off-the-shelf parts. The free parameters directly control the claimed savings and are not derived from any analysis.

free parameters (3)

Power threshold PwThresh = 700 W
Chosen as the trigger for switching to battery; no justification or sizing analysis is given, and it determines all peak-shaving behavior.
Time-of-day threshold PresMin = 6 (units undefined)
Algorithm 1 uses PresMin < 6 and PresMin > 6 to switch between charging and discharging; the basis for this value is not explained.
SoC limits = 20% and 100%
The 20% discharge floor and full-charge ceiling in Algorithm 1 are arbitrary; they are not linked to battery health or cost optimization.

assumptions (3)

domain assumption The PZEM-016 sensors provide accurate power measurement with ±0.1 W error at 1 s polling.
These sensors enforce the 700 W threshold; if their accuracy or polling speed is inadequate, switching decisions could be wrong.
domain assumption A utility TOU tariff with fixed peak/off-peak windows applies to the modelled household.
The MATLAB simulation assumes such a tariff, but real demand charges often use a 15-minute peak-window structure that the controller does not explicitly address.
domain assumption The rule-based controller can achieve demand-charge reduction without optimization or forecasting.
The paper never compares against a baseline or optimizes the threshold, yet it concludes the system minimizes cost and reduces demand charges.

invented entities (1)

Eco-mode (software mode)
purpose: Autonomous controller mode that evaluates live power measurements, SoC, and power factor to switch between grid and battery.
This is a named software state, not a new physical entity, and no falsifiable prediction outside the prototype is offered.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Demand Charge Management: Prototype Design and Testing." pith.science (2026). https://pith.science/paper/5LHYMQK5

@misc{pith2026250910713,
  author       = {Pith},
  title        = {Pith review of: Demand Charge Management: Prototype Design and Testing},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5LHYMQK5}},
  note         = {Machine review of arXiv:2509.10713}
}

abstract

This paper presents the design, implementation, and validation of a smart, low-cost Energy Management System (EMS) and Demand Charge Management (DCM) prototype, developed as part of an undergraduate senior design project. The system serves as both a practical solution for reducing electricity costs and a pedagogical tool for teaching real-time energy control concepts in power and embedded systems courses. Unlike conventional EMS/DCM solutions that rely on high-cost commercial hardware or purely theoretical models, the proposed system integrates grid power, lithium-iron phosphate (LiFePO4) battery storage, and real-time control into a unified, scalable platform constructed at a fraction of the cost, approximately $1,800 compared to over $16,000 for leading commercial options. The controller dynamically optimizes energy usage by switching between grid and battery sources based on real-time measurements of electricity prices, load power, and battery state of charge (SoC). This enables peak shaving, energy arbitrage, and backup power functionality, thereby enhancing cost efficiency and grid resilience for both residential and small commercial users. The architecture features a modular three-layer design comprising a sensing layer for electrical data acquisition, a control layer executing Python-based logic on a Raspberry Pi, and an actuator layer for seamless energy switching. Data is communicated via MQTT and visualized through the Blynk IoT platform, providing an intuitive and remotely accessible user interface. The prototype's effectiveness was validated through real-world testing, confirming its capability to reduce demand charges and ensure reliable energy delivery under varying operational conditions. Its affordability, open-source control logic, and educational versatility make it an ideal candidate for both deployment and instructional use.

Figures

Figures reproduced from arXiv: 2509.10713 by the authors.

**Figure 1.** Schematic of the proposed framework for Demand Charge Managemen. Inverter UPS Monitoring & Control Interface Rectifier Raspbery pi Relay DC Relay Common Ground Emergency Stop Fuse Voltage Bus AC Ground Neutral [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗

**Figure 2.** Prototype Demand Charge Managemen system. ensures both operational safety and optimal performance. The energy storage component consists of a lithium iron phosphate (LiFePO4) battery bank configured as 16 cells in series (16S), each with a nominal voltage of 3.2 V, resulting in a total nominal pack voltage of 51.2 V [28]. The BMS monitors individual cell voltages, overall pack current, and temperature. It also estim… view at source ↗

**Figure 3.** System switching from battery power to grid power 3) State of Charge Monitoring: State of Charge monitoring functionality was validated by comparing the SoC values reported on the battery’s built-in display with those displayed on the Raspberry Pi interface. As illustrated in [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Figures from the paper (3 more)

**Figure 4.** Figure 4: Battery mode to grid mode system with battery discharging and charging battery operation. Such modifications could help minimize reactive power contribution and enhance overall system efficiency. 0 5 10 15 20 25 30 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Grid and Load…

**Figure 6.** Figure 6: Average household power draw with and without battery support. V. CONCLUSION This paper has presented the development, implementation, and experimental validation of a smart, low-cost Energy Management System (EMS) and Demand Charge Management (DCM) prototype. Designe…

**Figure 5.** Figure 5: Grid and load power factor under battery and grid operation A. Performance The proposed system employs a stacked logic flow architecture for decision-making, leveraging real-time data streams to enable reliable, adaptive power distribution. Unlike conventional system…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references

[1]

Demand response and pricing strategies for energy consumption management,

B. Zhao, X. Zhang, and J. Chen, “Demand response and pricing strategies for energy consumption management,” Energy Economics , vol. 65, pp. 112–125, 2017

2017
[2]

Optimal energy management strategies considering time-of-use pricing,

Y . Li and H. He, “Optimal energy management strategies considering time-of-use pricing,” IEEE Transactions on Smart Grid , vol. 10, no. 2, pp. 234–245, 2019

2019
[3]

Time-of-use pricing and its impact on resi- dential demand response,

A. Al-Wakeel and J. Wu, “Time-of-use pricing and its impact on resi- dential demand response,” Renewable and Sustainable Energy Reviews , vol. 76, pp. 556–564, 2017

2017
[4]

Dynamic energy pricing for demand response optimization,

Z. Xu and X. Wang, “Dynamic energy pricing for demand response optimization,” Applied Energy , vol. 263, p. 114683, 2020

2020
[5]

Real-time pricing and demand response for efﬁcient energy management,

E. Bitar and Y . Xu, “Real-time pricing and demand response for efﬁcient energy management,” IEEE Transactions on Smart Grid , vol. 8, no. 2, pp. 997–1005, 2017

2017
[6]

Impact of seasonal electricity pricing on energy consumption behavior,

X. Ma and L. Sun, “Impact of seasonal electricity pricing on energy consumption behavior,” IEEE Access , vol. 10, pp. 48 231–48 240, 2022

2022
[7]

Optimal sizing of energy storage units in demand charge management and pv utilization applications,

B. Asghari, R. Sharma, and M. R. Narimani, “Optimal sizing of energy storage units in demand charge management and pv utilization applications,” Jan. 3 2019, uS Patent App. 15/992,955

2019
[8]

Optimal sizing and operation of energy storage for demand charge management and pv uti- lization,

M. R. Narimani, B. Asghari, and R. Sharma, “Optimal sizing and operation of energy storage for demand charge management and pv uti- lization,” in 2018 IEEE/PES Transmission and Distribution Conference and Exposition (T&D) . IEEE, 2018, pp. 1–5

2018

Show all 33 references

[9]

Energy storage control methods for demand charge reduction and pv utilization improvement,

——, “Energy storage control methods for demand charge reduction and pv utilization improvement,” in 2017 IEEE PES Asia-Paciﬁc Power and Energy Engineering Conference (APPEEC) . IEEE, 2017, pp. 1–5

2017
[10]

The effect of demand response on distribution system operation,

M. R. Narimani, J.-Y . Joo, and M. L. Crow, “The effect of demand response on distribution system operation,” in 2015 IEEE Power and Energy Conference at Illinois (PECI) . IEEE, 2015, pp. 1–6

2015
[11]

An efﬁcient scenario-based stochastic programming framework for multi- objective optimal micro-grid operation,

T. Niknam, R. Azizipanah-Abarghooee, and M. R. Narimani, “An efﬁcient scenario-based stochastic programming framework for multi- objective optimal micro-grid operation,” Applied Energy , vol. 99, pp. 455–470, 2012

2012
[12]

Dynamic economic dispatch with demand side management of individual residential loads,

M. R. Narimani, J.-Y . Joo, and M. L. Crow, “Dynamic economic dispatch with demand side management of individual residential loads,” in 2015 North American Power Symposium (NAPS) . IEEE, 2015, pp. 1–6

2015
[13]

A new bi-objective approach to energy management in distribution net- works with energy storage systems,

A. Azizivahed, E. Naderi, H. Narimani, M. Fathi, and M. R. Narimani, “A new bi-objective approach to energy management in distribution net- works with energy storage systems,” IEEE Transactions on Sustainable Energy, vol. 9, no. 1, pp. 56–64, 2017

2017
[14]

Reliability as- sesment of power system at the presence of demand side management,

M. R. Narimani, P . J. Nauert, J.-Y . Joo, and M. L. Crow, “Reliability as- sesment of power system at the presence of demand side management,” in 2016 IEEE Power and Energy Conference at Illinois (PECI) . IEEE, 2016, pp. 1–5

2016
[15]

Battery-integrated demand-side management for residential energy efﬁciency,

R. Wang and Y . Zhang, “Battery-integrated demand-side management for residential energy efﬁciency,” IEEE Transactions on Sustainable Energy, vol. 11, no. 4, pp. 2305–2315, 2020

2020
[16]

Battery storage and demand response integration for cost-effective energy management,

R. Chauhan and J. Raj, “Battery storage and demand response integration for cost-effective energy management,” IEEE Transactions on Smart Grid, vol. 9, no. 6, pp. 6231–6240, 2018

2018
[17]

Assessment of voltage regulation methods in distribution networks,

P . Paudyal, P . Ciufo, and A. Ghosh, “Assessment of voltage regulation methods in distribution networks,” IEEE Transactions on Power Systems, vol. 26, no. 3, pp. 1714–1723, 2011

2011
[18]

Optimization-based electric vehicle charging strategies for grid stability,

T. Huang, W. Zhang, and Y . Wang, “Optimization-based electric vehicle charging strategies for grid stability,” IEEE Transactions on Smart Grid , vol. 10, no. 3, pp. 3125–3136, 2019

2019
[19]

Iot-based real-time energy management for smart grids and demand-side optimization,

A. Mohamed, I. Khan, and S. Ahmed, “Iot-based real-time energy management for smart grids and demand-side optimization,” IEEE Internet of Things Journal , vol. 7, no. 6, pp. 5023–5035, 2020

2020
[20]

Demand side management for homes in smart grids,

M. R. Narimani, “Demand side management for homes in smart grids,” in 2019 North American Power Symposium (NAPS) . IEEE, 2019, pp. 1–6

2019
[21]

Method for operation of energy storage systems to reduce demand charges and increase photovoltaic (pv) utilization,

B. Asghari, M. R. Narimani, and R. Sharma, “Method for operation of energy storage systems to reduce demand charges and increase photovoltaic (pv) utilization,” Jan. 31 2019, uS Patent App. 16/006,239

2019
[22]

Optimization-based demand response strategy for smart grid with renewable energy integration,

Y . Tang, X. Li, and Z. Chen, “Optimization-based demand response strategy for smart grid with renewable energy integration,” IEEE Trans- actions on Smart Grid , vol. 12, no. 3, pp. 2154–2165, 2021

2021
[23]

Optimized energy management for demand charge reduction in smart grids: A deep reinforcement learning approach,

X. Zhou, Y . Wang, J. Zhang, and C. Liu, “Optimized energy management for demand charge reduction in smart grids: A deep reinforcement learning approach,” IEEE Transactions on Smart Grid , vol. 13, no. 2, pp. 1582–1594, 2022

2022
[24]

Home energy management systems for demand response optimization,

M. Beaudin and H. Zareipour, “Home energy management systems for demand response optimization,” Energy Policy , vol. 82, pp. 146–159, 2015

2015
[25]

Optimal scheduling strategy for demand charge management in commercial buildings,

Y . Lin, C. Wang, and Y . Chen, “Optimal scheduling strategy for demand charge management in commercial buildings,” Energy and Buildings , vol. 158, pp. 1106–1115, 2018

2018
[26]

Iot-based smart metering and demand-side opti- mization in microgrids,

J. Chen and Y . Xu, “Iot-based smart metering and demand-side opti- mization in microgrids,” IEEE Internet of Things Journal , vol. 7, no. 5, pp. 4224–4233, 2020

2020
[27]

Powerwall 3 location based purchase price,

Tesla, “Powerwall 3 location based purchase price,” https://www.tesla. com/powerwall/design/overview, 2024, accessed: 2025-04-17

2024
[28]

51.2v 100ah server rack lithium lifepo4 battery,

AoLithium, “51.2v 100ah server rack lithium lifepo4 battery,” https://aolithium.com/products/ 51-2v-100ah-server-rack-lithium-lifepo4-battery , 2024, accessed: 2025-04-17

2024
[29]

Raspberry pi 5,

Raspberry Pi Ltd., “Raspberry pi 5,” https://www.raspberrypi.com/ products/raspberry-pi-5/ , 2025, accessed: 2025-04-17

2025
[30]

Introduction to python,

GeeksforGeeks, “Introduction to python,” https://www.geeksforgeeks. org/introduction-to-python/, 2025, accessed: 2025-04-17

2025
[31]

Why is can still relevant in the automotive industry?

Nikhil Agnihotri, “Why is can still relevant in the automotive industry?” https://www.engineersgarage.com/ can-comparison-lin-most-ethernet-ﬂexray-mqtt-dds/ , 2024, accessed: 2025-04-17

2024
[32]

Pylontechusx00c_can2mqtt,

M. Plewka, “Pylontechusx00c_can2mqtt,” https://github.com/plewka/ PylontechUSX00C_CAN2MQTT, 2024, gitHub repository, Accessed: 2025-04-17

2024
[33]

Pzem-016 ac communication module - multifunction voltage, current, power, energy, frequency meter,

LiebeWH, “Pzem-016 ac communication module - multifunction voltage, current, power, energy, frequency meter,” https://www. amazon.com/Function-V oltage-Frequency-V oltmeter-PZEM-016/dp/ B0BSV3VT17?ref_=ast_sto_dp&th=1, 2023, accessed: Feb. 20, 2025

2023

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.

[1] [1]

Demand response and pricing strategies for energy consumption management,

B. Zhao, X. Zhang, and J. Chen, “Demand response and pricing strategies for energy consumption management,” Energy Economics , vol. 65, pp. 112–125, 2017

2017

[2] [2]

Optimal energy management strategies considering time-of-use pricing,

Y . Li and H. He, “Optimal energy management strategies considering time-of-use pricing,” IEEE Transactions on Smart Grid , vol. 10, no. 2, pp. 234–245, 2019

2019

[3] [3]

Time-of-use pricing and its impact on resi- dential demand response,

A. Al-Wakeel and J. Wu, “Time-of-use pricing and its impact on resi- dential demand response,” Renewable and Sustainable Energy Reviews , vol. 76, pp. 556–564, 2017

2017

[4] [4]

Dynamic energy pricing for demand response optimization,

Z. Xu and X. Wang, “Dynamic energy pricing for demand response optimization,” Applied Energy , vol. 263, p. 114683, 2020

2020

[5] [5]

Real-time pricing and demand response for efﬁcient energy management,

E. Bitar and Y . Xu, “Real-time pricing and demand response for efﬁcient energy management,” IEEE Transactions on Smart Grid , vol. 8, no. 2, pp. 997–1005, 2017

2017

[6] [6]

Impact of seasonal electricity pricing on energy consumption behavior,

X. Ma and L. Sun, “Impact of seasonal electricity pricing on energy consumption behavior,” IEEE Access , vol. 10, pp. 48 231–48 240, 2022

2022

[7] [7]

Optimal sizing of energy storage units in demand charge management and pv utilization applications,

B. Asghari, R. Sharma, and M. R. Narimani, “Optimal sizing of energy storage units in demand charge management and pv utilization applications,” Jan. 3 2019, uS Patent App. 15/992,955

2019

[8] [8]

Optimal sizing and operation of energy storage for demand charge management and pv uti- lization,

M. R. Narimani, B. Asghari, and R. Sharma, “Optimal sizing and operation of energy storage for demand charge management and pv uti- lization,” in 2018 IEEE/PES Transmission and Distribution Conference and Exposition (T&D) . IEEE, 2018, pp. 1–5

2018