In this study, machine learning techniques including Linear Regression, Extreme Gradient Boosting, and Genetic Algorithm Optimization are
*et al.* Multi-objective optimization of daylighting performance and solar radiation for building geometry using a hybrid evolutionary algorithm. # Multi-objective optimization of daylighting performance and solar radiation for building geometry using a hybrid evolutionary algorithm. Parametric modeling based on the additive and subtractive design generation algorithms included in EvoMass on the Grasshopper platform, with the goal of minimizing the solar radiation variation between summer and winter on building envelopes and maximizing useful daylight illuminance (UDI). First, parametric modeling is conducted using EvoMass’s design generative algorithm, followed by building performance simulations based on two key performance metrics: solar radiation and daylighting. Considering the climatic characteristics of HSCW regions, the optimization results indicate that a moderate atrium scale, shallow floor depth, south orientation, self-shading geometry (such as cantilevers and elevated floors), and the utilization of shading generated by surrounding buildings can significantly improve the building’s solar radiation and daylighting performance. Multi-objective optimization of daylighting performance and solar radiation for building geometry using a hybrid evolutionary algorithm.
As a result, we proposed hybrid solar irradiance forecasting models in which the cuckoo search algorithm (CSA) and adaptive moment estimation (ADAM) are used to optimize the weights assigned to the ANN’s edges. The literature review reveals that ANN based solar irradiance forecasting models experience a decrease in prediction accuracy when confronted with highly fluctuating solar irradiance data, and that the random assignment of initial weights to the edges of the ANN further affects the prediction performance of the model. The proposed solar irradiance prediction model, shown in Fig 11, depicts the functionalities of machine learning (ML), artificial neural networks (ANN) with a meta-heuristic algorithm known as cuckoo search algorithm (CSA) and gradient-based optimization technique ADAM. In this paper, a CSA-ADAM optimized ANN model for solar energy prediction is proposed, in which the initial weights at the edges of the ANN between the different layers are determined using the cuckoo search algorithm, a meta-heuristic method. It is observed that our proposed hybrid CSA-ADAM optimized ANN model performs better in terms of solar irradiance prediction accuracy.
Solar radiation modification (SRM), also called solar geoengineering, is a group of large-scale approaches to reduce global warming by increasing the amount
This study focuses on improving ANN-based techniques for precise solar irradiance prediction as the prediction accuracy of an artificial neural
This article presents a review of different hybrid SVM models with SOA applied to obtain the best parameters to reduce the prediction error of solar radiation.
SRM attempts to modify the amount of solar energy that reaches the Earth's surface, effectively offsetting some of the warming effects of greenhouse gasses.
zg7}O5Tq/ѐ\30'b=|Am['uO9>/c+ :Q/o"o~Wz\*^ r -M]EY|~ scjc24ϹVW;y;𮁽I3[HD'O(){wcwVf>ӏؒ'D[#)t6N ޕA}<b xu=/C+X q\*&h '+ ( ( .4cy, T:msom{Z@:P}&@uA/jkX}wBfT)`{\n֑GQ;!(p(\OS\_|~}5\_xq6؞7 DMX72іA5oP40Ҩ>l܀u"/}@iz8ZyO5J5&u )/bT<va\=B Ky9xom`QMQ/bl/%AAxS2<d3::o8dGCхޓg5|NZD#.o={oFsTܯZJ#ׯR,A"aM^>oٻT ;yww;$t` ^ 4CJ1~ cͼhz|g/k۱]9˸n~sU};v>r4|JrlV1d i|F&BPPy:(b9>ʻ'h~H ]`)ȦD A;n(Anڣ@cEX-WhPu\_ٙZ9J1qb;|T\*1W5gE$ &W JuP4ܡbC\_cW5fSg @{j V q-ֿhb \*O<1 yb u%c`Sk3Kx6t#\* {\_{\'l: X&-1N;>x`. I@Y B>μ]/<\_~Г@!}~UkSk} PcBmKsifΪ&{U/ nu P㼃NJ^ %uh揾n=Ǚ\_W3B,QaߎÁ@0Ϯ9cW{坵%8/|߯6Bg==&<cgٗ{ӿfoN@8cAro{ETCJ!/x tУ~klvJD[yeAްIE>Rx\_aЉEٓκ? Ÿ\!8%Q<4귎K/&NMgvWW7{5 q endstream endobj 119 0 obj <>stream HlUKo6WlOi]t R+[Y4Hځ;á83|j/V6E{N&C[Trѻatqɲm\*Y8oiw\_vW~ j݇X~^ɺP]+|{f;dA$}<OYpZraՇGx=wo[J5j$NLȭ [QHȕH=Sɻބ`{7hf^ʋVY`un'Ɇ~;.K=$t66P59&ϖ%&4k-/m}|4 nN:QPLr\*xCIB] b0ZIi"lfFh&v!I]Bϳ4Kr(ex-au|'taMQnHn.vLzO̓iƿD KXV0zF1R6DV&u j:)op6 Xt e%G =Ɛ-=Bf=+hti٥1}Ε킸Di0}@iU"h/yCJ\rс,L%>stream HTMs WpT:5 $qML;Ӊoq"1 'm~}Y·{tr0v}YAr^}ದ2<,\*t,htö7.ГeTR& e⍸c =\*3-J C[oZ{~RR.e~9x<N+uk[e/.V'XP[y!̯xu\*ߍ=D4s붻ɏʷ}/w "ص ΛʰqJOaǭ1mJ\_UZ:E^'8jx8.l m1I8k`H Z`q%LqugSE a:ÌƛWpo1a~WaYqӋFLUIFX)yAK! rs>ps<w̷7t~ګԗ ;;9; L|JxJpgŻ33nG =ʕJOGu~g7n-r ޠ](W).ʎes}}E/) endstream endobj 126 0 obj [/Separation/Color1 124 0 R 157 0 R] endobj 157 0 obj <>stream hS\_Ѹ\*+ymukýHPPD@qEA\`ܳp8ǎݾmo7uc ZlkM:rźl0Ylиf[\]7f~:ls&ˬɢ[ZWfѬ^4O8><6o3fΨgvQ1 j j=^Z虜9Ĝd|V<D\*]jsdCmWjۆ5 uB<4 SI(`xmX VFJB2R.V%aHRp:`Q} @vknK?2z{P( R@JH2%|q\_։(u.H$Uw&WuĊ6/oM(o/kflIs, Sa G] |ㇰy08Ī d 5ЇQ̪ͨnr (K~JLrQ8\_S@rgr'#}r<~C`N̅xkBk<E<9.6)$862CwA wDݎL #ԛaДk!!+$x9x)x10^H:O8O8G8G8x ??p;Dtb0P(# endstream endobj 127 0 obj [/Separation/Color3 124 0 R 158 0 R] endobj 158 0 obj <>stream hdr0Ќw^{!` H>l40®жY6j\-\_jh)-hokҴ(3ɨ8hp . RH|wm75& \_u;l\_|pYSկˣ\*rrvsՆ[՚obS!=vxO^(- endstream endobj 128 0 obj [/Separation/crosslinkcolor 124 0 R 156 0 R] endobj 129 0 obj [/Separation/LargeCUPcolor 124 0 R 159 0 R] endobj 159 0 obj <>stream h;qu{<l!-$2sB(ȨQQ2 qy^?/v3h{zwۻ+r{w}k-ŕd+d>9<> Fӑtxz~pr Fѩ=<90l@pSwmj ):vW^QC7ZjSHR-Q6VܲR,We L! 3epjQ.X esR/EKcя# ywJ@Τ=!