Enhancing Solar Air Heater Performance with Quatrefoil Artificial Roughness: An Adaptive Neuro-Fuzzy System Approach

: Solar air Collector is used to transform the solar energy into heat energy. Solar heating technologies use only free, renewable and clean energy. More number of previous studies was done. The roughness element is changed to enhance the airflow's capacity to absorb heat. In this study Experimental investigates the glazed solar air heater with quatrefoil shaped artificial roughness is placed above the absorber plate in the solar air heater. The solar air heaters (SAH) with quatrefoil shaped artificial roughness and conventional solar air heater are tested. The outcome indicates that the maximum temperatures for conventional SAH and purposely roughened SAH are 67 0 C and 47 0 C, respectively. By the increased of heat transfer area and increased the turbulence lead to higher heat recovery rate of quatrefoil shaped artificial roughness is 47% more compared to conventional type. The solar air heater is predicted by using Adaptive Neuro Fuzzy Inference System (ANFIS) and accuracy is measured as 90%.


Introduction
In recent years various types of solar air heater has been developed to increase it to thermal efficiency.Most commonly the artificial roughness in different sizes has been investigated in the experiments.Solar air heater with different artificial roughness has been used as a reference in this project.Solar air heater replacing flate plate absorber plate with adjoining tubes and the same direction of air flow through SAH.The results reveal that tubular solar Air heater provides more accuracy than flate plate solar air heater [1].The S shaped air heaters are implemented, and the results shows that a solar heater has an enhancement in heat transfer with the lowest pressure.[2].The roughness to be implemented in bottom of plate provides more accuracy and occurs to improve the performance with implementing V ribs.The experiment has been conducted to collect data on heat transfer.[3].discrete multiple shaped arc roughness has problem investigated roughness is provided below the absorber plate to increase the efficiency.The fulfilled performance of air heaters is improved by changing parameters in roughness [4] The rectangular fins and absorber play an important role in improve the efficiency to 81%.[5].The comparison is made to check the solar air heater in the presence and absence of flow restricting device [6].The analysis of transferring heat and properties of fluid of SAH was implemented using artificial roughness.The rib roughness in the form of discrete with reverse arc is been utilized [7].The more thermal efficiency was occurred by using serpentine wavy Wire mesh.It provides 80% of thermal efficiency [8].The solar air heater is predicted by using multilayer perceptron.The accuracy is predicted by using Artificial Neural Network (ANN) [9][10][11].This methodology can also predicted and bring together by particle swarm optimization [12].(Elbreki.et.al) was experimentally tested the cooling of PV module using fins and planar reflector.results indicate that, the best performance is demonstrated by passive cooling with scalloped fins, with a mean PV module temperature of 24.6 °C.and electrical efficiency and power output are 10.68% and 37.1 W, respectively [13].has experimentally investigated the efficiency of photovoltaic panels by using lengthwise fins.Different types of lengthwise aluminum fins were fixed to the nethermost face of a PV panel.The power improvement by 2.45 W. The effectiveness value of the PV panel with cooling was about15.3 against 15 for conventional PV panel [14].(Parkunam.et.al) has experimental analyzed the cooling flat photovoltaic panel with various heat sinks and wick structure with copper and aluminum fins.the efficiency of the copper fins is higher [15].Performance study carried out by using fin-based cooling of pv panel.The cooling impact of up to 5°C was demonstrated to be a fair expectation for this technique under particular operating conditions.[16].(Savvakis.et.al)Found fin cooling minimize the efficiency [17].Sangeetha and al. have by experimentation investigated the nano fluid based mostly electrical phenomenon thermal (PV/T) system for better electrical potency and chemical element production.3 totally different nanofluids are flow in the back of the PV panel outcome indicated that muti wall carbon nano tubes, Aluminum oxides, and Titanium oxides revealed 41%, 39%, and 22% gains in power generation and 48%, 31%, and 25% elevations in electrical potential, respectively, suggesting great promise [18].When compared to a conventional PV system, the effects of using serpentine tubes with three different cross-sections the overall energy and exergy efficiencies are improved.by 6.7% and 0.8%, employing 2.2 % volume concentration nanoparticles for the example of the rectangular serpentine tube with a flow velocity of 20 kg/h.employing a water/silver nanofluid-based optical filter, the effectiveness of a hybrid PV/thermal system was experimentally investigated.At 25 °C, Electrical and thermal efficiency are displayed by the hybrid system of 10.7 % and 57.7 %, respectively, Compared to the solo system's electrical efficiency of 11%.The study's findings show that a hybrid PV/OF system can effectively replace a single PV system, and that this system's potential is increased by high air temperatures and sun concentrations It was investigated how to increase the efficiency of solar PV systems using nanofluid, and an inverter topology with TiO2 nanofluid as the working fluid was designed.The new design panel can absorb 30% more heat than a conventional panel, according to the results.Its efficiency has been increased by around 1.17%, and the integrated system is functional with 16.38% of panel efficiency and decreased converter losses aluminum oxide and titanium oxide water-based mixtures at different ratio are taken as 0.02%, 0.06%, and 0.1%.Additionally, three In addition to using water and fresh air to cool the PV panels, nano fluids were also used.Study found that when compared to water and ambient air, nanofluids for cooling dramatically increased heat transfer rate.The results from using TiO2 nanofluids at the intended concentration (0.1 wt%) were the best.The solar heater using quatrefoil shaped artificial roughness is shown in Fig. 1 The setup is mad with 1654 mm of air passage with 1645mm x 600mm x 12mm cross sections and 20 mm thickness of ply panel.The air inlet with 445 mm and for the test it has 1200 mm and for exit it have 300 mm.The experiments were carried out for 8 hours sunshine duration.The absorber Plate surface was made up of with aluminum with dimensional of 600x1200x1 mm.The quatrefoil shaped artificial roughness was designed with outer diameter of 50mm and inner diameter of 30mm and are arranged over the absorber surface in a linear arrangement.The experiments were taken from 09.00 a.m. to 18.00 p.m. under the sunny weather conditions for both Conventional and modified solar air heater (SAH) .The testing was carried out with different slope of the solar air heaters was taken as 0°,15°,30°,45 ° facing, BOSCH air blower (GPL620 to pass the atmospheric air in to the test section.due south.One hole were made at the middle of the collector to measure the Air temperature by using K-type thermocouples.Two thermocouples were attached at the inlet and exit to measure the air inlet and exit temperatures.

Analysis of Parameters
The accuracy of solar air heater (SAH) achieved higher level using a quatrefoil shaped artificial roughness in the collector area and compared with solar air heater (Old Method without any roughness.The comparison is made on the same day which is tested in the clear sunny day.day goes nearby noon it provides the more temperature to the heaters.It shows the maximum value at 2.00 PM from this outlet temperature of SAH with quatrefoil shape have higher compared to Conventional SAH.The average values are measured for different air flow rate.Both Conventional and Quatrefoil measurements are made.Ti, TM and Te are measured at different air flow rate.Initially it is tested using at the airflow rate of 3.0 m 3 /m in Table 1 and it is increased to 3.5 m 3 /m and also for 4.0 m 3 /m mentioned in Table 2

PREDICTION OF SOLAR AIR HEATER USING ADAPTIVE NEURO FUZZY INTERFERENCE SYSTEM (ANFIS)
An ANN called ANFIS uses the Takagi-Sugeno system for fuzzy inference as its foundation.It integrates neural and fuzzy systems in a particular way.It maps the input membership function to the set of fuzzy rules.It is a system with continuous function with greater accuracy.Fuzz Inference System is designed, and parameters are adjusted using Back Propagation Algorithm either by using Least square methods.It is combination of fuzzy layers, normalization layer and de-fuzzification layer and its output layer [11] Fig.

Conclusion
In this study experimental investigated the performance of the quatrefoil shaped artificial roughness SAH and the flat conventional solar air heater.The final temperature of artificial roughened SAH and Conventional SAH are 670C and 470C respectively.Due to increase in thermal area and increase the turbulence lead to higher heat recovery rate of quatrefoil shaped artificial roughness is 47% more compared to conventional type.The Performance of solar air heaters depends on radiation and features of plate which absorbs heat.In future the project will be implemented using Double pass with different dimension of quatrefoil shaped artificial roughened SAH.The Prediction of SAH using ANFIS is done and obtained accuracy is 90% with the testing error value of 1.1789.The Accuracy will be increased by using different machine learning algorithm in the future work.

Table 1 Average
. Table III.increases Air Flow Rate Of 4.0 M 3 /Mi & Table IV.increases for Different Solar Irradiation Values of Parameters for the Air Flow Rate Of 3.0 M 3 /Mi

Table 2 Average
Values of Parameters for the Air Flow Rate Of 3.5 M 3 /Mi

Table 3 Average
Values of Parameters for the Air Flow Rate Of 4.0 M 3 /Mi

Table 4 .
Average Values of Parameters for Different Solar Irradiation

3 .
Training and Testing using ANFIS ANFIS is mainly chosen for nonlinear structure, and it is mainly used in regression or S predicting the experimental output with a greater accuracy.In this proposed methodology the ANFIS is implemented to predict the measures of solar air heater.It tends to predict the efficiency of solar air heater efficiency by implementing input values to the system is shown in Table5.It describes the training data by dot and testing data by asterisk.Six inputs are given in the model and it generates output is shown in Fig 3.For each individual input, 6 Gaussian membership function are generated which leads to an output.The Gaussian function is implemented by using Fuzzification process by generating Sub clustering-FIS System

Table 5
By using the FIS System, the training and testing are done using sub clustering methods.The training is denoted by *(Asterisk-Red) and testing is done by dot (Blue Color).The average testing error is 1.1728 which is shown in Figure3.The training Epoch is 100 and optimization method is hybrid.The files are included for training and testing using workspace in MATLAB.ANFIS Usually predicts the soar air heater experimentation result by combining learning ability of neural network and set of rules provided by fuzz logic.Table6measures the hyper parameters.The MSE of testing value is 1.1728 and accuracy is measured as 90.In this by using Measures of accuracy include Correct Positive (TP), False Positive (FP), True Negative (TN), False Negative (FN), Sensitivity, Specificity, Precision, Recall, and F-Measure Values.The accuracy of Prediction value is 90, Sensitivity is 94.4,Precision is 95, recall is 95 and F Measures is 95.