Utilising Neutrosophic Logic in the Design of a Smart Air-Conditioning System

Abstract

Air conditioners, which have become the most widely used cooling system, have already been employed in both household and commercial environments. Automation in air conditioning is a very complex and demanding task today. When a sensor in an automatic human detection air conditioner detects the motion and activity of people, it will automatically turn on and set the temperature accordingly. However, in situations in which there are no humans around or in which human presence is unstable for an extended period, energy and power are wasted. In this regard, a control system was created utilising a Neutrosophic logic controller to regulate the AC temperature to a specific level by reducing the compressor and fan speeds without considering other parameters. Since neutrosophic logic handles the truth, ambiguity, and untruth of people in a closed environment, a more intelligent air-conditioning system is created by the suggested approach. As a result, massive quantities of energy savings are achieved. To accomplish the desired outcome, a MATLAB simulation is applied.

1. Introduction

Several clever technologies made by scientists and engineers have made human life more comfortable and less burdensome. Since then, as automation has increased, human involvement in appliances has been reduced. Automatic devices work best when they use a good sensor, transducer, actuator, and control system. The intelligence of these gadgets is characterised by their capacity to make decisions in response to events that are influenced by their shifting environment. This research is increasingly focused on developing environments that can react to the people who living in them. Smart gadgets are those that are aware of their surroundings, situation, task, and position. They are also aware of their environment and surrounds [1,2,3,4,5].

Cooling systems are now an essential component of a pleasant lifestyle, especially in warm environments. They are employed to regulate the temperature, dew point, level of freshness, and airflow velocity of the interior space. Human connection is necessary for a traditional process. While the generator and fans are running, we must turn the device on, or off, and choose the airflow intensity (lower, midrange, or severe). In today’s airconditioning systems, automation controllers are crucial components that guarantee consistent performance, higher quality, lower operating costs, and improved security. Therefore, the precise details of the based mechanism are needed for the realization, design, and integration of the controller systems [6,7,8].

A thermally comfortable environment has always been a goal of human endeavours. From ancient times to the present, this has been reflected in global building customs and constructions, the clothing worn according to the season, and the usage of various heating and cooling systems [9,10,11]. Temperature and humidity interact to determine our level of comfort at home. By lowering the air’s temperature as well as the amount of moisture in the air, our air conditioners control these factors.

Interior settings are affected by outdoor climate variables in different ways and through different channels. For instance, in the summer, intense solar radiation can raise building interior temperatures, which increases the need for cooling. The assessment, creation, and maintenance of indoor comfort settings for urban residents have recently been the subject of multiple independent research in a variety of disciplines. These studies have concentrated on specific problems and challenges [12,13,14,15]. The province illustrates how difficult it is to analyse TC (Thermal Comfort) in the built environment. The aim of experts should be to create comfy and healthful indoor spaces. In order to establish new approaches and tools for the detection and measurement of TC, researchers are continuously looking into this problem. Deep analysis of user perception is currently popular. Applying Internet of Things (IoT) solutions and cutting-edge algorithms are replacing conventional approaches for both tracking and calculating [16,17,18].

No system or mathematical model is necessary for intelligent control systems. With the aim of achieving performance enhancement over conventional controls, recent discussions have focused on the hands-on applications of intellectual controllers for heating, ventilation, and air conditioning (HVAC) systems. Among these are genetic algorithms, fuzzy logic, neural networks, expert systems, and fuzzy logic [19,20,21,22]. Many researchers have proposed that a fuzzy logic controller could be more useful for managing a load’s power usage in a way that optimises energy consumption without sacrificing user enjoyment [23,24].

Making human-like decisions from a variety of options requires the use of fuzzy logic. Zadeh established the idea of fuzzy sets in 1965, and since then, several extensions of fuzzy sets have been presented, including intuitionistic fuzzy sets, which deal with membership and non-membership. The limitation on the sum of belongingness and non-belongingness, however, limits the options for membership and non-membership standards. To solve this problem, Smarandache established neutrosophy, a distinct field of philosophy, in 1995. Neutrosophy is the spirit of the neutrosophic set (NS) and neutrosophic logic (NL). Compared to fuzzy systems and intuitionistic fuzzy systems, NS simultaneously considers truth, ambiguity, and untruth membership, which are additionally efficient and steady. Extensions to the NS include the single-valued neutrosophic set (SVNS) [25,26].

In this study, we contributed by developing a model for an intelligent air-conditioning system utilising a software for the Neutrosophic logic processor. The sensor triggers the need for conditioned air whenever it notices human presence. Fully automated systems will start and stop on their own. The algorithm considers the truth, uncertainty, and falsehood of human presence and activity while estimating the power consumption of the compressor speed and fan speed. In addition to controlling excessive power consumption, the system can maintain a comfortable temperature in the enclosed environment of diverse circumstances. The tools for fuzzy logic in MATLAB are employed to boost a smart system’s performance.

The following describes this paper’s structure: Associated work of this suggested system is described in Section 2. Preliminaries are discussed in Section 3. The proposed approach is defined in Section 4. Results and discussion are offered in Section 5. This study conclusion is found in Section 6.

2. Related Work

Rajani Kumari et al. presented a new design for an air-conditioning system in 2012 by combining fuzzy logic and a neuro-fuzzy algorithm. Finally, they compared the output from each system’s simulation and discovered a higher output [27]. Chin-Chi Cheng and Dasheng Lee introduced a smart air conditioner with wearable devices in 2014 for assessing the sleeping state and flexibly altering the sleeping function. These strategies could detect human temperature and action all through sleep. By utilizing wearable technology to improve the sleeping function, smart air conditioners can reduce energy consumption by up to 46.9% while maintaining human healthiness [28]. The comfort zone in residential buildings, as specified by the ASHRAE code, was researched in 2015 by Sohair F. Rezeka et al. They found that the proposed control method of air conditioning meets the space load and achieves both goals. PID control has been shown to be less effective and energy-efficient than fuzzy controller operation [19]. The strategy of a server ambient temperature and moisture controller system employing fuzzy logic grounded on a microcontroller with average AC temperature and average AC mode set was proposed by Febryan Hari Purwanto et al. in 2018 [23].

Lei Hang and Do-Hyeun Kim suggested an Advanced Design Predictive Control Method Employing Fuzzy Logic for Retaining Comfortability in IoT Smart Space in 2018. In this study, an improved design predictive supervision convenient structure for preserving the interior temperature is presented. It includes multiple regression analysis linear predictive models and an authentic fuzzy controller that considers both the PMV index and the requirements of the open environment [12]. For a compressor cooling system, Huaxia Yan, Yan Pan, Zhao Li, and Shiming Deng proposed in 2018 that more advancements in a thermally pleasant fuzzy controller be made. This study simplifies a previously established fuzzy logic controller while maintaining system reliability, hence PMV was employed as the fuzzy controller’s limiter. The PMV index included factors such as temperature, precipitation, and regional air velocity. By adding a localized fan, 4–7.6% can be saved in energy [29].

The fuzzy logic Controller for the classroom air conditioner was proposed by Safa Riyadh Waheed et al. in 2019. The creation and performance assessment of an FL-based air-conditioning unit controller that can be used in a classroom context is described in this study. This FL-based control system can simplify programming, making it possible to use common microcontrollers used in control panels for classroom air conditioners [30]. Fuzzy Logic Improvement of a Cooling Water Management System to generate Polymer Pipes was the idea put out by Marin Kochev and Malinka Ivanova in 2021. Its goal is more efficient freon expander load regulation to achieve those implications on energy consumption economics. Simulation in the FisPro environment is used to validate the produced model where fuzzification functions and the level of expertise are modelled [31].

A hereditary ambiguous controller for a divided air conditioner constructed on Fanger’s PMV Index was proposed by Chandrakant Balkrishna Patil and R.R. Mudholkar in 2019 [23]. Sanjeev Kumar Sinha and Chiranjeev Kumar validated neutrosophic logic’s efficacy for healthcare document retrieval in 2020, and their suggested work has been validated for documents pertaining to cancer care [32]. For the early detection of eight symptoms of heart disease in pregnant women, Shaisha Habib et al. established a new neutrosophic logic in 2021. In this work, a unique single-valued neutrosophic decision-making model for the care of pregnant women with heart illness was provided [25].

The development of a neutrosophic self-tuning control system for an AC lasting magnet synchronal mechanical built on neutrosophic theory was proposed by Zhongliang Fu et al. in 2021 [33]. The neutrosophic PID tuning control algorithm is examined in this study and has good control accuracy under various response characteristics. Juan Manuel Belman Flores et al. compiled the fuzzy logic control techniques used for the refrigeration system in 2022 [34].

3. Preliminaries

(i) Fuzzy Set [35]: A fuzzy set Ầ of a non-empty set X can be defined as

$$Ầ=\left\{\right(x,{\mu }_{Ầ}\left(x\right)):x\in X\}$$

where ${\mu }_{Ầ}$ (x): X; → [0,1], membership function of the fuzzy set Ầ and ${\mu }_{Ầ}$ (x) is the membership value of the element x.

(ii) Intuitionistic fuzzy set [35]: An intuitionistic fuzzy set Ầ of non-empty set X can be defined as

Ầ = {(x, μ_Ầ (x), ν_Ầ (x)): x ∈ X},

where μ_Ầ (x): X → [0,1], membership function of the intuitionistic fuzzy set Ầ and μ_Ầ (x) is the membership value of the element x. ν_Ầ (x): X → [0,1], non-membership function of the intuitionistic fuzzy set Ầ and ν_Ầ (x) is the non-membership value of the element x. 0 ≤ μ_Ầ (x) + ν_Ầ (x) ≤ 1 for all x ∈ X.

(iii) Neutrosophic set [36]: A truth, indeterminacy, and falsity membership function, which are represented as T_Ầ, I_Ầ, and F_Ầ, respectively, are characteristics of a neutrosophic set in X and are defined by

$$Ầ=\{⟨x,{\mathrm{T}}_{Ầ}(x),{\mathrm{I}}_{Ầ}(x),{\mathrm{F}}_{Ầ}(x)⟩\mid x\in X\}$$

The functions T_Ầ (x), I_Ầ (x), and F_Ầ (x) are real standard or non-standard subsets of [0⁻, 1⁺]. That is, T_Ầ: X → [0⁻, 1⁺], I_Ầ: X → [0⁻, 1⁺], and F_Ầ: X → [0⁻, 1⁺]. There is no restriction on the sum of T_Ầ (x), I_Ầ (x), and F_Ầ (x), so 0⁻ ≤ sup T_Ầ (x) + sup I_Ầ (x) + sup F_Ầ (x) ≤ 3⁺.

(iv) Single-valued neutrosophic set (SVNS) [36]: To apply the neutrosophic components in real-world paradigms, SVNS defines them over the closed interval [0,1] rather than the unconventional interval [0⁻, 1⁺].

Let X be a space of points, let x be a general element within X. A SVNS Ầ in X is defined by the truth, indeterminacy, and falsity membership function, which are denoted as T_Ầ, I_Ầ, and F_Ầ. For each x ∈ X, T_Ầ (x), I_Ầ (x), F_Ầ (x) ∈ [0,1]. Thus, it can be written as Ầ = {⟨x, T_Ầ (x_i), I_Ầ (x_i), F_Ầ (x_i)⟩| x_i ∈ X}, where 0 ≤ T_Ầ (x) + I_Ầ (x) + F_Ầ (x) ≤ 3.

4. Proposed System Design and Simulation

There are four basic components to the suggested system. The first section, called neutrosophication, contains the truth, ambiguity, and untruth membership of the input parameters’ occupant count and type of activity. The second component is a neutrosophication rule base, which includes every rule that could possibly apply to the system. The inference engine, which is the third component, includes active decision-making rules. De-neutrosophication is the final step, and the fuzzy system’s centroid approach is used. The planned system’s diagram appears in Figure 1.

4.1. Working Principle of Sensor

A human body passing in front of the PIR sensor causes a change in the ambient temperature, which the sensor uses to detect motion. The electrical devices to which it is linked are then turned ON. The voltage stabiliser of the AC has previously been connected to the power supply. The power supply will now be connected through a contactor, because the motion sensor cannot handle a larger electrical load directly. The motion sensor will operate as a switch between the contactor and the stabiliser, since the contactor will be connected to the stabiliser through the motion sensor. Only when a human body moves, according to the PIR sensor, will the contactor turn on the power supply to the stabiliser [37]. The sensor is depicted diagrammatically in the following Figure 2.

4.2. Variables That Describe the Proposed System

4.2.1. Input Variables

Outside Temperature (OTE)

The difference in temperature between the outside air and the refrigerant determines how much heat is transferred and how quickly. More cooling is accomplished via the heat exchanger rather than the compressor when the outside air temperature drops. When the temperature outside rises, the compressor of the air conditioner has to work harder to cool the house. We can categorise the outside temperature as being low, moderate, high, or very high.

Humidity(H)

Water vapour content in the air is referred to as humidity. Temperature and the relative humidity of the air are both important factors in determining humidity. Humidity levels will be high if there is a lot of water vapour present. The type of weather we experience—dry, normal, or sticky—depends on the amount of water vapour in the air.

Occupants (O)

The number of persons who are exposed to air conditioners is known as occupancy. Depending on the individual, bodily radiant heat has a different effect, especially in large, densely populated spaces like auditoriums, where a somewhat lower effective temperature is needed. The number of persons will determine if the occupancy is low, medium, or high. We have considered the state of a medium-sized room as well as the truth, indeterminacy, and falsity of occupants.

Activity Type (A)

Depending on activity level, the human body emits between 100 and 450 W per person through metabolism. Heat production increases as physical activity increases. The actions of people in a closed setting will determine the heat rise in a closed environment. We also considered the membership of truth, indeterminacy, and falsity in this variable.

Period of the Day (TOD)

The period of the day refers to the time of day when the AC would be running. After midnight and early in the morning, the relative humidity is frequently at its maximum. Once the sun rises, humidity quickly declines until it reaches its lowest level right about midday. Its intensity then increases once more until midnight, increases in pace in the early and late afternoon, and then levels off around midnight [38].

4.2.2. Output Variables

Temperature (T)

The temperature of the room, which relates to the compressor, is measured by a thermostat. Air conditioning requires the ability to regulate temperature. While the temperature outside is changing, the internal temperature is kept constant, which is the desired outcome. Since we are dealing with an automatic human detection air-conditioning system, the temperature is automatically regulated to low, medium, or high depending on the people present in the enclosed space.

Compressor Speed (CS)

The compressor, which introduces a refrigerant into the system, is regarded as the air conditioner’s heart. It distinguishes between the systems with high and low pressure. No cooling will occur if the compressor malfunctions. The compressor speed is labelled as low, medium, or high depending on the humidity level and the number of people present in the space.

Fan Speed (FS)

The fan aids in drawing in ambient or fresh air from the outside and distributing it evenly across the evaporator coil. It is located near the evaporator coil or behind it. Depending on the amount of moisture in the room, the fan speed is set to low, medium, or high. The

Table 1. Variable classifications and their membership functions.

Variables in the Proposed System		Membership Functions
		TFNs (Triangular Fuzzy Numbers)
I-1 (I-Input)	Outside temperature Range: [0–100%]	Low [0–30]	Moderate [30–60]	High [50–80]	Very high [75–100]
I-2	Humidity Range: [0–100%]	Dry [0–50]	Normal [50–70]	Sticky [70–100]	-
I-3	Occupants Range: [0–8]	Low [0–4]	Medium [3–6]	High [6–8]	-
I-4	Activity Range: [0–100%]	Less [0–40]	Medium [40–75]	More [75–100]	-
I-5	TOD Range: [0–24 h]	Morning [0–11:30]	Afternoon [11–17]	Night [17–24]	-
OT-1 (OT-Output)	Temperature Range: [18–30 ℃]	Low [18–22]	Medium [22–27]	High [27–30]	-
OT-2	Compressor Speed Range: [0–100%]	Low [0–40]	Medium [38–70]	High [65–100]	-
OT-3	Fan Speed Range: [0–100%]	Off [0–10]	Minimum [10–40]	Average [40–70]	Maximum [68–100]

shows the classifications of inputs and outputs variables of the proposed study.

4.3. Neutrosophication

The proposed framework replaces fuzzy logic statements as predecessors and consequents of the “IF-THEN” commands with neutrosophic logic expressions. By transforming brittle inputs using the three membership functions of neutrosophic logic, i.e., truth, falsehood, and indeterminacy, one can create the neutrosophic knowledge base. Inference rules, which support the system’s computation capabilities, are one of the core ideas behind the Mamdani method [39]. These guidelines may be influenced by previous information, observations, and expert insight. Two concepts—the if-then statements and the linguistic variables—make up each inference rule. The ability to make any automated choice as well as the interactions of the input and output are taken into consideration when creating the proposed work.

In this study, the number of rules depends on the potential membership functions of inputs such as the outdoor temperature, humidity, the number of people inside, the type of activity being performed, and the time of day. However, outputs are things such as temperature, fan speed, and compressor speed. For truth, twenty-three rules are required, but indeterminacy and falsehood require nine rules each, as demonstrated in

Table 2. Rules for truth membership of neutrosophic logic represented as (T-O, T-A).

Inputs					Outputs
OTE	H	T-O	T-A	TOD	Temp.	CS	FS
Low	Dry	Less	More	Morning	High	Slow	Medium
Low	Normal	Medium	More	Morning	Low	Slow	High
Very high	Sticky	More	More	Morning	Low	Fast	Low
Very high	Sticky	More	Medium	Afternoon	Low	Fast	Low

,

Table 3. Rules for indeterminacy membership of neutrosophic logic represented as (ID-O, ID-A).

Inputs		Outputs
ID-O	ID-A	Temp.	CS	FS
Low	Low	Low	Fast	Low
Medium	Low	Medium	Medium	Low
High	High	High	Slow	High

and

Table 4. Rules for falsity membership of neutrosophic logic represented as (F-O, F-A).

Inputs		Outputs
F-O	F-A	Temp.	CS	FS
High	Low	High	Slow	Slow
Low	Low	Medium	Medium	Slow
Medium	High	High	Slow	Slow

. The neutrosophication is the last phase, when we can obtain the precise value of each output. The most popular defuzzification technique we employed in this study is the centre of area approach (COA), and we also acknowledged the centroid technique. With the help of this method, one may locate the fuzzy set’s centre and obtain the related crisp value.

5. Results and Discussions

When one output depends on one or more inputs, the surface viewer is used to show this relationship. A surface viewer (Figure 3) that displays the output in three dimensions with XYZ axes—i.e., two inputs and one output—allows users to analyse the outcomes of this suggested system. Given the inputs OTE = 50%, H = 50%, O = 4, A = 50%, and TOD = 12:00 h, we obtained a crisp output value for truth membership for CS = 50%, FS = 50%, and T = 24 °C. When the temperature is 24 degrees Celsius, which is optimal for electricity savings, it is fair to use an automated air conditioner. The level of interior humidity can also be significantly reduced by reducing the fan’s speed. Aimed at the purposes of occupant luxury and energy saving, the clever air conditioner may actively alter the fan speed and compressor output as the temperature and humidity rise.

We have only taken human stability and activity into account while considering indeterminacy. With the help of the results, we can therefore draw the conclusion that the temperature should be set at 28.4 °C when human presence is unstable. In this case, the fan continues to run while the compressor operates at a reduced power to preserve the closed-environment cooling effect. This was shown by Figure 4. When the human presence is stable, this will minimise the effort that the compressor must perform.

We have regarded the presence of humans and their activities as being false for this reason. In this scenario, an air conditioner will operate for a while before turning off. As a result, we may conclude that the temperature should be set at 28.4 °C, which is the point at which the fan and compressor work at their slowest speeds. This will result in significant reductions in both power usage and electricity costs. This was shown by Figure 5.

Furthermore, it is justified in light of the findings that the neutrosophic controller system was created with the capability of intelligently optimising compressor motor consumption. As a result, cutting back on unnecessary energy use increases the efficiency of the air-conditioning system. This idea is presented in the

Table 5. Findings from neutrosophic logic.

Neutrosophic Logic	Compressor Speed (%)	Fan Speed (%)	Temperature (℃)
Truth of occupants	50	50	24
Indeterminacy of occupants	20	55	28.4
Falsity of occupants	20	23.6	28.4

below.

The average body temperature of a person is between 36 and 37 degrees Celsius. As a result, people in tropical countries like India are advised to take their temperature near to 24 degrees. According to the ministry’s comfort table, values up to 25 degrees, together with the required moisture and air movement, are extremely acceptable for a human body. The thermal comfort zone, as defined by ASHRAE Guideline 55–2013 Thermal Ambient Circumstances for Human Occupancy, is the frame of mind that indicates comfort with temperature and humidity. In addition to specific factors like clothes and metabolic activity, this condition is defined by the warmth, wetness, and wind that the human body experiences. According to technical study, the temperature setting can be 24–25 ℃ to obtain the ideal comfort level at a stable condition [40].

According to the findings of Table 5, the suggested system of an automatic human detecting air conditioner using a neutrosophic logic controller will meet the ASHRAE standard. Furthermore, without sacrificing the comfort of the people within, fan and compressor speeds are also at their medium levels. In the situation where human is stable running fan speed and compressor speed with 50% does not consume more power at 24 °C whereas for unstable and untruth situation the temperature drops to 28.4 °C by lowering the compressor speed.

6. Conclusions

This work has integrated the idea of fuzzy logic with neutrosophic logic to construct a smart air conditioner system, giving it the benefits of both fuzzy logic and neutrosophic logic. Electricity conservation is crucial for both emerging and industrialised nations. Neutrosophic logic has a role in research because there is so much uncertainty in the world. This study clearly demonstrates how neutrosophic logic has been used in the auto-human detection air-conditioning system to handle a challenging issue without having to consider complex interactions between physical variables. Results from MATLAB simulations have demonstrated that the suggested system can lower energy consumption according to the guidelines of ASHRAE. In order to conserve energy while individuals are present, the system will set an ideal temperature of 24 °C. This may vary according to the number of occupants. When humans are absent for an extended period of time or are unstable in a closed environment, this controller will set the temperature to 28.4 °C by default. In the future, we can develop this research to determine energy optimisation using significant variables, such as residents’ ages and genders.