Introduction

As per a recent report presented by the Electric Power Research Institute (EPRI) in the United States, large industrial units lose around $119 billion to power problems every year, mainly due to voltage disturbances and supply variations1. The light flickering and load-related disturbances often lead to troublesome experiences for the residents in commercial as well as residential units2. In general, power quality is defined as the unwanted deviation in waveform shapes and nominal values of voltage, current, and frequency that may lead to malfunction or failure of electrical appliances on the customer’s end. Since this has become a sensitive issue over the years, experts across the world are looking forward to utilizing advanced technologies for efficient monitoring and assessment of power quality fluctuations. Recent researchers in the domain have proposed Power Quality Analysers (PQA) as a relevant solution to observe, identify, and evaluate the power quality in large electrical networks3,4. These systems are designed to monitor all essential power supply parameters such as voltage, current, power factor, and frequency to identify disturbances, including harmonics, flickers, transients, swells, voltage sags, and interruptions on time5,6.

Power Quality fluctuations can happen due to several reasons; they can cover grounding issues, wiring problems, harmonic generation, and load variations7. When poor power quality goes undetected, it can even end up damaging some expensive equipment. As per European Standards (EN 50160) for electricity quality, there are some set parameters defined for national grid codes to ensure adequate delivery of power supply to consumers1. If there is a disturbance in supply voltage, the end devices may draw a non-sinusoidal signal, and it may create several technical issues, such as malfunction, overheating, and premature ageing of the equipment as well8. The non-sinusoidal currents also lead to insulation and thermal stress in network devices like feeder cables and transformers. Repeated fluctuations in the supply voltage and current ultimately lead to huge financial losses for individuals and governments. It leads to equipment downtime and an enhanced number of maintenance activities, and can even cause shorter lifetimes while posing a huge burden on individuals, businesses, and governments9.

As per a survey conducted in the United States, 85% of power quality incidents happen due to voltage swells or dips, grounding issues, wiring, and harmonics1. Furthermore, the EU-25 countries experience an estimated financial loss of $160 billion every year due to power quality problems1. Since the electric equipment is designed to operate in a specific supply range, the sudden fluctuations result in process outages and shutdowns. In order to deal with these issues, many business houses and industries have started installing local renewable energy generation sources such as wind and solar10. However, in order to operate equipment with this distributed range of sources, it is important to have switch mode power supply arrangements. The increased adoption of switching power supplies and power electronics principles has further introduced harmonics generation as a common problem in industrial equipment11. These power supplies inject harmonics into electrical lines that further lead to degradation of power quality, and every equipment connected to the supply network gets affected including cables and transformers. Facility managers can observe the problem associated with large harmonic currents as the network components become overloaded. In several cases, the extended burden on the network components can even lead to the tripping of essential protection devices. Furthermore, experts are also integrating ML algorithms to get enhanced control of the possible disturbances in the electrical network12. These SPQA systems can be easily integrated into Android applications or cloud data storage platforms for better energy management. Published studies show an improvement of 20 to 30% in the power quality assurance standards with enhanced satisfaction ratings from customers. The detailed literature review is provided in the related work section for an in-depth understanding of the domain.

Related work

With the advancements in technologies, researchers these days are using the Internet of Things (IoT), and Artificial Intelligence (AI) to improve the monitoring and assessment capabilities in real-time scenarios10. These IoT-based Smart PQAs (SPQAs) have revolutionized features in comparison to conventional monitoring systems. These SPQA systems can promise end users enhanced data accessibility, automation, analytics, and maintenance opportunities13. The researchers are now working on design of smart metering mechanism for monitoring distributed networks of electric systems as it may ease in power system planning and management as well14,15. The integration of IoT and AI in real time energy system can lead to real time data analytics and assessment of critical scenarios. These sophisticated systems can help to analyse power quality remotely while enabling an alerting mechanism for timely decision-making. These systems can utilize voltage and current sensor networks to measure essential power quality parameters for quick assessment and the data can be stored on the cloud for uninterrupted access from any corner of the world. Furthermore, machine learning (ML) and deep learning (DL) algorithms can be used to forecast critical conditions or possible fluctuations in advance. This could further contribute to predictive maintenance and anomaly detection in real-time. Unlike conventional meters, their device features an integrated web dashboard and bidirectional communication, making it ideal for end-user applications in smart grids. The authors in11 presented design of IoT enabled smart energy meter with ability to monitor voltage, current and power factor parameters in real time. The system focuses on use of low cost sensors with microcontroller and cloud interface to ensure remote data access. The proposed system is recommended for smart scale energy management and household electricity disturbance monitoring. The researchers in16 used a wireless sensor network with a smart microgrid connected to the IoT system to monitor power quality issues such as grid instability and voltage violations. They also used Advanced Encryption Standards (AES) for improved security of the data online so that intrusion of hackers can be prevented on the grid or supply data. This Arduino-based system was claimed to offer an uninterrupted power supply with efficient monitoring around the clock. The authors in17 also designed a portable power meter and protection circuit using IoT for handling low voltage distribution issues. The system was designed using a current transformer, a voltage transformer, and a microcontroller unit to provide real-time monitoring and notifications for the measured field parameters. The authors in3 present an FPGA-based design and implementation of a real-time power quality analyzer capable of detecting and classifying disturbances in electrical signals with high accuracy. Their work demonstrates the effectiveness of hardware-level modeling in enhancing processing speed and reliability for power quality monitoring applications. In18 explored the performance of a hybrid energy system (solar PV, wind, BESS) integrated with a three-phase UPQC to address sag, swell, and harmonic disturbances. Authors in19 further contributed to this domain by proposing the 3Ph-oZm, a three-phase open-hardware power quality analyzer capable of advanced computations including symmetrical components and harmonic distortion. Researchers in20 introduced a time-domain analysis approach for classifying and compensating current components in non-sinusoidal, non-linear systems. Authors in21 developed a novel IoT-based power monitoring system using non-invasive sensors and NodeMCU for real-time energy tracking. Their system enables home automation and cloud-based monitoring, offering a low-cost solution for domestic power management and theft detection. The researchers in22 proposed an intelligent algorithm for power quality management that improves data accumulation process for better data logging with pre & post value analysis. They conducted a simulator based analysis using LabVIEW software and used IoT based Firebase power quality monitoring system for real time tracking of power quality disturbances. The system was designed with a focus to meet industrial drive application with cloud based storage. The authors in23 worked on designing IoT hardware with heterogeneous multi sensors to monitoring power grid systems. However, their main focus was to detect partial discharge in the grids to avoid any damage to the insulating parts of the machineries. The system was designed using multi sensor interface while following Bluetooth low energy communication. The proposed system also conducts denoising of the signal using FPGA based logic. The summary & highlights of contribution made by these papers along with their limitations is provided in the Table 1. Other than this, IEEE has already published detailed information on instrumentation choices, measurement techniques and preferred analytics tools for power quality analysis in real time systems24. Another research paper from IEEE discusses about definitions and standards for the measurement of electric power parameters under non sinusoidal, sinusoidal, unbalanced and balanced conditions25. The document provides detailed information on well-established concepts that can work as a guideline for beginner researchers while providing metrics for the performance evaluation of new generation equipments.

Table 1 Summary & highlights from existing studies in literature.
Full size table

However, in all these cases, the overall cost involved in system design was considerably high and they have focused on limited range of field parameters for measurements. Considering the challenges in the real time settings, the authors conducted this hardware based experiment to enable remote monitoring and management of electrical networks. The main goals of this study are:

  • Monitoring power quality fluctuations by focusing on the primary parameters such as voltage, current, power, frequency, and power factor.

  • Comparing performance of an IoT based low cost monitoring system against traditional FPGA based system to demonstrate the advantages of IoT based monitoring and real-time data analysis.

  • Designing a system that can ensure real-time monitoring as well as storage of the data on the cloud platform for easy access. The stored data in the future can be used for forecasting applications to have improved decision-making.

  • Calibrating and validating IoT enabled PQA system against conventional FPGA based setup with experimental demonstration using low cost sensors.

The proposed SPQA system can promise enhanced reliability, live monitoring, and quality control on the electrical networks. Although this experimental analysis is conducted in the lab settings, but the approach can be further utilized for the real-time settings with enhanced scalability using high end IoT sensors. In future, it can be helpful for residential, commercial as well as industrial settings to have a reliable monitoring and assessment mechanism. The second section of this paper provides detailed information on various materials and methods used for the development of the proposed SPQA system. It provides descriptions of hardware components, circuit diagrams, and system architecture to measure target parameters from the electrical network. The third section focuses on the results obtained from the experimental analysis conducted using IoT-based SPQA system hardware. This section also provides detailed discussions on how these outcomes may be beneficial for the end users. Ultimately, section four includes a conclusion along with the limitations and future scope of this study. The comparative analysis of power quality methods are tabulated in Table 2.

Table 2 Comparative analysis of power quality methods.
Full size table

Materials and methods

This experimental analysis focused on the development of IoT-based SPQA for a phase system. In order to conduct continuous monitoring of electrical networks, the IoT system was designed using an Arduino Mega microcontroller. The adequate integration of hardware and software can help in the efficient reading of power supply parameters from the electrical network. The design process started with the selection of relevant hardware units. The experimental analysis was conducted using an Arduino Mega microcontroller and ESP32 Wroom was chosen as gateway unit. For the voltage and current measurement, the authors used ZMPT101B and ACS712, respectively. ZMPT101B is an AC single-phase voltage sensor that uses high high-precision voltage transformer inside. As per manufacturer specifications, the operating voltage for this sensor is between 5 V and 30 V and it can measure voltage up to 250 V AC. However, the output signal ranges between 0 and 5 V and the rated input current is 2 mA. On the other side, the ACS712 sensor was designed to work over the supply voltage between 4.5 V and 5.5 V with a current range of up to 20 mA. Since both these sensors can be used over single-phase measurements, the circuit was designed using a set of three voltage and three current sensors to measure power quality in the three-phase signal. The design architecture for the proposed design is given in Fig. 1.

Fig. 1
figure 1

Design architecture for the proposed SPQA system.

Full size image

The Arduino Mega microcontroller helps to collect live readings from the electrical network via an analog channel where all voltage and current sensors are connected in sequence. As seen in Fig. 1, the voltage sensors for phase 1, phase 2, and phase 3 measurements are connected to the A0, A1, and A2 pins of Arduino Mega, respectively. Whereas, current sensors for phase 1, phase 2, and phase 3 are connected to A3, A4, and A5 pins, respectively. The coding for the proposed IoT system was done using the Arduino IDE software platform. The monitoring system was installed in a laboratory environment of the Electrical & Electronics Engineering Department of an Engineering College in Hyderabad, India to ensure live data collection over a three-phase supply. In order to observe the variations in supply voltage more precisely, different resistive and inductive load settings were also applied to the circuit for experimental analysis. The Fig. 2 below provides hardware circuit diagram for the proposed system and Fig. 3 displays the actual hardware deployed in the laboratory environment. Figure 3 also display use of GSM 900 Module for alerting mechanism; the analysis regarding this alerting system is already published in26. The circuit diagram clearly displays the connection of voltage and current sensors with Arduino Mega 2560. The provided circuit diagram displays initial connections without load; the variable resistive and inductive loads are further added as per requirements to conduct experimental analysis. The cost involved in designing this hardware interface is also provided in the Table 3. It can be observed that the cost of designing this IoT based SPQA system was 5414.00/- INR only whereas the convention FPGA based PQA system design cost was approximately 2,00,000/- INR. With this considerable price difference in the system design, the authors further conducted multiple experiments in the lab settings to measure real-time field parameters with varying load settings. The performance of both conventional and IoT based PQA system were compared under similar lab settings over multiple inductive and resistive load conditions.

Fig. 2
figure 2

Hardware circuit diagram for the proposed system.

Full size image
Fig. 3
figure 3

Hardware Setup deployed in the laboratory environment for real-time power measurement.

Full size image
Table 3 Cost analysis of proposed IoT based PQA System.
Full size table

The circuit was designed to measure 25 different parameters from three three-phase supply; the list of target parameters is provided in Table 4. This experimental analysis was conducted for several days while varying the load settings between different ampere ratings. During resistive load connections, the experimental analysis was conducted on No Load, 0.8amp load, 1.6 amp load, and 3.2 amp load conditions. However, the second test was conducted by combining a fixed inductive load of 100mAh with the above-mentioned settings of resistive load conditions. In order to ensure accuracy in the parameter measurements, the authors used manufacturer-specified calibration procedures for all voltage and current sensors connected to the circuit. The calibration was conducted in both no load and with load conditions to ensure reliable outcomes and the accuracy of the system was measured against a conventional PQA available in the laboratory. The main goal was to measure the fluctuations in the power supply with sudden load variations and ensure adequate data collection for future analysis.

Table 4 Parameters measured by the proposed IoT-based SPQA system.
Full size table

The description of the selected parameters in terms of mathematical formulas is given by the equations below:

$$Vrms\,={\raise0.7ex\hbox{${{V_p}}$} \!\mathord{\left/ {\vphantom {{{V_p}} {\sqrt 2 }}}\right.\kern-0pt}\!\lower0.7ex\hbox{${\sqrt 2 }$}}$$
(1)
$$Irms\,=\,{\raise0.7ex\hbox{${{I_p}}$} \!\mathord{\left/ {\vphantom {{{I_p}} {\sqrt 2 }}}\right.\kern-0pt}\!\lower0.7ex\hbox{${\sqrt 2 }$}}$$
(2)
$$Apparent~Power=Vrms~ \times ~Irms$$
(3)
$$Real~Power=Vrms~ \times ~Irms~ \times ~cos~\varphi$$
(4)
$$~Reactive~Power=Vrms~ \times ~Irms~ \times ~sin~\varphi$$
(5)
$$Power~Factor=\frac{{True~Power}}{{Apparent~Power}}$$
(6)
$$Frequency{\text{ }}\left( {Hz} \right)\,=\,1/T$$
(7)
$$Vlab{\text{ }}=\sqrt 3 *{\text{ }}Vphas{e_1}$$
(8)
$$Vlbc{\text{ }}=\sqrt 3 *{\text{ }}Vphas{e_2}$$
(9)
$$Vlca{\text{ }}=\sqrt 3 *{\text{ }}Vphas{e_2}$$
(10)
$$Average{\text{ }}Vrms{\text{ }}={\text{ }}{\raise0.7ex\hbox{${\left( {Vrm{s_1}\,+\,Vrm{s_2}\,+\,Vrm{s_3}} \right)}$} \!\mathord{\left/ {\vphantom {{\left( {Vrm{s_1}\,+\,Vrm{s_2}\,+\,Vrm{s_3}} \right)} 3}}\right.\kern-0pt}\!\lower0.7ex\hbox{$3$}}{\text{ }}$$
(11)
$$Average{\text{ }}Irms{\text{ }}={\text{ }}{\raise0.7ex\hbox{${\left( {Irm{s_1}\,+\,Irm{s_2}\,+\,Irm{s_3}} \right)}$} \!\mathord{\left/ {\vphantom {{\left( {Irm{s_1}\,+\,Irm{s_2}\,+\,Irm{s_3}} \right)} 3}}\right.\kern-0pt}\!\lower0.7ex\hbox{$3$}}$$
(12)
$$Total{\text{ }}kW\,=\,k{W_1}\,+\,k{W_2}\,+\,k{W_3}$$
(13)
$$Total{\text{ }}kVA\,=\,kV{A_1}\,+\,kV{A_2}\,+\,kV{A_3}$$
(14)
$$Total{\text{ }}kVAr\,=\,kVA{r_1}\,+\,kVA{r_2}\,+\,kVA{r_3}$$
(15)
$$Average{\text{ }}PF{\text{ }}={\text{ }}{\raise0.7ex\hbox{${\left( {P{F_1}\,+\,P{F_2}\,+\,P{F_3}} \right)}$} \!\mathord{\left/ {\vphantom {{\left( {P{F_1}\,+\,P{F_2}\,+\,P{F_3}} \right)} 3}}\right.\kern-0pt}\!\lower0.7ex\hbox{$3$}}{\text{ }}$$
(16)
$$Average{\text{ }}Hz{\text{ }}={\text{ }}{\raise0.7ex\hbox{${\left( {H{z_1}\,+\,H{z_2}\,+\,H{z_3}} \right)}$} \!\mathord{\left/ {\vphantom {{\left( {H{z_1}\,+\,H{z_2}\,+\,H{z_3}} \right)} 3}}\right.\kern-0pt}\!\lower0.7ex\hbox{$3$}}{\text{ }}$$
(17)

Since the IoT-based systems are expected to provide access to field data online, either via some mobile application or a cloud service, the proposed system used the ThingSpeak IoT framework for live data collecting on the web. In order to send field data to this cloud platform, the authors used the ESP32 Wroom chip as a gateway. The Arduino Mega microcontroller used in this experimental analysis has a 10 bit resolution and the sampling time for the data collection from sensors was 2 min. Every sample from Arduino Mega was transmitted to the ESP32 chip via dedicated Transmit and Receive lines as shown in Fig. 1. The ESP32 Wi-Fi module further helps to send this dataset to the ThingSpeak cloud platform by using the Wi-Fi network available on the campus. The cloud update using ESP32 gateway is observed to have a latency of less than 1s which ensures reliable measurement from the field environment. The results obtained from this experimental analysis are presented in the next section.

Results and discussions

Power quality disturbances cause detrimental impacts on industrial, commercial, and residential equipment. The proposed SPQA system is expected to measure such fluctuations live via an IoT sensor network as described in the previous section. The performance of this SPQA system was tested by using multiple experiments with varying load settings. The outcomes were compared against conventional FPGA-based PQA systems to verify the accuracy of the system. Figure 4 shows laboratory settings for the experiments conducted using the proposed system. Furthermore, the flowchart of the methodology used in this design process is provided in Fig. 5.

Fig. 4
figure 4

Experimental setup for the IoT-based SPQA system.

Full size image
Fig. 5
figure 5

Flowchart of the design methodology used this experimental analysis.

Full size image

The IoT-based SPQA system was used to measure Vrms, Irms, Real Power, Reactive Power, Apparent Power, Power Factor, Frequency, and Line-to-Line Voltages from all three phases. During experimental analysis, authors collected all these parameters for individual phases and also obtained average and total values for various elements as listed in Table 3. The first set of measurements was conducted using resistive load variations and the second set was with a combination of resistive and inductive loads. Figures 6 and 7, and 8 represent the average and total values of various power quality parameters for resistive load settings. The readings were collected with five different load settings that include no load, 0.8amp load, 1.6amp load, and 3.2amp load as shown in the figures below. It is observed that the readings obtained from the SPQA system for all parameters were comparable to the readings obtained by the CPQA system, within tested load conditions.

Fig. 6
figure 6

Comparison between 3-Phase Average Vrms & Irms measurements from Conventional and Smart Power Quality Analysers (Resistive Load).

Full size image

Figure 7 shows readings for real, apparent, and reactive power supply. These readings are derived from the voltage and current sensor data by following the equations provided in the previous section.

Fig. 7
figure 7

Comparison between different 3-phase Power Measurements from Conventional and Smart Power Quality Analysers (Resistive Load).

Full size image

The Arduino Mega platform follows an ADC resolution of 10 bits that leads to the quick mapping of input voltages and currents to the desired three-phase supply signal. The ADC resolution on the proposed SPQA system plays an important role in maintaining accuracy in field observations. It can further help in the careful adjustment of sensitivity levels for the voltage and current sensors so that calibration tasks can be performed easily. Figure 8 provides comparison between measured power factor and frequency using conventional PQA and Smart PQA system.

Fig. 8
figure 8

Comparison between 3-Phase Average Power Factor and Frequency Measurements from Conventional and Smart Power Quality Analysers (Resistive Load).

Full size image

Based on the multiple measurements obtained from both IoT based SPQA and conventional PQA, authors conducted error comparison between both systems for in-depth result analysis. The comparative outcomes are listed in Tables 5 and 6.

Table 5 Error comparison between SPQA and CPQA readings – Resistive load Only.
Full size table
Table 6 Error comparison between SPQA and CPQA readings – Resistive + Inductive Load.
Full size table

This error analysis was conducted after collecting readings of SPQA and CPQA in different load settings with 120 different observations for each resistive and resistive + inductive load conditions. For each parameter, the error parameters were calculated as:

$${\text{MAE }}\left( {{\text{Mean Absolute Error}}} \right)\,=\,{\text{Average of }}|{\text{CPQA}}--{\text{SPQA}}|$$
$${\text{RMSE }}\left( {{\text{Root Mean Squared Error}}} \right)\,=\,{\text{Square root of average of }}{({\text{CPQA}}--{\text{SPQA}})^{\text{2}}}$$
$$\begin{gathered} {\text{Max }}\% {\text{ Error}}\,=\,{\text{Maximum of }}\left( {\left| {{\text{CPQA }} - {\text{ SPQA}}} \right|{\text{ }}/{\text{ CPQA}}} \right){\text{ }} \times \hfill \\ {\text{ 1}}00(with{\text{ }}small{\text{ }}value{\text{ }}of{\text{ }}epsilon{\text{ }}(e){\text{ }}if{\text{ }}the{\text{ }}actual{\text{ }}value{\text{ }}nears{\text{ }}0) \hfill \\ \end{gathered}$$

It is generally observed that for industrial applications such as motor control, machine handling and sensitive manufacturing tasks, the error margin should be less than 1% for the critical parameters such as Vrms and Irms. However, for the residential, commercial settings or other low cost deployment sites, the errors between 5 and 10% can be accepted. Moreover, the predictive monitoring applications and smart grids can tolerate slightly higher errors if real time observations are more critical than precision. Based on the error comparison, it is observed that the Vrms in both cases is within acceptable range (1.24% and 1.31%); however, Irms error (9.16% and 8.86%) is slightly high but can be tolerated in case of basic monitoring system and can be further improved with recalibration of sensors. Power factor with a variation between 7 and 9% is tolerable for low cost IoT applications. The real power error (approx. 6 to 7%) is also in the moderate range which can be accepted to some extent in non-billing grade systems. However, the more attention is required for reactive and apparent power measurements since they have higher error range (20.03 to 21.32% & 15.10 to 10.01%, respectively). Since reactive power is more affected by non-linear loads and phase shift sensitivity; it can be improved in future by replacing low cost sensors with factory calibrated version of sensors. The present analysis can be accepted with low cost system design perspective using basic sensors; however, it needs further improvements in future for real time applications. On the other side, the Irms and Vrms values can be further improved using ADC resolution or by using more precise voltage and current sensors. With reference to the error margins observed in the experimental analysis, the proposed low cost, IoT based cloud monitoring system can be suitable for educational settings, labs, predictive maintenance applications, early warning systems and various non-critical industry setups. However, the proposed system cannot be used for grid level controls, energy billing and high precision industrial diagnostics. In future, the system outcomes can be improved testing calibrated sensors under different load conditions. It is possible to use hall-effect sensors or split core current transformers for better measurements. The error margin can be further reduced by improving ADC resolution by using ESP32 (12 bit ADC) as main microcontroller unit instead of Arduino Mega (10 bit ADC). The authors are further working on the testing of this system under fluctuating loads to check its behaviour against real world noisy signals. The edge based processing can be deployed for noise filtering as well as for auto calibration of sensors time to time, especially in the fluctuating conditions. Other than this, the future researchers can also think of using Kalman filters or digital filtering mechanism to stabilize readings. Furthermore, the future designs can be further improved with shielded enclosures to have better usability in the field environments.

The readings obtained from the SPQA system were transferred to ThingSpeak by using the ESP32 Gateway unit. The data collected on this third-party cloud interface can help end users check field measurements in real-time. This information can be further utilized to generate quick alerts for adequate decision-making. Glimpse of few readings stored on ThingSpeak platform are provided in Figs. 9 and 10. The Fig. 9 shows Vrms and Irms readings from all three phases along with average calculations of both these parameters displayed on Channel 1 of ThingSpeak platform. Similarly, Fig. 10 shows Real and Apparent Power readings from three phase system on Channel 2. In the similar manner, remaining readings from electrical network are stored on Channel 3 and 4 of ThingSpeak platform. The variations in the power supply parameters can be observed due to switching between resistive and inductive loads. The data can be accessed by end users from anywhere in the world by simply using their ThingSpeak login credentials. Moreover, if required, these channels can be activated with public access mode so that data is publically available for anyone to visualize.

Fig. 9
figure 9

Channel 1 Readings on ThingSpeak Platform (Vrms1, Vrms2, Vrms3, Avg Vrms, Irms1, Irms2, Irms3, Avg Irms).

Full size image
Fig. 10
figure 10

Channel 2 Readings on ThingSpeak Platform (Real and Apparent Power from Three Phase Supply).

Full size image

The stored data on the ThingSpeak platform can be accessed in the form of CSV files to conduct pattern analysis or forecasting tasks. The proposed system can be installed in various building environments to ensure continuous monitoring of power supply conditions so that they can take preventive measures whenever required. Such systems can help small industry owners and facility managers avoid sudden downtime and equipment failures due to disturbances in the power supply. However, the system can be further improved in future by using calibrated sensors to achieve enhanced accuracy to make it suitable for critical monitoring zones as well.

Conclusion

Over the years, power quality management has become a complex issue for facility managers. Sudden disturbances may lead to difficulty in handling varying loads on building premises. Timely identification of interruptions can help to enhance service quality while boosting the overall productivity of the system. The IoT-based SPQA system allows real-time monitoring and assessment of power quality parameters. The voltage and current sensors capture signal from three phase supply and this data is further utilized to calculate remaining power quality parameters. The field data is further transferred to the ThingSpeak platform using the ESP32 Wroom module. The performance analysis for the SPQA system was conducted against an FPGA-based conventional PQA system. The tests were conducted in different combinations of resistive and inductive loads. Observations state that the proposed experimental analysis offers a real world deployment feasibility with slight improvement in the system design. The proposed system is also cost effective in comparison to the conventional FPGA based PQA (as discussed in Materials & Methods section) that makes it suitable for small to medium scale industries, residential settings and small commercial units as well. The modular hardware design approach offers enhanced reliability and transferability with potential to enhance outcomes with the integration of artificial intelligence based methods in future.

Nevertheless, the proposed SPQA system has a few limitations that the authors plan to improve with the future design. The main issue observed during this experimental study was the repeated calibration requirement for sensors. The input sensors were not able to follow large changes in the applied voltage and current. However, when the sensitivity levels of these sensors were adequately adjusted after careful calibration, the readings became more accurate in comparison to the CPQA system. However, the current study is limited to the laboratory settings with resistive and inductive loads, in the future, authors will focus on testing system performance under variable frequency sources and harmonic distortions. The further validations based on field deployments will be also conducted by replacing low cost sensors with calibrated voltage & current sensors. In the future, the proposed module can be converted into a portable system with a chip-level design. This easy-to-use monitoring system can ensure quick decision-making for managing fluctuations in the electrical network. The data gathered on the ThingSpeak platform can be further utilized for pattern analysis and forecasting of electrical conditions in the long run.